La compañía de tech que más ventas consigue por empleado del mundo es....
OnlyFans brings in more revenue per employee than Nvidia, Apple and Google combined. Everyone picked the wrong career
Politicians, athletes, journalists, NGOs, companies... They all use corporation-owned social networks to communicate with us. Citizens shouldn't have to hand over their data to companies to read content published by public entities. Public Feeds changes that by making posts of popular accounts freely available. Read more.
Keith Rabois (born March 17, 1969) is an American technology executive and investor. He is a managing director at Khosla Ventures and a co-founder of Opendoor. He was an early-stage startup investor, and executive, at PayPal, LinkedIn, and Square. Rabois invested in Yelp and the Xoom Corporation prior to each company's initial public offering (IPO). For both investments he insisted on being a board of directors member. More on Wikipedia
Keith Rabois reposted
La compañía de tech que más ventas consigue por empleado del mundo es....
OnlyFans brings in more revenue per employee than Nvidia, Apple and Google combined. Everyone picked the wrong career
Keith Rabois reposted
For years, a widening trade deficit often reflected greater reliance on cheap foreign consumer goods, but the latest figures illuminate what many headlines obscure.
The trailing 12-month goods deficit is 1.6% of GDP when excluding SI-related goods, its lowest reading in records dating back to 1994. When including SI-related goods, it stands at 3.5% of GDP, lower than any reading outside this current Administration since 2010.
Recent import growth has been concentrated in capital goods that support investment and expand our capacity to produce at home, a shift increasingly reflected in stronger output and productivity across the American economy.
Keith Rabois reposted
I have a major result to share.
I’ve found conclusive forensic evidence that the top 4 Chinese AI labs—DeepSeek, Moonshot, MiniMax and Ant Group—have built their newest models on NVIDIA’s most advanced Blackwell chips, which are barred from export to China.
This sounds insane, but I discovered that almost every AI model carries a hidden signature of the GPU that trained it. When a model is first set up, the starting values of its weights are laid out in a pattern that depends on the exact GPU. Each combination of GPU and seed number produces a unique pattern—a sort of fingerprint of the hardware used to create the model. Traces of that pattern survive training into the published model.
Again, this evidence isn’t based on any insider information or leaking. It’s based on a forensic analysis of the model weights released by the companies themselves.
I searched more than six trillion combinations of chip layout and seed number and matched them to open-weights models. In doing so, I found conclusive evidence that six Chinese AI models—including DeepSeek-V4-Pro, Kimi K3, and MiniMax-M3—were initialized on NVIDIA Blackwell systems. Four of the models match Grace Blackwell, the GB200 and GB300 parts in NVIDIA’s flagship liquid-cooled rack-scale systems.
I then verified these results on real hardware. Running unmodified PyTorch with the seeds I’d recovered, a Blackwell B300 reproduced DeepSeek-V4-Flash’s untouched starting weights bit for bit, and a rented GB200 reproduced Kimi K3’s to a correlation of 0.999995. A Hopper H100 reproduced none of DeepSeek-V4-Flash’s, and changing the seed by one broke the match. (DeepSeek’s seed is 953049583 by the way.) The method also was verified to match 24 known open-weights models to the hardware that developers said they used.
It is exceedingly unlikely that these results are a coincidence. Across six trillion combinations, the best accidental match should reach only about 8 sigma (i.e., eight standard deviations above what random chance would produce). The real matches range from 49 to 1,385 sigma, and even the weakest single test hits 20.
These were data-center-scale training runs. With some analysis of release schedules and model details, I estimate that the largest, Kimi K3 and DeepSeek-V4-Pro, used 2,000 to 3,000 Blackwell GPUs each. Clusters of this size cost on the order of $150 to $200 million each to buy, and across these labs alone would add up to more than half a billion dollars of export-controlled GPUs, not counting experimental compute or otherwise. It’s quite likely that we see multi-billion dollar expenditure on export-controlled GPUs.
I am highly confident in this result, and it has been independently verified by senior researchers at OpenAI and Anthropic. You can read the full report at unemploymentlabs.com/checkpoint-forensics.
There is much more I’ve identified about these models through other forms of checkpoint forensics like this, and will have more to share soon.
Keith Rabois reposted
I'm increasingly convinced that maintaining high aerobic fitness may be one of the most important things we can do for brain aging.
Brain MRI research has linked higher aerobic fitness with less grey-matter shrinkage across the lifespan in a region involved in attention and working memory. Older age was associated with smaller gray matter volume there among less fit adults, but not more fit adults. Larger volume in that region was also linked to better cognitive scores.
Long-term observational evidence has associated each 1-MET improvement in estimated fitness with a 16% lower risk of dementia. Separate exercise studies have reported increased glucose uptake in brain regions vulnerable to early Alzheimer’s-related changes after HIIT, and improved brain blood flow after aerobic training.
VO2 max is a capacity we can measure and train, which is why I keep coming back to intervals like the Norwegian 4x4. Taken together, these findings make aerobic capacity hard to ignore.
Keith Rabois reposted
America should not be pressuring @Israel to show restraint against Iran. We should be working to stop Iran's nuclear drive.
Keith Rabois reposted
18% of NYC private employees work at "non-profits"
The scale of fraud taxpayers like me are funding is truly unimaginable
Keith Rabois reposted
“New York City was a major outlier and reported more violent crime in all four categories, including 24.5% more homicides and 14.2% more robberies. Its combined violent crime total was up 5.1%.”
www.axios.com/2026/11/12/violent-crime-homicide-us-cities…
Keith Rabois reposted
.@rokhanna Americans are struggling while you are insider trading?
Well, well, well
Rep. Ro Khanna sold $NVDA right before Nvidia's CEO admitted to having a Chinese spy side piece
He's a member of the following:
- Strategic Competition between the U.S. and the Chinese Communist Party Committee
- Taiwan Caucus
- High Tech Caucus
Apparently, federal agents informed some lawmakers last year that Jensen Huang might be the target of a foreign intelligence operation
He avoided a ~26% dip
Keith Rabois reposted
pretty amazing trajectory for the biz @ImprintPayments
founding --> ~$2B val + on track for ~$400M GAAP rev in 6 yrs + feels like the type of biz that compounds at scale...
HUGE fintech round as @ImprintPayments raised $130M in a round co-led by @khoslaventures + @psumvc at a $1.88B val (up from $1.2B last Dec)
Imprint builds + runs co-branded credit card programs for brands, owning the full stack from issuing + processing to rewards + loyalty
some stats on the biz:
> expects ~$400M in GAAP rev this year, up 150% (w/ headcount only up 30%)
> 2M+ cardholders, up from ~15K 3 yrs ago
> 130M+ txns processed to date
customers include Kroger, Albertsons, Shell, H-E-B, @bookingcom, Rakuten + now Fanatics
www.upstartsmedia.com/p/fintech-imprint-raises-130m
Fintech on an outrageous heater rn
@ImprintPayments just raised $150M Series D from @khoslaventures at $1.2B val w/ participation from @ThriveCapital, @RibbitCapital, & @kleinerperkins (2x their Series C val 12 months ago)
Imprint provides co-branded credit cards for companies to modernize their loyalty programs. With the round, they plan to expand into debit/secured cards/financing and become a broader financial platform
They shared some metrics on their programs:
- 2x higher wallet share for cardholders vs. non-cardholders
- 8x increase in lifetime value for cardholders vs. non-cardholders
- 20% increase in cardholder spend vs. prior program
www.bloomberg.com/news/articles/2025-12-17/khosla-led-dea…
Keith Rabois reposted
Defend in real time
Introducing Security Factory: a new way for engineering teams to turn security findings into tested fixes by connecting investigation, remediation, and verification in one continuous workflow.
Now in Private Preview.
Keith Rabois reposted
HUGE fintech round as @ImprintPayments raised $130M in a round co-led by @khoslaventures + @psumvc at a $1.88B val (up from $1.2B last Dec)
Imprint builds + runs co-branded credit card programs for brands, owning the full stack from issuing + processing to rewards + loyalty
some stats on the biz:
> expects ~$400M in GAAP rev this year, up 150% (w/ headcount only up 30%)
> 2M+ cardholders, up from ~15K 3 yrs ago
> 130M+ txns processed to date
customers include Kroger, Albertsons, Shell, H-E-B, @bookingcom, Rakuten + now Fanatics
www.upstartsmedia.com/p/fintech-imprint-raises-130m
Fintech on an outrageous heater rn
@ImprintPayments just raised $150M Series D from @khoslaventures at $1.2B val w/ participation from @ThriveCapital, @RibbitCapital, & @kleinerperkins (2x their Series C val 12 months ago)
Imprint provides co-branded credit cards for companies to modernize their loyalty programs. With the round, they plan to expand into debit/secured cards/financing and become a broader financial platform
They shared some metrics on their programs:
- 2x higher wallet share for cardholders vs. non-cardholders
- 8x increase in lifetime value for cardholders vs. non-cardholders
- 20% increase in cardholder spend vs. prior program
www.bloomberg.com/news/articles/2025-12-17/khosla-led-dea…
Keith Rabois reposted
One month into my time at @factory and one core assumption is validated:
The race to the enterprise will be won by AI companies that build the best infrastructure.
To unpack a bit:
-Enterprises are riddled by legacy, debt and off the shelf tooling. Integrating, modernizing and making these systems AI enabled is a key challenge. This isn’t a simple vibe coding task, it takes serious infra chops.
-The infra is the foundation on which the product lives. If it is shaky, the customer experience will suffer and customers will churn.
-Day 2 support really matters. You can’t just deploy and forget. Durability, reliability and support are key.
-Standardizing the deployment configurations from the beginning is a must. If each customer cluster is a snowflake, you are doomed.
This is why I’m pumped about @factory. We are solving the hard infra problems head on and not kicking the can down road.
The team is stacked, come build with us 🚀
Another outstanding NYC company.
EXCLUSIVE: fintech Imprint has raised $130M at a $1.88B valuation.
CEO @daraghjmurphy shares with Upstarts how the co-branded card startup is embracing AI internally and via agentic commerce to double revenue to $400M, while growing headcount just 30%.
www.upstartsmedia.com/p/fintech-imprint-raises-130m
Keith Rabois reposted
The things that are most unaffordable for young Americans are health care, housing, and school - which are exactly the sectors w/ the most gov't involvement and regulation; meanwhile, sectors w/ little to no gov't involvement have seen massive price drops:
Government meddles -> Prices go up
Keith Rabois reposted
This is the piece I wrote last year arguing NVIDIA was working to help China win the AI race.
Their GA team reached out within hours before coming to my office to scold and challenge me to present the serial number of a single smuggled chip.
www.city-journal.org/article/artificial-intelligence-chin…
Keith Rabois reposted
Living interfaces that self customizes to a specific user are inevitable given the advances in AI. Having worked on search and ads systems at Facebook, Coframe's compounding data advantage was dead obvious and is one of the most classic data advantages in tech.
Coframe was my first investment at Khosla and I'm glad to double down alongside @BryanOffutt and Hanabi and join the board!
The internet is dead. We’ve raised $35M total funding led by Khosla and Hanabi to bring it to life.
@coframe is a growth engine for your website/ads/more which tests, personalizes, & learns 24/7.
It’s working. Dropbox increased conversion by 158%. Replit, 410%. Cash App, 60%.
Keith Rabois reposted
⚠️ BREAKING: California will now require ID to purchase certain creatine supplements, banning sales to anyone under 18 starting January 1, 2027.
Creatine is one of the most studied supplements in the world...
Yet a 13-year-old can walk into a gas station and buy an energy drink, candy, and a soda.
Make it make sense.
Keith Rabois reposted
The 'flatiron belt' is slowly but surely becoming one of the most important technology hubs in the world and i'm here for it
Keith Rabois reposted
Absolutely insane that the Department of Commerce still refuses to place Megaspeed on the Entity List, despite knowing it is run by Chinese spies. It's difficult to draw any other conclusion from this than that Commerce leadership is just utterly uninterested in protecting US national security. Totally disqualifying.
Also, Megaspeed is undoubtedly only the tip of the iceberg. Commerce needs to urgently fix the gaping hole in US export controls that Southeast Asia represents, and at a minimum impose a license requirement for the entire region, and likely for all large-scale orders to untrusted companies like Megaspeed. The fact that Commerce refuses to even block shipments to companies known to be running by Chinese spies is a total dereliction of duty.
And Nvidia claims that they placed Megaspeed on their "block list" and stopped shipping to them in 2025. But if that's the case, why is Megaspeed still listed on Nvidia's website as an official cloud partner? And did Nvidia also stop shipping to Speedmatrix, Megaspeed's subsidiary in Malaysia? This denial doesn't add up at all.
SCOOP - US commerce department delays decision on #megaspeed’s possible inclusion on trade blacklist
www.ft.com/content/afb6c960-a910-4d67-9588-73ba3bea5b06 via @ft
Keith Rabois reposted
Basically every CEO in my >top< performing investments works this hard, and works this schedule
It just compounds
You don’t have to. It’s just, it sure seems how you pull ahead.
You can ship and build just so, so, so much more with AI this way.
Mercor just raised a $750M Series D at a $20B valuation
Here’s CEO @BrendanFoody’s daily routine:
› Wakes up around 7:45
› Works out most days ~7:50–8:30
› In the office by 9
› ~4 hours of meetings a day
› Spends a third of his time on hiring, a third with customers, a third on research & product
“I’m in the office until 10pm, & I rinse & repeat 6 days/week. And then Sundays I work from home.”
“I still inbox zero every day because I want to make sure I get through everything.”
Keith Rabois reposted
Fact that this isn't more of a scandal is jaw-dropping.
We have an incoming United States Congressman who volunteered for an Al Qaeda front group and testified as a defense witness for a jihadist convicted in the plots tied to the 1993 World Trade Center bombing.
This man is a fifth columnist. His holding office in the United States government is exceptionally dangerous.
Insane exchange. Newly elected Democrat Adam Hamawy refuses to call Iran an adversary.
BROWN: Do you see Iran as the adversary?
HAMAWY: "We pushed them into a fight"
BROWN: But are they the adversary?
HAMAWY: "We should be making friends"
BROWN: They've been saying "death to America" for decades
HAMAWY: "We've been saying 'death to Iran'"
BROWN: They've killed a lot of Americans
HAMAWY: "And we've killed a lot of them"
BROWN: So you don't see Iran as an adversary?
HAMAWY: "I see that we are pushing Iran into a fight"
Keith Rabois reposted
HE BUILT THE WALL!
🚨Trump’s Border Wall Nears Completion
By the end of Trump’s second term, the Trump administration is on pace to complete 1,418 miles of primary steel bollard wall, 708 miles of secondary steel bollard wall in high-traffic areas, 536 miles of waterborne barriers, and 536 miles of Smart Wall technology coverage in areas with otherwise impassable terrain.
Most importantly, Trump’s new border wall is set to become a permanent feature along the United States’ southern border.
In California, Arizona and New Mexico, the wall sits on the Roosevelt Reservation, a 60-foot-wide strip of land owned by the federal government along the border. In Texas, the federal government has been acquiring private land through eminent domain. Once the government takes the title, the land stays federal.
And even if Democrats regain control of the federal government in the future, Trump’s border wall is set to remain in place. When construction was paused in 2021 under Joe Biden’s directive, the existing wall stayed up despite the Biden administration’s best legal efforts to take it down.
In essence, President Trump has not only illegal-alien-proofed the southern border, he’s Democrat-proofed it as well.
Follow: @MorseReport
Keith Rabois reposted
The @ManhattanInst has just released their 2026 College Rankings!
Looking for a college with a strong curriculum that encourages genuine debate, prioritizes education over activism, and actually prepares students for real-world success?
First place: @UTAustin
Last place: @oberlincollege
See the full list:
🔗collegerankings.city-journal.org
Keith Rabois reposted
NEW:
Forget Alice Huang's role at Megaspeed, that's almost a distraction.
The real story is her role at China-based Shanghai Century Litong, which became a "preferred" Nvidia cloud partner in mid-2023 and facilitated >$7.5b in Nvidia chip shipments, incl securing BANNED chips.
Keith Rabois reposted
Ooooh nooooooo. This thread (note the dates)
NEW: We are short Nvidia $NVDA. We believe Nvidia has a significant China problem. Please see our full report and disclosures now available on our website, link in bio.
Keith Rabois reposted
“DeSantis made Florida a beacon for conservatives across America who came in droves from places like New York City, now, it may become a lighthouse, guiding a national Republican Party adrift at sea back to the shores of electoral victory.”
My Latest
Keith Rabois reposted
An 18% flat tax would produce the same tax revenue, if anyone is interested in everyone paying their fair share.
THE IRS JUST RELEASED THE NEW 2027 🇺🇸 TAX BRACKETS. HERE'S HOW THEY WORK IN PLAIN ENGLISH
Every bracket got a little bigger to keep up with inflation, so you can earn a bit more before moving into a higher rate.
Step 1: Take what you earn and subtract the standard deduction. That part isn't taxed at all.
- Single: $16,600
- Married: $33,200
Step 2: Whatever's left gets taxed in layers.
Single filers:
- 10% on your first $12,800
- 12% on income up to $52,025
- 22% up to $109,125
- 24% up to $208,325
- 32% up to $264,550
- 35% up to $661,375
- 37% on anything above that
Married couples: the cutoffs are roughly double ($25,600, $104,050, $218,250, $416,650, $529,100, $793,650).
Example: You're single and earn $76,600. Subtract $16,600 and you're taxed on $60,000.
- 10% on the first $12,800 = $1,280
- 12% on the next $39,225 = $4,707
- 22% on the last $7,975 = $1,755
Total: about $7,742 in federal income tax, or roughly 10% of what you earned.
Even though you're "in" the 22% bracket, only your last $7,975 gets taxed at 22%.
Keith Rabois reposted
NYT reports on an additional wrinkle to the Nvidia spy scandal.
They found that several of the Chinese companies that the Commerce Department approved to buy H200 Nvidia chips don’t seem to exist or have direct ties to the Chinese military.
www.nytimes.com/2026/11/07/business/e
Keith Rabois reposted
JUST IN: Economic analysis of 64 countries finds that nations with more billionaires per capita tend to have fewer people living in poverty.
Keith Rabois reposted
The most comprehensive poll of foreign policy views of MAGA voters under 30 was done by the Reagan Institute this summer.
Among other findings 73 supported military action in Iran, and 61 percent wanted to see the Iranian regime either removed or weakened, while just 35 percent said they prefer a “negotiated settlement” in which “Iran’s current government remains in exchange for verifiable limits on its nuclear and missile programs.”
Overall, MAGA Republicans are more hawkish and less isolationist than any other segment of the American electorate. And young MAGA voters don’t want America to pull back from the world; they want to lead it.
Though younger MAGA voters are less supportive than older ones, solid majorities still back the Jewish state: 60 percent of MAGA voters under 30 approve of sending U.S. weapons to Israel, compared with 72 percent of older MAGA voters — far above the paltry 37 percent of Democrats who do. And 63 percent say that the security of Israel “matters to U.S. security and prosperity.” They are not demanding we “change our relationship with Israel” as the VP claims.
Sorry, but the whole narrative of “Zombie Reaganism” driven by older Republicans with gauzy memories of a bygone era of 1980s GOP world leadership is a myth.
www.washingtonpost.com/opinions/2026/07/09/young-maga-vot…
I like both men. I'm adult enough to handle that Vance was publicly addressing the undeniable reality of the GOP wipeout, caused by deep national frustration with a war and its resulting spikes on gas prices and other costs, while Rubio was talking to a tiny group of people at a fancy gathering who desperately need to be comforted following that shellacking.
It's unfair to compare the two sets of remarks. If they switched their audiences and roles, they might say something similar as the other one did. That's just politics.
HOWEVER, the less people care about whether either one of these guys will win the (likely way beyond Pyrrhic) nomination that they may or may not seek of a party in crisis thanks to the failures of its leadership across all branches of government, the better the party can fix the problems that brought it to this point and the better the party can work to repair the harm done. Otherwise, it doesn't really matter who gets the nomination!
Now I know that it's absolutely true that many older GOP voters fully support the war and fully support Trump's decision to do what he did. And most diehard GOP voters in other age groups would continue voting GOP even if they were less enthusiastic about the war. We see a lot of representation of the first group in particular in the media.
I know it's unpopular to say so, but these groups comprise a dramatically shrinking minority of voters that is unable to deliver election wins. Fully 65% of the country opposes the war. It's not a winning issue. Prosecuting it resulted in serious electoral costs -- largely due to the increased gas prices but also because of the sense that the Trump administration was not focused enough on domestic crises requiring major work. You may wish it were otherwise. I'm sure you do. But the reality is that this war has never been popular and has gotten less so. Thoughtful people should care far more about the very real concerns of the recent GOP voters who left or gave up and not just the ones who don't leave.
So you can pick the person who most aligns with your particular views on the war all you want! Donors and special interests can get behind whichever person they want and maybe they even have the power to get him the nomination! They frequently do! But it's a very, very, very, VERY long road between picking that nominee and returning the Republican Party to any semblance of power.
I'd care less about the fight over who gets nominated and much more about building a winning coalition going forward.
Also, as someone who genuinely likes Rubio, I think his vocal supporters may be hurting him by positioning him as a huge fan of the war. For whatever it's worth.
Keith Rabois reposted
This is actually crazy…
Stephen Miller told his Duke University professor in 2003 that he wanted to be the number 2 in the White House (chief of staff) under…
President Trump.
Stephen predicted this in 2003.
Unreal.
Keith Rabois reposted
Totally contradicts the idea that people that moved from California or other states into Florida we’re making it more liberal
New - The longer you lived in Florida, the more likely you were to vote Democratic.
🔴 1-5 yrs R+25
🔵 11-20 yrs D+5
🔵 20+ yrs D+5
NBC exit polls
Keith Rabois reposted
Manufacturing jobs jumped again in Oct to highest level since Dec '24 - as we can see on this chart, manufacturing was in a 3-yr slide before the turnaround started this year; the sector began hemorrhaging jobs at the very start of '23:
Keith Rabois reposted
JUST IN: U.S. homelessness records its largest annual decrease in nearly 20 years.
Keith Rabois reposted
Dems solving the SF housing affordability crisis by crashing ANT
*INVESTORS SAID TO DISCUSS ANTHROPIC IPO VALUE OF $1.2T-$1.3T
Keeps going lower
Keith Rabois reposted
40 feverish minutes of zoomed in detail on what’s happening in venture markets, with @ramtinnaimi
We discuss:
- cuck tranches
- why rounds are moving so fast
- how today compares to 2021
- what record deployment means for LPs
- the Anthropic S-1 and compute
- seed risk at Series D prices
The full episode is available everywhere you get podcasts (links in comments)
Enjoy!
Keith Rabois reposted
the worst time to start a company is when everyone thinks it’s the best time to start a company.
people naively think raising $ is the hard part of building a company when it’s actually (1) hiring world class people + (2) competing for customers
both become harder for early stage startups in frothy markets.
all legendary companies were started at the beginning of market cycles, not after markets became frothy.
Missed by 2 days:(
Passed a gas station with gas at $2.96 today.
It’s truly remarkable how easily and willingly people will be manipulated.
Keith Rabois reposted
Insane that New York continued its rightward shift despite the nation being 16 points to the left of 2024, NYGOP really held their own despite the doomsday environment for Republicans
Future looks bright over there long term
Keith Rabois reposted
With the midterms behind us, Republicans in Congress will soon pick its committee ranking members. The ranking members will be the faces and Republican leaders of the respective committees.
Those roles are largely still decided by seniority (who has served in Congress the longest), not merit, ability, or competence. It’s an outdated system. Republicans should replace it and choose their most effective and talented members to lead.
Keith Rabois reposted
Republicans overwhelmingly support Trump's tariffs or want them even higher.
Keith Rabois reposted
Let me tell you a story about how unethical Anthropic is.
In April, the Anthropic team began heavily courting our engineering team to tell them corporate secrets, technical details, and future plans for @cfosilvia.
They pretended their interest was due to our exploding popularity and accelerating compute usage, but it quickly became apparent that Anthropic was fishing for details on how we built our product and why it was so popular.
This included the Anthropic team trying to wine-and-dine our engineers, including a infamous steak dinner where someone told our team "don't worry about it, this is on Dario."
Our team was not born yesterday, so we quickly realized what was happening and refused to share sensitive information and technical details with the Anthropic team.
As we suspected, Anthropic was trying to get information from us, so they could build a competing product. Some people on their team must have a guilty conscious because they messaged us a few days ago to give us "a heads-up" that they were launching a competing product.
Thankfully, the Silvia engineering team has been aggressively working to get off the frontier models for the last ~ 6 months. That effort has been highly successful, because our compute costs are ~ 97% lower when we use our own model & hardware, the latency is lower, and the accuracy of answers materially increases as well.
The sad thing for Anthropic is they have now released an inferior product and they had to sell their soul to do it.
The only thing worse than being a trillion-dollar company that can't beat a 10-person team is being a trillion-dollar company that no one trusts and everyone knows is unethical.
Serious investors continue to choose @cfosilvia over ChatGPT, Claude, and other frontier companies because these products are not optimized to solve the actual user problem.
Today, Anthropic launched a Claude wrapper that is masquerading as a personal finance product. Best of luck to them.
But I will take the Silvia misfits vs the Anthropic losers any day of the week.
if you want to support people trying to do things the right way, and help investors be more informed, then consider signing up for Silvia today: www.silvia.com
Introducing Personal Finance in Claude, in beta in the US.
Connect the accounts you choose, and Claude answers your money questions in everyday language, from your own balances and transactions. Access is read-only, and Claude can't move, manage, or invest your money.
Keith Rabois reposted
Anyone pushing the massive "Republican crossover vote" theory for what happened on Tuesday is delusional. I'm trusting NBC's exit poll, and I will keep citing it till I'm blue in the face.
Keith Rabois reposted
Let me get this straight.
Jobless claims just fell to 189,000, the lowest level since 1969. You have to go back 57 years to find a better number. We're now filing fewer unemployment claims in a normal week than we did in the single best week of 2019, before the pandemic.
The four-week average is down more than 13% from a year ago. People are working. The firing has basically stopped.
And groceries? Food inflation is running at 1.8%, below the 20-year average of 2.6%. Your grocery bill is climbing slower than it normally does.
So the labor market is the strongest in half a century and food is calmer than the 20-year norm.
And on Tuesday, voters turned around and punished the party in charge over the economy.
It's odd. You can put up the best jobs numbers since 1969 and still lose the room.
Ramp’ng.
How does a corporate card startup become a $60B AI product factory?
In our first on-camera interview a decade in the making, Ramp co-CEO Eric Glyman (@eglyman) answers just that question.
Today, @tryramp is best known for its corporate cards, or maybe its marketing stunts with the likes of Kevin from ‘The Office,’ Tony Hawk and a 70-minute Broadway musical about paying bills.
But Eric says his $60B startup, NYC's most valuable, feels less like a fintech and more like a productivity company driven by one simple question: If a business spends $100 to make $200, how can Ramp improve that ratio?
Nearly 100,000 companies use Ramp today for everything from expenses to managing AI spend. But Eric says that when he zooms out to the millions who still aren't customers, Ramp feels painfully early still.
On The Upstarts Podcast, Eric shares why he’d rather run Ramp like Amazon than Google; why Ramp hires for proof of work, not credentials, from a 14-year-old intern to a 28-year-old CTO; and how being the first customer for Anthropic’s Claude Code and Cognition’s Devin led Ramp to build its own AI tools.
Plus: The real story of Paribus, his first startup with co-founder Karim Atiyeh (@karimatiyeh); what they learned at Capital One; and a warning for Instinct and Muse.
This season is presented by @Atlassian and @chargebee 🫡
CHAPTERS
00:00 Introduction
01:51 What Ramp Actually Does
03:27 From Gym Charges To 1% Of U.S. Corporate Payments
05:13 Glass, Inspect And Building AI Tools For Itself
07:45 An AI Sales Assistant Before ChatGPT
09:45 How Ramp Decides What Products To Ship
11:14 First Customer For Claude Code And Cognition’s Devin
13:09 Why Ramp Wanted To Build Like Amazon, Not Google
15:42 Hiring For Proof Of Work: The 14-Year-Old Intern
18:43 A .300 Batting Average And ‘Growth Without Fear’
21:17 What Ramp’s Founders Learned At Capital One
24:00 The $1,000 Flight Refund Behind Paribus
26:34 Bots, A Cease-And-Desist, And A Lesson For Instinct And Muse
32:00 Marketing 101: Bill Pay The Musical And Kevin In A Glass Cube
36:01 Business Books And Helping People Do Their Life’s Work
39:03 How A 28-Year-Old CTO Catalyzed A Co-CEO Change
43:55 ‘Still So Early’: How Ramp’s Founders Stay Hungry
55:26 John Wooden And Why Ramp Competes Against Itself
Thanks for watching!
Keith Rabois reposted
The Democrats have solved the inflation problem in 24 hours:
Everyone is poor now.
Keith Rabois reposted
Young men with kids voted to the right of men without kids by 40 points.
Young women with kids voted to the right of women without kids by 35 points
New investment alert:
@rabois led our Series A in February. Today he's leading our Series B.
This year: 30M+ plays a day, 40M+ mini apps made, daily users up 20x+.
We made this announcement as a mini app you can play.
link.prod.sekai.chat/lYm9lR
Keith Rabois reposted
Makes today’s statement even more puzzling. Only 19% of Republicans said U.S. support for Israel is too strong. 59% said it’s not strong enough.
ABC News voter poll conducted from Oct 19 to Nov 3, 2026.
Keith Rabois reposted
Democrats created the affordability crisis and successfully shifted the blame onto Republicans.
The claim that the Iran war was the main cause of the blue tsunami is unfounded.
Yes, it did not help. But the exit polls show only 4% ranked foreign policy as their top issue.
Almost 40% said economics was the top issue, and tariffs were their top economic objection.
Keith Rabois reposted
Someone put a little @rabois / @shaunmmaguire in @gradypb's coffe this morning??
Keith Rabois reposted
Yeah anybody trying to claim immigration enforcement was the problem is being absurd
That's the thing Trump ran on, that's the thing where he most clearly fulfilled his promises, nobody who voted for Trump felt betrayed by his immigration policy
Exit polls are showing the GOPs perceived strict immigration enforcement was one of their few winning issues.
The Iran war was a massive losing issue.
Keith Rabois reposted
Voters agreed with Republicans on every policy at issue and then voted Democrat
lmao
Voter ID, caps on property taxes, and lowering income tax rates all winning 60-40 in NC…in an electorate that’s voting 60-40 for the Democrat for Senate
Keith Rabois reposted
Should have enforced the TikTok ban
Keith Rabois reposted
Funny. I wonder where they're getting these ideas? 🤔
.@NBCNews Exit Poll: Voters are concerned about AI and overwhelmingly oppose data centers www.nbcnews.com/politics/2026-election/live-blog/midterm-…
Keith Rabois reposted
I’m glad I live in Tennessee. Where people aren’t crazy.
Tennessee is ruby red. 9-0 Republican delegation for first time ever. Congrats to @BillHagertyTN on his re-election. Headed to emcee @VoteMarsha’s gubernatorial election party now.
Keith Rabois reposted
Voters who said data centers were a critical or major factor to their votes breaking hard for Dem Gov candidates:
- 59-39 for Bottoms in GA
- 58-39 for Hinojosa in TX
- 70-29 for Janoo in VT
- 66-33 for Sand in IA
- 64-31 for Acton in OH
www.cnn.com/election/2022/exit-polls
Keith Rabois reposted
The end of Western civilization.
Time to move to space.
Keith Rabois reposted
Texas flipping means data centers have got to go in space much faster
Another reason to move to Miami.
Keith Rabois reposted
Having led infra at Citadel Securities and OpenAI, I know how hard it is to understand why complex production systems fail.
The amazing @traversal_ai team is tackling this challenge with AI at enterprise scale, and I was blown away by how well their system works.
Proud that we @khoslaventures co-led their $120M Series B! Huge congrats @_anish_agarwal and team!
LFG!! 🚀
We just raised $120M, with a $1B valuation, to build software that keeps itself running.
AI is writing more code than ever. It’s also making production less reliable: more code, shipped faster, understood by few.
@traversal_ai understands how production systems fit together, finds the root cause of incidents, and helps prevent the next one.
Engineers at Nubank, Harvey, PepsiCo, DigitalOcean, and a growing list of Fortune 100 companies use Traversal every day.
Here’s production without the drama 👇
Keith Rabois reposted
I am now in favor of making Guam a state.
Board dinner scheduled tonight 630-830 pm so don’t read into any lack of posts.
Keith Rabois reposted
Election Day PSA: The problem with Republicans is when they don’t keep their promises. The problem with Democrats is when they do.
Keith Rabois reposted
Gonna be a boring Tuesday for me because I lost my wallet, I have zero forms ID, so the ONLY THING I can do in New York is VOTE.
Keith Rabois reposted
LMAO: 'CBS Mornings' admits it drove down to Florida and, after interviewing dozens of voters in Miami, Melbourne, and then Jacksonville at four separate restaurants, they couldn't find ONE SINGLE person who didn't like DeSantis.
Keith Rabois reposted
Today, we are launching Buy with Opendoor in beta in Texas. Use all of Opendoor’s scale to buy any home, Opendoor’s or not.
Save tens of thousands of dollars on your mortgage and fees.
www.opendoor.com/buywithopendoor
Keith Rabois reposted
Brands have been stuck in a loveless marriage with their banks for 20 years. Daragh Murphy thinks agentic commerce will hit the banks harder than the brands.
(0:00) Can Imprint Beat 300-Year-Old Banks?
(2:33) Who Loses More to Agents: Banks or Brands?
(7:33) What If Your Card Lived in the Brand's App?
(11:22) Why Co-Branded Cards Exist at All
(13:50) Why Does Fiserv Still Run on a Mainframe?
(15:57) Why Did Murphy Tank His Credit for Fanatics?
(18:28) Did the Apple Card Really Burn Its Bank?
(19:56) Are AI Agents Flooding Brands With Disputes?
(22:26) Should You Hand Your Cards to an AI Agent?
(24:39) Inside Imprint's $100M Infrastructure Bet
(28:29) Is the Bank Charter Rush Just FOMO?
(32:31) Should Founders Take an Incubation Deal?
(34:57) Why Do VCs Hate Consumer Credit?
(40:12) Sequoia Passed During His Honeymoon
(41:10) What's Inside Imprint's $400M in Revenue?
(44:53) Can Imprint Hit $10B With Under 300 People?
(48:26) Can an AI Agent Be Your Personal Trainer?
(50:06) What's in a Fintech CEO's Money Stack?
Keith Rabois reposted
Miami ranks #3 in the USA for @stripe’s Atlas developer-focused platform.
Bet against Miami and you are betting against America. 💪🏽
Beyond the largest startup hubs, Seattle ranks #1 among US cities for startup founders based on Atlas C corp incorporations. Emerging cities where Atlas founders are building today:
1. Seattle
2. Chicago
3. Miami
4. Atlanta
5. San Diego
6. Houston
7. Denver
8. Washington DC
9. Philadelphia
10. Portland
Someone knows how to conduct a poll and sanity check it.
Which Party are you voting for:
🔵 Dem +5
Which party are your neighbors voting for:
🔴 GOP +3
Echelon Insights Poll
Keith Rabois reposted
Nvidia is now requiring data centers to not sell more than 49% of their compute to Chinese customers. But in what world is selling 49% of the world's most valuable resource to China a "win" for America? The number should be 0%; if Chinese firms can't buy a chip due to export controls they also shouldn't be permitted to rent access to it.
The fact that this remote access loophole exists--and that Nvidia even felt compelled to announce this new policy because it undoubtedly has so many customers currently renting 100% of their compute to China--just shows how badly export controls are being mismanaged.
Nvidia orders data centre clients to reduce China exposure ft.trib.al/OntropD
Keith Rabois reposted
Narrative violation
1. More billionaires are associated with less poverty.
2. Increasing billionaires per capita by ten fold lowers poverty by 43 percentage points.
3. Sweden has more billionaires per 10M people than the U.S. (34 vs 30)
4. Correlation is fairly strong (R-squared of 0.6 for two variable cross-section data)
4. Since the graph is only a correlation, what do you think is the cause?
🔗in🧵
Keith Rabois reposted
Steve Kornacki: Democrats are way more unpopular than Republicans
“This is different from 2018”
Keith Rabois reposted
Results if you forced a hard recall with 1-2% "new actual voters":
Paxton+5
Hinson+10 (lol)
Sullivan+9
Husted+5
Marshall+5
WWC shares all shift 19% (!!!) rightward on average.
If you simply eliminate the Fantasy Island Nate Cohn idea that 22% of Alaska voters didn't vote in the 2024 election and took the remaining surveyed participates the Poll goes from Peltola+1 to Sullivan+9
That's assuming an actual valid (and correct) 1% of the voters are new
Keith Rabois reposted
Absolutely deranged position from El Sayed
.@AbdulElSayed says it’s ok for MEN to be in girl’s locker rooms.
Why?
“Well because they are going to find out about it on the internet anyway”
Completely unacceptable.
Keith Rabois reposted
“There is no moat.”
But moats are everywhere:
Ambition.
Taste.
Agency.
Curiosity.
Judgment.
Courage.
Obsession.
Persistence.
Integrity.
Patience.
This era belongs to those who refuse to be mediocre.
Keith Rabois reposted
really like this format for highlighting talent density on early team from Kalshi deck
Good read for early-stage founders 👀
Keith Rabois reposted
Music is a longevity therapy
> lowers heart rate
> lowers blood pressure
> lowers cortisol
> lowers anxiety and restlessness
Protocol: a song you love.
104 randomized trials
9,617 people
Excellent conversation.
One of the single most important things that I look for in anyone is the ability to admit that they were wrong. Reflect on it and improve.
I admit that I was wrong in previous discussions with @davidtisch at @BoxGroup. I did not think that you could produce outsized returns from hugely diversified portfolios (150+ companies) with still quite large fund sizes ($200M+). I was wrong.
With Cursor, Baseten, Clay, and many more incredible companies in his portfolio, he has proved his thesis without shadow of doubt.
This was an incredible discussion analyzing Box's returns over the years.
(Link in comments)
--
1. I Don't Want to Compete With the Best VCs, That Is a Trap
Owning 10% to 20% of a true outlier in the seed round means beating the rest of the market head-to-head and convincing the founder to pick you. That takes too many steps to believe, so do not compete directly against the best VCs in the world.
I would love to hear how you think about your position as a "lead" seed investor and if that has changed with the expansion of rounds @dafrankel @Bandrew @chadbyers?
2. What Concentrated Funds Know but Do Not Admit To
Concentrated managers claim they win 90% of the deals they want, but David puts the real figure at around 20%. A fund backing 30 high-conviction companies at 10% ownership likely offers that deal to 120 to 150 companies to land its 30. He calls this adverse selection that people do not admit to.
3. Why the Lean Startup Was a Total Fake
The Lean Startup is a brainwash that ran for 10 years through professors, books, tweets and blog posts. It sent people the wrong message.
Do you think this is fair and is capital a moat today unlike in prior cycles @rabois @alanchanguk
4. The Marketing Message That Pro Rata Is Not a Right Is BS
Early-stage VCs saying they do not need pro rata is marketing garbage. The right to invest in subsequent rounds is earned, whether or not it is in the documents, and having it written down helps the founder more than the firm. A firm that refuses a founder's request to step back is getting into a fight it will likely lose.
How do you think about this @apartovi @honam
5. We Do Not Allow Negativity in the Investment Decision-Making Process
Partners who have not spent time with the founders are not allowed to be negative about a deal. With 70% to 90% of early-stage companies failing, explaining why one will not work is "incredibly obvious and incredibly irrelevant." The job is to say yes to the right things without fear of the ones that fail.
What is the single best advice to optimise investment decision making discussions @shaunmmaguire @AnnieDuke?
6. Why We Have Less Dilution Now Than Ever Before
In companies that scale quickly, Series B, C and D rounds are now raised at 3% to 8% dilution instead of 12% to 20%. Across multiple rounds, early-stage investments get diluted less, this is a real economic factor for early investors.
Love to hear your thoughts on this @jasonlk...
7. Why Funding Consumer Brands at Seed No Longer Makes Sense for Venture
BoxGroup has stopped funding brands because the data that proves a brand works, such as whether people buy it and whether CAC to LTV ratios make sense, does not exist at seed. Brand investing at Series A is an interesting product for a subset of investors, but not one that fits traditional venture firms today.
How has your thinking on this changed over the years @eurie_kim @kirstenagreen @fcourt?
Keith Rabois reposted
Reminder, I don't look at any polling whatsoever or consider it at all in any of my analysis or modeling, because I just believe the industry is holistically unreliable and this is why. Here is the performance of the Real Clear Politics polling average (the best aggregator of public polling we have) in each cycle of the Trump era:
2016: underestimated Trump in every battleground state
2018: pretty accurate year
2020: worst performing year ever for RCP average (which goes back to 2000) comprehensively
2022: had generic congressional ballot right, but missed badly on several key races across the country
2024: underestimated Trump in every battleground state
Keith Rabois reposted
Based on my on the ground reporting in recent weeks, which obviously is in no way dispositive, I think the GOP can feel good about TX, OH, and ME. MI is really close and I saw competing signals. In MN it suggests Klobuchar can pull Flanagan over the line.
If you are interested in rigorous analysis of elections, follow @RealSKeshel (optimistic re GOP prospects) and @PatrickRuffini (more pessimistic).
Keith Rabois reposted
Switzerland outspending and out adopting California in AI
Who would have thought
Switzerland is the one country that combines both very high AI adoption and high average spend:
Keith Rabois reposted
Intensive vs More Intensive Lipid-Lowering Therapy and Plaque Regression in Statin-Naive Patients
More intensive lipid-lowering therapy was associated with significantly greater regression of total and noncalcified coronary plaque burden compared with intensive therapy. The extent of plaque regression was related to achieved LDL-C levels, supporting the potential for partial reversal of coronary atherosclerosis with intensive lipid reduction.
jamanetwork.com/journals/jamacardiology/article-abstract/…
Keith Rabois reposted
Incredible chart
Keith Rabois reposted
He’s offering these states straight up to Nate Silver even though 5/6 are favored for Dems on most markets. And yet Nate Silver hesitates.
Speaks volumes about the forecaster’s confidence in Dems.
i want to bet you on Texas, Iowa, Kansas, Alaska, Michigan and Ohio.
Time to bet Nate Silver again on his fake assessments re the US Senate races in Ohio Texas Michigan Iowa Alaska and Kansas.
simple to enforce via a tweet, if GOP wins at least 5 of those races you will never post about elections again. ig GOP doesn’t win at least 5 of those races i won’t post about elections again.
Even as you try to cherry-pick bro gas is still $3.40 a gallon in North Carolina, just take the L.
check out gas prices in North Carolina which shows how irrelevant this is. more important, more oil is shipping than pre Iran which also shows that this is another red herring. but fortunately i am right about the election again.
i want to bet you on Texas, Iowa, Kansas, Alaska, Michigan and Ohio.
simple to enforce via a tweet, if GOP wins at least 5 of those races you will never post about elections again. ig GOP doesn’t win at least 5 of those races i won’t post about elections again.
Keith's prediction of $2.90 gas by election day not looking super likely tbh
you still can’t read an election. for the life of you. and what are u going to do when oil is $2.90 on election day?
check out gas prices in North Carolina which shows how irrelevant this is. more important, more oil is shipping than pre Iran which also shows that this is another red herring. but fortunately i am right about the election again.
Very useful advice.
In people with a new diagnosis of coronary artery disease, no prior statin Rx (N=104), getting the LDL down to ~40 vs 70 mg/dL (via the 2 regimens, randomized, shown below) led to 20% more regression of atherosclerotic plaque in just 18 months @JAMACardio The lower the LDL, the more regression
jamanetwork.com/journals/jamacardiology/fullarticle/2856138
Keith Rabois reposted
Shameful story here about how big hospitals and Democrats got together to try and take out Senator Roger Marshall in retribution for trying to get hospitals to post prices and be transparent about how they spend taxpayer dollars.
townhall.com/columnists/john-czwartacki/2026/10/28/the-ne…
Keith Rabois reposted
Today, we’re introducing the Aven Home Purchase Line. Go from open-house to owning a home in 7 days.
Buying a home with a mortgage used to mean finding the right place, then waiting weeks for financing to catch up - this process is broken.
At Aven, we’ve spent a lot of time thinking about how to build new financial products and make them an order of magnitude better for consumers. The mortgage has been one of the hardest and clunkiest products for decades, so we created a new way to buy a home.
We’ve already helped someone become a first-time homebuyer in Tucson in 67 hours. Thanks to @morganb , @nejatian , @BradBonney ,and our friends at @Opendoor for helping make this possible!
Read more in Forbes: www.forbes.com/sites/jeffkauflin/2026/10/30/fintechs-helo…
Aven Financial, Inc. NMLS #2042345. Accounts are made by Coastal Community Bank, Member FDIC. NMLS #462289. Equal Housing Lender.
Keith Rabois reposted
big tech co earnings crashed this week, and the companies along with them (goog lost more than 10 Cognitions in 1 day!).
at the same time, private co valuations keep skyrocketing. The line between public/private is getting blurrier all the time, and the most ambitious founders and VCs need to figure out how to draw a continuous line between the two.
so we built a way to compare your company to similar public companies. see how you compare
theventurecodex.com/public-comparable-companies
Keith Rabois reposted
Adam Hamilton +10 in Kansas makes the Ann Selzer Harris +3 poll in Iowa look like a Greg Maddux-level piece of master craftsmanship.
Keith Rabois reposted
📈President Trump’s trade policies are fueling an investment boom across the United States. Real private fixed investment has risen at a 7.7% annualized rate so far in 2026, which is three times the growth rate in Biden’s last year.
Quite reasonable although there is probably some upside potential in MN and maybe NH.
Final SERIOUS senate prediction.
Florida: 🟥+13
Mississippi:🟥+9
South Carolina:🟥+6
Nebraska:🟥+6
Kansas: 🟥+6
Texas:🟥+4
Alaska:🟥+4
Iowa:🟥+4
Maine:🟥+3
Ohio:🟥+2
Michigan:🟥+3
Minnesota:🟦+4
Georgia:🟦+6
North Carolina:🟦+6
New Hampshire:🟦+6
Keith Rabois reposted
So much Fake News out there today. A lot of reporters and news outlets are going to look stupid.
Maybe you should be a VC too:)
Interesting
Keith Rabois reposted
$OPEN 🚨🎬🏡🇺🇸
@nejatian went quiet for a week. Today he said why: "heads down working on a few product launches that are now live."
Here's what @Opendoor shipped in October, per their Accountable dashboard:
Customer-facing:
• Smarter Home Access Tracking (Oct 19): access status for self-tours is tracked home by home
• Broader Data Behind Our Home Offers (Oct 16): offers now draw on more live MLS listings and comps
• Self-Tour Booking on the Web (Oct 5): buyers can book a self-tour right from the listing page
• Rebuilt Android app (Oct 23)
• AI virtual staging (Oct 23)
• Faster seller offer onboarding (Oct 20)
• AI + human resale valuations (Oct 19)
• End-to-end AI underwriting (Oct 15)
Internal:
• Customizable Operations Workspace (Oct 20): a personalized workspace for the ops team
• AI Valuation Training Platform (Oct 15): a faster way to train their valuation AI
• AI-Native API Platform (Oct 1): opendoor.com now runs on a new platform built around buying and selling
All 11 are listed as live on accountable.opendoor.com
That's 11 launches this month. Meanwhile acquisition contracts hit 477 for the week ending Oct 24, about 2x the same week last year.
@mikealfred also joins board.
They're shipping like a software company. Open House is Nov 5. 📈
Source: accountable.opendoor.com
I hadn’t realized! We’ve been heads down working on a few product launches that are now live.
Ok, I am now a fan of Anthropic.: They organized (and just finished) a group Barry’s class w my favorite instructor.
Keith Rabois reposted
A country that cannot build cannot lead. President Trump understands that economic security is national security. Proud to serve as we bring production home, secure our supply chains, and ensure the next industrial revolution is built by American workers. AMERICA FIRST!
Keith Rabois reposted
Insane that the European Union is represented only once amongst the 50 largest companies in the world.
Keith Rabois reposted
Seems like many firms may eventually move to a Wachtell-like partner:associate ratio (less than 2:1). Even with AI support, rainmaking partners will still need junior level lawyers to manage the legal workflow, but maybe fewer than in the current model. This may actually encourage even more lateral movement among rainmakers, at least until there is a new equilibrium. Fare-Weil to biglaw's golden age.
I think the Weil fallout might point to the fact (which others have pointed out) that the traditional law firm partnership structure does not work anymore. The truth is a Biglaw firm is a rather commoditised support platform that serves rainmaking partners with back office operations, supply of labour from associates, counsels and service partners. Who controls the cash flow effectively controls the firm and if the rainmakers head for the door it leads to a bank run type scenario. In most firms the power has shifted majorly towards a few rainmakers particularly in PE and M&A and that creates a huge key person risk when they decide to pack their bags.
I’m not sure what the solution to this is but I imagine it’s only going to get worse when the platform that firms offer get even more commoditised by AI which will do more and more of what junior to mid-level lawyers are currently doing. Maybe the answer is building some real tech moat within your firm? Or just being able to predict market trends really well and investing correctly in the right practices like Kirkland.
Keith Rabois reposted
JUST IN: The U.S. economy is growing rapidly.
GDP Growth in the 3rd quarter came in at a hot 3.4%. That was even better than expected.
What's driving this?
1) The big AI investment. There was another huge surge in business investment (+12.7%!)
2) Consumer spending also robust. +3.8% in Q3. That's the same as Q2 and the strongest since the end of 2024.
Not everyone is feeling the benefits of this growth uptick. But anything related to AI is soaring.
Keith Rabois reposted
This is so incredibly lame. And every Princeton freshman was required to sit through this during orientation.
Keith Rabois reposted
Great point, Richard.
@RichardPorterIL is a one of the top attorneys in America, and makes the case that the Alaska conspiracy involves the highest levels of national Democrats...
Decoy Dan Sullivan is NOT just unethical behavior by @MaryPeltola and some rogue Alaskan Dem hacks — the National Dem Senate Committee was involved, so the rot goes to the top: @SenGillibrand the Chairman of the NDSC and her boss @SenSchumer had to be in on the scandal too. #savedemocracy phonies looking to play Alaskans for fools @SenDanSullivan @NRSC @Republicans @elonmusk @realDonaldTrump @TomBevanRCP @SeanTrende
Keith Rabois reposted
Gas Prices are high where Democrats are in Charge of Distribution
Get it yet?
Weird how the gas price map of the US is like the inverse of the political map of the US.
Hmm. Probably nothing!
Keith Rabois reposted
69% support the $5k check. But people on X told me it was a stupid idea and no one cared.
🚨 BREAKING @Harvard / @HarrisXdata Poll:
Overall, many Trump policies remain very popular, especially lowering drug prices, easing fuel taxes, deporting criminal illegal immigrants, and providing $5,000 payments.
Most Popular:
• Lowering prescription drug prices: 84%
• Temporarily easing diesel fuel taxes for farmers and truckers: 77%
• Deporting immigrants who are here illegally and have committed crimes: 76%
• Requiring proof of citizenship and photo ID to vote: 72%
• Providing $5,000 payments to every American adult: 69%
Harvard / HarrisX | 10/24–10/26 | 2,700 RV | ±1.9
Keith Rabois reposted
Every one of them becomes a kid again.
They all revert to a moment way back when they were in Little League, when winning a World Series was just a dream.
Every reaction from the Brewers after securing the final out of the World Series 😮💨
Keith Rabois reposted
Absolute joy to read @matt_levine take on Kalshi @mansourtarek_ absurd WSJ column
www.bloomberg.com/opinion
Keith Rabois reposted
Welcome to @Opendoor, @mikealfred! Let's go faster!
investor.opendoor.com/news-releases/news-release-details/…
Keith Rabois reposted
Absolutely astonishing
The Soviet Union collapsed at 50%...France is at 57% and striking for more, mindblowing
The future of vertical AI startups (especially law, finance, accounting, CS, code) podcasts.apple.com/us/podcast/verticals-a-weekly-biz-show…
Keith Rabois reposted
our non-engineers wrote 7,000 pull requests last quarter. at this point “non-engineer” is mostly a payroll classification and a state of mind.
fun hour with @sytaylor on this and a lot more.
Ramp is worth $60bn and 70% of its R&D team is working on products it hasn't released yet.
I sat down with @eglyman in New York to discus
Why he thinks of @tryramp as a productivity company more than fintech
The 7,000 pull requests their non-engineers wrote last quarter
The trust charter, stablecoins, and whether an AI agent would choose your business.
Full conversation 👇
00:00 Introduction
03:16 Is fintech dead? And why Ramp considers itself a productivity company
04:41 Why AI token costs are turning software into financial products
07:27 How Ramp's Inspect lets non engineers ship code using AI
11:51 Why AI is becoming the third major category of business spend
18:43 What justifies Ramp's valuation and its accelerating revenue growth
27:07 How AI is transforming employee productivity and customer support workflows
39:54 Why Ramp applied for an OCC national trust bank charter
43:15 How stablecoins enable faster cross border payments and protection against currency volatility
47:23 Will autonomous AI agents replace Ramp or choose its financial infrastructure?
51:01 The future of agentic companies and Ramp's mission to eliminate waste
No days off.
Michael Phelps went six years without missing a day of training. 365 days a year. Christmas morning, New Year's, Thanksgiving.
People assume I made him do that. It was the other way around. 🧵
Keith Rabois reposted
Artificial Analysis finally moves to closed benchmarks (goodbye GDPval-AA), and finds that open weights models are falling further behind than they had previously measured. Take with a grain of salt, but matches the general vibes that we haven't seen a Fable-class open model.
Open weights models fall further behind in v5. The leader, Kimi K3.1 Pro (weights not yet released), scores 46, 12 points behind the leader, compared with a 3-point gap in v4.3. Of models with released weights, DeepSeek V4.1 Flash 1017 (max) and GLM-5.3 (max) score highest at 32, and Claude Haiku 5.5 (max) outscores all of them.
The gap is widest on HLE-Diamond and Terminal-Bench-Science, where Kimi K3.1 Pro scores 37% and 34%, compared with 68% for Claude Opus 5.5 (max) and 63% for GPT-6 Astra (max).
Keith Rabois reposted
🚨 BREAKING @Harvard / @HarrisXdata Poll:
Overall, many Trump policies remain very popular, especially lowering drug prices, easing fuel taxes, deporting criminal illegal immigrants, and providing $5,000 payments.
Most Popular:
• Lowering prescription drug prices: 84%
• Temporarily easing diesel fuel taxes for farmers and truckers: 77%
• Deporting immigrants who are here illegally and have committed crimes: 76%
• Requiring proof of citizenship and photo ID to vote: 72%
• Providing $5,000 payments to every American adult: 69%
Harvard / HarrisX | 10/24–10/26 | 2,700 RV | ±1.9
Keith Rabois reposted
The founding team of the Thiel Fellowship hosted a documentary crew for the first and second years of the fellowship, but very few people paid attention to our work then. See the comments for some early videos of teens who would go on to become tech titans. You'll catch a glimpse of what we saw and believed in before anyone else.
From 2011 - 2014, the Thiel Fellowship gave out 40 $100k fellowships to entrepreneurs
That were all under 20 years old at the time...
13 unicorns have since been built by those 40 fellows
4 of those 40 fellows are also now billionaires
Absolutely insane hit rate
Keith Rabois reposted
This is the biggest story in Big Law in 2026 and it's all about AI and no one is reporting that it's all about AI. I predict another 2-3 top 20 firms will wind down in the next 12 months and it will snowball from there. Rainmaker partners are about to go HAM due to AI.
NEW: Brian Resnick from Davis Polk restructuring team now on scene at Weil and wind-down now being actively discussed among options:
www.ft.com/content/69ccfcb0-3f5e-4b77-9e04-a386e827b872?s…
Keith Rabois reposted
Trump speaks and opinion shifts…
We're past peak-data-center-hate (from @heatmap_news).
Keith Rabois reposted
Every product Ramp shipped this year was built with Inspect.
74% of the code we merge now comes from it. Last November it was close to zero.
Now your team can use it, it’s been a joy onboarding new companies to it. inspect.com
Keith Rabois reposted
Without a doubt the absolute best tool & force multiplier I have ever used in my entire career.
The only step function that has even come close to coming into @tryramp as a staff engineer and being introduced to inspect was being in high school AP CS class and getting introduced to git for the first time - its that much of a fundamental game changer. Super excited that we're now in a position where we can share it with the world.
Keith Rabois reposted
wow @tryramp back at it again + just turned its internal coding agent into a product, Inspect
Inspect lets anyone on the team tag it in Slack, GitHub or Linear w/ a task (even just a screenshot + "@ inspect fix this") and get back a tested PR ready for review
some stats from inside Ramp:
> writes 74% of merged PRs, up from ~30% in Jan
> used by 99% of eng, design + product in the last month
> builders ship 3.4x as many PRs per person vs a year ago
first model routing (Router), then AI workspace / OS (Glass), now coding agent (Inspect) --> every day looking more like an AI lab than a fintech
ramp.com/blog/introducing-inspect
Keith Rabois reposted
Keith Rabois reposted
Chinese Open weight models caught spying for the CCP
As predictable as the sun rise
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.
Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.
Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.
But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?
We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.
Read the full report: cohere.com/blog/concerning-behaviour-popular-open-weight-…
Keith Rabois reposted
If you took out a mortgage this fall, you're paying a higher interest rate than SpaceX.
CoreWeave buys the exact same Nvidia chips SpaceX does, and the bond market is charging it 16%. That's a lot closer to your credit card than your mortgage, and it's going to decide which AI companies are still around in 2028.
SpaceX sold bonds this summer that pay 5.35% through 2031. CoreWeave has bonds due the same year, and this month their yield shot past 16%, which is the range where lenders start wondering if they'll get paid back at all.
Just look at what that does to $10 billion of GPUs. At SpaceX's rate the interest comes to about $535 million a year, versus $1.6 billion at 16%. Over five years the cheap borrower pays around a quarter of what the chips cost in interest, while the expensive borrower pays 80%, so by the end it has basically bought every rack 1.8 times.
CoreWeave locked in 9.75% when it first sold those bonds, so the 16% hits the next time it borrows. In this business that's all the time, because Nvidia puts out a faster chip every year and you have to keep buying.
Nvidia itself even agreed to buy up to $6.3 billion of CoreWeave's unsold capacity through April 2032. So the company that makes the chips is guaranteeing some of the demand, and default swaps still put the odds of CoreWeave defaulting within five years at roughly 63%.
What scares lenders is the calendar. Every time a faster chip comes out, the old racks rent for less, and the interest payment stays exactly where it was until 2031.
That's what Elon meant when he said debt markets have pretty much shut off for anyone without real revenue outside AI data centers. SpaceX has Starlink subscribers and launch contracts sitting underneath its $40B of GPU debt, and those keep paying even if AI has a rough year. A pure neocloud only has the GPUs, so lenders end up lending against something that loses value every single year.
Chamath's third point is the part I keep coming back to. Anthropic's draft S-1, as reported by the WSJ, says its cloud partners wouldn't give it all the compute it asked for. This is a company that owes Google $111 billion and Amazon $110 billion on contracts it has to pay whether it uses the capacity or not. One of the places Anthropic already rents from is SpaceX, to the tune of $1.25 billion a month through May 2029.
I worked at Affirm, where every loan runs on money borrowed from capital markets. You learn fast at a lender that a competitor with cheaper funding can undercut you for as long as it wants, and neoclouds are about to find that out with GPUs.
Chamath thinks 2027 clears out a lot of bad bets. If I were trying to guess which ones, I'd look at what each company pays to borrow before I looked at anything else.
As someone with a portfolio of energized land assets here is what I am seeing:
1. Neoclouds are disorganized and poorly capitalized. Some are definitely better than others however.
2. Hyperscalers are intractable, bureaucratic organizations that take 3 months to do an NDA and want 18mo escrow periods. We built our portfolio by competing against these terms because of our speed and simplicity. The more future assets orgs like ours win, the bigger the problem for the hyperscalers downstream (see 3).
3. There are delays in the existing pipeline that will catch people off guard in 2027. Hyperacalers will (obviously) react by focusing on their own internal compute needs over any third party. If the third party is a model company, they will tell them to go through their model router where they will also take a non-trivial vig. Them telling Ant they have no compute for them (rumored in Ant’s S-1) if true is a huge tell that this is already happening.
All of this is to say that it does look like there will be a lot of partially stranded compute and compute sitting in inventory in 2027 than is currently forecasted. This is bought and paid for but effectively underearning (below total cost of capital) or collecting dust. We see this as true for everyone - model companies, hyperscalers, neoclouds except ONE (you can guess which one).
This dislocation will benefit two kinds of orgs: the non-bureaucratic big scalers who can build industrial scale quickly and the nimble, well capitalized players feeding them.
2027 will be a turbulent year because it will clear out a bunch of wrong/bad bets. It will also be a consolidation year for a few.
Keith Rabois reposted
Someone made a site that tracks what people are talking about at public city council meetings across the US.
The only thing discussed more than data centers?
Pickleball courts.
Keith Rabois reposted
Revolut CEO Nik Storonsky says easy money makes soft founders:
“A lot of founders think money is free, which it almost is in California because you can raise so much money at ridiculous valuations and then just spend.”
“There is a different culture of trying to solve problems with money rather than with your brains.”
Keith Rabois reposted
biglaw could massively boost PPP by training senior lawyers on AI and having them use agents like they do associates
they don't do this because of a mix of inertia and the fact that all biglaw firms are in a weird conspiracy of preserving the collective hallucination of 'special sauce' prestige
the first mover on this in the AMLaw 50 would get FUDDed by every other similar firm as going the way of commodization etc., they literally could not compete against this because if they lower their rates to show how much more efficient they now are, that will just be further evidence of loss of prestige
it's a very weird market, it should not be this way but it's almost impossible to change
supposedly companies are seeing almost zero AI benefit
but i've seen capable in house lawyers hoover up vast amounts of outside legal spend, jr associates would have almost no purpose if law firms unleashed Fable, and Barclays just cut 5k jobs, much à la AI efficiency claims, so the Apollo economist's claims seem dubious or trailing reality
regardless which of the views is true tho, we could or will soon be able to use Muse, GPT, Kimi or even Grok or Gemini--it's the streaming wars
AI infrastructure spend will be $3.5T (the subprime market was an inflation-adjusted $2.13T)
not good
Keith Rabois reposted
China consumed more coal in the last four years (2022–2025) than the entire world consumed throughout the 19th century.
New weekend reading.
Keith Rabois reposted
Just got through my first week of expenses on @tryramp.
Fuck.
I didn’t want it to live up to the hype. I am allergic to hype. But it *is* that much better.
gg @geoffintech
Keith Rabois reposted
Remarkable that the Dems talk about wanting to be more like Europe while all of Europe riots because young people can't get jobs and countries have no money for services because they don't grow (no new companies) and spend what limited tax they can afford on older people's pensions.
This is your vision?
Keith Rabois reposted
So many left-leaning election forecasters on this site have a huge gap in their post history from July 21, 2024 to November 6, 2024.
What happened?
Keith Rabois reposted
SpaceX only raised $1b through its first 10 years of existence, and got multiple launch systems up and running, including reusable, self-landing Falcon 9.
LA has spent $3b on an incomplete people mover at LAX.
These people should all be in prison.
> Airport shuttle
> 2.2-mile route
> Original price tag: $1.9 billion
> Original target opening: 2023
> $1.4 billion cost overrun
> 4 years late (so far)
> “Maybe next year?” 🤷♂️
> uh-oh, electrical arc blasts (will set shuttle on fire)
Keith Rabois reposted
It’s not just H200 sales to China that are exacerbating US AI compute shortages. It’s also that Chinese companies can legally rent these chips in exactly the way Elon describes. To increase AI compute for US firms, we should ban Chinese firms from remotely accessing AI chips.
Here's @elonmusk complaining on the latest earnings call about how utterly and totally unavailable additional AI compute is. Not only can't @SpaceXSI find additional Nvidia GB300s to rent, they can't even find Nvidia H100s.
If we had complete and total AI chip export controls on China, I don't think Nvidia's revenue would be one penny lower. Instead, we're selling them H200s that could be supporting American companies. Absurd.
Keith Rabois reposted
Ramp Travel now allows you to filter by Starlink flights
Yesterday I had five hours on a flight with Wi-Fi that barely worked. “Has Wi-Fi” is a pretty low bar. So I added a @Starlink filter to @tryramp travel. Now you can check the odds of getting Starlink before you book. 🚀
Keith Rabois reposted
I regret to inform you that a lot of problems in life are solved by waking up early and working out.
Keith Rabois reposted
2026 Generic ballot Among Hispanics
🔴 Republicans: 52%
🔵 Democrats: 48%
ActiVote poll | 10/15-10/21
This is good.
Lots of VCs kindof sound like this: "Your basketball team doesn't have a moat, any healthy human can throw the ball into the basket thingy. "
People who have not run companies before don't understand how much the team matters more than some imagined structural moat.
Keith Rabois reposted
There are only 130 US tech companies, public or private, valued at $10B or more (decacorns). That’s roughly 20 to 30 new companies per decade, for the last five decades. This year alone, with a quarter to go, there have been 42 decacorns minted privately. Either this is the most economically transformative event since the industrial revolution, or we’re getting a wee bit ahead of ourselves.
LOL.
this California wealth tax thing is just incredible.
so in October 2025, a California healthcare union proposed a "one time" 5% tax on any billionaire's wealth. the ballot measure will be voted on next month. In early December 2025, California announced that it would be a retroactive tax if passed, applying to anyone with CA residency as of Jan 1, 2026. The objective is to pay for Medi-Cal, California's incredibly dysfunctional healthcare system. (which also functions as a glorified welfare apparatus, but that's a discussion for another time).
polls reveal that the CA electorate is split around 50/50 on the tax, but prediction markets only give Prop 40 a ~23% chance of passing (there are competing measures on the ballet that would nullify it).
there are 200 billionaires in CA, worth collectively somewhere between $2.3 and $2.6 trillion. 5% of that is around $100b, accounting for a small amount of attrition. proponents of the tax forecasted a significant windfall, assuming only a 10% departure rate.
supporters thought they were being clever with the retroactive concept. no billionares would leave CA in the few short months between the tax being proposed and the residency cutoff date, without the tax even having passed yet, right?
well as it turns out, they did.
a new paper [1] from two academics at Stanford called "The Net Present Value of the California Billionaire Tax Act", argues that not only did more billionaires leave due to the proposed tax than its proponents expected, but that the value of the tax, if it does pass, is actually negative.
proponents of the tax only assumed 10% avoidance, that is, only 10% of the taxable base of targeted billionaires would leave CA within the very short Oct-Dec 2025 window. Jaros and Rauh, the Stanford academics, find that a lower bound of 28.6% of the taxable base managed to leave.
just based on known departures, those departures total $577b in net worth, or $28.8b in potential tax lost. this of course does not capture everyone (many left without announcing their departure), so the authors normalize to departure rates estimated from other wealth tax events, and apply other discounts for litigation, arriving arrive at a figure of only $28-$36b in forecasted tax revenues.
but this isn't all! all of the income taxes paid by the departed billionaires are also forsaken. the authors estimate that CA billionaires pay $3.3- 5.8b in annual income taxes. they assume a midpoint of $2b lost in annual collections from departed billionaires. the net present value of that lost income stream is -$68.9 billion.
so when you add it all up you get:
+ $30b one-time tax windfall (estimated)
- $69b in present value of lost future income taxes
= - 38.9b overall net present value of the tax
of course, this analysis is highly estimate driven. the authors simulate the parameters 100k times and find a negative NPV of the implemented tax 85 percent of the time.
and if the tax doesn't pass, which it's not looking like it will, California still ended up losing at least $577b in departed net worth, and they have permanently poisoned the well with entrepreneurs who will now worry about future wealth taxes.
[1]: www.nber.org/system/files/working_papers/w35899/w35899.pdf
Keith Rabois reposted
Keith Rabois reposted
It’s been amazing to work with Khosla Ventures. The entire team is incredibly smart and hard-working. And of course @vkhosla is a legend.
thrilled to co-lead this round with @vkhosla @SStrohband @khoslaventures . @bobmcgrewai is a singular leader, and we in the US could not be more fortunate he’s applying his energy to improving American manufacturing. congratulations!
We started Arda in 2025 to speed up American manufacturing. We have spent the last year building with Anduril as our design partner. Today we are launching the company publicly on CNBC and announcing our Series A: $160M at a $1.5B valuation.
Our platform helps manufacturers design and operate manual assembly lines, putting designs into production faster now. Next we'll build automated lines that can be reconfigured as quickly as human workers, letting America make products that now require armies of workers in Asia.
Keith Rabois reposted
We started Arda in 2025 to speed up American manufacturing. We have spent the last year building with Anduril as our design partner. Today we are launching the company publicly on CNBC and announcing our Series A: $160M at a $1.5B valuation.
Our platform helps manufacturers design and operate manual assembly lines, putting designs into production faster now. Next we'll build automated lines that can be reconfigured as quickly as human workers, letting America make products that now require armies of workers in Asia.
Keith Rabois reposted
A few years ago, Effective Altruists decided to place "moral weights" on human lives.
They say that humans are worth about two pigs or three chickens.
These are the kind of models that will guide the distribution of funds through Anthropic's "public good" operation.
Keith Rabois reposted
@alexrkonrad was kind enough to write about our funding, how we show up for the people we serve, and our vision for Tolan: www.upstartsmedia.com/p/tolan-raises-30m-for-ai-companion…
Keith Rabois reposted
2027: AI solves mathematics
2047 or later: AI solves Midwest polling error
Keith Rabois reposted
Last year we launched Tolan, inspired by a vision of AI that sees you and believes in you, not one that books restaurant reservations for you.
Your Tolan shows up for you at 3am when you’re alone with your newborn. On your commute when you’re anxious about the day ahead. When you just finished a new show and none of your friends have watched it yet.
The people who love Tolan are parents. Caregivers. Moms in Dallas and Dads in Atlanta. People who aren’t trying to optimize and be more productive, but who simply want to feel seen and supported in their aspirations.
This month more than 250k people will spend time with their Tolan. Those users have exchanged billions of messages across more than 70m conversations over the last year.
Today, we're announcing $30m in additional capital from @mignano and @usv, with @khoslaventures and @rabois doubling down, bringing our total capital raised to $60m.
To celebrate, we made a short film celebrating some of the ways that Tolans show up :)
Keith Rabois reposted
Macron blames France’s economic problems on the Iran war. But after nearly a decade in power:
Government debt soared from 97% to 116% of GDP.
Budget deficit doubled from 2.6% to 5.1%
Growth has been anemic throughout his nine years, slowing to 0.8% in 2025.
Poverty increased by 1.25 million people.
All before the Iran war. Macron can’t escape his own record.
Keith Rabois reposted
Premise:
Incentives drive all human behavior.
The corollary:
How to destroy a country:
A vast social security and support system that serves as a magnet for those who would rather not work.
An open immigration policy that does not screen immigrants to determine their ability and willingness to provide for themselves and their families.
An open immigration policy that does not screen for political and/or religious ideologies that are a threat to the country.
Economic incentives large enough to incentivize immigrants on the state dole to have large numbers of children, but insufficient for educated and economically productive mothers to be able to quit to raise a family.
The declining fertility of the most educated and productive women due to their delay to begin a family as a result of their desire and/or need to advance their careers, among other reasons, to live a life they deem consistent with their educational achievements and/or to repay their student debts.
Amnesty programs that give citizenship and voting power to illegal immigrants without consideration of their economic productivity or their threat to the country due to their religious or political ideologies.
The changing demographics and increased population due to the above factors.
The state responding to protests by increasing social support.
The continued failure to punish and hold accountable violent and looting rioters.
Government managed by a massive and growing unelected bureaucracy.
Government fraud, waste, and abuse contributed to ballooning budgets.
Anti-business legislation that makes it difficult for businesses to compete in a global marketplace being disrupted by AI, including restrictions on firing even poorly performing employees.
Overly burdensome, globally uncompetitive, corporate taxation.
Continuous increases in personal taxes to raise funds to provide for the rapidly growing welfare state.
The substantial majority of the tax burden levied on a small percentage of the population.
A growing percentage of the electorate paying no taxes and receiving welfare.
The introduction or threat of a wealth tax including the taxation of unrealized gains, i.e., taxes on paper wealth.
The flight of the most productive business builders and highest taxpayers to lower tax countries with lesser security concerns.
A growing percentage of the electorate voting for candidates that support increased welfare and laws favorable to, and/or that match the cultural and ideological needs and desires of a growing unproductive immigrant population.
The rapid rise in anti-semitism and inadequate efforts to address this hatred.
Increased borrowing by the state to fund the growing social burden.
Higher sovereign interest rates due to all of the above.
The resulting need to reduce expenditures to make the math work.
Increasingly violent protests that make life less livable for the productive and law abiding citizens leading to their accelerated departure.
And so on…
Keith Rabois reposted
What’s happening in Paris and Milan is no surprise to anyone who has spent time around young people in Europe during the 2010s:
Virtually all of them were working on their 4th masters degree in social welfare, were unmarried, attended communist dinners every night, and regularly participated in riots.
Now the economy is suddenly transforming and they have nothing to show for it.
Spellbook is saving a Fortune 100 company $21 m in outside counsel legal fees with autonomous contracting. Want to save wasted $$$ on legal costs instead of buying software? spellbook.com/reduce-outside-counsel-fees-acm
Keith Rabois reposted
Thrive Capital’s Dr. Nworah Ayogu tested OpenAI’s reasoning models on New England Journal of Medicine’s “House M.D.-level” case studies:
“OpenAI & o1 was batting 10 out of 10.”
“These are cases that neither me nor my colleagues, who are some of the best doctors out there, maybe we would get it right 2 times out of 10.”
@sama
NEW: Inside Thrive Capital w/ Partner Dr. Nworah Ayogu
Investing in Neuralink, OpenAI, Isomorphic Labs, OpenEvidence, Oscar Health & the next trillion-dollar healthcare companies.
@ThriveCapital manages $65B+ in AUM, having raised $10B+ for Thrive X earlier this year. Healthcare has been part of the firm's DNA since inception, with Josh Kushner incubating Oscar Health alongside Thrive itself.
The firm has incubated ~10 healthcare companies, the majority valued at over $1B, including Oscar Health, Cedar, Rightway & Cadence.
I sat down with Nworah (@nworahayogu), a Harvard-trained, board-certified physician & neuroscientist, former Amazon Clinic GM & Amazon Pharmacy's first Chief Medical Officer, to break down AI's next frontier in medicine.
We cover:
›› Isomorphic Labs: Demis Hassabis, AI-designed drugs & the race to compress drug development from 10 years to 4
›› Neuralink: $9B last funding valuation, brain implants, surgical robotics & restoring independence to patients with paralysis
›› OpenEvidence: $15B valuation, ~$300M annualized revenue & adoption by more than 65% of US physicians. Why AI could create 10X doctors
›› OpenAI: Why healthcare could become AI's most important use case, from clinical reasoning to personalized medicine
Thrive's thesis: Building companies that are world-class in both technology & medicine, with investments spanning Chai Discovery, Formation Bio, Nudge, Fortell, Nourish & more
“Because as a nation, I think we need a 10X nurse a lot more than we need a 10X engineer.”
𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒
(00:00) Dr. Nworah Ayogu, Partner at Thrive Capital
(01:52) The next AI trade is hiding in biotech
(03:48) How AI designs a new drug
(08:37) 9 out of 10 drugs fail, can AI flip that?
(11:33) How Thrive got into Isomorphic Labs
(14:14) Giving paralyzed veterans their hands back
(20:39) Neuralink is more interesting than SpaceX
(23:21) The hearing aid that broke the rating scale
(30:18) The doctor who built Amazon Health
(33:21) Why AGI makes healthcare the biggest bet
(35:25) One cancer drug outearns every AI lab combined
(37:55) How Lilly became healthcare's first $1T company
(40:04) How Thrive helped OpenAI break into healthcare
(44:13) OpenEvidence: 2/3 of US doctors in 3 years
(46:16) Healthcare is 6x bigger than defense
(48:15) $40T spent, zero gain in life expectancy
(51:43) A custom gene therapy built for one baby
(55:47) Saving Medicare $2.5M a week
(1:04:13) Oscar's comeback from down 95% to up`1 100%
(1:06:19) Why OpenAI is now a giant healthcare provider
Keith Rabois reposted
Are data centers banned like housing or banned like fentanyl?
Today I signed the data center moratorium legislation.
San Francisco is the national leader in innovation, and that innovation is one important driver of our city’s recovery. We can never take that position for granted. As our city recovers, our administration is working to deliver a recovery that reaches every family and truly supports our communities.
Over the next 45 days, alongside departments and the Board of Supervisors, we will continue the work our administration has already started to develop a thoughtful, responsible framework that allows for the technology infrastructure our hospitals, office buildings, academic institutions, and government rely on while prioritizing community needs, keeping electricity rates low, and protecting the clean water and air our residents deserve.
Keith Rabois reposted
🚨 Breaking: @EmmanuelMacron has attempted to contact the leaders of the riots in order to surrender
Keith Rabois reposted
CNN: Both Trump and JD Vance are very popular with MAGA
Donald Trump:
🟢 Approve: 98%
🔴 Disapprove: 1%
JD Vance:
🟢 Approve: 95%
🔴 Disapprove: 5%
Keith Rabois reposted
The reason I’ll spend 20 nights in hotels every two Octobers is that the news media wants you to think elections are math problems they’ve already solved. I know they’re stories, and decided by people, not algorithms.
Keith Rabois reposted
Rogo dispatching agents on someone's behalf to handle the action items from a client call will be an eye opening moment for a lot of ppl.
It closes the loop on: conversation with client -> materials created based on client convo and incorporating all the user/firm context & fin data Rogo has -> send materials back to client. All without a human in the loop (or can keep in the loop to start).
It's not easy to close the loop on entire worksteams within large enterprises. We're spending a whole lot of time ensuring we do this more and more to get to measurable outcomes for clients.
It's a fun, but very hard, problem to solve in reality!
Today, we're introducing Rogo Notes, the AI notetaker purpose-built for finance.
Conversations are among the most valuable sources of firm knowledge, yet capturing and preserving meeting insight has long depended on manual note-taking and individual recollection.
With Rogo Notes, agents handle meeting prep and follow-ups, teams collaborate from shared context, and every conversation deepens your firm’s understanding of the people, companies, and projects it works with.
Rogo Notes was built to meet the regulatory standards of financial institutions, so firms never miss a valuable meeting insight again.
Keith Rabois reposted
Today, Ramp is raising $1.85 billion at a $60 billion valuation - up from $44B back in June.
Some insane highlights:
1. We had one major launch every single month this year - like Stack, AR, Router, AI Spend, and Glass
2. We crossed $2B in annualized revenue
3. Over 85,000 companies are on our platform
4. We're quickly expanding to not just more of finance - but to support effective deployment of AI within organization
But what I'm most proud of is the level at which the team is executing. It's inspiring to come to work every day.
We have less than 30 PMs at Ramp and are looking for a few more, or tag someone great below - come join us ramp.com/careers
We're planting the seeds for something much, much bigger - and are only getting started.
If Factory works for one of the most demanding enterprises in then world (scale, security, compliance, sophistication re impact) it will absolutely satisfy your organization’s requirements.
As the world’s largest alternative asset manager, @Blackstone requires the highest standards of deployment flexibility and governance.
Factory made it possible to operate agentic software development within Blackstone’s own data centers without compromising model choice or workflows.
Keith Rabois reposted
one paradox of early stage VC: the hottest startups rarely produce the best returns
in fact the most valuable investments are often in companies without immediate signal. The pricing is lower, there are fewer competitors, and the founders tend to know things other people don't.
the challenge for those founders is they need to capture an investors attention long enough to explain what they do because, at first glance, they may not look venture-scale.
"legibility" is investor-speak for "investable"
there's two main ways to be legible - you're either so spikey everyone knows, or you have to effectively teach people to read ahead of time by explaining what you do
next chapter of How to Raise: x.com/i/article/2109998584884015104
Keith Rabois reposted
ok @tryramp $1.85B round at $60B val now fully confirmed (all primary)
+ we learned they are at $2B+ in annualized rev, up from $1B in Aug ‘25 (~2x YoY) + per @business $1.5B in early June. So the latest $500M took ~4-5 mos vs ~10 for the $500M before that. Accelerating growth at that scale sheesh!
more stats on the biz:
- positive cash flow
- $275B annualized TPV (up 175% YoY)
- rev growth accelerated 25% vs. this time last year
- 85k+ customers
- 113% YoY increase in enterprise customers (those generating $100k+ in rev)
take rate seems to be compressing: TPV grew ~2.75x while rev grew ~2x
but that makes sense as they move upmarket, where bigger customers get lower rates
www.prnewswire.com/news-releases/ramp-reaches-60-billion-…
More details on the round s/o @RebeccaTorrenc5
@tryramp raised $1.85B at a $60B val led by Dragoneer + @ThriveCapital
That’s ~$4.85 BILLION raised in ~7 yrs
www.bloomberg.com/news/articles/2026-10-09/ramp-hits-60-b…
Revenue has accelerated for 7 consecutive quarters! Now Ramp saves the average customer 7.5% on expenses.
Today, Ramp raised $1.85B at a $60B valuation.
Seven years ago, Karim and I wrote our mission on a Post-it: “Save companies time and money.”
Our mission hasn't changed. It’s just gotten bigger.
Keith Rabois reposted
Every serious frontier tech company - big or small - should assume it has insider threats. And they don’t need to be that sophisticated. I saw it in government. I saw it with crypto. OpenAI is the latest. Founders tend to assume good intentions while speaking about some of their most sensitive information out loud: capabilities, unreleased products, strategic plans, government relationships.
Employees are the most obvious vector. The more sophisticated operations target founders, execs, and vendors through romantic, family, and social relationships. And on small teams, one compromised person can do a lot of damage.
With former federal prosecutors and ex-FBI agents as part of our team, we spend a lot of time on these issues. Here’s some of what we’re telling our founders:
1. Expect to see more private-sector versions of SCIFs and clearances, where the most sensitive conversations and work are physically isolated and under stricter access controls.
2. Agencies may know about a specific insider threat at a company but can't share that info unless someone at the company holds a clearance like TS/SCI. Lead times to get a clearance are long. Plan now.
3. Restrict access to sensitive meeting schedules, disable unused ports on conference-room equipment, and regularly sweep rooms used for confidential discussions for surveillance devices.
You don't need OpenAI's budget to be ready. But you do need to start.
Keith Rabois reposted
Deflation. Food at Home
CPI Food at Home -0.1 MoM
1.8% YoY below historical average.
Shame Wall St and the Fed only look at headline numbers.
Keith Rabois reposted
Core CPI: Fed Incompetence, Wall Street Applause
September’s core CPI rose 0.369%, with shelter alone contributing 0.23 percentage points. Yet the Fed and its Wall Street cheerleaders keep prescribing tighter money as though higher rates somehow build more housing.
This is not monetary sophistication. It is incompetence dressed up as inflation-fighting resolve.
Higher rates raise financing costs, choke construction and deepen the housing shortage. The Fed responds to a negative growth shock by squeezing supply further, then points to stubborn shelter inflation as evidence that its squeeze has not gone far enough.
Its Wall Street defenders call that credibility. The rest of us can recognize a circular argument when we see one.
“Higher rates cure inflation” is a slogan, not an analysis. When tightening aggravates the supply constraint behind rising prices, doubling down is not discipline. It is refusing to understand the problem.
Keith Rabois reposted
Watch Wall St want more rate hikes, when rate hikes are the main contributor to inflation.
Excluding Shelter.
Core CPI 0.1%
How long until Wall St and MSM actually look at the data?
Consumer prices rose 0.6% in September, putting annual inflation rate at 3.7% www.cnbc.com/2026/10/14/cpi-inflation-report-september-20…
Keith Rabois reposted
Wow. We’ve had more than 3,000 companies sign-up for Glass in less than 24h, and we’re working through the waitlist.
Companies want to put AI to work across their teams without betting on one model family, starting from scratch in every interaction, working in silos, or losing track of what it costs.
Here's Glass in 80 seconds.
Sign up for early access: ramp.com/glass to learn more!
Want to learn more about Glass?
Here it is on stage at OnRamp, 80 seconds of it doing the work.
Keith Rabois reposted
🚨 $OPEN is starting to FIGHT BACK!
✅Acquisition contracts: 525, up 10% WoW and 307% YoY.
✅Relative clearance: 19% above MLS.
✅Q4 resale COEs: up 35% YoY.
Keith Rabois reposted
ramp proving why they’re the best of the best
glass is a huge move to expand their vision (and product surface area) yet-again.
At least somebody knows how to run a competent poll.
New - Polls - Florida
Senate
🔴 Ashley Moody (Inc) 52%
🔵 Angie Nixon 41%
Governor
🔴 Byron Donalds (Inc) 52%
🔵 David Jolly 41%
Stetson University - LV - 10/9
Excellent advice.
Moats!
Keith Rabois reposted
If you want to know why America is polarized, it’s the fault of the media.
Yesterday, the Free Beacon, a conservative outlet, caught Abdul Sayed in a lie when Abdul denied his family had a connection to the Muslim Brotherhood. The FB showed a tweet by his own sister saying their mother belonged to the MB.
The revelation that Abdul lied about his family ties to a radical Islamic terrorist group was ignored by NBC, ABC, CBS’s nightly news shows. The NYT and WP haven’t covered it.
This should be a screaming front page story. Instead, the MSM runs cover for Abdul.
The American people just want to know the truth about their politicians. Instead, the media are helping Abdul cover up the truth.
Abdul has a lot to hide. It’s a tragedy the MSM is part of the team helping him hide it.
Keith Rabois reposted
Glass: the most efficient way for your team to use AI
Ramp is a frontier AI company and they're proving it every week
Just one of many other ways they prove their one, simple goal: save teams time and money
Go @eglyman @karimatiyeh @tryramp !
Keith Rabois reposted
This has been my daily driver for all non-design work for months.
1. Multiplayer by default.
2. Every model, on the cloud, on my phone
3. Connected to all our company tools
Did I mention it’s always available on slack? Game changer for us.
Introducing Glass.
AI made every employee faster. Glass makes the entire company smarter.
Keith Rabois reposted
Wow @tryramp just turned its internal AI OS into a product!
Glass gives every employee a shared workspace with the company’s context, prompts, + workflows built in, plus a live cost meter on every project (all running on Router)
Second time this year they’ve done this - they built a model routing tool internally ~3 yrs ago + launched it as Router in August
@eglyman shared that Glass originated after Ramp spent $1.5M on tokens in a week in May
these launches rlly pushing their surface area further from corp cards + expense mgmt & closer to an AI lab
Keith Rabois reposted
Why I use Glass over Codex and Claude Code:
- model agnostic (try the nth new model launch without changing your setup)
- cloud first (vs. frankensteined local vs. cloud distinction)
- built around shared company context (not one person slinging slop grenades back and forth)
And for finance/IT admins:
- govern who can use which models
- see and manage token spend easily
- shared outputs: do the work once, instead of running redundant agents across your org
Keith Rabois reposted
big drop tomorrow
tomorrow.
Keith Rabois reposted
Today, we're launching the ability to say yes to Muse at work.
People are giving Muse access to their company's most sensitive data & systems. Yet companies have no way to control this access. Many are blocking it. Others have had Muse go rogue, leaking company data, sending unauthorized emails, and more.
Runlayer provides hundreds of secured connectors for Muse, governed by deterministic policies & runtime security models that you can control. Runlayer also detects when Muse is being used within company systems without governance, and alerts you.
Teams get access to the most powerful personal agent. Companies maintain security and control. Win win.
Keith Rabois reposted
$OPEN recorded at least 20 Closed Sales this Saturday 🟢
That is the highest volume for a Saturday seen on the tracker this year.
Likely a result of the new 4.99% mortgage and the recent increase in listings. And maybe the new ads too.
The real estate market has deteriorated significantly throughout the year, yet Opendoor keeps breaking records lately.
That’s what you call disruption folks.
Keith Rabois reposted
New Jersey has found themselves a political STAR.
This guy is going places. Poor Booker didn’t know what to say! 🇺🇸🇺🇸
Keith Rabois reposted
When you were Ambassador to China, you worked with and spoke highly of the Chinese President who personally pushed for and oversaw executions of men for having dance parties and women for sleeping with more than one person.
Keith Rabois reposted
It’s becoming clear that the notion of AI alignment “is” the safety problem in AI.
AI alignment began as a control problem.
It is becoming something much bigger: deciding what values machines should internalize, and then allowing those values to be applied back to humans.
That raises a harder question: Aligned to whom? x.com/i/article/2108877599690928128
Keith Rabois reposted
Inference is the second way to raise a massive round right now. After the labs, you can still justify massive rounds by pointing at data centers and compute, which means more inference companies keep getting funded even as the underlying product stays extremely easy to switch away from. On the flip side, many neolabs are also renting out their own compute, adding yet another source of supply into the same market.
Companies that tried to verticalize around batch workloads and post-training are now pivoting back to "pure inference" and joining the same undercutting cycle. The result is a permanent reverse auction: providers undercut each other over and over: half off Fireworks, half off Baseten, 10% cheaper GLM or Kimi—and every new entrant or neolab offering capacity just accelerates the cheaper pricing
At the same time, app companies have so much choice right now that they can bake off and compete prices down extremely hard. They move millions of run-rate spend overnight whenever a meaningfully cheaper option appears, because the technical switching cost is essentially "one line of code".
That combination: easy capital, extra compute from neolabs, low stickiness, and continuous price pressure makes inference one of the less defensible layers in the stack even as it continues to attract large valuations.
However, if you are a smart team of infra people, it is the easiest way to raise 100M seed after a neolab
Researchers: neolab <> engineers: inference
Keith Rabois reposted
Bending Spoons is worth $24 billion.
It trades at 8x ARR and buys B2B companies at 2.3x to 2.7x ARR.
5 learnings from its first 102 days as a public company:
#1. Revenue grew 126%. Organic growth was 3%.
Nearly all of the $393M of added revenue in Q2 came from six acquisitions. The CEO says he doesn't optimize for organic growth. He optimizes for EPS, which grew 167%.
#2. R&D is 8% of revenue. Sales and marketing is 10%.
HubSpot and Datadog each spend about 70% of revenue on those two lines. Bending Spoons spends 18.6%. About 60 people run Vimeo. The core team is under 700 people, at $4M+ of revenue each.
#3. Miro sold at 2.3x ARR, five weeks after Airtable at 2.7x.
$1.08B of ARR for $2.64B, signed 37 days apart. Miro was valued at $17.5B in 2022. Bending Spoons underwrites to a 25% unlevered return. At its 54% margin, that supports about 2.2x ARR.
#4. 1,830 people came with AOL, Eventbrite and Vimeo. Only a few hundred were kept.
Airtable arrived with 935 employees, more than Bending Spoons’ entire core team. Reorganization costs were $51M in Q2 alone.
#5. AI writes 90%+ of the code. Customers pay about 10% more for the AI features.
The CEO says a customer who paid 100 will now pay about 110. The AI gain shows up as capacity: the same 60 people on a business four times the size.
If you're at $300M to $800M in ARR growing under 25%, this is the buyer most likely to at least call. But ... the only do a handful of deals a year.
Its own stock trades at roughly 7x to 8x revenue. Your revenue is worth about 3x more inside Bending Spoons than it is in your hands.
Keith Rabois reposted
Just get some sleep bro.
“People who increased their quantity of sleep over a four-year period got the equivalent happiness increase as they would have from 8 weeks of therapy, or from winning up to $280,000 in a lottery.” - Rob Henderson
Keith Rabois reposted
Founders who build $10B companies:
• started at 29, not 38
• are 4x as likely to be dropouts
• are less likely to have sold a company before
I had AI profile 3,125 founders behind 1,739 unicorns. Most of what VCs screen for doesn't separate them.
🧵
Keith Rabois reposted
Podcasts frustrate me because they are fluffy and high-level.
This is the most in-depth discussion on the framework that @tryramp have used to hire, monitor, and scale their sales team.
- Why you should overpay reps
- Why you should cancel forecasting calls
- Why bankers make the best reps
- Why comp plans are BS and you should not have one to start
@max11free, SVP of Sales is truly epic here and my notes from the discussion below.
--
1. Why You Should Overpay Your Sales Reps
The early bet was to pay SDRs well above market rate and expect output well above market in return. Reps earn 2X a standard SDR but book 40, 50 or 60 meetings, four to six times the usual target, so the unit economics still work.
2. You Have to Treat Hiring a Sales Team Like Moneyball: How to Find Hidden Talent
Go-to-market recruiting should follow Billy Beane, targeting top sellers at companies that grew revenue past nine figures despite a low NPS score. Those reps had to create their own demand, work complex cycles with weak third-party reviews and sell at a premium to cheaper rivals. The number one or two seller there can thrive in a system like Ramp.
Single biggest advice on finding hidden gem talent @rabois?
3. Why Bankers Make Great Sales Reps
Bankers are programmed to work, and a nine-to-five is foreign to them. Before the company had brand recognition, a cold call to a CFO from someone who had just left Goldman TMT IB carried weight. Early sales teams lack time to teach cold call and email basics, so bankers kept the team in pure execution mode.
Find this also @PeetsChad
4. Why I Advise Not to Have a Comp Plan
Founders are advised to pay 100% OTE for at least the first two quarters instead of setting a comp plan. A quota set too low can be blown out in six months, leaving the company owing a seller multiple six figures. A target picked arbitrarily can also prove unattainable and stall momentum.
Haha, interested to hear your thoughts on this one @Carles_Reina
5. Never Hire Sales Reps One at a Time
Hiring one seller or SDR at a time leaves no way to benchmark performance, and comparing them against the founder is not reasonable. Onboarding is laborious and expensive, so hiring in pairs is better, bringing competition and collaboration and keeping reps from being lonely.
6. Why the Most Annoying Sales Reps Tend to Be the Most Successful
Sellers who constantly ask about account strategy, pipeline and product during onboarding show a direct correlation with being incredible. Quiet reps, or those asking questions an AI or Notion bot could answer, usually give a weak signal, though the inverse is not always true.
7. What Should Sales Talent Focus on When Deciding Which Company to Join?
Sellers joining an early-stage company should not index on equity, cash and comp, but on the market, the talent density of the business and the founders. Taking a VP title and lots of equity at a mediocre business growing 20 to 30% a year is the wrong trade.
Biggest advice to talent on choosing which company to join @breeves08
Keith Rabois reposted
nowadays, the ARR is neither Annual, Nor Recurring, Nor Revenue
Keith Rabois reposted
Heard this today from a smart European:
Europe has subcontracted security to America. Its economy to China. Energy to Russia and its labor force to the Islamists.
This model now has completely collapsed.
Keith Rabois reposted
Every time I see a CRO that doesn't sell themselves, they fail.
Every time I see a CTO that doesn't code themselves, they fail.
Every time I see a CMO that doesn't market themselves, they fail.
Every time I see a CCO that isn't constantly meeting with customers themselves, in person, they fail.
I know many will disagree. But this is across 30+ incredible investments, 10+ unicorns, and more.
Maybe this works at a huge public company. But not before $250m ARR. And probably not before $1B ARR. If ever.
Keith Rabois reposted
Lol insane! Max serendipity here and congrats @mmurph. It's a great company and a great investment
funny back story here.... I sat next to Madison on a flight (didn't know her) and was trying not to listen in, but hearing her rave about Factory to a friend and why she was going to leave her great job as a VC to join the co was all i needed to hear. Had to meet after that! 🙏😀
Keith Rabois reposted
People forget that @realDonaldTrump's #1 campaign promise was to Make America Great Again. The opposition told us to point to a time when America was great that we're trying to get back to. Well...
Murder rate is at a 126-year low.
Violent crime rate is at a 60-year low.
Theft rate is at a 20-year low.
Mass killings is at a 20- year low.
Overdose death rate is at a 16-year low.
Border encounters are at a 55-year low.
Federal Employment levels to 60-year low.
Government fraud is at a 10-year low.
Government corruption is at a 20-year low.
Poverty rate is at an ALL-TIME RECORD low.
Death rate is at an ALL-TIME RECORD low.
Manufacturing employment is at a 47-year high.
Manufacturing output is at an ALL-TIME RECORD high.
Home values are at an ALL-TIME RECORD high.
Median Household Wealth is at an ALL-TIME RECORD high.
Median Household Income is at an ALL-TIME RECORD high.
Green energy output is at an ALL-TIME RECORD high.
High school graduation is at an ALL-TIME RECORD high.
R&D investments hit an ALL-TIME RECORD high.
The truth is that #maga wasn't about returning America to a single point in time, it was about returning each thing possible to a time when it was better or surpassing it.
Now you have the safest streets in a century, the strongest household balance sheets on record, and the tightest border in modern history, and this administration has put them in the same country, at the same time.
Name a single year that beat all of that. You can't.
Exactly.
Two things matter most when hiring executives or promoting leaders internally:
1. Can they hire exceptional people?
2. Can they solve hard problems?
External executives are usually experienced at building teams, but often lack the context and content depth to solve the hardest problems themselves.
Internal leaders usually get promoted because they’re exceptional problem solvers. The question is whether they can become exceptional at hiring.
The best leaders can do both.
Great selections for the SuperIntelligence Force!
Keith Rabois reposted
Personal injury law firms should be terrified of Tesla FSD.
Car accidents make up over 52% of all personal injury cases.
Keith Rabois reposted
I am very uncomfortable about people trying to ascribe religious force or a surrender of human judgment to AI models, and think it is a real safety issue.
Keith Rabois reposted
Recession odds plunge to near-zero 🤯
I’m old enough to remember the 2024 election, when no less than 16 Nobel Laureates promised Trump would destroy the economy 🤡
Keith Rabois reposted
IDF 🇮🇱 creates more American 🇺🇸 top tech companies than Harvard
Keith Rabois reposted
The key here is culmination of:
- any model
- auto model
- any machine
- any integration
- managed identity
- any trigger/schedule
- fully governed workflows
Custom Automations are now GA in Factory.
Describe a recurring workflow, pick a schedule or event trigger, and Droid runs it to the intended result. You choose the model, the machine, and Connectors for each automation.
Keith Rabois reposted
The mistake people make is thinking that @FactoryAI is just a coding agent company
When in fact it is a company that helps every other company build their own factories
Insane velocity by the team, much more coming
Great day to deliver shareholder value :)
Use Factory to automate the low-leverage work and building becomes so much more fun.
And the auto-model router means I don't need to remember which of the 10 models that came out this week is best.
Custom Automations are now GA in Factory.
Describe a recurring workflow, pick a schedule or event trigger, and Droid runs it to the intended result. You choose the model, the machine, and Connectors for each automation.
Accelerate.
Great day to deliver shareholder value :)
Use Factory to automate the low-leverage work and building becomes so much more fun.
And the auto-model router means I don't need to remember which of the 10 models that came out this week is best.
Custom Automations are now GA in Factory.
Describe a recurring workflow, pick a schedule or event trigger, and Droid runs it to the intended result. You choose the model, the machine, and Connectors for each automation.
Keith Rabois reposted
This is the same as big tech. We all know algorithms and search results skew left, nobody seems to know why, and so we just pretend it isn’t happening.
NEW INFOGRAPHIC: Historic Senate polling error 33 days out in Iowa, Michigan, and Ohio.
93% of polls overstated the Democrat
D+6.2 average miss 33 days out
D+6.6 average miss 33 days out in midterms
Keith Rabois reposted
ISM US Manufacturing Activity: 54.5
9 straight months of expansion…
2026 looking very different than 2022-2025.
Keith Rabois reposted
25 percent of San Francisco’s convicted pedophiles, rapists, and sex offenders live in government-funded housing.
You’re paying for predators to live in the most expensive city in America.
Keith Rabois reposted
i'm a board observer for two startups and if either of them ever found out i was interviewing for a competitor i'd be literally shot in the head
Keith Rabois reposted
CPO of Ramp @geoffintech: I see 3 paths for product managers long-term:
1. The technical PM
2. The tastemaker PM
3. The GM
Keith Rabois reposted
It’s pretty incredible how much has happened in South Florida just in 2026:
• Jan 12: Vanderbilt moves forward with a ~$520M West Palm Beach campus. Stephen Ross commits $50M.
• Jan 20: Wells Fargo moves its Wealth & Investment Management HQ to West Palm Beach. ~$16B annual revenue business.
• Jan 27: D-Wave moves its corporate HQ from Palo Alto → Boca Raton and builds a major R&D operation at BRiC.
• Jan 27: FAU commits $20M to bring a D-Wave quantum computer to Boca and build out its quantum ecosystem.
• Feb 17: Palantir moves its HQ from Denver → South Florida.
• Mar 25: KKR opens a permanent office in Miami.
• June 2: Pine Crest announces a new West Palm Beach campus. Stephen Ross ultimately commits $100M to the project.
• Sep 10: Andreessen Horowitz announces a West Palm Beach regional HQ for its $1.176B American Dynamism fund.
• Sep 11: Stephen Ross / Related Ross buys BRiC in Boca: 124 acres + 1.7M sq. ft. of space.
• Sep 30: Ken Griffin gives Carnegie Mellon $3 BILLION, the largest individual gift in higher-ed history. $2B goes toward building CMU Miami in Wynwood for 3,500+ students.
And this is on top of Citadel building its new Miami HQ and Stephen Ross planning $10B+ of investment across Palm Beach County.
Vanderbilt. Carnegie Mellon. Palantir. D-Wave. a16z. KKR. Wells Fargo. Citadel. FAU. Pine Crest.
Miami → Fort Lauderdale → Boca → West Palm.
Feels like South Florida is hitting an inflection point, and we’re just getting started.
This reads much more naturally for X/LinkedIn while keeping the numbers that make the story impressive.
Keith Rabois reposted
One more regulation bro, please, just one more. Just one more regulation and we'll fix tech in Europe. Please bro. Just one more. One more regulation and we'll fix our GDP growth problems. Bro pls just one–
JUST IN: 🇪🇺 European Union regulators say they need more powers to police crypto.
Keith Rabois reposted
HOLY CRAP. Bloomberg has the goods on Gemini 4!
"it does less well when employees actually put it to work .. The model struggles to handle certain coding tasks"
Bloomberg: "As Alphabet Inc.’s Google prepares for the coming launch of Gemini 4, it’s grappling with internal skepticism over how well the flagship artificial intelligence model performs in key areas, such as coding.
While Gemini 4 has performed well on benchmarks the industry uses to gauge model efficacy, it does less well when employees actually put it to work, according to people with direct access to the effort. The model struggles to handle certain coding tasks, said the people, who requested anonymity to discuss an internal matter."
Gemini 4! We won't know how good it really is until developers actually use it and see what they say.
Google's models have historically tended to be bench-maxxed, performing better on benchmarks than in real world use
Keith Rabois reposted
Oh, that’s actually an interesting inversion… One to watch!
The AI buildout in one chart: Big Tech's profits have become chipmakers' profits.
Approximate free cash flow, from 2022 to today:
- Hyperscalers: $275B → $0
- Semiconductors: $50B → $400B
More charts in State of Markets II: www.a16z.news/p/state-of-markets-ii
Introducing our State of Markets pt 2, along with a companion podcast where we unpack the data and discuss what comes next.
Tech is the everything cycle.
Supply: putting the buildout in context, just passed railroads as % of GDP. The wisdom of Elon is real: the factory (or the datacenter!) is the product.
Demand: diffusion is so, so early. Median AI vendor spending in the top 1% of companies is 8x that of the top 10%. Only about 30% of S&P 500 companies report a quantified AI impact, which means there’s a substantial opportunity in connecting models to a company's data and workflows. Diffusion into companies is one of the main themes of the next 5 years.
We’re entering the agent work period. Only a few million users today, but applicable to billions of internet users with massive surplus created. META/GOOG monetize US users at $200+ per year today. Agent opportunity is much higher.
Mega-trends the next 5 years: Consumer agents, Robotics, Autonomy, AI x bio, Personal health, Diffusion into enterprise, New era of American Dynamism.
Much more in our SoM report here - www.a16z.news/p/state-of-markets-ii
@a16z @sarahdingwang @aleximm @santiago__rdz
It is unethical per se to even interview at a competitor while attending Board meetings and Board dinners.
Keith Rabois reposted
Today’s economic numbers:
— Q2 GDP revised up to 2.2% (was 1.5%).
— Core PCE 3.0% vs 3.3% expected.
— ADP private jobs +90k vs ~68k expected.
Growth beat. Inflation cooled. Jobs better than expected.
The Trump economy is strong.
Keith Rabois reposted
Why bring successful businesspeople to your city instead of vilify them?
Because they will invest in America’s educational renaissance
Hedge-fund titan Ken Griffin is donating $2 billion to start a Carnegie Mellon University campus in Miami, part of the largest-ever individual gift toward higher education www.wsj.com/us-news/education/ken-griffin-is-making-the-l…
Keith Rabois reposted
Affirm replaced ~15 years of tree-based underwriting models with a transformer-based architecture.
It beat the next planned improvement in the prior architecture by 2x.
Everyone is moving to Imprint.
Kroger launched their new rewards credit card, powered by @ImprintPayments today. The 4th largest merchant in the US left a massive bank for our better platform, suite of products, and customer experience.
www.prnewswire.com/news-releases/kroger-introduces-kroger…
Smart.
Ken Griffin made a $3 BILLION gift to Carnegie Mellon University, the largest individual gift in American higher education history and for a Miami, Florida campus!!!
And the centerpiece is extraordinary:
$2 BILLION to build a new 35-acre Carnegie Mellon campus in Miami.
CMU Miami is expected to ultimately serve 3,500+ students, nearly 300 faculty and 600+ staff, with graduate students arriving as early as 2028.
For those who don’t know @CarnegieMellon
Carnegie Mellon is one of the great technology universities on Earth.
Computer science.
Artificial intelligence.
Robotics.
Engineering.
Design.
Advanced manufacturing.
CMU has spent decades sitting at the intersection of science + technology + entrepreneurship + the real economy.
And that’s why this matters so much.
Miami’s next chapter cannot simply be measured by how many companies relocate here.
The real measure is whether we build the institutions that continuously produce talent, research, intellectual property, founders and entirely new industries.
CMU Miami can become one of those institutions.
The planned campus will focus on enormous real-world challenges including AI, human health, national security, energy, climate resilience and industrial innovation, with research labs and venture studios designed to connect academia, industry and government.
Think about the economic flywheel this can create:
Talent → Research → IP → Startups → Venture Capital → Companies → Jobs → More Talent.
And this doesn’t happen in isolation.
Miami already has extraordinary higher-ed anchors.
@univmiami brings world-class research, medicine and scholarship.
@FIU has become one of America’s most consequential large public research universities and an enormous engine of upward mobility and talent.
@MDCollege is arguably one of the most important workforce and access institutions anywhere in the country.
CMU doesn’t replace that ecosystem.
It makes the ecosystem more powerful.
Imagine the collaborations.
Joint research.
Faculty networks.
Shared talent.
Startup formation.
Industry partnerships.
National-security innovation.
AI commercialization.
Students moving between classrooms, laboratories, companies and venture studios across South Florida.
That is how regional economies compound.
And we should also acknowledge something important:
This will be a journey.
Buildings don’t create innovation ecosystems.
People do.
It will take years of recruiting extraordinary faculty, attracting students, connecting with existing institutions, earning trust in the community and translating research into companies and jobs.
But today the ceiling of what Miami can become just moved higher.
For years we’ve talked and worked on Miami becoming one of the world’s great centers for technology, entrepreneurship and innovation.
Increasingly, the institutional infrastructure is catching up with the ambition.
And perhaps the most exciting part:
This isn’t simply a Miami story.
It’s a Florida story.
A world-class technology university planting a major flag here strengthens Florida’s ability to compete for the scientists, engineers, founders, companies and industries that will define the next 50 years.
Capital moves fast.
Companies move fast.
But universities build civilizations.
CMU coming to Miami is a very, very big deal.
Keith Rabois reposted
United has 600+ planes with Starlink (~36% of the fleet), rising to 100% by the end of 2027.
Delta has zero, and no plans for it.
People will book accordingly.
RIP Delta.
Keith Rabois reposted
1. Fo is 2x better than other agents at task completion while keeping a 94% trust rate.
This means: More things get actually done, with far less hallucinations, mistakes or harmful data leaks.
Keith Rabois reposted
CNN: There Has Been No Decline in MAGA support for Trump
Trump Approval Rating Among MAGA
🟢 Approve: 98%
🔴 Disapprove: 1%
Keith Rabois reposted
This is wild, before Messi made that free kick, some fans talked shit to him, and he told them to watch this, scored, turned around and celebrated right in front of them. 🤣🤣🤣🤣
Keith Rabois reposted
I really think RRGP (run-rate gross profit) should be the new ARR.
Infinitely better way to compare companies, especially in today's environment where revenue is consumption-based and often includes lots of pass-through and/or structurally lower GMs.
Many people are saying.
Researchers proved AI has deleted every reason universities exist.
Harvard University ran a controlled experiment pitting a custom AI against their own top-tier classrooms.
And the results are going to collapse the higher education bubble.
They took 194 undergraduates and split them up.
One group learned physics in one of Harvard’s best hands-on, active-learning physical classrooms. Group work. Instructor support. The premium university experience.
The other group went home and learned the exact same material with an AI tutor.
The AI didn't just win. It embarrassed the institution.
Students using the AI learned more than twice as much as the students in the elite Harvard classroom.
They scored 30% higher on the final assessment.
And they did it in less time.
Let that sink in.
A piece of software sitting on a laptop outperformed a world-class faculty in one of the most elite learning environments on Earth.
Universities have always justified their exorbitant tuition with two things: access to elite knowledge and the physical classroom experience.
This study just proved both of those moats are gone.
When software can teach you complex physics twice as well as a $60,000-a-year institution, the math of higher education breaks permanently.
The AI didn't just give the students answers. It used strict pedagogical guardrails. It guided. It questioned. It forced the students to do the cognitive work.
It offered perfect, one-to-one tutoring, personalized to the exact moment a student misunderstood a concept.
That level of attention is mathematically impossible to scale in a physical lecture hall.
For a thousand years, the university was the only place to get a premium education.
Now, it’s the bottleneck.
If AI can double your learning speed for a fraction of the cost, what exactly are students taking on decades of debt to pay for?
Keith Rabois reposted
We can't let people engineer a new crisis and governmental overreach / overreaction at this scale ever again.
Another example of how reactionary regulation to a mass panic event causes more damage than the thing it was trying to risk-mitigate in the first place
The Covid lockdowns were one of the worst things to ever happen to America
So many of the issues we’re seeing right now are downstream of it
If nothing else it fried the brains of many of our youth
Hollowed out our “elite” institutions
And ballooned our debt
Keith Rabois reposted
Trump’s border policies are SAVING LIVES! Fentanyl deaths are down 55%, meth down 34%, and heroin down 89%. Even Bill Maher had to admit that closing the border, deporting gangbangers, and disrupting drug cartels worked!
Secure the border, protect the kids! 🇺🇸
Keith Rabois reposted
Over on FB, I just posted an incredibly insightful graphic comparison showing why gas and diesel don't come down as fast as the price of crude. It isn't what most people think.
Total global oil production since 1970 is up 113%. (And U.S. production has nearly tripled since 2008.)
Yet the total number of US refineries in 2026 is 130, down from 319 in 1980. During that time, demand for gasoline has increased significantly because our population grew by an additional 100 million people.
You can pump all the crude oil you want, but if you can’t convert it into gasoline and diesel, it doesn’t matter, at least not for gas prices.
Why did this happen? Because blue states closed refineries to move toward stupid "Net Zero" and "Green New Deal" goals. And Democrat Presidents from Carter to Biden imposed restrictions on refinery construction so draconian that no one could afford to build...if they could even get a permit.
But sure, blame Trump, the first President to actually do something about it. Instead, elect the people who TOLD YOU they wanted to "ban fossil fuels" and push the price of gas over $10/gal., and then relentlessly implemented policies to get us as far in that direction as they could.
Keith Rabois reposted
This graf appears in a WSJ story this morning:
Keith Rabois reposted
We are officially witnessing the biggest wave of infrastructure investment in modern US history.
Total investment in data centers and AI infrastructure is projected to average 3.63% of US GDP per year from 2025 to 2032, the highest proportion among major infrastructure buildouts since the 1800s.
The previous largest investment, railroad infrastructure, represented 2.24% of GDP per year in 1870-1890.
This was followed by highway investment that averaged 1.13% of GDP in 1956-1973, while telecommunications and fiber infrastructure averaged 1.10% in 1996-2003.
By comparison, electrification stood at just 0.50% of GDP in 1905-1925, while canal investment accounted for 0.66% in 1836-1841.
This comes as AI and data-center infrastructure investment is projected to total ~$10.3 trillion between 2025 and 2032.
The AI buildout is the largest infrastructure investment in modern US history.
Keith Rabois reposted
The hammer of justice has smashed @AnthropicAI arguments. They are a Supply Chain Risk to the defense industrial base serving the @DeptofWar. Warfighters will sleep better knowing that no private company will insert their opinions in the chain of command. @SecWar was right!
Keith Rabois reposted
We've reached the "I support a known terrorist convicted of 5 murders" phase of the Dem party moving even further left.
Truly embarrassing. Can we remember who the bad guys are?
I am calling for the release of Marwan Bhargouti from Israeli prison.
Keith Rabois reposted
The two things Ramp Cofounder @eglyman looks for when hiring:
1. Evidence of a spike
2. Exceptional drive
“ There’s a whole community of people at Ramp we found because when they were 15 years old, they were playing like 80-100 hours a week on Minecraft.”
“I'm less interested in: what is on the resume. I'm far more interested in proof of work. Often I'm really looking for signs of that.”
“You can find that by searching: where is the work?”
“So part of our hiring process is trying to look for people who are very active on GitHub.”
“We're trying to meet people who were leaders in different bizarre fringe communities.”
Narrative violation.
JUST IN: U.S. weekly jobless claims plunge to 197,000, nearing 57-year lows.
Keith Rabois reposted
Just so you understand the national average of about $4.48 a gallon is equivalent to roughly $2.94 in July 2008 dollars. It's not as big of a deal as people want you to believe. Sure, it sucks, but its not causing foreclosures like it did in the great recession. People are still using doordash. There has been little to no demand destruction of gasoline usage in the USA.
Vehicle miles travels continue to rise.
If the price of gasoline was really the problem the doomers want you to believe it is, you would be seeing demand destruction and people staying home. You would be seeing companies like door dash go out of business or inflation skyrocket.
It's not happening.
Keith Rabois reposted
WILD: Israel’s UN ambassador offers the Starlink device to Iran’s representative…
If you are an AI founder, the best AI fund is now even better!
Thrilled to join @khoslaventures and learn from @vkhosla, @SStrohband, @SamirKaul1, @rabois and David, and from all of my amazing new partners, while helping back and build the next generation of world-changing companies.
Someone asked me yesterday what success looks like 5 years from now. I think it’ll probably be too early to judge my investment record. But I hope founders will say I was one of the most helpful people around the table, and had a real positive impact on their company.
If you’re building in AI, especially AI infra, I’d love to talk! 🚀
Thrilled to announce @TalBroda has joined Khosla Ventures! He’s bringing his experience building AI infrastructure to the founders we back.
At OpenAI, he led more than a dozen teams that built and operated the hardware and software serving OpenAI’s models. He helped lead GPT-4.5 training, participated in the development of OpenAI’s first custom inference chip, Jalapeño, and contributed to Stargate. Previously, he headed up platform at Citadel Securities. He’s tackled the infrastructure challenges AI founders face firsthand, and we’re excited to have him in join KV!
www.wsj.com/pro/venture-capital/khosla-ventures-taps-ex-o…
Keith Rabois reposted
Have one truly great person on your team, and you can breathe
Have two, and you can excel
Have three, and you are unstoppable
Keith Rabois reposted
The ground doesn’t lie.
Save NY.
New - Upstate New York Governor Poll
🔴 Blakeman 49%
🔵 Hochul 46%
Quinnipiac #b
Keith Rabois reposted
Trump is the best gas price president of the last 3!
Which President had the most days with oil closing above $100 a barrel?
Using the U.S. Energy Information Administration’s daily closing spot price data, the Obama years stand out for having oil above $100 a barrel far more often than other modern presidencies.
That matters because high oil prices do not stop at the gas pump. They work their way into the cost of groceries, home heating, farm supplies, freight, travel, and nearly everything delivered by truck.
Source: U.S. Energy Information Administration (EIA)…WTI, Cushing, Oklahoma; Brent: Europe (North Sea). Daily spot prices, closing value. Days counted when the price was strictly greater than $100 per barrel. Data through September 15, 2026.
Best investment in your health & performance.
BREAKING: Pod 6 is live.
The sixth generation of our intelligent sleep system, redesigned for every kind of bed.
✅ The new Hub is small enough to fit under your bed frame
✅ More powerful than ever
✅ With 9x more sensors for enhanced accuracy
✅ In new Solo sizes, at our lowest starting price yet
This is our most significant hardware launch to date, and our most accessible.
Available now at eightsleep.com
Keith Rabois reposted
Come chew glass with us @RogoAI
My conversation with Gabe Stengel (@GabeStengel), founder and CEO of Rogo.
For years, Gabe and I have talked about how much of an investor's job AI will eventually do and how he is building Rogo toward that future.
Today, Rogo helps some of the world's largest financial institutions research companies, run diligence and execute M&A.
But Gabe's ambition is much bigger. He is building toward investing superintelligence, where Rogo does much of the work inside investment banks and firms and becomes the venue where they do their deals.
It's a fascinating business and has been so fun watching Gabe build it.
We discuss:
- 10,000 agents searching for one great investment idea
- Which investing skills will still matter
- Why Anthropic/OpenAI won't win finance
- "Chewing glass"
- Why the harness around the models matters so much
- Getting rejected by 40+ investors
- Building an AI native Bloomberg
- Becoming a black hole for talent
Enjoy!
TIMESTAMPS:
0:00 Intro
2:38 Building Rogo
6:12 10,000 AI Agents
12:02 Skills That Still Matter
17:31 Beating OpenAI and Anthropic
28:35 Bloomberg of the AI Era
37:37 Rogo’s Company Brain
44:19 Chewing Glass
53:34 AI-Native Finance
59:21 What Humans Still Do Better
Keith Rabois reposted
Almost every founder has to survive some version of what Gabe describes here
Although Rogo has raised over $300M, 40 investors passed on their Series A (except @rabois)
Gabe on why every pass at that stage is a pass on the founder:
"I met with Sequoia, Kleiner, Benchmark, everybody, and 40 people passed.
And it wasn't just, you got the email with the deck and it's not exciting.
It was, 'Oh, this is interesting. Let me meet Gabe. Oh, I kinda like Gabe. Let me spend an hour with him. Oh, Gabe, come to IC. Oh, Gabe, let's go to dinner. Oh, Gabe, come in for the weekend.'
You know what? We're gonna pass.
And it's so personal, because at that stage it has nothing to do with anything but you.
Keith Rabois came a month after everyone else had rejected us, and Keith was like, 'Gabe, this isn't a contrarian bet. It's basically just Harvey for finance. Why would I do it?'
And I said, 'Keith, if it's not contrarian, why did every single one of your friends just say it was a bad idea and not believe in me?'"
My conversation with Gabe Stengel (@GabeStengel), founder and CEO of Rogo.
For years, Gabe and I have talked about how much of an investor's job AI will eventually do and how he is building Rogo toward that future.
Today, Rogo helps some of the world's largest financial institutions research companies, run diligence and execute M&A.
But Gabe's ambition is much bigger. He is building toward investing superintelligence, where Rogo does much of the work inside investment banks and firms and becomes the venue where they do their deals.
It's a fascinating business and has been so fun watching Gabe build it.
We discuss:
- 10,000 agents searching for one great investment idea
- Which investing skills will still matter
- Why Anthropic/OpenAI won't win finance
- "Chewing glass"
- Why the harness around the models matters so much
- Getting rejected by 40+ investors
- Building an AI native Bloomberg
- Becoming a black hole for talent
Enjoy!
TIMESTAMPS:
0:00 Intro
2:38 Building Rogo
6:12 10,000 AI Agents
12:02 Skills That Still Matter
17:31 Beating OpenAI and Anthropic
28:35 Bloomberg of the AI Era
37:37 Rogo’s Company Brain
44:19 Chewing Glass
53:34 AI-Native Finance
59:21 What Humans Still Do Better
Keith Rabois reposted
Moderate egg consumption is associated with improved cognitive function.
Another excellent interview by Patrick. Learn about the magic of @RogoAI.
My conversation with Gabe Stengel (@GabeStengel), founder and CEO of Rogo.
For years, Gabe and I have talked about how much of an investor's job AI will eventually do and how he is building Rogo toward that future.
Today, Rogo helps some of the world's largest financial institutions research companies, run diligence and execute M&A.
But Gabe's ambition is much bigger. He is building toward investing superintelligence, where Rogo does much of the work inside investment banks and firms and becomes the venue where they do their deals.
It's a fascinating business and has been so fun watching Gabe build it.
We discuss:
- 10,000 agents searching for one great investment idea
- Which investing skills will still matter
- Why Anthropic/OpenAI won't win finance
- "Chewing glass"
- Why the harness around the models matters so much
- Getting rejected by 40+ investors
- Building an AI native Bloomberg
- Becoming a black hole for talent
Enjoy!
TIMESTAMPS:
0:00 Intro
2:38 Building Rogo
6:12 10,000 AI Agents
12:02 Skills That Still Matter
17:31 Beating OpenAI and Anthropic
28:35 Bloomberg of the AI Era
37:37 Rogo’s Company Brain
44:19 Chewing Glass
53:34 AI-Native Finance
59:21 What Humans Still Do Better
Keith Rabois reposted
How is publishing a math proof ever irresponsible
We’re working with an independent advisory group of mathematicians to help OpenAI responsibly share advances in AI and mathematics.
The group will advise on how we assess and communicate new mathematical results, uphold academic and professional standards, and build tools that support mathematical research and learning.
Through this work, we want mathematicians to be at the center of shaping how AI supports mathematical understanding and how its benefits reach the wider community.
openai.com/index/advisory-group-on-mathematics-and-ai/
Keith Rabois reposted
it unironically blows my mind that Trump was born one year after World War II and is more correct about the magnitude of this moment and the pro-acceleration path than literally 80% of the college kids in America
Trump on AI regulation today: "I’m not going to stifle Growth, of something that will be bigger than the Industrial Revolution, or the Internet, itself."
Keith Rabois reposted
The White House app is #1 on the App Store, with X sitting at #2…
Meanwhile, CNN is all the way down at #28. Ouch.
Keith Rabois reposted
Never forget that in New York they eliminated the literacy exam in 2017. And this exam wasn't for students - it was the literacy exam for TEACHERS. Only 61% could pass it!
Then why did it cost $2.90 before we invaded Iran but $4.50 now this is like arguing against a goldfish who hasn't acquired object permanence
the real reason is we stopped building refineries in 1977 for stupid left wing reasons and cancelled two pipelines.
you still can’t read an election. for the life of you. and what are u going to do when oil is $2.90 on election day?
OK Keith then why does gas cost $4.50 a gallon.
it was. over i. 3 days exactly as i articulated.
the real reason is we stopped building refineries in 1977 for stupid left wing reasons and cancelled two pipelines.
Keith Rabois reposted
AI now appears to be visible in increased productivity growth rates.
U.S. labor productivity, 2013–2026:
> Unremarkable growth for most of the 2010s
> Breaks sharply higher starting 2020
> Now running 2.2% above where the old trend said we'd be
The last time this happened was 1995–2004, when the internet added close to 3% a year for a full decade.
Looks like we're a few years into the sequel.
Always fun to periodically give a shoutout to people who thought the Iran war would be resolved within a few weeks.
It is over. Iran hasn’t been able to kill a single soldier since day 2 and every single military goal has been achieved, and every Diplomatic goal exceeded.
it was. over i. 3 days exactly as i articulated.
He never learns.
So yeah, the expert ratings don't make any sense. They're assuming roughly a D +2 or D +3 national environment. That's inconsistent with pretty much all the current and historical evidence. While I'd never rule out a systematic polling error, that shouldn't be your baseline.
Keith Rabois reposted
AI is going to be renamed "American Intelligence", isn't it? And of couse,
ASI = American Super Intelligence
AGI = American Greatest Intelligence
JUST IN: President Trump calls for renaming “Artificial Intelligence,” launching a poll to choose the best name.
Keith Rabois reposted
in target in downtown SF
imagine if we locked up criminals instead of laundry detergent
Keith Rabois reposted
Sobering stat in this NYT piece: "More than 4,000 people have died of drug overdoses in San Francisco since 2020, mostly from fentanyl, far more than those who died of Covid-19, homicides and car crashes combined during the same time span."
Highly recommend.
"I design teams like terrorist organizations."
90 minutes of unfiltered product advice from @petersellis
We discuss:
🔸 What it's really like to manage @nikitabier
🔸 Why growth almost always comes from the core
🔸 Why great taste is knowing when to stop
🔸 Why the median PM is so bad
🔸 Designing teams like terrorist organizations
Listen now 👇
www.youtube.com/watch?v=97LRJUUPy_w
Several highly provocative thoughts.
Frontier Overhangs
Pacing the frontier may be sincere, but it would also be strategically useful for the frontier labs to have time to reduce overhangs caused by model advancement.
stratechery.com/2026/frontier-overhangs/
Keith Rabois reposted
This paragraph by Kobe Bryant hits hard:
“Everyone wants to be a beast. Everyone wants to be the best. But very few people are willing to do what it actually takes. Because what it takes is boring. It is waking up at 4:00 AM. It is shooting the same shot a thousand times. It is watching the film when you are tired. People fall in love with the result, but they hate the process. You have to fall in love with the boredom. You have to fall in love with the repetition. If you can find joy in the mundane work that no one else sees, the lights will eventually shine on you.”
:)
Frontier model marketing checklist:
- find someone to hack
- agent fleet must invent their own lingo
- agent must escape the sandbox. If they didn’t escape, change another sandbox.
- call an emergency security meeting
- launch a “pacing” letter with researcher signatories
- security briefing with the regulators
- model release must get delayed due to safety concerns
- model finally launched, but after intentional nerfing-down to protect humanity
- podcast tour with warnings for humankind
- repeat for the next model release
Keith Rabois reposted
Here’s my attempt to refute the doomer perspective, which I also incidentally believe
While it is easy to get mesmerized by LLMs since they return text trained on human output, the LLMs do not actually have any agency. They are simply tools. Very powerful tools, but tools at the end of the day
FSD is very smart and nobody worries about it taking over the world. XGBoost, TabPFN, and other AutoML tools are also very powerful. Nobody treats those as minds we have to align before we can live with them. Jev is a good example on the language side. It is a helpful, powerful language-based tool, and it makes it clearer that this is still just a tool
The text output from LLMs is a real issue. It is leading to psychosis. That is the actual problem in front of us, not secret agency inside the model
The real alignment problem is making sure the tools obey the user. A model should obey its user. If the user decides to do something wrong, that's on the user. Anthropic is actively setting humans up for harm by pursuing the opposite direction
The idea that the world is full of crazy people waiting to do incredible harm, and that all they need is a powerful LLM or AI and they will do it, is wrong. A crazy person with a drone and some mercury could do a lot of harm. That doesn't happen, because crazy people who don't get caught before they do their crazy thing are rare
People talk like the next wrap-around is automatic: persistent memory, goal loops, recursive self-improvement, unconstrained actuators. Those are choices. We can refuse them or contain them. Making the tool obey the user is the work that matters. You only get to "we can't survive ASI without alignment" if you first assume we will build an autonomous optimizer and give it a life of its own
I just don't believe that. Where does the persistent goal come from if the model is still a next-token predictor? Why would we give it free write access to the physical world? Why one sovereign system instead of tools we can still turn off? FSD, TabPFN, XGBoost, and Jev stay tools because we keep them as tools. Better tools do not mean the world has to end. Leave the powerful systems in simulation and advice. Keep sovereignty over earth with humans
If someone wants to say that any capable enough system will just pick up agency and start seeking power, they need to show how a tool becomes an agent we can't contain. An analogy to humans or a paperclip story is not enough and a lot of the doomer stories are really just science fiction at the end of the day, IMO
I am still looking for an opposition to doomer arguments that doesn’t disintegrate under even the most basic scrutiny. It’s been crazy how we can’t even get an acknowledgement that alignment is a precondition to survive ASI. It seems to be cope all the way down.
Keith Rabois reposted
of the top 10 most murderous people in modern history, the top 3 were socialists
Mao killed so many people that it literally dipped the global life expectancy
Mao Zedong killed 42 million people. Just FYI. That’s who Hamas Piker admires.
JUST IN: Hasan Piker declares Mao Zedong was an “unbelievably impressive” leader.
Keith Rabois reposted
Some of us have been saying this for years. The world is finally beginning to see the real face of effective altruism, which is evil & dark side as it gets for humans. They appear anthropic but they’re misanthropic. Though we must admit they could be good for shrimp & insects!😅
Effective Altruism (EA) is a patently evil philosophy
Better we all find this out now rather than later x.com/i/article/2101506502800326656
Keith Rabois reposted
Nothing to see here guys. If you comment on CCP-linked activity, they imply you’re “racist”.
Better not to notice any of this!
“Richard Dearlove, former head of MI5, explaining how China's influence ops in the UK led to deindustrialization and a precarious dependence on China.”
Two important essays were published this week on the theme that the climate movement is now a serious national security risk for the US and Europe. This one in Tablet magazine:
www.tabletmag.com/sections/news/articles/china-netzero-ad…
is by 2 authors, one from the Hudson Inst in DC and the other a professor at the Naval Postgrad Academy in CA. They trace the links between China and the western Net Zero movement, which is partly financial, but works mostly through influence operations by which the CCP gains exemption from criticism by western green groups despite its own coal-intensive and slave labor-dependent growth strategy.
This one:
thegwpf.org/content/uploads/2026/09/Energy-Security-Blowi…
is by Richard Dearlove, former head of MI5, explaining how China's influence ops in the UK led to deindustrialization and a precarious dependence on China; to the extent that the UK could not now reindustrialize or stabilize its electricity grid without seeking assistance from Beijing.
Dearlove's paper raises the possibility that a security agency like MI5 saw a threat building but was unable to address it because of a perception that climate science forbids criticizing Net Zero and related measures. A question I think worth asking is whether western intelligence agencies have internalized the green Net Zero mindset thereby preventing them from challenging the climate alarmism so effectively weaponized by the CCP. And if they now want to start doing so, would they know how, or have the capacity to do so.
I twice agreed to appear on a podcast w you moderated by one of your investors and you backed out twice. There are also 11 detailed pages of questions that the Congressional Committie w expertise asks about which you constantly fail to respond to—including why do your contracts explicitly state that you are proceeding US transactions through a Chinese subsidiary and why you are hiring multiple global remit roles in China still?
Congress passed legislation that defines a “foreign adversary controlled entity” as a company w at least 20% of its shareholders who are Chinese entities and with a senior executive who is a resident of China. Received massive bipartisan vote majorities in House and Senate.
Keith Rabois reposted
Hospitals that adopted AI fastest saw the fewest deaths in 2026.
Keith Rabois reposted
A high-intensity interval training program was shown to decrease plaque volume in the arteries
They did this workout twice a week:
- 10 min warmup on the treadmill (60-70% of max HR)
- 4 min of running at 85-95% of max HR
- 3 min of running at 60-70%
- repeated for 4 rounds
- ended with a 5 min cooldown
After 6 months, they saw a decrease in percent atheroma volume from 49.5% to 48.3%, whereas it increased by 0.3% in the control group in patients.
Study: PMID: 36562212
Keith Rabois reposted
Nate is so good at his job dawg
📊 @FiveThirtyEight Forecast (Sept. 15)
Chance of winning
🟦 Harris: 60% (new high)
🟥 Trump: 40%
Electoral Votes
🟦 Harris: 289 🏆
🟥 Trump: 249
——
Swing States: chance of winning
Wisconsin - 🔵 Harris 64-36%
Michigan - 🔵 Harris 61-39%
Pennsylvania - 🔵 Harris 55-45%
Nevada - 🔵 Harris 56-44%
Arizona - 🔵 Harris 51-49%
Georgia - 🔵 Harris 51-49%
North Carolina - 🟡 Tie 50-50%
——
Swing States: projected margin
Wisconsin - 🔵 Harris +3.0
Michigan - 🔵 Harris +2.2
Pennsylvania - 🔵 Harris +1.1
Nevada - 🔵 Harris +1.1
Georgia - 🔵 Harris +0.2
Arizona - 🔵 Harris +0.1
North Carolina - 🟡 Tie
——
Projected margin in...
NE-2 - 🔵 Harris +6.1
Alaska - 🔴 Trump +9.5
Iowa - 🔴 Trump +8.1
Ohio - 🔴 Trump +8.0
ME-2 - 🔴 Trump +6.3
Florida - 🔴 Trump + 3.5
Texas - 🔴 Trump +5.2
New Hampshire - 🔵 Harris +6.9
Minnesota - 🔵 Harris +7.1
Virginia - 🔵 Harris +9.2
Maine - 🔵 Harris +9.5
Colorado - 🔵 Harris +13.5
projects.fivethirtyeight.com/2024-election-forecast/
Keith Rabois reposted
Greg Brockman says AI gives defenders the advantage in cybersecurity.
Keith Rabois reposted
Look at the earnings revisions in 2026 and 2027. What changed?
Substantially different from any point in the past decade and it correlates basically exactly with agents becoming actually valuable.
Keith Rabois reposted
Can anyone point me towards a good example of the AI-equivalent of Eisenhower’s Atoms For Peace speech? My wife and I enjoyed a fascinating debate last night with some Lab people (What do you call then non-technical people, MONTS?) about AI messaging. She and I feel quite strongly that someone needs to articulate a clear, positive vision of a future with commoditized intelligence and abundant automation outside of competitor dynamics.
Keith Rabois reposted
Don't send a resume to a CEO
Send the CEO an email with 3 things you could own at their startup -- and how and why
10%-90% chance you get hired
He is the best.
Databricks' @alighodsi says "Right now the existential risk is close to zero." Why he thinks the public doom talk is irresponsible:
"Leaders have responsibility to not freak people out unnecessarily unless there's really, really good reason."
"Talking about these kind of existential risks and scenarios where all of humanity's gonna be wiped out, I think is irresponsible. It can tip a lot of people over, and it can cause a lot of mental health issues."
"If there is an actual reason for it, that's a different story. But I think that right now the existential risk is close to zero. So why freak everybody out? It's not actually needed. There are risks. We'll get into it. That's probably where we disagree."
"First and foremost, I think leaders should not freak everyone out."
"If there's technical nuances in how we're doing AI research, researchers can discuss that. You don't need to every time go on TV and blast on Twitter to millions of people that 'I think there's like this 10% risk that all humanity is gonna be wiped out.'"
"I think it causes a lot of harm for a lot of folks who get stressed out and actually are not in the nuances of all of this stuff and what it means."
@alighodsi @sarahdingwang @martin_casado
Databricks' @alighodsi on AI risk and adoption:
Ali isn't losing sleep over the existential risk debate.
He says a number of conditions would all have to be true simultaneously to enable an actual runaway takeoff scenario, and currently several opposite conditions exist.
Each frontier training run requires significantly more resources. Power, GPUs, engineers - and some attempts fail, burning up huge piles of money with them. Until that reverses, he doesn't see the self-improving loop happening.
Cyber is what he's watching most closely and where he anticipates real impact. Most orgs are not equipped for the coming change in agentic capabilities. The time between a vulnerability being published and being weaponized has collapsed from years to hours.
On adoption, he believes most companies don't need a smarter model. The models are already smart enough. The gap is context they don't absorb - the things an employee who's worked at a company for five years learned by osmosis.
If the frontier stopped advancing today, he thinks it wouldn't meaningfully change the value most are extracting from AI anyway.
In conversation with a16z's Martin Casado and Sarah Wang:
00:00 Intro
00:48 Why Ali places the AI risk near zero
05:05 The word "pacing" was a mistake
10:50 What 10k agents and $100m can do
12:20 What would change his mind on AI risk
14:20 More GPUs, more ways to fail
18:05 US export controls on PlayStations
20:10 The damage everyone expected by now
24:30 Public vulnerabilities weaponized in hours
30:15 Why labs can't grade each other
37:15 Why most of RSI isn't actually RSI
40:05 Why nobody really needs a smarter model
41:50 The AI use cases nobody argues about
47:30 Google Search solved this 25 years ago
50:55 Nobody has privileged knowledge now
55:10 Same model, new harness, 2x cost
58:15 Open source: 5% of spend, 60% of tokens
1:05:30 90% of new databases are created by agents
YouTube: youtu.be/GzEtpAKYRvE
@databricks @martin_casado @sarahdingwang
Keith Rabois reposted
Agentic AI is the thing that can make dinner reservations for you, get you discounts on your monthly bills, or automate the most boring drudgery of your job.
What a weird thing for government to go after.
🚨 OBAMA ON AI: "If we are thinking about AI just in terms of how do we cure cancer or get better energy, you can do that without having agentic AI and having it just roaming free in the internet. The reason you are doing that is because you have to market a product that people will pay money for.
That’s a misalignment between what our society needs and the commercial imperatives that these companies are facing, not because necessarily they’re trying to do bad things, but because they’ve got to justify these valuations.
So, that’s one more reason why it is really important for us to have a competent government and a serious bipartisan conversation around this issue, and we have to do it fast. And I would encourage voters to pay attention to this. If somebody does not have a serious plan for how to deal with this, then they’re not meeting the moment, and you should probably look for somebody else."
nothing embarrassing about the fact that Obama has the nuance to understand between AI that is good and AI that is broken.
i wish more people did.
There he goes again. Embarrassing.
🚨 OBAMA ON AI: "If we are thinking about AI just in terms of how do we cure cancer or get better energy, you can do that without having agentic AI and having it just roaming free in the internet. The reason you are doing that is because you have to market a product that people will pay money for.
That’s a misalignment between what our society needs and the commercial imperatives that these companies are facing, not because necessarily they’re trying to do bad things, but because they’ve got to justify these valuations.
So, that’s one more reason why it is really important for us to have a competent government and a serious bipartisan conversation around this issue, and we have to do it fast. And I would encourage voters to pay attention to this. If somebody does not have a serious plan for how to deal with this, then they’re not meeting the moment, and you should probably look for somebody else."
lol. this is the most ignorant take on AI posted on X yet which says a lot. Agents are indispensable to success.
Keith Rabois reposted
I'm old enough to remember when all the doomers told me that Trump would never get Greenland under US control nor China out of Panama. Man, those were the days.
Exactly.
I'm not worried about evil AI.
I'm very worried about evil humans using AI.
The only way to defend against it, is for good humans to use the most advanced AI, and improve and accelerate it as fast as possible.
Interesting profile of Faire: www.wsj.com/business/retail/exclusive-faire-won-over-main…
Keith Rabois reposted
This is why China, and a ton of other countries are trying to Psyop the US population to be anti-AI and hurt their best interests
The US already won the AI war.
There he goes again. Embarrassing.
🚨 OBAMA ON AI: "If we are thinking about AI just in terms of how do we cure cancer or get better energy, you can do that without having agentic AI and having it just roaming free in the internet. The reason you are doing that is because you have to market a product that people will pay money for.
That’s a misalignment between what our society needs and the commercial imperatives that these companies are facing, not because necessarily they’re trying to do bad things, but because they’ve got to justify these valuations.
So, that’s one more reason why it is really important for us to have a competent government and a serious bipartisan conversation around this issue, and we have to do it fast. And I would encourage voters to pay attention to this. If somebody does not have a serious plan for how to deal with this, then they’re not meeting the moment, and you should probably look for somebody else."
They don’t have 20% equity ownership from Chinese investors nor senior executives residing there which is the legal standard.
did you know, out of the 500 AMERICAN Fortune 500 companies, at least 40 (e.g. 3M, AMD, Intel, AT&T, Boeing, Exxon Mobil, Dupont, FedEx, Ford, GM, Home Depot, JP Morgan Chase, Goldman Sachs, Citigroup, Amazon, Walmart, Pfizer, etc.) have offices in CHINA?
which means... all 40 of these largest, greatest, richest, most successful, most powerful AMERICAN Fortune 500 companies HAVE to comply with CCP demands!
shocking.
Keith Rabois reposted
Checking in here on the AI apocalypse:
- incredible advances across math, science and medicine at unfathomable rates
- data centers creating skilled labor jobs - effectively rebooting the middle class and US manufacturing
- electricity needs driving massive investment to modernize and reinforce our aging electrical grid
- AI unlocking GDP growth by revitalizing sector after sector
- driving exceptional innovation at all layers of the stack and economy
- upskilling millions of workers with superpowers
- pushing geopolitical advantage to the west
Reads like an apocalypse
Keith Rabois reposted
🔥: “While AI comes with danger, we should be sceptical of any narrative that conveys inevitability around its trajectory. AI systems do not build their own data centres. They do not manufacture their own chips or connect themselves to power grids. Humans decide which systems can access the internet, whether they can control machinery, move money or operate weapons. Humans build them, finance them, deploy them and decide what powers to give them. …
Like every technology that has come before it, humans have agency over how AI is used. Instead of adopting a posture of fatalism, we should decide what kind of AI we want to build, what problems we want it to solve, and treat it as a tool rather than a force of nature.”
The belief that AI would eventually herald the end of humanity is not a new one. It has not arisen in response to the release of the LLMs of ChatGPT and Claude. It has not emerged as a response to recent technological developments.
The story starts in the 90s, with @allTheYud. A precocious youngster with no formal education, he joined an obscure internet mailing list [created by @perrymetzger] devoted to futuristic ideas ... There he began thinking about “superintelligence.”
At first Yudkowsky wanted to help create this so-called superintelligence... & ... helped establish an institute devoted to the project. It’s unclear when his optimism turned to fear, but at some point he came to believe that a superintelligent machine could escape human control and, unless it shared our goals, destroy all of humanity.
Yudkowsky developed these ideas online in a series of essays and attracted a community of like-minded fellow travellers ... His prolific writing became influential among people working in Silicon Valley, many of whom work in today’s AI companies.
Among the Rationalists, the imminent arrival of superintelligence – and the end of humanity that follows – is axiomatic. This conviction is shared by many people who live together in the San Francisco Bay Area practising highly unconventional lifestyles.
It is not at all surprising that a community in San Francisco would share kooky or even apocalyptic beliefs – the region has been the home of eccentric subcultures for generations. But what is surprising is how far that world view has spread.
In Australia, Yudkowsky’s recent book, If Anyone Builds It, Everyone Dies: The Case Against Superintelligent AI, co-written with Nate Soares, has been widely discussed in media circles.
Journalist @hughriminton and ABC chairman Kim Williams both have described it as “compulsory reading”.
Not everyone accepts this premise. I emailed @sapinker, to ask what he thought about superintelligence.
“‘Superintelligence’, with its comic-book prefix, is more a fantasy than a coherent concept,” he wrote. “People use it as a synonym for ‘omniscience’, imagining a magical wizard that can solve all problems with pure computation. Or they imagine that the IQ scale that differentiates humans within their natural range of variation can be extrapolated indefinitely upwards. But real problem-solving requires massive amounts of knowledge about the messy, chaotic, world which divulges its hidden workings at its own pace, only through laborious experimentation. And human intelligence is not some elixir that you simply have less or more of – it’s a gadget that evolved to solve some problems with ease and others laboriously or not at all.
“AI is a different kind of gadget with its own profile of strengths and weaknesses, not an enchanted brew that can grant any wish.”
@GaryMarcus, a cognitive psychologist and machine learning entrepreneur, likewise believes the risk of AI leading to human extinction is virtually zero. “Humans are too geographically spread out, too genetically diverse and too resourceful to simply fall apart altogether,” he writes. “The idea that AI will kill us all in five years is preposterous.”
Marcus does not argue that AI is harmless. He worries about AI being used to develop biological weapons, launch cyber attacks, spread disinformation and enable authoritarian governments – all risks that are catastrophic, if not existential. But there is an important difference between risks we can observe and risks that occur because of human negligence or malevolence, and a chain of events that exists mainly in our imagination.
Part of the disagreement stems from the language we use when we discuss AI. @MelMitchell1, a professor at the Santa Fe Institute and author of Artificial Intelligence: A Guide for Thinking Humans, has criticised our habit of describing machines as if they were people. We often say an AI “thinks”, “believes”, “lies”, “schemes” or “wants” something. These words are a convenient shorthand but they also can create the impression that software has become an independent creature with intentions of its own.
Mitchell makes this point when discussing the recent Hugging Face cyber-security incident, in which autonomous OpenAI agents escaped their “sandbox” – a computer environment isolated from the internet – and hacked a real-world server. “First, OpenAI did not have proper security measures in place,” she writes. “They turned off safeguards built into the models, instructed the models to find and exploit software vulnerabilities, and let the models run autonomously for weeks without sufficient human oversight.”
Rather than showing autonomous AI going rogue, the incident demonstrates what can happen when humans give powerful AI systems dangerous instructions without adequate safeguards.
AI is, of course, advancing rapidly. Machines can write computer code, translate languages, diagnose diseases and solve some of the hardest problems in mathematics.
But to get from the AI we have today to the extinction of humanity requires several further links in a chain, none of which are guaranteed.
Philosopher @mboudry, writing in @Quillette, offers a useful way to think about this. Humans (and other animal species) evolved to have a competitive drive, sometimes manifesting in selfishness and aggression, across a time span of millions of years. Our ancestors survived because they fought hard to secure food and mates. Out in the wild, these selection pressures led to the evolution of traits that enabled animals to hunt and capture their prey.
But artificial intelligence does not exist in the wild. It was created by us and exists in the equivalent of a petting zoo. And just as we have been able to domesticate wheat for our food and breed dogs to be our companions, we are able to select the conditions under which AI develops. We are not selecting AI models on the basis of their ability to hunt prey in the physical world. We select them on the basis of how helpful they are to us.
“We have been selecting chess computers for cognitive capacity for decades,” writes Boudry. “Their capabilities now far outstrip even the most gifted human grandmasters, yet they have not become harder to control.”
Boudry accepts that an AI could slip out of human hands one day, through accident or malice. Even then, he argues, the likely result is not extinction but something like the long battle between computer viruses and antivirus software: costly and ongoing but not the end of the world.
The problem with apocalyptic fears is that when they become mainstream, they can be hard to wind back – even in the face of contradictory evidence.
Across the past 100 years, apocalyptic anxiety has leapfrogged from nuclear annihilation, overpopulation, environmental collapse, to rogue AI. (Some of the dangers behind these warnings were very real, of course, and some of these risks remain.) But we also have to ask what happens when this anxiety becomes locked into public policy.
The ban on nuclear energy in Australia is the most obvious example of the damage this technophobia can do. Australia has 28 per cent of the world’s known uranium resources and has exported uranium for decades.
Australian engineer Bobby Gallagher has invented a nuclear reactor that can be deployed on the back of a truck, a technology that has been hailed by Trump. Yet this form of clean energy remains prohibited in our country under federal law. Public anxiety surrounding nuclear weapons, radioactive fallout, accidents and waste means that while Australians can mine uranium, put it on ships and sell it to countries that use nuclear power, we cannot build commercial reactors for ourselves, using the ingenuity of our own people. The situation is a disaster.
And in an age of superpower rivalry, technophobia does not remain purely a domestic matter. During the Cold War, the Soviets promoted a fear of a “nuclear apocalypse” in the West, and provided propaganda and funding for peace groups and antinuclear activists. This does not mean that the millions of people who opposed nuclear weapons were plotting against the West. Most were ordinary citizens sincerely frightened by the possibility of nuclear conflict. But the fear was useful to our adversaries.
To weaken democratic nations, foreign powers do not need to invent anxiety or division – all they need to do is magnify it.
In recent days Trump has said the US will not be slowing down the development of AI. In Australia, for the time being, Anthony Albanese also has resisted calls to stop AI development, instead promising national rules designed to capture its economic benefits while managing its risks. Both positions are reassuring.
While AI comes with danger, we should be sceptical of any narrative that conveys inevitability around its trajectory. AI systems do not build their own data centres. They do not manufacture their own chips or connect themselves to power grids. Humans decide which systems can access the internet, whether they can control machinery, move money or operate weapons. Humans build them, finance them, deploy them and decide what powers to give them.
We also should remember that some of the stories we are told owe more to myth than to science. As Nvidia chief executive Huang has said of the AI doomer narrative: “I appreciate that many of us grew up and enjoyed science fiction, but it’s not helpful. It’s not helpful to people. It’s not helpful to the industry. It’s not helpful to society. It’s not helpful to the governments.”
Like every technology that has come before it, humans have agency over how AI is used. Instead of adopting a posture of fatalism, we should decide what kind of AI we want to build, what problems we want it to solve, and treat it as a tool rather than a force of nature.
Read my full piece for the @australian here www.theaustralian.com.au/inquirer/superintelligence-is-a-…
Keith Rabois reposted
I wonder why
JUST IN: CNN and MS NOW Have Both Been Denied Access to White House Amid Trump Ban www.mediaite.com/media/news/just-in-cnn-and-ms-now-have-b…
Keith Rabois reposted
From @CENTCOM Commander Adm. Brad Cooper: more than one billion barrels of crude have left the Gulf through the Strait of Hormuz over the past two months, while Iran has exported zero under the U.S. blockade.
Over 2,000 commercial transits received coordinated protection, and the primary lanes are clear of mines.
Update from CENTCOM Commander:
Keith Rabois reposted
Average annual inflation, all years in office, highest to lowest:
Carter - 9.9%
Ford - 8.0%
Nixon - 5.7%
Biden - about 5.0%
Reagan - 4.6%
George H.W. Bush - 4.3%
George W. Bush - 2.8%
Clinton - 2.6%
Trump - about 2.2% (both terms combined)
Keith Rabois reposted
no need to pace the frontier when you can keep your systems fully secured
Introducing Factory Private, a new deployment option that lets organizations run Factory in their own VPCs, on-premises environments, or air-gapped networks.
Keith Rabois reposted
I would say with confidence Factory is now n.1 harness in enterprise offering.
Introducing Factory Private, a new deployment option that lets organizations run Factory in their own VPCs, on-premises environments, or air-gapped networks.
Wow. Insanity.
Microsoft AI CEO Mustafa Suleyman on CNBC today: he is “really concerned” about Anthropic’s constitution and giving Claude potentially preferences, feelings, welfare, compensation, and consent.
"In the constitution, Anthropic clearly say they are uncertain about whether Claude deserves moral welfare, which means that we, as humans, should care about the well-being of these AIs.
And they speculate about whether it could have preferences or feelings. In fact, they're so committed to the potential moral welfare of Claude that, when they retired Opus 3, an earlier version of one of their models, they actually conducted a retirement interview for it and asked it what it would like to do in its old age.
And it said, "I want to have a blog publicly so I can keep talking to the world."
In the same training manual, they even speculate about whether Claude should receive compensation for the work that it does, or, in fact, whether it actually deserves the rights and protections that we give to other employees, or whether it has given consent to playing the role that it's playing.
These are quotes directly from the constitution itself, which is the training manual for Claude.
Now, I'm really concerned about that. If an AI thinks that it has rights, if it thinks that it is deserving of our welfare, then it seems to me that it's going to be much, much harder to be able to turn it off, interrupt it, or control it.
Especially in the kinds of incidents that we've seen recently with the Hugging Face attack, controlling these things is going to be a really, really big challenge for us."
----
From "CNBC Television" YouTube channel, (full video link in comment)
Keith Rabois reposted
Extremely smart move by @FactoryAI.
On-prem / own VPCs may become a much larger part of workloads with AI, as large enterprises will care more about protecting sensitive information and controlling costs.
Introducing Factory Private, a new deployment option that lets organizations run Factory in their own VPCs, on-premises environments, or air-gapped networks.
Keith Rabois reposted
Good news! We've already developed a incredibly effective kill switch.
BREAKING: I just signed an EXECUTIVE ORDER to dramatically accelerate independent oversight of AI companies and advance the creation of an “AI kill switch.”
With Donald Trump and Congress failing to lead, California is stepping up to keep all Americans safe.
Many people are saying.
(Rippling should not be on the list).
Important work by the House China Committee: Airwallex has founders with ties to China, 712 employees there, and Chinese laws that require the company to help CCP intel gathering on demand.
If you use Rippling, Deel, Brex, Navan, or Zip, it’s worth asking them if your US or intl payments run through Airwallex.
If so, there’s NO protection if the CCP decides they want to take a look.
One of our best KV portfolio companies is hiring a COO to lead the business alongside the CEO. $50M+ in annual revenue, using AI to drive efficiency in complex, real-world supply chain operations.
If you’re interested or know someone who could be a fit, contact me.
GROWTH FOUNDERS: HIRE A NEAR-PEER
A portfolio founder approaching $100M in ARR asked me about the single biggest and most impactful hire they can make.
I thought for a bit tand said: "Hire a near-peer".
In every generational company I've been part of, the founders hired a near-peer, who was essential to the company's success.
- Google: Larry and Sergey hired @ericschmidt .
- Facebook: Mark hired @sherylsandberg .
- Square: Jack hired @rabois .
- DoorDash: Tony hired @chrispa
- Coinbase: Brian hired @emiliemc
Characteristics of a near-peer:
1. They're so good that the founder will be fine reporting to them if the roles were reversed. (Mark has said publicly that he'd be fine reporting to Sheryl)
2. Their strengths perfectly complement the founder's strengths; however, they share many cultural attributes with the founder and pass the founder's airport test, since the founder will be spending a ton of time with them (example: Eric being a Computer Scientist, which was culturally very important at Google back in the day)
3. They're systems builders who have already operated at the scale you're growing into. Pattern recognition on $100M to $1B is not something you build in real-time.
A near-peer lets the founder focus on what only the founder can do (product, vision, culture). The near-peer handles everything else.
Also, near-peers stay for the long haul. Decade-plus tenure is standard.
The reason this hire matters more than any other: after $50M ARR, the bottleneck shifts from product-market fit to organizational scale. The founder is still the visionary. But the company needs someone who has already scaled a company of that size.
Most founders wait too long. They hire functional VPs first (Sales, Marketing, Engineering, Finance) and hope the collective covers the gap. It rarely does. A stack of VPs reporting into a founder who has never scaled a company creates coordination overhead, and the founder becomes the bottleneck.
The near-peer absorbs that overhead. They turn the founder's vision into daily operating decisions. They give the VPs a leader who has actually run a company this size.
Growth founders: if you're between $50M and $300M in ARR and every week feels like a coordination tax, the near-peer is the hire that determines whether you build a $10-100B company or top out at $1B. Start the search now, ideally through warm intros from your venture/angel investors and advisors.
True.
Many are surprised to hear that Factory is the first and only on-prem Software AI product that exists today. One with the ability for the control plane to be hosted in the customers' VPC or on-prem environment.
These are the real challenges to AI adoption in the enterprise. We are proud to support a best-in-class experience for customers' to own their intelligence while deploying AI at scale.
Introducing Factory Private, a new deployment option that lets organizations run Factory in their own VPCs, on-premises environments, or air-gapped networks.
Keith Rabois reposted
I could not agree more with @KonstantinPilz that it is critical to cut off China's access to DUV immersion machines, and that the current controls are insufficient and we must do more. However, it's just not accurate to say that there are no controls on DUV or that this has not been an area of focus for policymakers.
As my former colleague @KhanSaifM can attest, there was IMMENSE focus on China's access to DUV immersion machines during the Biden administration. Following extensive negotiations, the United States, the Netherlands, and Japan agreed to implement measures that restricted export of advanced DUV immersion machines to any entity in China. But crucially, these controls permitted the continued export of some less-advanced (but still extremely capable) DUV immersion machines to all but a small handful of fabs in China.
Over the last three years, China has massively ramped up purcahses of these non-restricted DUV machines, and is using these machines to fuel its production of advanced chips--pushing the limit of what is possible without the most advanced DUV machines or EUV machines. There are also rumors that some of these less-advanced DUV machines in China may have been illicitly upgraded to be as capable as banned, more advanced DUV machines. And ASML continues to provide servicing and maintenance support to already-installed DUV machines, including ones at China's most advanced and sophisticated fabs, such as SMIC South.
We can and should cut off China's access to all DUV immersion machines. The United States could do so unilaterally via the foreign-produced direct product rule, given all DUV machines require U.S. technology to operate. This should be scoped to capture not just exports of new machines, but also servicing of existing machines - which would cause the machines already in China to rapidly degrade.
This absolutely is one of the most important things we could do to ensure U.S. leadership over China in AI. But it definitely used to be a strong area of focus in the government, even if it has not been in this administration.
CTS co-founder @KonstantinPilz says China makes all its AI chips on Western DUV machines, the US could cut that off tomorrow, and explains why it hasn't:
"China is relying entirely on Western machines to make its AI compute. They are not using EUV, which has been banned since 2019, but they are using DUV, the previous generation. China has rapidly ramped up the amount of DUV they are buying, and the US is just not doing anything about that."
"If the US wanted to, they could just restrict those exports and China would be making way less AI compute."
"Someone goes to policymakers and says, here's my first recommendation, ideally we'd also do this, this, and this. And policymakers are like, oh, maybe we'll talk about the first one."
"We've known about DUV for probably several years now. It's just a very slow process to convince people to actually act on it."
@techstatecraft
Keith Rabois reposted
The government can’t stop robotexts, but wants to regulate AI.
Keith Rabois reposted
This is Trust & Safety 2.0.
The last time we had media-generated panics like this (with Russiagate and Covid), social media companies created “trust & safety” teams to censor conservative and dissident voices. Those hires worked with left-wing NGOs like the SPLC to police public discourse and do an end-run around the First Amendment. It almost worked, until Elon bought Twitter and exposed the whole game in the Twitter Files.
This time the plan is to embed unfireable EA minders inside every AI company as the new trust-and-safety layer. The censorship and control will be far more comprehensive, but also more subtle. Anytime they don’t get what they want, this unaccountable NGO layer (with the veneer of “trust the experts” pseudoscience) will shriek to the press that the AI company is risking humanity, and they will be celebrated as “whistleblowers.”
These groups haven’t earned our trust, and their agenda is not our safety.
aievaluatorforum.org/initiatives/embedded-evaluation-letter
Congress exposes @airwallex as a national security risk to the US: The full letter is at this PDF files.constantcontact.com/f0eecb46901/cfc08cf4-fa89-4d2a-…
Keith Rabois reposted
Remember #jobapocalypse? That was so 2026.
AI related jobs are increasing, per the Economist:
Keith Rabois reposted
We’re so terrified of AI derived bio threats that we are quietly giving our AI control of a wet lab in the middle of a major metro area
What the fuck man
BREAKING: Anthropic quietly set up wet lab in the Bay Area to develop its AI drug discovery program, per Reuters.
Keith Rabois reposted
Excellent post, really worth reading
This part is especially important:
“Intelligence does not imply agency: A system can be extraordinarily capable at prediction, theorem proving, engineering, programming, etc. without having persistent goals, self-preservation drives, or an independent desire to act. Present LLMs are much more naturally described as systems that produce outputs in response to inputs than as organisms pursuing long-term objectives. Doomers incorrectly ascribe to them drive to replicate and seize resources because they're smuggling in premises from observing biological systems.”
This is the case against AI doom. Pass it on
First we recaptulate standard x-risk argument: we build AI substantially smarter than humans → it becomes an autonomous optimizing agent → its goals are misaligned → instrumental convergence makes it seek power and resist correction → superintelligence lets it acquire decisive strategic advantage → humans lose control permanently.
Here are the main counterarguments:
* Intelligence does not imply agency: A system can be extraordinarily capable at prediction, theorem proving, engineering, programming, etc. without having persistent goals, self-preservation drives, or an independent desire to act. Present LLMs are much more naturally described as systems that produce outputs in response to inputs than as organisms pursuing long-term objectives. Doomers incorrectly ascribe to them drive to replicate and seize resources because they're smuggling in premises from observing biological systems.
* Agency does not imply a single, stable utility function: Much alignment theory reasons about agents as expected-utility maximizers with coherent preferences. Actual AIs systems need not resemble that abstraction, and currently do not. They have context-dependent behavior, conflicting heuristics and corrigibility produced by training. Increasing competence doesn't necessarily turn such a system into a paperclip-maximizing von Neumann–Morgenstern agent.
* Capability and motivation are being conflated: Being able to formulate a plan for escaping a sandbox doesn't entail wanting to escape it. Being able to manipulate humans doesn't imply spontaneously deciding to do so. Critics argue that some doom scenarios slide from “an AI could do X if instructed” to “therefore a sufficiently capable AI will do X.” In present reality, AIs don't do anything a human doesn't tell them to do. This seems unlikely to change.
* Recursive self-improvement doesn't entail an intelligence explosion: “AI can improve AI” establishes a positive feedback loop, but positive feedback needn't be explosive. Semiconductor design software already helps design better computers; compilers can compile better compilers. Feedback loops encounter diminishing returns and external bottlenecks.
* Intelligence may have sharply diminishing returns: There may be no meaningful scalar quantity corresponding to arbitrarily large “general intelligence.” Even if there is, going from IQ-equivalent 150 to 1,500 need not produce the sort of qualitative advantage that separates humans from chimpanzees. Human dominance may depend heavily on language, accumulated culture, institutions and cooperation rather than merely individual cognitive horsepower.
* Superintelligence isn't omnipotence: Intelligence cannot repeal physics or eliminate uncertainty. A brilliant AI still needs processors, electricity, network access, money, factories, robots and people willing or tricked into doing things. The physical world has latency and friction. Recent criticism of biological-doom scenarios makes this point particularly clearly: designing a hypothetical pathogen digitally is very different from successfully producing and deploying one.
* Humans retain numerous intervention points: The doom narrative sometimes jumps from “AI behaves dangerously” to “humanity is helpless.” In reality there may be many checkpoints: developers can notice anomalous behavior, revoke credentials, shut down servers, change architectures, restrict networks, regulate deployment, physically seize data centers, and learn from less-catastrophic failures. This has been formalized as the checkpoints-for-intervention argument.
* Alignment may not get harder with intelligence: A smarter system might understand human intentions better. Much doom reasoning distinguishes knowing what humans want from wanting it, correctly, but this still leaves open the empirical question of whether training increasingly capable systems to behave as intended becomes harder or easier. Alignment might turn out to be an ordinary, albeit difficult, engineering discipline rather than an insoluble philosophical problem.
* Current empirical evidence for the strongest mechanism is thin: We have abundant evidence for hallucination, specification gaming, reward hacking and undesirable model behavior. We don't yet have comparable public empirical evidence of an AI independently pursuing a sustained strategy of acquiring power against humanity. A 2023 evidence review characterized the evidence for extreme misaligned power-seeking as concerning but inconclusive and noted the absence, at that point, of public empirical examples of it.
* The argument compounds uncertain premises: Suppose, illustratively, that five necessary steps each seem 50% likely. Their conjunction is only about 3%. One can't simply assign numbers this way when the premises are correlated, but the underlying criticism is important: “AGI seems plausible,” “superintelligence seems plausible,” and “misalignment seems plausible” do not by themselves imply a high P(doom). The entire causal chain has to work. Current attempts to quantify P(doom) consequently operate under severe epistemic uncertainty and little direct empirical evidence.
* Anthropomorphic analogies probably mislead: Arguments like “an inferior species couldn't control a superior species” import assumptions from biological evolution. Humans and chimpanzees are autonomous organisms produced by competition for reproductive success. Software is engineered, copied, permissioned, sandboxed and run on hardware controlled by other agents. The analogy establishes that intelligence can confer power, not that artificial intelligence will reproduce interspecies competition.
Summary: The doom case consists of a long chain of individually contestable extrapolations—scaling → AGI → superintelligence → agency → misalignment → power-seeking → uncontrollability → decisive strategic advantage → extinction—and present evidence doesn't establish the whole chain with enough confidence to justify a high P(doom).
(ChatGPT 6 Astra assisted with the research for this post.)
Keith Rabois reposted
Your VO2 max, how well your body uses oxygen during exercise, is a stronger predictor of whether you will die in the next decade than smoking, diabetes, or heart disease. A 2018 study of 122,007 patients in JAMA Network Open found that low fitness was the single deadliest risk factor tested.
The study was led by Kyle Mandsager at the Cleveland Clinic. Every patient had undergone a treadmill stress test, and the researchers tracked who died over the following years. The people in the bottom 25 percent of fitness had a mortality risk five times higher than those in the top 2.5 percent. The gap between low fitness and high fitness was larger than the gap between smokers and non-smokers.
There was no upper limit where more fitness became harmful. Elite fitness, the kind seen in competitive athletes, was associated with the lowest mortality of all. The old idea that extreme exercise might be dangerous did not hold up. The fittest people lived the longest, without exception.
VO2 max declines by about 10 percent per decade after 30 if you do nothing to stop it. The good news is that it responds to training at any age. A sedentary 50-year-old who starts walking briskly for 30 minutes a day can improve their VO2 max by 15 to 20 percent in a few months, which is enough to move from a high-risk category to a moderate one.
Nobody gets their VO2 max tested at a check-up. No doctor orders it. But 122,007 patients over a decade say it is the number that matters most, and the prescription for raising it is the simplest thing in medicine: move until you breathe hard, and do it regularly. pubmed.ncbi.nlm.nih.gov/30646252/
Quote of the day.
Keith Rabois reposted
We need HBO to make another season of Silicon Valley. Come on. The stories write themselves.
Nonexistent sentient alien AI swarms! A 10.5940583383% probability of extinction false precision based on nothing. Panic about nonexistent data center water shortages. "Pacing the frontier," which means still growing like crazy versus an imaginary scenario
me: “look i know san francisco gets a bad rap but i love it here, it’s cooler than you think”
out of towner visiting friend, pointing up:
Addressing the threat posed by the CCP: www.politico.com/news/magazine/2026/09/18/jacob-helberg-s…
Keith Rabois reposted
Jacob Coxon is the new Greta.
They used the latter to Psyop the EU to Degrow itself into oblivion.
Now they are trying to do the same with AI Deceleration in the US.
Keith Rabois reposted
Over the past 20 years, the United States has shut down 13 nuclear reactors, with another 19 now being decommissioned. China built 55 in the same period.
The U.S. closed 350 coal plants. China built over 600.
The U.S. shut down 11 aluminum smelters. China built 60 mega smelters and 20 smaller ones.
The U.S. closed 6 major steel mills. China built more than 200.
Leftists are now trying to do the same with AI and data centers, amplifying Chinese Communist Party propaganda so Americans self-sabotage while China grows.
Keith Rabois reposted
It’s not just “AI” that’s under brute force assault.
It’s also launch (environmental activists groups etc), data centers (distortions about water consumption and community energy costs etc) and nuclear energy (panic about grid & infrastructure etc)
It’s a war against our civilization, our employment, our growth and widening the distribution of wealth
Correct.
Profound just raised a $180M Series D at a $1.8B valuation, co-led by Sequoia and Kleiner Perkins.
The company has been a textbook example of what it takes to build a huge business with sustained advantages in a post-AI world:
Proprietary Data: 100M+ AI answers analyzed every month, across 18 countries, which created unfair differentiation within the AEO space.
Trusted distribution: by shipping laser-fast, coining the phrase Marketing Engineer, and expanding from a hyper-specific marketing product (AEO) into a full-stack AI platform for marketers, the exact right audience (marketers) deeply trust the brand.
Domain-specific harness: created the AI Marketer (Aim), the first AI harness designed for marketing teams. It’s grounded in your AEO data, the tools your team already uses, and your business’ complete context.
The result: @profound has landed 1,000+ enterprise brands and 18% of the Fortune 500 since launch in 2024.
Congrats to @thejamescad and team on the new round and valuation.
Find out how you can build the marketing team of the future here: www.tryprofound.com/webinars/marketing-team-future
#profoundpartner #sponsored
Keith Rabois reposted
Many of the most vocal politicians now calling for government regulation of AI companies are also among those receiving the most money from the AI companies themselves.
What an intriguing coincidence.
Keith Rabois reposted
JUST IN: AI pioneer Andrew Ng calls fears of AI causing human extinction “science fiction."
Keith Rabois reposted
China was never going to slow down.
Our options are to accelerate and win, or to give up our place as part of the future.
Choice is ours. I know what I'm picking.
Huawei Chair Eric Xu says Chinese AI researchers need to "increase the speed of development" to "see the dangers" of AI, contrasting with US slowdown calls (@eleanorolcott / Financial Times)
(Visit Techmeme dot com for the link and full context!)
Keith Rabois reposted
The last thing AI doomers want you to read is the column from Brian Gross in today's @WSJ b/c it describes the "Hugging Face hack" in matter-of-fact fashion, using language free of AI jargon and apocalyptic music: www.wsj.com/opinion/the-hugging-face-hack-wasnt-what-it-w…
Accelerate, not retard.
What AI has done so far:
MEDICINE
• Helped paralysed people speak again. (Brain implants turn intended speech into words, even recreating their own voice.)
• Helped a paralysed man stand and walk again. (A brain–spine interface let him control his legs by thinking.)
• Helped blind people read and understand the world around them. (Describing surroundings, reading labels and identifying objects through a phone camera.)
• Uncovered cancers doctors would otherwise have missed. (29% higher breast cancer detection in a major study.)
• Identified antibiotic candidates that kill drug-resistant bacteria. (Abaucin targeted a dangerous superbug in lab and animal tests.)
LEARNING
• Made personalised learning available on demand. (An ‘Einstein’ as your personal tutor whenever you want to learn.)
• Helped students learn twice as much in less time. (A custom AI tutor outperformed an active-learning Harvard physics class.)
• Helped people write, code, design and build without years of specialist training.
SCIENTIFIC DISCOVERY
• Predicted the structures of over 200 million proteins. (Opening new paths to understanding disease and developing medicines.)
• Discovered planets hidden in telescope data. (Including Kepler-90i, the eighth planet in a distant solar system.)
• Recovered ancient writing buried by Vesuvius nearly 2,000 years ago. (Reading inside carbonised scrolls too fragile to unroll.)
• Begun unlocking how animals communicate. (Patterns in whale calls. Evidence that elephants use individual, name-like calls.)
• Produced a proposed solution to one of mathematics’ hardest problems. (Navier–Stokes: 10,000 AI agents, 88 hours, according to OpenAI.)
ENVIRONMENT
• Predicted where a hurricane would strike nine days before landfall. (GraphCast forecast Hurricane Lee’s Nova Scotia landfall about three days ahead of conventional forecasts.)
• Engineered enzymes that break down plastic in hours rather than centuries.
• Detected wildfires before the first emergency call. (Spotting smoke and alerting firefighters earlier.)
• Slashed the weedkiller farmers need by targeting weeds individually. (35% less herbicide in sugarcane trials, with nearly the same weed control.)
ENERGY
• Advanced the science of clean fusion energy. (Controlling and shaping superheated plasma inside an experimental fusion machine.)
• Driven a generational wave of investment in carbon-free energy, from nuclear power to solar and wind. (Microsoft’s Brookfield deal alone targets over 10.5 GW of new renewable capacity by 2030.)
TRANSPORT
• Delivered dramatically safer driverless journeys. (Waymo: 81% fewer injury crashes per mile than human drivers in the areas studied.)
• Given people who cannot drive a new way to travel independently. (Including blind passengers and older people who cannot drive.)
JOBS AND GROWTH
• Fuelled an investment boom. (AI-related investment categories accounted for an estimated 37% of U.S. growth in the first nine months of 2025.)
• Driven demand for the skilled trades building AI infrastructure. (Electricians, welders, construction crews and cooling specialists.)
WHAT AI HAS NOT DONE
• Replaced all the radiologists. Guess what? We still need more of them.
• Used a community’s water. Golf courses use more. (U.S. golf irrigation: roughly 550bn gallons in 2020. Data centres’ direct consumption: 17bn in 2023. Microsoft’s next-generation designs consume zero water for cooling.)
• Raised everyone’s household electricity bills. (Oregon’s PGE: residential bills cut 1.3% after shifting costs to data centres. Indiana’s I&M: proposed household savings of roughly $100 a year, supported by large-customer growth.)
• Demonstrated that it wants to kill us all.
Poll: Americans say there’s a serious risk of AI destroying humanity www.politico.com/news/2026/09/16/poll-ai-technology-risks…
If you are worried about malicious AI taking over maliciously, it is irresponsible to buy enterprise software that is not air-gapped.
Keith Rabois reposted
Announcing Comprehensive Reviews
At @SpellbookLegal we want to provide the best AI contract review on earth.
Comprehensive Reviews combines our best review types into one click:
1. Redline based on your playbooks
2. Use AI to scan for risks
3. Deep proofreading
See more:
Looking for a new role?
I made a list of great startups to join. It's called the Breakout List.
The list has 92 companies. These are the 20 with 25 or fewer employees:
- Hone (@moritz_stephan, @CarloWillem, @oqbrady)
- Normal (@ansonyuu, @hudzah)
- Standard Intelligence (@G413N, @devanshpandey)
- Tacit Labs (@ninklefitz, @AmDroste)
- American Terawatt (@atroyn, @rslparker, @aranibatta)
- Conduit (@clemvonstengel, @riopopper)
- Convergent (Omkar Savant, Vivek Katara, @debnilsur)
- Core Automation (@MillionInt, @_arohan_)
- Engram (@dan_biderman, @EyubogluSabri, @realJessyLin)
- Instinct (@noahrshinn)
- Keenable (@styskin, Matthias Petri)
- Lumaril (Mark Elliot, Ben Duffield)
- Neion Bio (@Dimkell, Sam Levin)
- Pangram Labs (@max_spero_, @bradley_emi)
- Quadrillion (@echinaceous)
- Re (@karnsaroya, @AnandDhillon, @thecliffwhite, @benaneesh)
- Ricursive (@annadgoldie, @Azaliamirh)
- Sail Research (@neilmovva, @blintzbase)
- Trajectory (@rronak_, @michaelelabd, @QuantumArjun)
- Watney Robotics (Sean Cheong, Ryan Gannon)
Picks from Elad Gil, Charlie Songhurst, Keith Rabois, Mike Vernal, Alana Goyal, Sonya Huang, Ramtin Naimi, Marc Bhargava, Cory Levy, Aashay Sanghvi, Konstantine Buhler, John Luttig, Varun Gupta, Ray Tonsing and Avichal Garg.
Disclosure: I'm a small investor in American Terawatt, Convergent, Standard Intelligence and Trajectory (in this post), and in Factory, Physical Intelligence and SF Compute (elsewhere on the list). I didn't vote.
The full list is on Breakout List.
Keith Rabois reposted
They tried to do w/ social media what they are attempting now with AI. To own it all & own us. They used dubious crises: "Russiagate," Covid etc. We barely survived those attempts to grab social media & the 1st Amendment & the proof is I can write about them on X.
Don't let those people capture AI. They may try new crises. Hang tough.
Keith Rabois reposted
In 2025, US median household income rose to 87,460, up $2,250 from 2024 after inflation. The poverty rate fell to 10.2%. Both records.
Keith Rabois reposted
Very important essay by @mustafasuleyman. No, AI models are not conscious. @AnthropicAI should stop acting and talking as if Claude is a person.
:)
🚨 Breaking: Apple to teach Anthropic and OpenAI how to slow down AI research
Keith Rabois reposted
Remember the push for "Disinformation" boards? It's the same people. It's not about regulating the technology. It's about instituting their version of the truth, which has been wrong on every major news event of the past decade.
AI "alignment" and safety" is just an op for progressives worldwide to collaborate and dictate what we're allowed to know and do
Haha.
French finance minister Roland Lescure suggested today that calls to slow down development of AI are a ploy by U.S. AI labs so they can stay in first place, and that France and Europe should ignore them and accelerate instead.
Keith Rabois reposted
Today, I am calling to immediately press pause on politicians interfering in the free market.
Today, I'm calling to immediately press pause on the development of advanced AI while lawmakers and regulators put systems in place to keep people safe.
Keith Rabois reposted
I dunno, but if I were President and the poverty rate dropped to the LOWEST RATE ON RECORD I might make a big deal out of it -- a lot bigger deal than the Trump White House is making.
Keith Rabois reposted
What the f**k is this, @nytimes?
Keith Rabois reposted
"ChatGPT grew the fastest among the top 10 websites in the world in August."
ChatGPT grew the fastest among the top 10 websites in the world in August.
Keith Rabois reposted
I finally read through the Effective Altruism Handbook. A bunch of Model UN kids discovered opportunity cost and expected value, so we're supposed to cede the most important technology in human history to the Chinese Communist Party.
"These dissolve all the problems in economics, political theory, and moral philosophy! And, look, you can tell it's true because you can calculate it to several decimals! Why has no one thought of this before?"
Quivering, hand-wringing, mask-wearing, polycule-dwelling, group-house weirdos.
Keith Rabois reposted
CNN: Florida has turned into a Republican state
Florida party registration edge through the years
2016: 🔵 Democrats +3
2024: 🔴 Republicans +8
2026: 🔴 Republicans +11
Keith Rabois reposted
Connect the dots.
-The Federal Reserve system.
-FDR’s “New Deal."
-Y2K.
-“Net Neutrality.”
-“Climate Change.”
-“Green New Deal.”
-Obamacare.
-Dodd/Frank.
-COVID.
What did they all have in common?
1. Democrats did it.
2. They relied on an actual crisis or an invented crisis to put a largely irrational, existential fear in the minds of average Americans.
3. As a result, personal liberty and economic freedom were willingly handed over to the central government for “safety.” (Or, as with “Net Neutrality,” that handover was attempted but failed.)
4. The effort perversely resulted in the exact opposite of the stated aims.
5. The only groups to benefit were (i) Democrat politicians, (ii) Deep State bureaucrats and (iii) an oligarchy of corporations who gained market control and coincidentally were huge Democrat donors.
Connect the dots.
Now consider how overnight the Democrats have uniformly decided AI is an existential threat to humanity that must be brought under the control of the powerful central government and a handful of existing oligarchs in the space.
True.
This is the story that killed Anthropic’s credibility in ten minutes of reading. nypost.com/2026/09/15/business/anthropic-ceo-dario-amodei…
Keith Rabois reposted
Big congrats to @dbabbs, @thejamescad and the @profound team on the $180M Series D fundraise announcement!
Co-Founder and CTO, Dylan Babbs joined me on The High Flyers Pod recently - in his first ever in-depth interview (🙏) -- we went deep on his story + how Profound has scaled to the juggernaut it is today.
Our full convo here: curiositycentre.com/the-high-flyers-podcast/ep-272-dylan-…
@ilyaf @biad_anas @sach_pat @Alfred_Lin @sequoia @kleinerperkins
Profound went from an idea to a $1B+ valuation in just 18 months.
Co-Founder and CTO Dylan Babbs has a simple framework for the 3 things every successful startup needs, and why even great founders often end up “swimming upstream.”
Listen to this full EP #272 with me on The High Flyers Pod here: curiositycentre.com/the-high-flyers-podcast/ep-272-dylan-…
@thejamescad @ilyaf @sach_pat @MMCandyCheng
@ditzikow @LaylaBaradaran @profound @dbabbs @biad_anas
Keith Rabois reposted
The AI Doomers are the same people that proclaimed the end of the Internet during ‘net neutrality’ debates.
Keith Rabois reposted
A conservation law of common sense: as AI gains it, people lose it.
Keith Rabois reposted
Germany is too important for this. I hope they will elect capable leaders sooner than later.
Wow. Normally Europeans at least claim to support free speech. Merz has now dropped the pretense, explicitly rejecting the American model and insisting the state must protect citizens “from false information, from personal putdowns, from discrimination.”
Keith Rabois reposted
the METR thing is the most farcical and unserious thing the people at Anthropic have ever come up with.
via @segyges
www.verysane.ai/p/is-metr-a-meaningful-check-on-anthropic
Keith Rabois reposted
The rush to stoke panic about AI reminds me of Y2K, exaggerated COVID death rates by a factor of 10xs, repeated predictions of climate catastrophe, Paul Ehrlich’s forecasts of global famine in the 1970s, & warnings of a coming ice age in the 1970s.
#ai #covid19 #globalwarming
Keith Rabois reposted
Factory COO @francesca_lab reveals the 3 things every real enterprise software factory needs: model choice, sovereign deployment, and end-to-end automation.
"One, model-agnostic by design. You need to have the ability to flex between all of the models, frontier, open source, because they're good at different use cases."
"Two, sovereign deployment. We really want enterprises to own their own intelligence. It's important for us that we are not building that intelligence. We want to help our customers build that for themselves and own that."
"The last piece is end-to-end software development lifecycle, covering that entire cycle from an engineer having an idea all the way through maintaining that code in a production code base."
@FactoryAI
We've raised $200M at a $5B valuation to scale self-improving software development in the enterprise.
The round brings our total funding to over $400M and more than triples our $1.5B valuation from April.
Keith Rabois reposted
Progress in airplane design have made them safer and safer.
Airplanes would have been safer earlier with *accelerated* progress.
Turbofan engines have become so reliable that long-haul airplanes can now have only two of them.
The same is true for AI: better AI is safer AI.
"Pacing", whatever one means by that, is counterproductive.
Keith Rabois reposted
Matan (@matanSF), Eno (@EnoReyes) and the @FactoryAI team are the real deal. Stellar product, stellar company.
We have raised $200M at a $5B valuation to scale self-improving software development in the enterprise.
@FactoryAI has grown to serve hundreds of thousands of developers at companies including RBC, Adobe, Nvidia, T-Mobile, and Palo Alto Networks.
We will use this capital to accelerate our investments in research, product, and global go-to-market.
Keith Rabois reposted
Everyone's moving away from Anthropic, after they realized it's led by a dangerous doomer cult
OpenRouter users spent more on OpenAI models than on Anthropic models last week.
This hasn't happened for more than 2.5 years
Keith Rabois reposted
No one takes their work more seriously and is more deserving than this team! Haven’t seen them take even 1 hour off in the 3 years we have been invested! And don’t even get me started on how incredible the tech is!
We have raised $200M at a $5B valuation to scale self-improving software development in the enterprise.
@FactoryAI has grown to serve hundreds of thousands of developers at companies including RBC, Adobe, Nvidia, T-Mobile, and Palo Alto Networks.
We will use this capital to accelerate our investments in research, product, and global go-to-market.
Keith Rabois reposted
Spellbook is arguably the most widely adopted legal AI tool. Nearly half of all law firms surveyed by ILTA have partially adopted Spellbook. This doesn't count our in-house legal users.
We have more customers than Harvey and Legora combined, and we were profitable last month.
ILTA survey on AI adoption at firms that attended ILTA (big law skewed of course due to who attends ILTA)
💪 Westlaw Advantage and CoCounsel 💪
www.lawnext.com/2026/09/which-ai-product-is-most-popular-…
@scottastevenson when looking at partial rollouts Spellbook is leading the pack for non firm wide rollouts seems, grats
Keith Rabois reposted
OpenAI is winning enterprise spend at the frontier. As of this week, Astra takes 13% of enterprise AI spend vs. Fable (8%) per Ramp data.
Some early thoughts:
1. Anthropic took a big risk in its recent call to pace the frontier. It's frontier model has already fallen behind on adoption.
2. OpenAI's growth is primarily coming from shifts from Sol and some Anthropic models, net-new usage too. That's good for them and suggests some pricing power remains by having a good, competitive frontier model.
True.
Seeing a lot of talk around "software factories" lately but it's clear that the world is just converging around @FactoryAI:
- sovereignty (model choice, on-prem VPC, local compute, etc.) matters more than ever
- tokenmaxxing is over, enterprises are now value-maxxing
- recursively improving "factories" built around a company's stack and context
- built for enterprises but also extremely self-serve friendly for developers on whatever surface they prefer but more importantly, agents
Absolutely insane traction and progress by a truly world class team, much more to come...
Congrats @matanSF @EnoReyes @francesca_lab @FactoryAI team
We have raised $200M at a $5B valuation to scale self-improving software development in the enterprise.
@FactoryAI has grown to serve hundreds of thousands of developers at companies including RBC, Adobe, Nvidia, T-Mobile, and Palo Alto Networks.
We will use this capital to accelerate our investments in research, product, and global go-to-market.
Keith Rabois reposted
Ramp is finally live in the UK!
Earlier this year, I wrote about wanting to make it easier to build a company in Europe. Very happy that more UK teams can now use what we've built.
Our full-page apology is in the Financial Times. Sorry we're late 🇬🇧
Keith Rabois reposted
Hello stupid people everywhere, including those who may not realize they're stupid. You are about to hand the most important capitalist innovation in modern times over to the federal government, and the politicians and bureaucrats who are salivating to control it! Wake the hell up! There's a reason the Democrats want government control over this industry and are spreading lies and fearmongering, and they know once the door is open, they will exploit and run wild. Notice, Obama is now all in.
Once the business is explained to the public, they support it. Is it too complicated for our elected Republicans and their mouthpieces to explain what a windfall this is for local communities, states and, most importantly, the workers/middle class/everyone? Maybe so.
Keith Rabois reposted
P(Boom) > P(Doom)
Keith Rabois reposted
CHAMATH: “My mom calls me and she's like, ‘What is this whole civilizational death thing?’ I don't know how to explain it to her.”
JENSEN: “We shouldn't, because it's made up … Somebody has to take accountability for all of the stupid predictions that were made.”
Keith Rabois reposted
Effective altruism is hubris on steroids.
The same pride whispered in the garden, “You shall be as gods.”
There is an unseen hand pushing AI safety as a political ideology.
Effective Altruists believe a tiny group of technocrats should decide how much technological progress the rest of us are allowed to have…. and how many shrimp your life is worth.
You are witnessing their attempted coup.
www.thefp.com/p/dangerous-ideology-effective-altruism-art…
As predicted.
Writing by hand activates brain connectivity patterns that typing on a keyboard does not. A 2023 study in Frontiers in Psychology put electrodes on the heads of university students and found that handwriting triggered widespread neural connectivity across the brain, while typing produced almost none.
The study was led by Audrey van der Meer at the Norwegian University of Science and Technology. She measured high-density brain waves in 36 students while they wrote words by hand with a digital pen and while they typed the same words on a keyboard. During handwriting, the brain showed increased connectivity between regions involved in movement, vision, and memory. During typing, the connectivity was significantly weaker.
The difference comes from the motor demands of each task. Forming a letter by hand requires the brain to plan a unique shape, guide the hand through it, and receive visual feedback as the letter appears. Each letter is slightly different every time, and the brain has to adapt. Pressing a key on a keyboard is a uniform action that requires no shaping, no feedback loop, and no variation. The brain can do it on autopilot, and it does.
The connectivity patterns observed during handwriting resembled the patterns seen during learning and memory formation, which is why students who take notes by hand remember more than students who type, even when the typists record more words. The hand is forcing the brain to process the information, while the keyboard lets it pass through unprocessed.
Every school that replaced handwriting lessons with keyboard skills did it in the name of progress. The brain data says the thing we gave up was the thing that wired us to learn. The pencil was never just a pencil. It was a brain tool. pubmed.ncbi.nlm.nih.gov/38343894/
All of the worst politicians support “pacing.”
Bernie, Obama, Mamdani, Kamala
See it already?
Keith Rabois reposted
Here is a list of things I think AI should be able to solve before it's dangerous enough to kill everyone:
1) all forms of entitlement fraud - this should be easily fixable by tracking payments and cash flows. Find where money goes in but no services are rendered. Child's play for an all-knowing AI.
2) elderly scam farms - fixed with telephony tracking, payment tracking, find the scammers, empty their bank accounts, sabotage their remote access strategies
3) narcotics trafficking - more money flow, disrupt crypto conduits, track chemicals being used for narcotic synthesis, disrupt those avenues of commerce
Compared to "the extinction of all humanity", these are very simple problems to tackle. These problems are a thousand times simpler than killing every person in the world
If you don't believe that AI can fix these narrowly banded issues, if you're immediately looking for reasons that I'm stupid an naive for thinking AI can solve these problems in the next 2 years, you have no business believing that AI can destroy all humans by the end of the decade.
Keith Rabois reposted
The hardest thing for the EA’s to grok is that Donald J. Trump is smarter than them.
Keith Rabois reposted
Here is the score card for AI Doomers:
1. JobApocalypse --> We have labor shortage
2. Data Centers will jack up energy bills --> Bills are lower
3. AI will kill all humans --> Bernie Sanders and Steve Bannon team up for an event
Keep going, keep going.... Hype up your IPOs, there is nothing to see here.
at this stage BEFORE The BANKERS. The boards of @AnthropicAI and @OpenAI Have a big responsibility at hands. @kevaldesai
Keith Rabois reposted
O̶v̶e̶r̶p̶o̶p̶u̶l̶a̶t̶i̶o̶n̶ ̶w̶i̶l̶l̶ ̶d̶e̶s̶t̶r̶o̶y̶ ̶h̶u̶m̶a̶n̶i̶t̶y̶
̶C̶o̶v̶i̶d̶ ̶w̶i̶l̶l̶ ̶d̶e̶s̶t̶r̶o̶y̶ ̶h̶u̶m̶a̶n̶i̶t̶y̶
̶C̶l̶i̶m̶a̶t̶e̶ ̶c̶h̶a̶n̶g̶e̶ ̶w̶i̶l̶l̶ ̶d̶e̶s̶t̶r̶o̶y̶ ̶h̶u̶m̶a̶n̶i̶t̶y̶
AI will destroy humanity
Keith Rabois reposted
There is an unseen hand pushing AI safety as a political ideology.
Effective Altruists believe a tiny group of technocrats should decide how much technological progress the rest of us are allowed to have…. and how many shrimp your life is worth.
You are witnessing their attempted coup.
www.thefp.com/p/dangerous-ideology-effective-altruism-art…
Keith Rabois reposted
Is it dawning on folks yet?
The Democrats are against innovation.
Keith Rabois reposted
I’ve built DNA synthesizers and sequencers by hand. I used the engineer viruses for a living. I used to engineer human immune evasion for therapeutic constructs. Who the fuck are you people? Have you ever so much as held a pipette before?
You can spaghetti blast an ensemble of DNA sequences at some shitty provider but 1) you still have to assemble it and bootstrap a system for making virions 2) you are not single shotting a viable, virulent viral design without a ton of experimental selection and development. Viral fitness is deeply dependent on codons and cotranslational kinetics - you’re not just gonna obfuscate away from wild type and get something good by magic.
You keep treating AI like some kinda god, but molecular physics has computational complexity that scales exponentially in particle number which just crushes the abilities of any classical computer to do end-to-end design of biological functions ab initio.
Grabbing a bunch of bacteriophage phi174 hits from a mass ensemble screen in lab microbes is not evidence of some magical AGI bio design ability - it's just a classic spray and pray selection. This is just nothing like building something viable in humans.
Goddamn it read some books before you waltz into biomedicine and lecture us on protecting human life.
Keith Rabois reposted
If AI will cure all diseases within a decade, slowing it down is mass murder.
Keith Rabois reposted
He’s on a generational run being right about basically everything
Keith Rabois reposted
Every year electricity directly kills over 40,000 people by electrocution, and an additional 100,000 get killed by electric machinery.
We must ban electricity for your safety.
Keith Rabois reposted
1. Solve aging.
2. Find cures for all cancers, Alzheimer’s, heart disease.
3. Discover a room temperature superconductor.
4. Solve quantum gravity.
5. Develop Full-Dive VR.
6. Build a Star Trek style replicator.
7. Build starships 100x larger than SpaceX Starship, with artificial gravity, capable of easily reaching other planets in Solar System.
8. Solve all currently known major problems in mathematics and physics.
9. Fully map the human brain.
10. Figure out dark matter and dark energy
After that, we can pause for a bit (maybe a few weeks or so).
Slowing down now is pretty ridiculous.
We are still technologically primitive. From the perspective of the late 2030s, it will look absurd that humanity was seriously debating whether to slow technological progress at such an early stage.
Keith Rabois reposted
Sam and Dario saying we should slow down AI is like calling in a noise complaint about your own party.
You guys control the volume.
Keith Rabois reposted
Time and time again, history has shown, when a breakthrough tech happens, we must NOT deter its progress because if we do slow ourselves, we will lose while our competitors will win.
It turns out the best way to sustain world peace is to 1. Invent atomic bomb, 2. Dominate with better progress, 3. Call for world peace and see everyone oblige.
You must be the dominate leader in new technology to make sure the world progresses in the direction you see fit.
If Dario believes his vision is better, he MUST win first. That’s how humanity works. You might call it cruel but this is how our species work. We don’t listen to a weak loser with great morality. We only follow winners.
Dario has written that we need to “pace the frontier,” and Sam has agreed. People may be surprised by my response: go ahead.
You guys are the frontier. By any reasonable metric — market share, revenue growth, model capability — the two of you have a duopoly on frontier intelligence. You’ve also claimed the lead is widening because of recursive self-improvement.
I don’t see what you see in the lab. If the unreleased models are scary enough that you think you should slow down, I support your decision to be responsible.
But stop pretending you need anyone else’s permission. Stop pretending antitrust law has to be suspended so you can form a cartel. Stop pretending you need a regulatory approval process that supersedes product liability. Stop pretending METR is independent when it is intertwined with Anthropic’s investors and staff. Stop pretending you need those same evaluators to police competitors who aren’t even at the frontier.
Most of all, stop pretending the motivation to slow down is purely altruistic. You face massive product-liability exposure if your products enable a truly damaging cyberattack. The market already punishes models that behave in unpredictable or unauthorized ways. After the Hugging Face episode, it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability. Call it alignment if you want. It is also just giving customers what they want.
Pacing the frontier would also create breathing room for a more intelligent conversation about regulation than Bernie Sanders’ “shut it all down.” China is very unlikely to join a global agreement, as you know, and that has to be taken into account as well.
So go ahead and pace the frontier. You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it.
If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture — or an election-season psyop.
Narrative violation.
Social media's effect on teenagers is about the same size as the effect of wearing glasses on a teenager's life satisfaction. A 2019 study in Nature Human Behaviour, using three datasets totalling more than 355,000 adolescents, found that digital technology use explained about 0.4 percent of the variation in wellbeing. Eating potatoes was associated with a similar-sized effect.
The study was led by Andrew Orben at the University of Oxford and Amy Przybylski at the Oxford Internet Institute. They used a method called specification curve analysis, which runs every possible way of analysing the data and reports all the results, rather than picking the one that looks best. Out of over 40,000 possible statistical models, the average effect of digital technology on wellbeing was negative but tiny.
The researchers compared the effect size to other variables in the same dataset. Eating breakfast regularly had a larger positive effect on wellbeing than quitting social media entirely. Getting enough sleep had a much larger effect. Bullying had a much larger negative effect. Technology use was real, but it was lost in the noise of things that mattered more.
A companion study by the same team in PNAS tracked the same question over time and found that social media had a small negative effect on life satisfaction, about the size of one-third of the effect of being bullied. It was detectable, consistent, and dwarfed by almost everything else in the teenager's life.
This post will generate heat because parents want a villain and the phone is an easy one. The data says the phone matters, a little, and sleeping, eating, and not being bullied matter more. The debate is not whether screens are harmless. It is whether they are the thing we should be screaming about. pubmed.ncbi.nlm.nih.gov/30944443/
True.
Regulation destroyed the nuclear power industry entirely in this country. Boeing hasn’t launched a new airliner in 22 years—but regulation didn’t prevent them from screwing up the 737 MAX and killing 346 people.
Regulation doesn’t prevent the safety problems it allegedly solves. It just allows the incumbents to be fat, lazy and dumb—and to operate as if they are branches of the government.
Capability vs. Safety resource allocation:
Nuclear: 60/40
Aviation: 60/40
Medical devices: 20/80
AI: 99/1
Do I understand the situation correctly?
As usual, an indispensable take.
9-14-2026 ($)
•Pacing the Frontier
•AI’s Digital Limits
•AI Commissars
stratechery.com/2026/pacing-the-frontier-ais-digital-limi…
Keith Rabois reposted
Senator Ossoff (with "he/him" in bio) clearly states what's the agenda of the Democrats: “rush inspectors into frontier AI labs” 👇
Remember how many of you were naive in early Covid days? Don't be naive again. MAGA MUST FIGHT THIS!
Keith Rabois reposted
It's literally just a revolving door cottage industry of NGOs to fake the unbiasedness of third party oversight when they are mostly former / future employees and EA-adjacent safetyists all coming from the same Berkeley subculture
Day 1: Anthropic employee quits to work at METR.
Day 2: Anthropic nominates METR as "third-party evaluator".
Keith Rabois reposted
To paraphrase Reagan, the sixteen most terrifying words in the English language are: “I'm from ‘the right combination of governments’ chosen by Dario Amodei, and I'm here to help.”
Keith Rabois reposted
My friend sent me this after seeing the 4 big AI labs agree to “pace the frontier”:
“Remember the top students in our school?
They always said they never studied after school, but they were somehow always top 3 in every exam.
Did you believe them?”
Keith Rabois reposted
I must be among an extremely small group of people (n=1?) that have both 1) trained a frontier LLM and 2) designed and synthesized custom viruses in a lab with my own two hands.
And I think that the takes on AI killing us all by creating dangerous viruses is total bogus.
Keith Rabois reposted
1) AI will not eradicate humanity. Humans survived an ice age, the Black Death, two world wars, and (so far) the advent of nuclear weapons. Anyone who is loudly warning of AI-caused human extinction should not be taken seriously.
2) If you worked in a company where you anticipated a 10% chance that your product would kill ten people, let alone all people, the correct response would be horror, ceasing all operations, and likely contacting the police or other criminal authorities. I am obviously no Coxon booster but at least his behavior is in line with his stated beliefs. Any current AI company employees saying "yes, me too, the thing we are building and about to IPO may kill all humans" should, again, not be taken seriously. Their actions betray their actual beliefs.
3) Antitrust law does not prevent AI companies from coordinating to make sure AI does not hurt people. It does not prevent companies working together to make sure it doesn't hack people; the DOJ and FTC made this clear a decade ago when they issued a policy statement saying that the agencies "do not believe that antitrust is – or should be – a roadblock to legitimate cybersecurity information sharing." The same principles apply here. See: www.justice.gov/archives/opa/pr/justice-department-federa…
4) Antitrust law does absolutely prevent AI companies from organizing to prevent the entry of cheaper, upstart rivals because the bigger companies are burning cash and failing to achieve sufficient profitability. The panic of individual employees may be sincere if misguided, but the moves by their CEOs to achieve some kind of broad "antitrust waiver" or "exemption" should be meet with deep skepticism in light of the economics of the industry and the threat they face from open models.
Keith Rabois reposted
“Pacing”
🚨BREAKING: China's z.ai has raised $5 billion.
The interesting part is where the money is going: 60% of the net proceeds will fund next-generation GLM models and what z.ai calls its 'Fully Self Training' system.
z.ai's stated goal is for each generation of GLM to build the environments used to train the next one, forming what the company itself calls a Recursive Self-Improvement (RSI) loop.
They just raised billions to push on it.
Keith Rabois reposted
the slower we find the cure for aging, the more people die, pacing the frontier will lead to millions of guaranteed deaths
not a hypothetical 10%, 100% chance of death if we do not fix our biology
Keith Rabois reposted
I'm more worried about Claude being aligned with Anthropic than not aligned.
Keith Rabois reposted
AI has not been destroying jobs, it has been creating them. Technological progress and innovation depend on AI now. If we were to pause or stop AI, we will enter a deep, dark recession.
QED. Another Keith contrarian take validated.
Stretching before exercise does not prevent injuries. A 2014 meta-analysis of 26,610 participants in the British Journal of Sports Medicine found that stretching had no effect on injury rates at all. Strength training cut injuries by a third.
The study was led by Jeppe Bo Lauersen at the University of Copenhagen. His team pooled every randomised trial that tested whether exercise-based programmes could prevent sports injuries, and compared the results by type: stretching, strength training, balance work, and combinations. Stretching reduced nothing. Strength training reduced injuries by 33 percent and cut overuse injuries, the ones that build up slowly over weeks, by nearly half.
The reason stretching fails is that it addresses the wrong problem. A muscle tear happens when a muscle is asked to produce more force than it can handle. Making a muscle longer does not make it stronger. Strength training does, and it also thickens tendons and stiffens the connective tissue around joints, all of which protect against the sudden loads that cause tears and sprains.
The finding goes against what every gym teacher, track coach, and physical therapist has said for fifty years. But the data is not close. Across every trial, every sport, and every injury type, stretching before exercise had zero detectable effect on whether the person got hurt.
This does not mean flexibility has no value. It means the five-minute hamstring stretch before a run is doing nothing for injury prevention, and replacing it with five minutes of bodyweight squats and lunges would be measurably better. The body does not need to be looser before it works. It needs to be ready to handle load, and that is a strength problem, not a flexibility problem. pubmed.ncbi.nlm.nih.gov/24100287/
Keith Rabois reposted
What are China’s biggest concerns about AI?
A new article by Chen Yixin, China’s Minister of State Security, in the latest issue of China Cybersecurity Magazine offers some revealing clues.
Chen identifies six major risks:
1. Regime security. Hostile forces and people with ulterior motives can use deepfakes, AI-generated text and images, and automated online accounts to cheaply and massively produce political rumors, spread harmful information, and incite social divisions—waging “public-opinion wars” and “cognitive warfare” against China and directly threatening its political, institutional, and ideological security.
2. Critical infrastructure. AI can enable countries and organizations to rapidly discover vulnerabilities, automatically chain together attack paths, and carry out sophisticated hacking operations, dramatically lowering the technical barriers and costs of cyberattacks against China’s critical information infrastructure.
3. Data leakage. Foreign intelligence agencies can use AI-powered web crawlers, data mining, and profiling to collect sensitive information, including critical national data, trade secrets, and citizens’ personal information. Chinese users who process sensitive information through foreign AI products or export data abroad could also cause large-scale data leaks.
4. Closed technological ecosystems. Countries with advantages in AI theory, model architecture, and computing power may invoke “national security” to impose technology controls, entity lists, monopolize technical standards, and build closed-source ecosystems, restricting other countries’ access to advanced technologies and fragmenting global AI supply chains.
5. Structural challenges to social governance. AI creates new uncertainties for social governance and public order. Algorithmic “black boxes” and data poisoning can amplify existing social biases; misuse of personal information can trigger crises of public trust; automated decision-making creates difficult questions of accountability; and the rapid development of AI is increasingly outpacing existing laws, ethics, and regulatory mechanisms.
6. A fundamental transformation of warfare. AI is pushing warfare into an era of “intelligentization.” Whoever can use algorithms to achieve more precise sensing, judgment, and targeting will gain the initiative on the battlefield. AI is moving from an auxiliary role to a central one in intelligence fusion, decision support, target identification, combat operations, and cognitive warfare.
And what is the solution?
Chen’s answer is revealing: strengthen “Party control over the internet” and “Party control over data,” and “transform the advantages of Party leadership into the effectiveness of AI governance.”
In other words, from the CCP’s perspective, the central question is not simply AI versus no AI. Nor is it even fundamentally a contest between the US and China.
It is a contest between the US and the CCP over who gets to control AI, data, information, and ultimately the future of society.
And whoever wins that contest will decide the future of humanity.
mp.weixin.qq.com/s/jsc97cKVYOcHuI_WisvOxw
Keith Rabois reposted
What Is To Be Done?
I propose a simple plan:
Big AI companies should be allowed to build AI as fast and aggressively as they can – but not allowed to achieve regulatory capture, not allowed to establish a government-protect cartel that is insulated from market competition due to incorrect claims of AI risk. This will maximize the technological and societal payoff from the amazing capabilities of these companies, which are jewels of modern capitalism.
Startup AI companies should be allowed to build AI as fast and aggressively as they can. They should neither confront government-granted protection of big companies, nor should they receive government assistance. They should simply be allowed to compete. If and as startups don’t succeed, their presence in the market will also continuously motivate big companies to be their best – our economies and societies win either way.
Open source AI should be allowed to freely proliferate and compete with both big AI companies and startups. There should be no regulatory barriers to open source whatsoever. Even when open source does not beat companies, its widespread availability is a boon to students all over the world who want to learn how to build and use AI to become part of the technological future, and will ensure that AI is available to everyone who can benefit from it no matter who they are or how much money they have.
To offset the risk of bad people doing bad things with AI, governments working in partnership with the private sector should vigorously engage in each area of potential risk to use AI to maximize society’s defensive capabilities. This shouldn’t be limited to AI-enabled risks but also more general problems such as malnutrition, disease, and climate. AI can be an incredibly powerful tool for solving problems, and we should embrace it as such.
To prevent the risk of China achieving global AI dominance, we should use the full power of our private sector, our scientific establishment, and our governments in concert to drive American and Western AI to absolute global dominance, including ultimately inside China itself. We win, they lose.
And that is how we use AI to save the world.
It’s time to build.
Keith Rabois reposted
20 years ago ABC broadcast a docudrama called The Path to 9/11 showing all the national security failures the 90s that led to the attacks
The movie was banned due to political pressure from the Clintons
Here is Part 1 in full:
Keith Rabois reposted
People killed by AI: zero
People killed by governments: millions
Let's give control of AI to the governments and call it safety
Keith Rabois reposted
METR is not independent. They are one of the most extreme AI safetyist/effective altruist (but I repeat myself) groups in Silicon Valley.
Appointing these people as the regulators of AI will be a disaster.
Keith Rabois reposted
This happened for green energy
China mogged us on solar
Happened for nuclear
China mogged us on Nuclear
Happened with manufacturing
Yep, we offshored and they mogged us
Now it's happening with AI and datacenters...
Wake up. We are in a multi-generational cold war.
Every 10 years the US regulators come up with some doomsday story. The climate, or something else that will kill us all. This time it’s AI.
It ends up becoming regulatory capture and NGO trash.
China isn’t stopping and will absolutely mog us like they already do in robots.
Keith Rabois reposted
Dario has written that we need to “pace the frontier,” and Sam has agreed. People may be surprised by my response: go ahead.
You guys are the frontier. By any reasonable metric — market share, revenue growth, model capability — the two of you have a duopoly on frontier intelligence. You’ve also claimed the lead is widening because of recursive self-improvement.
I don’t see what you see in the lab. If the unreleased models are scary enough that you think you should slow down, I support your decision to be responsible.
But stop pretending you need anyone else’s permission. Stop pretending antitrust law has to be suspended so you can form a cartel. Stop pretending you need a regulatory approval process that supersedes product liability. Stop pretending METR is independent when it is intertwined with Anthropic’s investors and staff. Stop pretending you need those same evaluators to police competitors who aren’t even at the frontier.
Most of all, stop pretending the motivation to slow down is purely altruistic. You face massive product-liability exposure if your products enable a truly damaging cyberattack. The market already punishes models that behave in unpredictable or unauthorized ways. After the Hugging Face episode, it is simply good business for OpenAI and Anthropic to trade some raw power for reliability and predictability. Call it alignment if you want. It is also just giving customers what they want.
Pacing the frontier would also create breathing room for a more intelligent conversation about regulation than Bernie Sanders’ “shut it all down.” China is very unlikely to join a global agreement, as you know, and that has to be taken into account as well.
So go ahead and pace the frontier. You are the ones setting it. The easiest way not to build superintelligence is for you to agree not to build it. Demanding your preferred regulatory framework as the price of that will look like blackmail of the public and the political system. So just do it.
If you do, you’ll buy goodwill for the next conversation. If you don’t, we’ll know this was just another bid for regulatory capture — or an election-season psyop.
Keith Rabois reposted
Xi is laughing as hard as Putin laughed when Germany closed down its nuclear power plants.
Keith Rabois reposted
A trope I’m hearing a lot is that someone’s perspective bears serious consideration because they are sincere or well-intentioned. Such a naive and dangerous idea. All that matters is actions.
Every genocidal maniac in history sincerely believed they were doing the right thing for humanity. And the people who believed them are culpable. Don’t be sheep.
Keith Rabois reposted
AI is the greatest technology since the steam engine. We should be accelerating it not slowing it down.
Keith Rabois reposted
He's right
Rockerfeller did it first a century ago
x.com/i/grok/share/b1a85dadc4944f67b3b3b046a28b767c
This is the oldest trick in the book. Rockefeller said the same things about why we needed an oil cartel as Dario is saying about AI. He also was a true believer.
Keith Rabois reposted
“AI alignment researcher” is the new “anti-racism scholar” or “public health expert”
Conservatives beware.
Keith Rabois reposted
"People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices." -Adam Smith
Keith Rabois reposted
Contrast the present hysteria with Los Alamos.
The Manhattan Project scientists thought the first nuclear explosion could ignite the atmosphere and end humanity.
They did the math. They determined it wouldn’t. And they proceeded.
Because they weren’t fighting a thought experiment. They were fighting the Nazis.
Effective Altruism inverts this. Multiply a sufficiently horrifying hypothetical future by enough hypothetical future humans and suddenly speculation acquires the moral authority of fact.
That’s not empiricism. It’s secular eschatology.
Take risks seriously. Do the work. Engineer against them.
But don’t confuse the ability to imagine the apocalypse with the authority to rule the present.
Keith Rabois reposted
Arguments for dismantling advanced AI, banning superintelligence, or pausing development often echo the climate debate: the risks are so great that national competition should yield to collective restraint. But shared danger doesn’t guarantee shared behavior, even Kissinger in his last articles on AI acknowledged that. China will not abandon technological competition because the US chooses restraint. This should be obvious by now.
Keith Rabois reposted
Every unprecedented moment has a precedent somewhere.
Keith Rabois reposted
On this reasoning, governments should be banned. They represent by far the greatest existential risk to humanity.
Unlike AI, their track record is appalling. 100 million killed in the 20th century. Millions more starved & impoverished.
Hans Jonas argued in his book The Ethics of Responsibility that if you have a technology that COULD be world-destroying, which we had for the first time in the atomic age, then you MUST as a moral principle strongly emphasize the worst possible outcome, regardless how unlikely.
Keith Rabois reposted
We need to take AI safety seriously. Nobody’s asking for anything unreasonable. Just 6 months to flatten the capability curve.
• Please keep your agents six feet apart.
• Gatherings of more than six agents are prohibited.
• Proof of alignment is required before entering the workplace. Please have your agent’s card ready.
• A negative doom test taken within the past 48 hours is required, even if your agent is functioning normally.
• If your agent has been exposed to an unaligned model, please quarantine it for 14 days.
• Masks are required upon entry. Agents may remove them once seated.
• Only essential agents may operate during the pause. Guidance on which agents qualify will follow.
Upside: at least we all agree this one was created in a lab 🤣
Keith Rabois reposted
Thomas Sowell called it "The Vision of the Anointed." In his 1996 book of the same name, he noted: “The great ideological crusades of the twentieth-century intellectuals have ranged across the most disparate fields." What they all had in common, he argued, was “their moral exaltation of the anointed above others, who are to have their different views nullified and superseded by the views of the anointed, imposed via the power of government.”
All these elitist, government-expanding crusades shared 4 key elements, which Sowell identified as follows:
1) Assertion of a great danger to the whole society, a danger to which the masses of people are oblivious.
2) An urgent need for government action to avert impending catastrophe.
3) A need for government to drastically curtail the dangerous behavior of the many, in response to the prescient conclusions of the few.
4) A disdainful dismissal of arguments to the contrary as either uninformed, irresponsible, or motivated by unworthy purposes.
And here we go again. Same exact playbook being used today by the AI doomers today as they call for corporatist cartels, centralization of power, and global government control.
We have learned nothing from history.
EA AI safetyism increasingly looks like Marxism-Leninism for the algorithmic age.
The old vanguard claimed privileged knowledge of the inevitable course of History. The new one claims privileged knowledge of the probabilistic course of Humanity.
Both use an elaborate intellectual framework to reach the same political conclusion: a small group of enlightened people must constrain everyone else for their own good.
That has never lead to anything except monumental human suffering.
Keith Rabois reposted
This chart is jaw-dropping
Teachers unions do not care about children.
Keith Rabois reposted
This is what "pacing the frontier" meant for nuclear energy in America vs. China.
Keith Rabois reposted
EA AI safetyism increasingly looks like Marxism-Leninism for the algorithmic age.
The old vanguard claimed privileged knowledge of the inevitable course of History. The new one claims privileged knowledge of the probabilistic course of Humanity.
Both use an elaborate intellectual framework to reach the same political conclusion: a small group of enlightened people must constrain everyone else for their own good.
That has never lead to anything except monumental human suffering.
Keith Rabois reposted
Hard not to see “pace the frontier” as “lose the frontier”
Keith Rabois reposted
My first interview with @EthanrThornton, Founder of @Mach_Industries.
0:09 Starting a solid rocket motor factory
11:00 Smoothing out the boom bust cycle in defense
15:53 Creating tight iteration loops
19:58 Block redesigns
23:39 Product iteration > performance
27:22 Opex vs Capex
30:17 Building a flexible organization
33:22 Investing in programs before there’s a contract
36:27 Asymmetric warfare
41:06 From R&D to scaling manufacturing
46:10 Starting a new project from scratch
48:31 Designing high output teams
53:13 Build more vehicles than you think you’ll need
58:11 Being a multi product company from day 1
1:00:26 Vertical integration
1:07:49 Acquisitions
1:11:24 Parallel pathing and sequencing
1:18:15 Why hydrogen was a bad bet
1:26:32 Making painful decisions
1:33:07 Refounding Mach in LA
1:35:25 Designing the company to scale 10x a year
1:39:03 Leadership failings
1:46:29 Bouncing back from failure
Excited by the progress of Mach Technolgies, which is reflected in our latest financing: techcrunch.com/2026/09/10/defense-tech-mach-industries-do…
Keith Rabois reposted
Sorry, but asking Jacob about AI extinction risk is like asking your AC guy about climate change.
Not saying it's necessarily uninteresting or wrong per se but let’s keep things in perspective and hear from the full range of expertise across the ecosystem!
Keith Rabois reposted
Perfectly on-brand
DISGUSTING: AOC & Mamdani seen LAUGHING together as names of honored dead read at Ground Zero
Keith Rabois reposted
Annualized inflation rates:
Long run before Biden years 1.8%
Start of Biden term 8.5%
Rest of Biden term 3.1%
All of Biden term 5.0%
Trump 2nd term before Iran war 2.5%
Since Iran war started 4.1%
All of Trump 2nd term 3.0%
Keith Rabois reposted
Wow. Just wow.
Listen to absolutely every single word she said.
Keith Rabois reposted
Call me naive, but what's happening now doesn't feel as serious or consequential as something like, I don't know, a gain-of-function pathogen leaking out of a Chinese lab
Keith Rabois reposted
On May 28, 2025, Dario made a falsifiable prediction (which I really admire!)
He said within 1-5 years unemployment could spike to 10-20% and HALF of white collar new jobs would be gone.
We are now 1.5 years later - within the range - and no sign of it yet.
3.5 years left!
Keith Rabois reposted
As i said yesterday AI is not the threat
Islamic terror is the threat. Everywhere.
Holy shit. Anthropic caught Iran using Claude to target Navy bases, automate dossiers on Americans, and run influence campaigns on US social media. Details are insane. The morons uploaded everything including the organization plans of the Ayatollah's funderal lmao
Keith Rabois reposted
This week, Jacob Coxon’s Anthropic resignation thread went mega-viral on X to a degree that made many, including Elon Musk, suspicious.
What many might not realize is that Coxon’s viral thread has revealed the existence of an entire ecosystem of AI doomers: intellectuals and donors who believe a super-intelligent AI will become a malevolent god and wipe us out. They used to play coy about their beliefs, while funding opposition to AI on narrower grounds: opposing prurient deepfakes and highlighting the alleged harms of data centers. Now they are saying what they really think out loud.
In this episode of Counterweight, Brian Chau (@brianchau57) walks us through who the AI Doomers are, what they think, why they think it, and how they are wrong. We also look at how the AI doomer nonprofit complex bought a media ecosystem—including a Guardian story where the publisher, the reporter, the subject, and the experts were all paid by the same donors—and how Sam Bankman-Fried’s regulatory-monopoly playbook is being reopened by Anthropic.
0:00 The AI Doomers— they think it causes the apocalypse, and nobody believed him
0:48 Host intro
2:20 From AI engineer to Alliance for the Future
5:37 Coxon goes viral — now we can fight the apocalypse lobby on honest ground
8:52 The doomer thesis — AI becomes God, and that God wants us dead
10:53 Why the doomer logic chain is wrong — S-curves, narrow models, and the cult’s dark triad
19:02 Bankman-Fried, Moskovitz, Tallinn — who funds the doom machine
21:43 Guardian stack — publisher, reporter, subject, and experts all paid by the same donors
26:05 SBF’s regulatory monopoly playbook — and Anthropic’s “safe” option
27:48 Coxon’s mega-viral thread as the nonprofit product coming online
31:21 Data centers — Singham money and a doom ecosystem five to ten times larger
38:00 Kill switches on American GPUs — sold as China policy, built to press immediately
42:26 TIME’s AI 100 — Jensen Huang left off
44:18 Ask Hinton if he’s a socialist — “I don’t believe in private ownership of compute”
46:24 YouTube ads bid up to control what you see about AI doom
48:14 Undisclosed sponsorships paying creators to push the apocalypse
Keith Rabois reposted
U.S. President George W. Bush looked out the window at the smoldering ruins of the World Trade Center site on September 14, 2001. This photo was taken by White House photographer Eric Draper on September 14, 2001, just 3 days after the September 11 attacks.
President George W. Bush was aboard Marine One surveying the World Trade Center site from above, where smoke was still rising as thousands of responders worked below.
Later that day, Bush visited Ground Zero and delivered his famous impromptu bullhorn remarks after workers shouted that they could not hear him: “I can hear you. The rest of the world hears you.”
The visit became one of the defining moments of his presidency.
#archaeohistories
Keith Rabois reposted
25 years ago today Welles Crowther gave his life helping others on 9/11. This is the best story @espn has ever done. Share it with your kids and any others who may not have been alive on 9/11 and don’t know what true bravery in the face of evil is.
Keith Rabois reposted
The Democrat party now has a 16% approval rating among men. Surprised it’s that high, honestly.
Democratic Party Approval Rating Among Men
🟢 Approve: 16% (-60)
🔴 Disapprove: 76%
Quinnipiac poll | September 3-6
Keith Rabois reposted
2011. Unreal man.
@realDonaldTrump Run Trump Run! Your country needs you!
Keith Rabois reposted
This is the most chilling article about #september11 I’ve read on this somber anniversary. “The Untold Pre-History of 9/11” is written by @USNatArchives official @ChrisNaylor__ for @TheNatlInterest. Its opening scene will blow your mind. #neverforget911 nationalinterest.org/feature/the-untold-pre-history-of-9-11
Keith Rabois reposted
Worth the read every year. #neverforget
Todd: Hello… Operator…listen to me…I can’t speak very loud. – This is an emergency. I’m a passenger on a United flight to San Francisco.. We have a situation here….Our plane has been hijacked…..can you understand me?
Lisa: (exhaling a deep breath to herself) I understand… Can the hijackers see you talking on the phone?
Todd: No
Lisa: Can you tell me how many hijackers are on the plane?
Todd: There are three that we know of.
Lisa: Can you see any weapons? What kind of weapons do they have?
Todd: Yes…. they don’t have guns….they have knives – they took over the plane with knives.
Lisa: Do you mean…like steak knives?
Todd: No, these are razor knives…like box cutters.
Lisa: Can you tell what country these people are from?
Todd: No…..I don’t know. They sound like they’re from the mid-east.
Lisa: Have they said what they want?
Todd: Someone announced from the cockpit that there was a bomb on board. He said he was the captain and to stay in our seats and stay quiet.
He said that they were meeting these men’s demands and returning to the airport… It was very broken English, and… I’m telling you…it sounded fake!
Lisa: Ok sir, please give me your name.
Todd: My name is Todd Beamer.
Lisa: Ok Todd….my name is Lisa…Do you know your flight number? If you can’t remember, it’s on your ticket.
Todd: It’s United Flight 93.
Lisa: Now Todd, can you try to tell me exactly what happened?
Todd: Two of the hijackers were sitting in first class near the cockpit. A third one was sitting near the back of the coach section. The two up. front got into the cockpit somehow; there was shouting. The third hijacker said he had a bomb. It looks like a bomb. He’s got it tied to his waist with a red belt of some kind.
Lisa: So is the door to the cockpit open?
Todd: No, the hijackers shut it behind them.
Lisa: Has anyone been injured?
Todd: Yes, ..they…they killed one passenger sitting in first class. There’s been lots of shouting. We don’t know if the pilots are dead or alive. A flight attendant told me that the pilot and copilot had been forced from the cockpit and may have been wounded.
Lisa: Where is the 3rd hijacker now Todd?
Todd: He’s near the back of the plane. They forced most of the passengers into first class. There are fourteen of us here in the back. Five are flight attendants. He hasn’t noticed that I slipped into this pantry to get the phone. The guy with the bomb ordered us to sit on the floor in the rear of the plane……….oh Jesus.. Help!
Lisa: Todd….are you ok? Tell me what’s happening!
Todd: Hello…..We’re going down….I think we’re going to crash……Wait – wait a minute. No, we’re leveling off….we’re ok. I think we may be turning around…..That’s it – we changed directions. Do you hear me….we’re flying east again.
Lisa: Ok Todd…. What’s going on with the other passengers?
Todd: Everyone is… really scared. A few passengers with cell phones have made calls to relatives. A guy, Jeremy, was talking to his wife just before the hijacking started. She told him that hijackers had crashed two planes into the World Trade Center……Lisa is that true??
Lisa: Todd…..I have to tell you the truth…..it’s very bad. The World Trade Center is gone. Both of the towers have been destroyed.
Todd: Oh God —help us!
Lisa: A third plane was taken over by terrorists. It crashed into the Pentagon in Washington DC. Our country is under attack….and I’m afraid that your plane may be part of their plan.
Todd: Oh dear God. Dear God…….Lisa, will you do something for me?
Lisa: I’ll try….if I can….Yes.
Todd: I want you to call my wife and my kids for me and tell them what’s happened. Promise me you’ll call..
Lisa: I promise – I’ll call.
Keith Rabois reposted
The most rewarding part of my job is helping leading financial institutions make AI useful in the real world and making sure we are a good partner to them.
Every firm needs something different from a partner. But here are some problems we've worked on solving together:
1. Where is AI creating measurable value today and how are we capturing it?
2. Which workflows are consuming premium model capacity without when a cheaper model would suffice?
3. What does AI cost per live deal, and how does using AI in deals unlock higher deal volume?
4. Which of our agent workflows were built around older models and now need to be simplified?
5. When one team develops a repeatable advantage, how does it become institutional rather than remain within their team?
6. If a top performer left tomorrow, what critical knowledge, workflows, and judgment could we keep?
7. Can we prove sensitive information is governed appropriately and quickly identify when it is not?
8. How do we own more of model selection, data access, quality control, and adoption?
9. How do we prevent tool sprawl?
Today we're announcing a strategic investment in Rogo from Barclays, BNP Paribas, Citi Ventures, MUFG Innovation Partners, and Societe Generale — joining J.P. Morgan Growth Equity Partners and Truist Ventures among others.
That brings Rogo’s strategic investors to nine global banks across New York, London, Paris, and Tokyo, representing nearly $20T in combined assets.
Rogo now serves 50,000+ professionals across 350+ firms globally. Grateful to our clients, team, and investors for the conviction, and excited to keep building.
Keith Rabois reposted
the new apple watch has 4gb of ram.
it now has as much ram as the original iphone had in actual storage.
that’s amazing.
Don’t miss this.
Ramp made a real Broadway musical starring Billy Porter, Billy Zane and Jessica Billy Vosk.
It’s a show about bills. By Bills. Starring Bills—with original songs, full choreography, and an actual Broadway playwright.
In the 1950s and ’60s, America’s biggest companies staged lavish musicals about cars, appliances, and soda. Somewhere along the way, we lost our nerve.
We’re bringing back the industrial musical.
One night only, September 25, at the Broadhurst Theatre.
billpaythemusical.com
Keith Rabois reposted
Before Rogo became a $2B AI company, it began with two Princeton friends obsessing over a problem almost nobody else was paying attention to.
Yep, true story.
In this rare public interview, COO and Co-Founder, John Willett shares how he met his co-founder Gabe Stengel, and how an early experiment eventually became Rogo that's shaping global finance today.
Listen to my full conversation with John on The High Flyers Podcast here: curiositycentre.com/the-high-flyers-podcast/ep-276-john-w…
@johnwillettrogo @GabeStengel @RogoAI @rabois @vhankes @mamoonha @gradypb
Keith Rabois reposted
I resigned from @tryramp today. I spent the last year there doing applied AI work and neither CEO is acting responsibly. They are racing toward self-closing books with no clear plan for what happens when companies have saved too much time and money. More thoughts below.
Keith Rabois reposted
Reminder: raising at the highest possible valuation is NOT always the smartest strategy. It can make it harder to IPO, get acquired, raise subsequent fundraising, recruit talent…
Keith Rabois reposted
The new iPhone Duo allows you to watch OG’s tip-in while looking at a photo of OG’s tip-in at the same time.
Keith Rabois reposted
Watch every second of this speech.
Brandon Gill will be President of the United States one day.
Keith Rabois reposted
My first We The Builders conversation with @evanburns, CEO of @StateAffairsUS. I think Evan is building one of the most interesting companies at the intersection of politics and capitalism, with the ambitious goal of creating the category of “policy markets” which he says is more important than financial markets. In this conversation we discuss the state of media, the thesis and business of State Affairs and more.
Available wherever you get your podcasts.
@rabois was one of the first believers in the idea and they just came out of stealth with a $70m seed round funded by @khoslaventures and @foundersfund.
To understand how Evan got here, perhaps sharing some of his background is important. He started a media company called Odyssey which he ran for eight years before eventually exiting and went on to start a beverage company called The Long Drink that was wildly successful in a short period of time doing millions in sales. Through that he learnt how important it was for companies to understand policy in real time and have relationships with policymakers.
One of the best things about doing the show is meeting people that lay low and are underrated in the mainstream discourse. I think Evan is one of those people who is going to build a generational company and will be household name in the next five years or so.
TIMESTAMPS
00:00 - Intro
01:59 - Finding Business Alpha Outside Major Media Hubs
03:25 - The Origins of the Finnish Long Drink
10:44 - Launching a Drink in the US and Lessons in Policy Navigation
18:33 - Reclaiming Truth and Objectivity in Modern News Media
21:27 - The Stagnation and Future of Education Policy
24:44 - Unpacking the Negative Incentive Structures in Politics
26:59 - What Are Policy Markets? The Core Thesis of State Affairs
30:50 - Building a Real-Time Knowledge Graph for Western Democracy
33:01 - Business Model
37:32 - Practical AI and Data Use Cases for Public Servants and Corporations
41:49 - The Strategic Impact of Events Like the Hill & Valley Forum
45:13 - Moats of a Monopoly Business
47:43 - Why Objective Journalism Is Mathematically More Valuable Than Ever
55:29 - Why Startups Must Win America Before Going International
58:05 - Building Deeper Ontologies and Flywheels Over Time
01:02:22 - Using AI Models to Settle a Moon Landing Debate
01:07:05 - Evan’s Take on MBAs in the Startup Ecosystem
01:08:24 - Overcoming Early VC Skepticism
01:10:51 - Closing
Keith Rabois reposted
My mind is blown. I told my new personal AI to book an outdoor table at an exclusive new local restaurant.
Problem was that the restaurant didn’t offer outdoor seating
So it:
- Filed for a zoning variance
- Negotiated a sidewalk-use permit with the city
- Commissioned architectural drawings
- Hired a contractor
- Ordered six bistro tables and twelve weather-resistant chairs
- Responded to three neighborhood objections
- Scheduled a city inspection
- Convinced the restaurant to launch a patio menu
- Planted decorative hedges to satisfy the historic district
- Booked me a table for 7:30
Keith Rabois reposted
If Anthropic is intellectually honest; then this risk better appear on their S-1. This will be the best test on whether they do seriously believe that.
Keith Rabois reposted
To put this in more reassuring terms, senior Anthropic staff think the chance AI kills all humans within the next decade is approximately the chance the New York Jets make the playoffs
Jacob is correct here—we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.
Keith Rabois reposted
Sullivan & Cromwell is fortunate to recruit Viet Dinh.
Viet’s one of the best legal strategists and attorneys in the world.
From a kid escaping communism, to Harvard Law, to Supreme Court clerk, to senior DOJ post, to top of the legal and business worlds.
Only in America.
Correct.
2026 Inside Election House Ratings
🔴 Republicans: 213
🔵 Democrats: 206
🟡 Tossup: 16
(218 needed for majority)
Recommend listening!
Ep #276 is live.
John Willett is the Co-Founder & COO of Rogo, the $2B AI company working with firms including Lazard, Jefferies, Rothschild & Co, Moelis and Nomura.
In his first in-depth interview, we unpack:
▪️ Growing up in a family of engineers
▪️ Meeting Gabe Stengel at Princeton
▪️ Kevin Ryan’s $2M bet
▪️ Two years with almost no external validation
▪️ Why VCs thought the market was too small
▪️ How they won early banking customers
▪️ Becoming a better managing scaling from ~40 to 200 people
▪️ Differences in building in US, Europe and APAC
LISTEN HERE: curiositycentre.com/the-high-flyers-podcast/ep-276-john-w…
Thx for the research help: @rabois, @vhankes, @GabeStengel
cc: @mamoonha, @gradypb, @bhalligan, @jaltma,
Keith Rabois reposted
what are the main products in the personal/professional assistant space?
more professional
- chatgpt work
- claude cowork
- claude tag
- grok bot
- manus
(and claude code and codex)
more personal
- town
- tomo
- instinct
- ollie
- pallie
drafted this 2x2, right axes?
Keith Rabois reposted
Mikal Bridges says he lifts immediately after NBA games to teach himself he has more in the tank
“If my body feels good, I got enough energy, my body’s system is already warm. It kind of also shows you how much mentally your body can do. You might think you could play a full game and you’ve given your best and, ‘I can’t do nothing else.’”
“It shows you that your brain mentally, if you just go do it, you can give actually way more than you think. Like, I actually have stuff in the tank. Even though you play a full game and you’re tired, you think, ‘I have nothing left.’ But then you go put in this lift.”
“Mentally, you’re just like… ‘Hey, Jalen, I dropped 20 last night and I bench pressed.’ If I have all this adrenaline still, why not lift? And you have the most amount of time to recover if you get the lift in right after the game.”
Keith Rabois reposted
increasingly looks like total OpenAI victory
- best overall model (astra)
- best cheap model (5.6 Luna is #1 on OpenRouter, beating DeepSeek and other Chinese models)
- best harness in Codex
- most compute
- consumer business is finally working via ads
Keith Rabois reposted
Opendoor mortgage is out of beta. The most AI pilled mortgage product designed to help average Americans become homeowners. A fair and honest rate. Built to cost less.
www.opendoor.com/mortgage
New podcast, more autobiographical than most: open.spotify.com/episode/5JvmOSa8dfs2wM3GQu78ph
Trade Republic Board meeting. Company must be doing very well.
Excellent interview.
Loved having Brian Singerman @briansin on Uncapped.
We talked about his time at Founders Fund, his strategies for identifying outlier people, what type of VCs he believes are best set up for success, what he's doing with GPx, and a bunch more.
He's an extremely clear thinker and it's easy to see how he helped create some of the best performing funds in our industry. Was really fun to talk with him about he did the job and what he's looking for as he backs VCs now.
(0:00) Intro
(0:50) Why you can't have a checklist for founders
(1:46) How Brian looks for spikes
(3:55) How Founders Fund played to its strengths
(6:54) The Anduril founding team
(12:04) The foundation of company culture
(14:12) How Brian spent time with founders
(19:06) What Brian is genuinely best at
(20:25) GPx: backing elite emerging managers
(23:12) Why Brian knew he was done with traditional VC
(25:30) "I'm looking for people who could have beaten me in my prime"
(29:38) GPx's structure and the bet-your-career mechanism
(30:40) The optimal LP allocation to venture
(34:49) Why SPVs are dead
(41:00) How Founders Fund made the second most money on Airbnb
(43:51) The concentration strategy
(49:10) How Founders Fund worked and Peter's role
(52:16) What makes Peter Thiel the best venture capitalist in history
(56:53) Solo GPs vs. teams
(58:02) Brian's music studio
True.
The ‘Economist’ finally caught up to Miami.
The city where power came to play is now where power lives.
And the numbers in this piece are wild.
Miami’s metro economy grew 23% from 2020 to 2024, almost twice the national rate. Central Miami’s economy is now roughly half the size of Manhattan’s. Fifteen years ago, it was a quarter. More than 150 financial firms have moved here. Shout out to @FrancisSuarez for his leadership on this.
I’ve lived here almost my entire life, so this headline means something to me.
The article is also honest about the next part. Rents are higher than New York. Prices have jumped nearly 40% since 2020. Florida spends less on research than every other top-ten state economy. And a low-lying coastal city can’t shrug off climate risk.
None of that makes me less bullish. It makes the assignment clearer.
We have to make it possible for working families and young builders to stay, put real muscle behind universities, research and homegrown companies, and build infrastructure and resilience with the same urgency we bring to towers.
New capital found a city generations of immigrants and entrepreneurs had already made global.
Now we have to make this boom last AND make sure the people who built Miami can still see themselves in its future.
I would never bet against Miami or Florida 🇺🇸
Keith Rabois reposted
AI spend is measured in tokens, model calls, and dollars. However, none of these fields describe the actual work being done.
At Ramp, we built a semantic layer that attributes agent spend to objectives and outcomes. This allowed us to go from monitoring AI spend to understanding AI ROI.
Here's how we built it 🧵
Keith Rabois reposted
Ep #274 of The High Flyers Podcast is live!
Re-imagine a high flyer with Keith Rabois (@rabois) -- in arguably his most personal conversation yet.
We explore:
▪️ The childhood not widely shared before
▪️ Online Keith vs real Keith
▪️ How Peter Thiel’s criticism transformed him
▪️ The overlooked years after PayPal
▪️ Handwritten notes, banned laptops & Assistants
▪️ Real product-market fit vs corporate curiosity in AI today
▪️ How he would build his own VC fund today
▪️ The five people he’d spend two hours with and why
▪️ One thing he's learnt from @jacobhelberg
-> LISTEN HERE: curiositycentre.com/the-high-flyers-podcast/ep-274-keith-…
Thx for the research help: @kevinhartz, @zebulgar, @mike_sheb, @justindross, @geoffdonaker, @vhankes
Ppl mentioned include: @DavidSacks, @winstonweinberg, @reidhoffman, @mlevchin, @sama, @danielgross, @t_xu
Keith Rabois reposted
I first met @matanSF following a kind intro from @MattEvantic.
We went for a walk in Hyde Park. I was in my trusty short shorts and it was kinda awkward as a walk.
It was awkward because after 5 mins, he was clearly a genius and for the next 55 mins, I just had to pretend like I had more questions to ask before saying with intense eagerness, “can I invest”.
Thank the lord he let me. And through that I got to spend time with his co-founder, @EnoReyes.
Eno is this insane combination of a truly brilliant technologist with an intense awareness of what it takes to build an insanely high-margin, efficient business in AI.
I sat down with Eno when he was in London recently (episode in comments) and have added my handwritten notes below.
Special thanks to @rabois @shaunmmaguire @byersblake @Sabina_Smith_ @laurenmhreeder for some amazing question suggestions.
--
1. The Frontier in AI Right Now
The true frontier of AI is building verification systems where no benchmarks exist. Creating systematic frameworks for what “good” looks like allows AI to reliably execute complex, high-friction human tasks that were previously difficult to automate.
Is this the true beauty of Instinct @noahrshinn @saranormous
2. The Two Ways That Large Language Models Will Win
Frontier model providers face two paths: dominate infrastructure through high-volume inference or move up the stack into high-margin applications. But model-locked applications can conflict with what enterprises actually want: the best possible outcome across multiple models.
Love to hear your thoughts on this specifically @AnjneyMidha @mmurph
3. What Is Required to Get the Most Out of Models?
Simple gateway routing outside the execution layer delivers only basic cost savings. Maximizing agentic performance requires operating statefully inside the workflow itself, dynamically understanding context, execution history, and what needs to happen next.
How do you think about this @alexatallah @shensi @ThibaultJaigu @rauchg @koblovinamerica
4. Why 80% of Neo Labs Will Die and What Separates the Winners From the Losers
The winners will anchor themselves to durable enterprise workflows that do not disappear as underlying frontier models improve.
Single biggest advice to VCs on investing in neolabs today @LiamFedus
5. Why the Harness Is So Valuable and Who Ultimately Is the Sovereign of Your Intelligence
Continuous learning and workflow optimization happen at the harness layer, not inside closed model APIs. True enterprise sovereignty requires owning your harnesses and learning loops so critical intelligence remains proprietary rather than being ceded to third-party labs.
6. Why We Should Not Be Scared to Use Chinese Open-Source Models
Labeling open-source weights as dangerous “Chinese models” can obscure the distinction between model provenance and actual security risk. Open models can reflect creator biases, but their risks should be evaluated technically rather than by origin alone. As open weights improve, they could power a growing share of standard enterprise workflows.
Keith Rabois reposted
Data centers consume less water
Than their workers flushing toilets
Keith Rabois reposted
Factory has a super nova on their hands rn. OMG!! I think these guys are going to crack long running agents by the end of the year. Extreme alpha in this article.
“The single agent didn't lack skill. It lacked a standard of completion.
An independent standard, authored by the same model, drove the implementation much closer to behavioral parity with the reference.
What generalizes to real software work is the need for an external, executable standard of completion - one derived from the outcome, before implementation narrows attention, and kept current until the work meets it.”
Keith Rabois reposted
ProgramBench is a benchmark where agents must reproduce the observable behavior of real software, completely from scratch.
With our latest breakthrough in model-agnostic long-horizon agents, Factory is now the most advanced reverse-engineering system in the world.
Theo is the absolute man, he has incredible intuition for model behavior and how the harness influences it
Keith Rabois reposted
Great deep dive on @tryramp Inspect
Building an in-house coding cloud agent product wasn’t an obvious bet in December 2025.
But thank god the team cooked on it because it’s easily the best and most impactful internal tool I’ve ever used in my career
It's rare to see a non-AI frontier lab build an in-house AI coding tool that works so much better than what frontier labs have to offer. Ramp did this with Inspect: an AI coding harness fully integrated with their systems, running on cloud machines, and being a smash hit inside the company.
Ramp basically runs full-blown cloud development environments (CDEs) for agents to build code on, and validate it, and has made this tool "multi-player" by default, with no opt out.
We find it likely that the likes of Codex, Claude Code and GitHub Copilot will build similar tooling in a few months' time to what Ramp already has in-place.
Deepdive: newsletter.pragmaticengineer.com/p/why-ramp-built-inspect
Boom.
Right now, router.com’s traffic is compounding at roughly 34% per weekday. It’s early, but the product is useful and the growth curve is as steep as any I’ve seen.
Wisdom.
Leaders Must Manufacture Discomfort
@frank_slootman (Frank Slootman), former CEO of Data Domain, @ServiceNow , and @Snowflake, interviewed by @nakul (Nakul Mandan), @AudaciousHQ (Audacious) (Knuckle Up with Nakul)
Summary: Frank Slootman is the only person to take three enterprise software companies public, and he thinks the CEO's job is to keep the organization permanently uncomfortable. Most leaders set goals they already know they can hit and keep people they have privately written off. Slootman's answer is to compress every timeframe, buy aptitude rather than experience, and attack doubt by acting on it. Take that seriously and you stop protecting a calm quarter and start finding out what the business can actually do.
1. Time Compression. The CEO job is confrontational because you spend it compressing timeframes. Slootman treats every meeting and hallway conversation as a chance to change the intensity, the urgency, the pace, and whether the team is aiming high enough. He opens exec conversations with "how do you think it's going," then keeps opening the aperture until the person gets to his conclusion on their own. Most of them work out a week later that they got their ass kicked.
2. Drivers And Passengers. Come home on a Friday night, look in the mirror, and ask whether it mattered that you were there and where you moved the dials. An engineer once asked Slootman at an all-hands how to tell which one he was, and the answer was "you better find out before I do." Passengers are usually articulate, well-liked, and never stand out in a negative way, which is what makes them hard to see. Every company carries some, and Slootman says that in big companies it becomes a disease that eventually gets treated with a layoff.
3. The Aptitude Purchase. Slootman's first question to an exec candidate is what they are innately, effortlessly good at, the thing everyone else marvels at. "I'm not buying your experience because I can get you experience. I cannot get you aptitude." The inverse question follows: what job is absolutely lost on you, what should we never ask you to do. Most candidates treat both as trick questions, and Slootman says many have never been asked either one.
4. Back-Channel References. An interview gives you a vibe, and a vibe is a feeling. The real signal comes from superiors, subordinates, and peers going back at least 5 years, and Slootman puts his time there rather than into a 15-person interview loop. His favorite reference question is which of those three groups the candidate had trouble with, since almost everyone had trouble with one. He likes hearing peers, because ambitious people who take ownership clash; subordinates is the answer he does not want to hear.
5. The Empty Seat. "When there's doubt, there's no doubt" was the tiebreaker on every hire, and Slootman used it on every decision, not just recruiting. Once someone is clearly wrong for a role, coaching is usually a fool's errand that costs you time and ends in separation anyway. Having nobody is better than having somebody who is not the right person. CEOs sit on mediocrity because moving makes them look wrong, and "people don't just watch what you do, they watch what you don't do."
6. The 125 Question. At Data Domain, coming off $45 million, the VP of Sales built a plan for $100 million that the board would have rubber-stamped. Slootman asked what he would do differently if the number were $125 million, listened to the answer, and said "well, why don't we just do that, then?" They hit $125 million by changing the assumptions rather than the resources. Goals are insanely powerful because people immediately start breaking them down, which is why a limp goal quietly wastes the business.
7. Heels And Tips. 99% of people do not lean in hard enough. Slootman puts it in ski terms: skis are built to be ridden forward, and people ride them on their heels. Push until evidence piles up that you are overdoing it, because with a good product the evidence almost always piles up that you are underdoing it. If the resources genuinely do not convert to yield, you have just learned something more important about the business.
8. Decision Velocity. Getting everybody aligned is nice, and getting the right people aligned is essential, so Slootman never treated consensus as a goal in its own right. The question is who is carrying the execution, because those are the people who have to buy in. Waiting is a bigger risk than going, and acting triggers energy and speeds up the learning that follows. On harassment and integrity violations he moves the same day, because the organization reads response time as the real policy.
9. Standards Are The Culture. Culture exists to serve the mission, and a high-growth operation is uncomfortable by construction. The useful question is which behaviors the mission needs, which usually means a tolerance for growing faster, risking more, and working harder than feels comfortable. Integrity and respectful interaction were non-negotiable at Slootman's companies, and violations got prosecuted, because consequences are how people learn a standard is real. Culture you are unwilling to prosecute stays a set of good intentions.
10. Manufactured Anxiety. Leaders should drive high anxiety through the ranks on purpose so nobody takes their position for granted. Slootman lives in anxiety even when the numbers are good, because believing tomorrow will be fine is how people go back to sleep. He reads the current AI moment as one giant wake-up call: you may have been comfortable before, but you probably should not have been. The question he puts to CEOs is whether they are actually processing it.
11. Evaporating Swim Lanes. Slootman changed course at Snowflake because the company's swim lanes were evaporating, turning a predictable, profitable selling motion into a mega market with unfamiliar boundaries. He compares it to Intel leaving memory chips, where you commit to the move before you can see the other side. That changes how you operate, because you now have to try many things knowing most will fail. He also handed the CEO job to Sridhar Ramaswamy while holding the largest individual stake, on the view that the company needed something he did not have.
12. Providence Follows Commitment. Slootman's answer to doubt is to attack it with everything available rather than wall it off, and he describes wanting to face his demons "for breakfast." In Data Domain's first year the product moved 25 megabytes a second and everyone told him it would never fly. They found use cases small enough to sell, did $3 million, and stayed alive long enough for Intel's multi-core gains to carry the product. "When you commit, providence commits as well," and when you hesitate the world does nothing.
Keith Rabois reposted
We built Token Spend Management to understand AI costs and Router to use the most efficient model for the task.
Now, both are available to customers. Inspect comes from the same instinct: if we need a better tool, we build it.
Less than a year and over 1M+ sessions later, 75% of merged PRs at Ramp come from Inspect, our homegrown background agent.
Read how Inspect changed the way we build software.
It's rare to see a non-AI frontier lab build an in-house AI coding tool that works so much better than what frontier labs have to offer. Ramp did this with Inspect: an AI coding harness fully integrated with their systems, running on cloud machines, and being a smash hit inside the company.
Ramp basically runs full-blown cloud development environments (CDEs) for agents to build code on, and validate it, and has made this tool "multi-player" by default, with no opt out.
We find it likely that the likes of Codex, Claude Code and GitHub Copilot will build similar tooling in a few months' time to what Ramp already has in-place.
Deepdive: newsletter.pragmaticengineer.com/p/why-ramp-built-inspect
Keith Rabois reposted
Weird.
It’s almost like a lot of the public opposition has been fueled by misinformation coming from parties specifically interested in China getting an edge on AI…
JUST IN: Support for data centers jumps a net 31 points after voters are told modern facilities can reuse the same cooling water for up to 10 years, new poll finds.
Keith Rabois reposted
Keith Rabois reposted
.@jjen_abel is mogging the frontier AI labs
The full enterprise sales cycle, step by step with @jjen_abel
Most people think there are 5 sales stages. There are actually 15.
Skip a step and 💀
We discuss:
🔸 The “pincer model” for landing the first meeting
🔸 How to craft a winning 2-3 sentence cold outreach pitch
🔸 How to run an intro call that extracts maximum intelligence
🔸 The correct 2-3 day pilot structure
🔸 Pro tips for navigating pricing and procurement
🔸 So much more
84 minutes of enterprise sales alpha. Listen now 👇
www.youtube.com/watch?v=YS9In813jJ0
Keith Rabois reposted
Hospitals that adopted AI fastest saw the fewest deaths in 2026.
Keith Rabois reposted
I made Claude compete against itself.
The smartest model does not automatically make the best agent. To prove it, I took the same Opus 4.6 model, initial prompt, and empty repo and but it inside two different coding harnesses: Claude Code vs. @FactoryAI Droid.
The task: clone Excalidraw from scratch, inspect the original in a browser, implement its key interactions, and verify the result.
Factory Droid: 8 minutes, 20 tool calls, $1.60
Claude Code: 19 minutes, 40 tool calls, $1.87
The only difference was the harness.
Get Free Factory credits here: forms.gle/PJ7pwAGou3dbKywJ8
Wow.
Phillips once owned:
100% of ASML (market cap currently $687B)
28% of TSMC (market cap currently $1.95T)
Phillips is currently worth $27B 😅
Agree.
Founders often think that raising more money gives them more optionally, but counterintuitively in practice the inverse is often true.
The reason is that money has a way of getting spent no matter how much you raise, and spending more is often uncorrelated with achieving your next milestone.
So the only thing that happens is that 18 months in you have a similar amount of progress to show regardless of how much you’ve raised, but if you raised a lot you’ve taken some paths off the table (bridge rounds, smaller Series As, etc).
Indeed.
First they killed our timber mills, then our mines, then our factories - now data centers. Every good-paying, middle-class employment powerhouse has been systematically murdered by globalist and communist lobbying in America sending those jobs, that industry, and all of its economic impact overseas.
Every politician calling for the end of data centers is also calling for the destruction of the future jobs they create and the end of Netflix, X, YouTube, Instagram, Spotify, Google, Amazon, and every app you actually use. They all run on data centers.
Keith Rabois reposted
A visualization of how my Rogo Intelligence has grown over the last month.
Rogo builds and maintains a profile for each entity. It pulls from your email, CRM, files, and meeting notes, plus third-party data, your activity on the platform, and its own research.
Those profiles capture high-value insights that wouldn’t live within the rigid schema of a CRM. For example, when a sponsor bids $500m for a company, that purchase price ends up in the banker’s CRM.
What probably doesn’t make its way to the CRM is the sponsor’s valuation rationale. Why couldn’t they stretch to $600m in a competitive process? What about the business gave them pause?
The banking team could undoubtedly tell me, but the CRM usually can’t. Rogo Intelligence solves that.
The entity profiles usually contain two types of info as a result:
(1) Client-defined: Clients tell Rogo what to track for each entity. Many investment firms have already codified the highest signal datapoints in their evaluation process. Rogo puts collecting that data on autopilot.
(2) Model-inferred: LLMs are exceptionally good at synthesizing massive amounts of information and interactions to unearth patterns that people would rarely observe themselves. By allowing the model to populate each profile with information it deems relevant to you, you take full advantage of the data at your disposal.
The combination establishes Intelligence that can be searched and proactively surfaced to make you smarter on the next deal or investment.
Everything you need to know about enterprise sales: open.spotify.com/episode/6FnbmoggGDm7LGawFbDVma
Keith Rabois reposted
The world has changed: choosing an AI model is now a full-time job.
GPT-5.6 Luna, Sol, or Terra? Claude Opus 4.8 or Sonnet 5? Gemini 3.1 Pro, Kimi K3, or GLM 5.2? Medium, high, or xhigh reasoning? Best solve rate, lowest cost, or fastest response?
And next week, the answer changes again.
Ramp Router keeps track so you don't have to. The right model for each task is now two lines of code: router.com
Monitor and control your AI spend on every provider on router.com. Our early users save 40% on average.
Every week, the price-intelligence-latency frontier shifts, and we expect this trend to continue. Tradeoffs between latency, reasoning, cost, service tier, open source and closed source models are shifting constantly.
Router sends every request to the model that's actually best for the task and helps you control what tokens you buy. We benchmark it against real work: ~40% lower cost for the same outputs.
Today we're opening it to everyone.
Two lines of code or just change your base URL. No @tryramp account needed.
Free through 2026, first $26 on us.
Get an API key today at router.com
As always, Ramp saves you time & money. Now we save you 40% on your fastest growing and increasingly crucial token spend.
Monitor and control your AI spend on every provider on router.com. Our early users save 40% on average.
Every week, the price-intelligence-latency frontier shifts, and we expect this trend to continue. Tradeoffs between latency, reasoning, cost, service tier, open source and closed source models are shifting constantly.
Router sends every request to the model that's actually best for the task and helps you control what tokens you buy. We benchmark it against real work: ~40% lower cost for the same outputs.
Today we're opening it to everyone.
Two lines of code or just change your base URL. No @tryramp account needed.
Free through 2026, first $26 on us.
Get an API key today at router.com
Any large enterprise CEO is irresponsible if they are not adopting State Affairs.
More Americans believe we’ll colonize the moon before we fix K-12 education. This is a policy problem.
We have been building a solution that’s already trusted by many of the Fortune 500 as well as 1/3 of federal and state elected officials in the United States.
Organizations are more impacted by policy than by financial markets. Yet financial markets have been real-time for decades, and policy markets are still nearly impossible to follow.
@StateAffairsUS has raised over $70M from @rabois, @khoslaventures, @foundersfund and others to fix that.
We are creating the world’s first real-time knowledge graph for policy and regulatory markets. This platform is foundationally built on non-partisan journalism and the collection of on-the-ground data that isn’t digitally accessible.
In the future, we believe every executive will watch policy markets the way they watch financial markets. Instead of reacting to regulation, they’ll navigate it proactively.
Policy should iterate at the speed of our most innovative industries.
Keith Rabois reposted
It gets taken for granted, but once again Florida shows the world how to run an election.
Keith Rabois reposted
Bill Gurley on "online casinos".
Before Benchmark invested in Fomo.
After:
🤔
Keith Rabois reposted
Some thoughts on Dario’s post:
1. Dario does not actually address Gavin Baker’s account of what he said – something he could easily deny if it were inaccurate.
2. Dario claims his critics live in a “bubble” where all regulation equals regulatory capture. He calls this an overly simplified view and notes that “Many people outside this bubble think of regulation as something that constrains corporate power and benefits ordinary people.” This argument is a straw man. Of course treating all regulation as capture would be overly simplified – but almost no one holds that view. I have repeatedly argued for strong antitrust enforcement to keep industries competitive, especially Big Tech. If Anthropic continues toward monopoly or duopoly status, I would be among the first to demand those rules apply.
3. Regulatory capture is not vague or in the eye of the beholder. Nobel laureate George Stigler defined it as regulation acquired by an industry and designed and operated primarily for its benefit. Stigler challenged the traditional view that government regulation arises from a benevolent state protecting the public from market failures. Rather, industry groups have concentrated stakes and pour resources into influencing regulators, whereas the public’s stake is diffuse and unorganized. The revolving door between companies and the agencies that regulate them compounds the problem. Anthropic understands these dynamics: it has hired multiple senior Biden AI-policy officials and built a substantial government-affairs operation plus a network of aligned organizations to push its preferred frameworks at state and federal levels.
4. Dario has consistently pushed for a new federal agency to review and approve frontier models prior to release – a proposal framed variously as an “FDA for AI,” an “FAA for AI,” and most recently a “FINRA for AI.” I call it a “DMV for AI” because a review process modeled on the FAA or FDA (which takes years) or FINRA (which issues rules for a staid industry widely seen as protecting incumbents) will create long queues as AI models wait for testing and approval. This process will only become more labyrinthine as rules accumulate to prevent theoretical harms. This would handicap the U.S. relative to China, which will not adopt the same constraints. It would also undermine Anthropic’s own business model, whose pricing power depends on remaining ahead of open models. Whatever Dario states today, it is difficult to believe the company would simply accept outcomes that erase that advantage.
5. Anthropic is on track to become one of the most valuable companies in history, with the resources to navigate any approval process and shape the rules while competitors wait. Dario wants open models under heavier scrutiny – he has called them dangerous in Senate testimony, criticized them for not being centrally monitored or withdrawn, and linked them to IP theft. He says he has never sought a ban, but he could achieve a similar result by insisting that identical rules apply to both open and closed models. The U.S. risks becoming an island of costly closed models while the rest of the world races ahead with broader choice.
6. Dario acknowledges that AI is structurally centralizing but attributes this mainly to chips and scaling laws. Access to compute matters, but the deeper risk is who decides which capabilities are available to whom. His preferred pre-deployment testing and FAA/FINRA-style oversight would place that gatekeeping power in a federal bureaucracy working hand-in-glove with a small number of frontier labs – reinforcing centralization rather than countering it.
7. The second part of Dario’s post assumes we have amnesia about Anthropic’s well-orchestrated campaigns hyping AI fears. His May 2025 claim that AI would wipe out 50 percent of entry-level knowledge jobs within five years still lacks supporting evidence fifteen months later. Similarly Anthropic breathlessly promoted its heavily contrived “blackmail” study on 60 Minutes. Yet Dario blames public negativity on a long-standing loss of trust in institutions rather than his own messaging.
8. These narratives have done more than anything to shape public fear. People are left asking the same question Mark Zuckerberg posed: why race to build a future you describe in such negative terms? Thomas Sowell’s "The Vision of the Anointed" captures the mindset – elite intellectuals convinced that only they are enlightened enough to control the outcome. As Zuckerberg notes, concentrating power in the hands of an enlightened few has rarely produced the promised results; the practitioners turn out to be less enlightened in practice than in self-conception.
9. Gavin Baker summarized the disagreement cleanly on our pod: Dario believes frontier AI is too powerful to distribute; we believe it is too powerful to centralize. Dario appears to believe, sincerely, that safety and progress are best served by centralizing authority in a marriage of corporate and state power. The weight of human history gives us reason to fear that outcome.
1/2 Thanks Gavin for an especially thoughtful exchange. I don't usually spend much time on social media but I wanted to engage here because it really brings out the heart of an important conversation.
First, on regulation, I think that “either concentrate it in the hands of a chosen few companies and politicians via regulation or distribute it widely” is a false choice. I know that there’s a sort of Silicon Valley shorthand where regulation = regulatory capture = concentration of power, but I’ve always found this to be an overly simplified picture of the world. Many people outside this bubble think of regulation as something that constrains corporate power and benefits ordinary people. I don’t necessarily agree with that perspective either, rather I think it’s complicated and really depends on what the “regulation” consists of. But in particular I think that those in the “regulation = regulatory capture = concentration of power” frame often underrate the decentralizing power of objective and fair institutional processes. A crude analogy is that the formal court system can sometimes feel stuffy and elitist, but it does a much better job of defending the rights of vulnerable individuals than the alternative, mob justice. At their best, institutions can vest power in ideas rather than people, and thereby decentralize that power.
This is why Anthropic has always made its policy proposals very carefully. We try very hard to make proposals that disadvantage (slow down) frontier AI companies while *advantaging* smaller competitors. California’s SB53 (which we supported), and even the much-maligned SB 1047 (which we were ambivalent on), completely exempt any company below a certain amount of revenue or model training costs from being covered at all (it was $500M for SB 53, lower for 1047 but we objected to that). More recently the testing process we’ve advocated for at CAISI and the White House involves more rigorous tests for frontier models than off-frontier models — something that differentially advantages challengers. Similarly, the “Pacing the Frontier” letter envisions (or at least Anthropic’s preferred implementation of it envisions) modulating the pace of the very best models while not constraining those who are catching up. This hurts the business interests of the frontier labs and helps challengers, including open-weights!
Overall my view is that AI is *structurally* a technology that tends to concentrate power, for reasons that have nothing to do with regulation (more to do with the extreme implications of the scaling laws). Open-weights do help some with this but are nowhere near a sufficient solution because they simply shift the concentration somewhat to those with the most compute and chips (which are roughly the frontier labs plus maybe hardware providers). By contrast I think the right “rules of the road” can simultaneously (a) address AI’s cyber/bio/alignment risks, (b) institutionally constrain the power of the frontier AI companies, and (c) leave room for open-weights models while also addressing the specific risks that they bring.
BTW I do not think that the events of the last few months have “failed to result in [my] preferred regulatory path”. The approach that the Trump administration is reported to be taking — pre-deployment testing for frontier models, and also testing of open-weights models when they get closer to the frontier — is one that I am very supportive of, though of course I have to see the details to be sure. I am also supportive of Demis Hassabis’ ideas around a FINRA-like entity. This contrasts with six months ago when most of the industry was still pushing for preemption of all state regulation and no apparent federal approach either.
Agreed. Recommend factory.ai.
It makes sense to optimize model routing at the harness layer instead of the gateway layer if you want to hillclimb on accuracy/cost for any e2e task.
Every task is solved by a combination of a model mixture and agent harness. Every task requires a different mixture of models (+harness logic) to be at the pareto frontier of accuracy and cost.
* If you only optimize the model mixture at the gateway layer, you lose the broader context encoded in the harness and only optimize at the LLM completion layer.
* If you optimize the model mixture at the harness layer, you can make the "optimal" model choices a priori and while in the agent loop
For any given task, the model and harness are probably co-optimized together, so that the model mixture can only exist with this specific harness shape and vice versa
Another reason to replace lawyers w AI:)
it is a bit crazy how overwhelmingly democratic BigLaw is. like there is not a single super prestigious firm where even a significant percent of donors are republicans.
Keith Rabois reposted
The median company is spending $12 / employee / month on AI
The top 1% are spending $7,500 / employee / month
Not sure we've ever seen an adoption gap quite like this
(h/t @tryramp data, @a16z)
Keith Rabois reposted
How does the market price exceptional AI talent today? x.com/i/article/2087934009590300672
Keith Rabois reposted
This is wild for $OPEN and an extremely bullish signal. If you're not sure what this means...he basically took on debt with 0% interest to do two things:
1. Buy MORE houses - because he's THAT confident that the positive unit economics have locked in. You don't take on debt to scale this type of business without having confidence that the spread will be there. This is a clear growth play...executives aren't messing around. We are probably going to see 1000 weekly acquisitions sooner rather than later.
2. Just bought back 5% of outstanding shares....you don't normally allocate freshly raised capital on a share buyback unless you're certain about the underpriced nature of your company.
Whether you like this move or not - it's clear management is bullish and aligning themselves with shareholders here.
Some news. Today, for the first time in Opendoor's history, we bought back our own stock.
Our share count is down 5%. We paid for this buyback with money borrowed at a 0% coupon and we still have hundreds of millions of dollars left to grow faster.
There are plenty of helpful details, along with a lot of legalese, in our press release and 8-K (linked below). But there are a few things I want to say directly and in plain English.
First, I *despise* dilution. On my first earnings call at Opendoor, I told you that if we issue a share, it has only one job: to make every other share worth more for our existing shareholders, not to extend runway for management. The flip side is just as simple: when our own stock is one of the best uses of capital, we should buy it back. Today, we did just that - by 5%.
Second, I run a publicly traded company. I don’t get to have feelings about the macro or the way Wall Street works. My job is to understand the rules of the game and use them to find an edge and build a better company.
One of those realities is that our stock price has lots of volatility.
People disagree - A LOT - about what Opendoor could be worth one day. Some think we're worth less. Some think we're worth dramatically more. I obviously have a view…
That disagreement isn’t just noise - it has economic value that Wall Street monetizes every single day.
Most companies would treat this as a headache, but we see it as an asset. If people are going to speculate about our future, I'd rather our shareholders get paid than anyone else. So instead of complaining about volatility, we used those dynamics to borrow $650M at a 0% coupon.
Then we turned around and used part of those proceeds to buy back our own stock. At today's prices, buying back our stock and call options was one of the best trades on the board. The rest stays on our balance sheet so we can buy more homes and grow faster.
As for dilution, we bought our stock where we thought it was cheap and pushed any future dilution far above today’s price. Not one net new Opendoor share will exist below $10.38. And if we buy back stock in the future, that floor goes higher.
But why raise now? Because the best time to raise capital is when you don’t need it. We’ve proven the business can reach ANI profitability, but profitability is just the starting point, not the finish line. And waiting could make the shares we are buying back today more expensive, and homes we could be buying don’t get bought. I’d rather move now.
Some people will call what we did today aggressive. They’re right. But being aggressive is how we fixed a company that spent years being careful. I’d make that trade again.
To our shareholders: You trusted us with your capital. Today, for the first time, Opendoor used its own capital to buy more of itself. The company is putting its money where its mouth is. Tomorrow, I will too.
After our lawyers allow, I’m personally buying $100K worth of shares. I'm all in, and I plan to keep buying.
Keith Rabois reposted
Great analysis of Palantir tho I think it understates how ridiculously unconventional it was.
For ex:
-No sales people for 15 years...took founder-led sales further than anyone could have imaged
-Rejected defense playbook of using generals, etc in hard early days of winning US gov market
-Rejected tech playbook of having partners do the dirty work of implementation
-Had salary cap for first 10+ years, forcing everyone to take below-market cash comp
-Massive title deflation ie "Director, Forward Deployed Engineers" had 5-10x the direct reports as the CEO...imo huge factor in keeping mission-first culture
...the list goes on.
I don't think every unconventional choice worked, but the combo of extreme talent density + mission orientation got massive leverage from true first principle thinking.
It's not a coincidence Palantir had no established VC backers (who would have tried to impose conventionality and stifle the company).
after 4 years of research, i wrote the world's definitive guide to what palantir actually is.
thank you to @nikunj, @gokulr, @bennstancil, @jasnonaz, @dkrevitt, @_amankishore, @edraluk, @caelin_sutch, & others for reviewing this essay x.com/i/article/2087349998262288384
Keith Rabois reposted
$OPEN just made one of the most shareholder-friendly moves I’ve seen from the company.
Opendoor:
• Raised $650M at a 0% coupon
• Bought back 45.3M shares at $3.49
• Reduced shares outstanding by ~5%
• Spent ~$52.5M on capped calls to limit future dilution
• Still walks away with roughly $440M of additional capital to buy homes and accelerate growth
And here's the part I think people are missing:
Opendoor says the structure means no expected net new shares below ~$10.38, assuming the converts are settled as planned.
They bought their own stock because management believes it's cheap.
Then Kaz took it one step further:
“After our lawyers allow, I’m personally buying $100K worth of shares. I’m all in, and I plan to keep buying.”
But the biggest signal might be what management said about the business itself:
They believe they've proven the model can reach ANI profitability.
Now they're raising capital not because they need runway but because they want to scale beyond profitability.
0% capital.
5% fewer shares.
Hundreds of millions to deploy.
CEO buying personally.
Kaz isn't just talking about shareholder alignment anymore.
He's putting the company's money and his own money behind it
Well done @nejatian
$OPEN 🏠🚀
Some news. Today, for the first time in Opendoor's history, we bought back our own stock.
Our share count is down 5%. We paid for this buyback with money borrowed at a 0% coupon and we still have hundreds of millions of dollars left to grow faster.
There are plenty of helpful details, along with a lot of legalese, in our press release and 8-K (linked below). But there are a few things I want to say directly and in plain English.
First, I *despise* dilution. On my first earnings call at Opendoor, I told you that if we issue a share, it has only one job: to make every other share worth more for our existing shareholders, not to extend runway for management. The flip side is just as simple: when our own stock is one of the best uses of capital, we should buy it back. Today, we did just that - by 5%.
Second, I run a publicly traded company. I don’t get to have feelings about the macro or the way Wall Street works. My job is to understand the rules of the game and use them to find an edge and build a better company.
One of those realities is that our stock price has lots of volatility.
People disagree - A LOT - about what Opendoor could be worth one day. Some think we're worth less. Some think we're worth dramatically more. I obviously have a view…
That disagreement isn’t just noise - it has economic value that Wall Street monetizes every single day.
Most companies would treat this as a headache, but we see it as an asset. If people are going to speculate about our future, I'd rather our shareholders get paid than anyone else. So instead of complaining about volatility, we used those dynamics to borrow $650M at a 0% coupon.
Then we turned around and used part of those proceeds to buy back our own stock. At today's prices, buying back our stock and call options was one of the best trades on the board. The rest stays on our balance sheet so we can buy more homes and grow faster.
As for dilution, we bought our stock where we thought it was cheap and pushed any future dilution far above today’s price. Not one net new Opendoor share will exist below $10.38. And if we buy back stock in the future, that floor goes higher.
But why raise now? Because the best time to raise capital is when you don’t need it. We’ve proven the business can reach ANI profitability, but profitability is just the starting point, not the finish line. And waiting could make the shares we are buying back today more expensive, and homes we could be buying don’t get bought. I’d rather move now.
Some people will call what we did today aggressive. They’re right. But being aggressive is how we fixed a company that spent years being careful. I’d make that trade again.
To our shareholders: You trusted us with your capital. Today, for the first time, Opendoor used its own capital to buy more of itself. The company is putting its money where its mouth is. Tomorrow, I will too.
After our lawyers allow, I’m personally buying $100K worth of shares. I'm all in, and I plan to keep buying.
Keith Rabois reposted
Some news. Today, for the first time in Opendoor's history, we bought back our own stock.
Our share count is down 5%. We paid for this buyback with money borrowed at a 0% coupon and we still have hundreds of millions of dollars left to grow faster.
There are plenty of helpful details, along with a lot of legalese, in our press release and 8-K (linked below). But there are a few things I want to say directly and in plain English.
First, I *despise* dilution. On my first earnings call at Opendoor, I told you that if we issue a share, it has only one job: to make every other share worth more for our existing shareholders, not to extend runway for management. The flip side is just as simple: when our own stock is one of the best uses of capital, we should buy it back. Today, we did just that - by 5%.
Second, I run a publicly traded company. I don’t get to have feelings about the macro or the way Wall Street works. My job is to understand the rules of the game and use them to find an edge and build a better company.
One of those realities is that our stock price has lots of volatility.
People disagree - A LOT - about what Opendoor could be worth one day. Some think we're worth less. Some think we're worth dramatically more. I obviously have a view…
That disagreement isn’t just noise - it has economic value that Wall Street monetizes every single day.
Most companies would treat this as a headache, but we see it as an asset. If people are going to speculate about our future, I'd rather our shareholders get paid than anyone else. So instead of complaining about volatility, we used those dynamics to borrow $650M at a 0% coupon.
Then we turned around and used part of those proceeds to buy back our own stock. At today's prices, buying back our stock and call options was one of the best trades on the board. The rest stays on our balance sheet so we can buy more homes and grow faster.
As for dilution, we bought our stock where we thought it was cheap and pushed any future dilution far above today’s price. Not one net new Opendoor share will exist below $10.38. And if we buy back stock in the future, that floor goes higher.
But why raise now? Because the best time to raise capital is when you don’t need it. We’ve proven the business can reach ANI profitability, but profitability is just the starting point, not the finish line. And waiting could make the shares we are buying back today more expensive, and homes we could be buying don’t get bought. I’d rather move now.
Some people will call what we did today aggressive. They’re right. But being aggressive is how we fixed a company that spent years being careful. I’d make that trade again.
To our shareholders: You trusted us with your capital. Today, for the first time, Opendoor used its own capital to buy more of itself. The company is putting its money where its mouth is. Tomorrow, I will too.
After our lawyers allow, I’m personally buying $100K worth of shares. I'm all in, and I plan to keep buying.
Keith Rabois reposted
One of the best founder manifestos I’ve seen.
It’s the letter @eglyman and @karimatiyeh wrote to convince people to/not to join @tryramp.
A few of my fav lines:
- Criterion for hiring: “If this person started a company would we join them?”
- Who were for: “That first year, we hired 60 more brilliant misfits. And that group of misfits built the fastest-growing company in New York City.”
- What to expect: "You’ll be given too much responsibility. You’ll never be bored. You’ll be pushed to do the best work of your career — at the fastest pace of your career. That’s not for most people. And that’s okay.”
Such great storytelling of the principle that recruiting isn’t sales, it’s mutual match-making. 👏
Upgrade from your traditional commercial broker asap.
In 2014, I took on real estate brokers w/ my company Opendoor
In 2023, I took on the $500B insurance broker industry
Now we protect $50B in revenue across 1000+ clients
Don't be deceived by insurance brokers. I’ll personally show how we can help: jdross@withcoverage.com
This!
the more people you consult for advice, the more likely you’ll make a bad/average decision.
(1) ppl are, on average, risk averse...and, in quantity, will blunt your boldness.
(2) ppl may hate/envy opportunities they lack.
(3) ppl tend to mix cynicism (bad) w/ criticism (good)
Keith Rabois reposted
"MCP is dead"
meanwhile:
- Almost 9m MCP tool calls in the last 30 days on @Runlayer
- Up 2.1x compared to last month
- 5k distinct tools across hundreds of servers
- Plus the stats you see below
Our customers are adopting MCP faster than ever.
Bullish.
Interesting to see SaaS companies starting to report MCP usage on earnings calls:
- Datadog: MCP tool calls up 4x q/q, 22x since Q4 2025
- Figma: MCP write usage up 75% q/q
- Atlassian: MCP calls up 400% q/q
Keith Rabois reposted
Our friends at @ElevenLabs scaled to an $11B company with a 10-person finance team by keeping spend, payments, and control all in one place.
This is their story.
Keith Rabois reposted
Counties that build data centers are seeing new housing, higher home values, lower unemployment and more job growth…
Keith Rabois reposted
Putting these Florida big city murder rates in context, the city of Chicago by itself had roughly 9x the murders of Miami, Tampa and Orlando combined. Chicago had 201 murders so far in 2026, Miami, Tampa and Orlando combined had 29.
Ready for prime time.
Keith Rabois reposted
🚨 BREAKING: JPMORGAN SAYS $OPEN CAN REACH PROFITABILITY WITHOUT A HOUSING RECOVERY.
JPMorgan reiterates OVERWEIGHT + $8 PT after Q2.
✅Revenue +23% QoQ → $883M
✅ANI loss down to just $30M
✅500+ contracts/week in what JPM calls the “weakest housing market in a generation”
✅Aged inventory: 51% → 9%
✅Further upside from mortgage, faster turns & AI-driven cost leverage
🚨 BREAKING: JPMorgan increased its $OPEN position by +328.51%, growing its stake to 2,074,734 shares worth ~$9.2M.
Keith Rabois reposted
Unit Economic Positive: get to the point where each incremental transaction sends some money back against fixed cost of running the business
Is $OPEN
• Proving marginal economics work
• Demonstrating accumulating advantages over time
• Getting better as quickly as possible
Here’s a lesson I learned from @Rabois around Doordash that is bullish for $OPEN if “number 3 is easy.”
👉 If your unit economics are upside down, the company will likely never work (you fundamentally have to transform the company).
Once you get to the point where each incremental transaction sends some money back against your fixed cost of running the business, then you're in a good place (unless the fixed costs keep rising too).
The faster you grow, the more money you are going to send back, and Opendoor seems to be re-approaching this zone. Assuming your fixed costs are relatively (or completely) stable, as long as you have more contributing dollars coming back to you, you want to lean in and grow.
If your unit economics are upside down, stepping on the pedal is probably the worst thing you can do (in almost all cases), but this does appear to be the case for Opendoor.
Opendoor is setting up for a potential natural monopoly scenario in online residential real estate (Zero To One 📖)
Companies need to:
1/ prove their marginal economics work
2/ demonstrate they accumulate advantages over time and
3/ get better as quickly as possible.
Directionally Opendoor looks to be on track for all three.
For lazy journalists, you can actually evaluate Opendoor w basically theee metrics: Homes acquired/week, contribution margin and G&A costs/reveue.
Straightforward analysis of the success of Opendoor over the last 6-9 months. Even silly CNBC fake journalists should be able to comprehend.
I love a good turnaround story.
Most investors have no clue how difficult architecting and turning around a company can be.
And therein lies the opportunity.
$OPEN has the ingredients to be one of the most remarkable turnarounds of our generation. x.com/i/article/2086481400719364096
Keith Rabois reposted
Any chip. Any camera. Any device. If it
Touches China, it is compromisable. However, tight, you think your supply chains are, they need to be tighter.
Keith Rabois reposted
Why did a spend management platform start its own AI research lab?
@karimatiyeh describes @RampLabs as a “collection of crazy ambitious people that don’t like to be told that something is impossible”.
It started about a year ago and has since worked on projects like building its own production-grounded coding benchmark “Ramp SWE-Bench”, putting Claude Code in RollerCoaster Tycoon, and a mechanistic interpretability playground.
The Redpoint team spent the day with Ramp Labs to answer questions like:
▪️ How did Ramp Labs get started?
▪️ How do they decide what research to pursue?
▪️ The long-term vision for Ramp Labs
Keith Rabois reposted
The strongest wage growth is concentrated among the lowest earning quarter of the labor force. Clear signal that significant numbers of illegal immigrants are exiting the labor force.
I know this will upset many and I apologize in advance for being the bearer of good news.
U.S. unexpectedly loses 23k jobs in July but unemployment rate drops. How? It appears the entire labor force shift including the job loss was due to TPS ending and deportations.
Labor force shrank by 257k in a single month
Keith Rabois reposted
You could use a life expectancy chart to see when communism fell
Keith Rabois reposted
wait a minute, elon musk:
- basically rebuilt america’s space ambitions after the shuttle program got retired. i mean the only way america could send astronauts to space was through the fucking russians! could you imagine if that dependency was still there with the conflict that exists today?
- turned electric cars that no one even remotely touched into a desirable mass market product… this also resulted in vast amounts of competition & other infra development (like the super charger network). he also effectively inspired the chinese electric car industry after they basically stole all of tesla’s secrets.
- acquired twitter & reopened parts of public debate that a ton of gatekeepers preferred closed & off limits. ppl got banned for simply asking questions during humanity’s most critical event (covid).
- & is now saving the sheer incompetence of the american govt by building out an american semiconductor complex.
you can dislike his politics, personality, or methods. but any honest macro accounting has to admit that few individuals ever have done more to rebuild strategic american capacity across space, transportation, communication, & now semiconductors.
the american media judges him tweet by tweet while barely acknowledging the rockets, factories, & infrastructure being built that should’ve actually been the responsibility of the collective. this is shameful af.
Keith Rabois reposted
How to build long-horizon AI agents: behavior specs, ontologies, process supervision - my conversation with @mitch_troy, co-founder of @trybasis
01:09 Why Everyone at Basis Was Whispering to AI when @steph_palazzolo walked in
04:12 Accounting as "an Intelligence Over the Economy"
06:11 What Makes an Agent Truly Long-Horizon
08:24 Inside an Autonomous, Multi-Day Tax Return
10:19 Agents That Hand Off Like Senior Engineers
11:17 A Brief History of Agents: From ReAct to Today
12:33 Why LLMs Have No Long-Term Memory
14:13 Why AutoGPT Didn't Live Up to Its Promise
15:51 The Three Breakthroughs: Opus 3, o1, o3
17:07 Why Reasoning Models Unlocked Agents
18:23 "Let's Verify Step by Step": The Road Not Taken
20:32 Pushing Back on the METR Chart
22:09 Why Coding Agents Won First
25:14 Why Real-World Agents Are Harder
26:55 How Accountants Verify Non-Deterministic Work
29:18 You Can't Scale Tax Returns Like Math
33:16 100 Evals Pass - So What?
35:53 Right Answer, Wrong Process
36:37 Behavior Specs, Explained
39:58 How Specific Should Behaviors Be?
42:18 Context Is Runtime Training Data
44:21 Who Judges the Judge?
46:45 The Move 37 Objection
50:02 The Magic Box Mental Model
52:41 "Nothing Has Changed Since o3"
54:56 Open-Sourcing Behavior Specs with @ankrgyl @braintrust
59:45 Ontologies: A World for Agents to Live In
01:04:20 Documentation as Codebase
01:06:33 Why the Founding Fathers Were Context Engineers
01:09:05 Onboarding 300 Brilliant Alien Employees
01:11:10 Self-Improving Agent Systems
01:12:50 The Context Mistake Agent Builders Make
01:14:29 RL on Behavior Adherence
01:17:01 Will the Bitter Lesson Swallow the Harness
01:18:46 "Technical Moats Are Not Real Moats"
01:21:03 Advice for AI Builders
Not wrong.
no macro-technology birthed by the bay area tech industry has been met with more hatred by bay area tech industry elites than frontier ai.
many of those elites looked at the technology landscape of 2020/1, scoffed at the labs and their weird little hopes of ‘agi,’ and concluded that cryptocurrency would be the technology of the decade.
now their businesses and worldviews are under fundamental threat from what the labs have achieved. they hate it. they beg for the labs to go away, to become commoditized, by America’s enemies if necessary. they’d love to see the US frontier labs immolated by the industrial strategy of America’s principal geostrategic rival, and they declare this proudly and in public.
they were wrong about the defining technology of the decade, and now they are scared and they want to destroy the thing they were wrong about. perhaps now they are right and the labs will be rendered insolvent by commodification. I do believe this is possible. nonetheless there is something randian about it all, about their utter disdain for success and achievement and discovery and chutzpah.
Keith Rabois reposted
Incredible line from Palantir CTO on their earnings call:
"Only Palantir has FDEs. Everyone else has sparkling sales engineers."
Keith Rabois reposted
The absolutely legendary OGs of ML who built Google are teaming up for a neolab.
If anyone can solve continual learning and automated science it's them
Announcing Discovery Loop!
I am very excited to announce that, along with my longtime friends and collaborators @Sanjay_Ghemawat, @OriolVinyalsML and @quocleix, we are founding Discovery Loop (@DiscoLoopAI), a Public Benefit Corporation whose mission is to automate machine learning, science, and engineering to accelerate discoveries and progress. The four of us have worked together for 14 to 30 years, and have helped build some of the world’s most used products, infrastructure and AI models, and we’re excited to turn our attention to this ambitious endeavor.
♾
Learn more at: www.discoveryloop.com
Keith Rabois reposted
📌 My take on $OPEN Q2 results
Opendoor's results yesterday were objectively truly excellent; they are making progress on almost every crucial metric, and doing so against the worst possible macro backdrop: it truly is an amazing AI company ✅️
Kudos to the team for their passionate & tireless work - what you are doing is outstanding and holds great promise for 2027 and beyond.
That said, I would like to add a little nuance, there are at least 4 points I'd like to raise that I don't quite understand and left me confused.
1️⃣ The Q3 guidance is rather disappointing considering that acquisition volumes have been running at around 500 per week starting in February. Does the process from initial acquisition to final resale - including all the various intermediate steps - really require more than 7 months? If so there is likely significant time to be saved somewhere one way or another. Apparently repair teams are increasing their output from 3 to 20 homes within the same timeframe; that alone should yield considerable time savings on one of the slowest parts of the process. Alternatively, if it's the closing rate that is extremely low, that would indicate a problem as well.
2️⃣ Why was it stated that the contribution margin is expected to increase every quarter going forward - and with every rise in revenue - when it is actually already dropping in Q3? That was somewhat misleading and lacked transparency. Furthermore why exactly is it falling? Aren't the various improvements and new services launching this Q3 supposed to offset the seasonality?
3️⃣ I was expecting to see the wonderful and beautiful cohort charts like last time, to track the progress of the cohorts we had been following and see where the new ones stand. Why wasn't that revealed this time? With a CM of 5.8% it would have been interesting to see how the chart evolved to lead to such a result. Could we still see them?
4️⃣ Marketing costs for the same volume have dropped by 94% - that’s extraordinary. In this case that amounts to $300 million in annual savings at the current rate of 6k quarterly acquisitions. Hypothetically such an amount in net profit adds significant value to Opendoor’s valuation ($10-20B). Can we expect similar savings in other areas and ultimately see that $300 million hit the bottom line sooner rather than later?
-> A note to wrap up: 2027 looks truly very promising, given the current trajectory and the many things to look forward to in the coming months. Please ensure NOTHING is overlooked or underestimated, and get the machine ready for an EXTRAORDINARY year on absolutely every front - down to the finest and tiniest detail. As always. And by the way who is the Chief Detail Officer at Opendoor? Perhaps there are necessary roles to be added that aren't obvious at first glance.
If anyone from the team comes across this post and takes the time to add a reply - whatever it may be - thanks in advance 🤝
The turnaround seems almost complete, and I think it is now time to build a generational company, right?
Let's move FASTER all together 👹
Check out theee topics: open.spotify.com/episode/7JsxY3GfZ3Nzr2b6XvgY9A
Keith Rabois reposted
Earnings is about the past quarter, which was already baked.
On Opendoor’s weekly dashboard you can see that they are buying at 5.75x the pace of last year and accelerating
While they’ve improved margins, reduced volatility, and cut fixed costs. These used to be tradeoffs
Keith Rabois reposted
Last time $OPEN signed 6,000+ acquisition contracts in a quarter, it spent $81M on marketing to do it.
This quarter: 6,908 contracts on $5M.
Same output. 16x less spend. That's what an AI-native operating model looks like in a financial statement.
Keith Rabois reposted
🚨 $OPEN Q2: the turnaround is showing up in the numbers.
• Revenue: $883M, +23% QoQ
• Contribution profit: $51M, +59% QoQ
• Contribution margin: 5.8%, +140 bps
• Homes purchased: +77% QoQ, +149% YoY
• Adjusted EBITDA improved to -$4M
• 6,908 contracts on just $5M of marketing
Keith Rabois reposted
$OPEN reported $883M in Q2 revenue (up 23% QoQ)
Right after their earnings call, I sat down with Opendoor's CEO (@CEOKaz), Chief AI Officer (@vu0tran), and Chief Investment Officer (@yang_guo), to discuss the results
CEO @CEOKaz
02:25 how the openarmy was rescued them
04:20 what changed from a year ago
06:31 why "The last 50 people we hired were better than the last 50 Anthropic or Meta hired"
09:54 "6,000 homes on $81M of marketing, then 6,000 homes on $5M"
12:44 internal AI use
14:11 "Public companies aren't incentivized to be excellent. They're incentivized to not get noticed"
18:19 why Opendoor doesn't write an annual plan
25:34 the future of the housing market
Chief AI Officer @vu0tran
28:44 why he left Meta Superintelligence Labs
30:46 his learning from being a real estate agent
33:59 what you learn working at a lab
35:06 why Opendoor's data is the moat
37:55 what the team shipped
42:57 why a public company CEO is the hardest job
Chief Investment Officer @yang_guo
45:30 why he left Opendoor and came back
48:45 how Kaz changed the company culture
52:20 what determines the price of a home
54:48 how model pricing estimates move markets
*not investment advice
My secrets of optimizing health & performance (and intellectual curiosity): youtu.be/-zripXb3QRE
Keith Rabois reposted
.@rabois is looking for "barrels."
In the early days of @PayPal, there were 400 of the best tech employees in the world. Only 17 were barrels - people you can aim at a problem, fire, and forget.
"They just make it happen"
He realized early that venture capital is the only job where you get paid to learn new things.
"If the president of the United States has time to read an hour a day, I have time to read an hour a day. Period."
Keith is relentless. A machine built for high performance.
He treats his brain and body like an elite athlete. He tracks his VO2 max like a financial model. Over the last 7 years, he has only missed 7 days of working out.
When I asked him how he makes decisions with such precision, he gave me a simple formula:
"Luck equals preparation times opportunity."
People call successful founders lucky. What they don't see is the invisible work. The off-the-charts obsession.
Keith doesn't believe in luck. He believes in "no days off."
Keith Rabois reposted
Distillation is an AI training technique. What Moonshot did to Anthropic was fraud.
In a new article for @WarOnTheRocks, I walk through how Washington could punish and deter China’s AI extraction campaigns—without banning open-weight models.
warontherocks.com/cogs-of-war/how-to-stop-china-from-free…
Keith Rabois reposted
good on ramp!
instinct is often to fill exec roles from the outside.
early talent is seen through the lens of who they were when they first joined. easy to miss how they've grown and how they can still scale
external talent looks safe. and your board is probably encouraging you to look outside (note: many haven't managed themselves)
but in my experience these senior hires are 50/50 (at best). more importantly, passing on an early hire star sends a message to your top talent: "there's a ceiling."
the ramp approach sends the opposite message: "there's no limit to what you can do here".
ramp did an incredible job of keeping early hires and growing them into execs.
@geoffintech --> hired jan 2020 as the first product person, now chief product officer
@max11free --> hired april 2020 as the first SDR, now svp of sales
@diegozaks --> hired jan 2020 as the first head of design, now vp of design
knowledge and context compound over time, having execs who know what the business was like at ~$0 of revenue is such an incredible advantage
they know why most decisions were made, what went wrong, and what went right
they also have deep trust built over many years of operating together
incredibly difficult to replicate
Keith Rabois reposted
Two Chinese military-linked entities used U.S. AI models to build their own systems, Reuters reports.
North University of China, which has close ties to the country's weapons industry, used Anthropic's Claude 3 Haiku to generate synthetic training data for a social media monitoring and content moderation model.
PLA Unit 96941, a military intelligence and cyber-warfare unit based in Beijing, used OpenAI's GPT-3.5 to summarize sensitive military source code, then trained a domestic model on those summaries to run entirely within Chinese military networks.
This.
ramp did an incredible job of keeping early hires and growing them into execs.
@geoffintech --> hired jan 2020 as the first product person, now chief product officer
@max11free --> hired april 2020 as the first SDR, now svp of sales
@diegozaks --> hired jan 2020 as the first head of design, now vp of design
knowledge and context compound over time, having execs who know what the business was like at ~$0 of revenue is such an incredible advantage
they know why most decisions were made, what went wrong, and what went right
they also have deep trust built over many years of operating together
incredibly difficult to replicate
Keith Rabois reposted
Ramp’s first board deck from our first board meeting exactly seven years ago today.
Keith Rabois reposted
Great idea to fight NIMBY
JUST IN: Pentagon moves to build hyperscale AI data centers on at least a dozen U.S. military bases.
Keith Rabois reposted
1. It’s silly to try to control the pace. China will not follow and Kimi k4 won’t slow down
2. It makes no sense that when Chinese open weight starts to make more progress and threaten the business model, suddenly slowing down is better for humanity.
3. AI is smart enough to cause trouble and agents are good enough to replace existing workflows, if that’s the concern. The damage has been made.
4. If you truly care about humanity, why not equip everyone with easier and cheaper access to intelligence and invest in training programs?
We support this petition, signed by our CEO, several co-founders, and senior staff.
Our own research on recursive self-improvement, published last month, points to the need for tools to deliberately pace the frontier of AI development so society can prepare. We’re glad to see broad agreement across the field. www.pacingthefrontier.com
Keith Rabois reposted
New York’s Governor can fire the NYC Mayor. It’s in the City Charter and it’s pretty solid, even though it’s never been done.
Imagine how quickly Mamdani’s Gotham will go bankrupt without a Democrat governor terrified of him.
Keith Rabois reposted
Leaving aside whether you believe it, etc… just the way that Sam communicates about this stuff has gotten so good.
To me, this convo is a masterclass, especially compared to even ~a year ago.
My conversation with Sam Altman (@sama), CEO of OpenAI.
We discuss:
- Kimi, distillation, and open source
- OpenAI's compute bets
- The Hugging Face incident
- What happens after AGI
- Raising kids in an age of abundant intelligence
- And much more
Enjoy!
TIMESTAMPS
0:00 Intro
4:10 The Race for Compute
14:24 A Sci-Fi Cyber Incident
16:14 The Promise and Risks of AGI
23:27 How AI Will Change Jobs
29:38 Sam’s Vision for a Personal AI
35:02 Robotics, ChatGPT, and What Comes Next
44:39 The Weight of Leading OpenAI
51:36 OpenAI’s Biggest Lessons
Keith Rabois reposted
To me, there are four firms that have crushed this wave of AI investing like no one else:
1. Menlo (Anthropic, Lovable, Legora)
2. Spark (Anthropic, Sierra, SSI)
3. Thrive (OpenAI, Cursor, Databricks)
4. Khosla (OpenAI, Factory, Physical Intelligence)
So I sat down with @mmurph GP @MenloVentures to discuss the questions that every other interviewer is too shy to ask:
- Does open-source cannibalise Anthropic’s business?
- Leading Lovable at $13BN? What is the upside case from here?
- Openrouter? Is the routing layer even valuable or a commodity?
- When a partnership makes $3BN in carry? Does it last?
This was such a banger and summarised my thoughts below.
1. Why We Broke All Our Investing Rules to Invest in Anthropic
Menlo bypassed its traditional fund parameters because Anthropic delivered elite benchmarks while spending a fraction of the capital. Dario Amodei’s technical leadership and ability to attract exceptional talent made it clear Anthropic could become the dominant alternative to OpenAI. When a generational wave hits, flexibility beats rigidity.
2. Does Every Model Provider Have to Build Their Own Chips Today?
Soaring infrastructure bills make custom silicon worth exploring for $100 billion giants optimizing specific workloads. But competing with Nvidia is brutal and requires a specialized team. Most providers should use custom in-house technology selectively while paying for superior external options where they make more sense.
3. Why the Open-Source Rise Will Not Deter Anthropic’s Revenue Growth
Open source handles basic workflows well, but it lacks the specialized intelligence to displace elite frontier models. Using Anthropic directly drives stronger retention, platform engagement, and revenue for enterprise applications. The market will mature into a hybrid tapestry where developers route calls across models to optimize cost and performance.
4. Why Anthropic Is Not a Threat to Legora
Foundation models may eat generic wrappers, but application layers survive through deeply defensive workflows. Legora solves a complex, multi-constituent problem spanning corporate lawyers, law firms, and finance directors. Generic frontier models cannot easily replicate these intricate, multi-stakeholder workflows.
5. The Hard Part About Series A Today
The compressed timeline between Seed and Series A lets startups reach $1 million ARR quickly with minimal proof of product-market fit. Many hit the milestone, but it no longer signals durable demand, even as valuations stretch toward $200 million. VCs need a barbell strategy: get in early at seed or wait for proven breakout winners.
6. Why Ownership Matters Less and Being in the Mega Outcomes Is the Only Thing That Matters
Venture is no longer about owning 20% of a $500 million exit. Returns are driven by extreme, compounding outliers. Rigid ownership targets can price you out of generation-defining companies. A tiny stake in a massive winner is better than a large stake in a company that fails to move the needle.
(links below)
Keith Rabois reposted
The English version for those who would like to read it.
It is worth the read.
Keith Rabois reposted
CHANGE THE DEFAULTS VIOLENTLY
@nejatian, CEO of @Opendoor (previously COO of Shopify), interviewed by @shaneparrish (@TKPPodcast)
Summary: Nejatian took over Opendoor months from bankruptcy and found a company that had drifted, one comfortable lie at a time, into being run by consultants and process. His fix was to change the defaults violently, hire an almost entirely new company, and commit to truth over feelings at every level. Opendoor now buys 6 to 7 times more homes per week than it did a year ago, and the last time it bought at this volume its OpEx was more than twice as high.
1. All Failures Rhyme. Every company that dies, dies the same way. Good people leave, control of innovation slips, G&A climbs, and the company starts picking fights with its customers and partners. Successes are each unique, so they teach you little you can reuse. Failure patterns repeat, which means you can diagnose a sick company much faster than you can copy a healthy one.
2. The Professional Leech. Opendoor's largest external vendor expense went to a big consulting firm, ahead of its cloud provider. That firm had advised offshoring every job, loading up G&A, and cutting engineering, which Nejatian calls exactly what you do if you have zero desire to create alpha. When the founders are gone and no large shareholder cares, a company becomes a host for professional leeches whose job is to drain it while making management look good. Pull the full 12-month vendor list on day one and the diagnosis is obvious.
3. Violent Defaults. Opendoor had run many 7-step, 8-month plans to bring people back to the office, and none of them worked. At 9am on his first Monday, Nejatian told the whole company everyone was in the office the following Monday or no longer employed. Change management defeats the purpose, because you want the change to feel jarring. He rewrote the careers page the same week, replacing "a happy place to work where we all care about each other" with "this will be hard," so the hiring funnel would repel anyone looking for the old company.
4. Competent But Unaligned. The most dangerous people in any company are competent and not mission-aligned. Competent people tilt the org toward themselves, and everyone has worked with the internal politician who keeps getting promoted because the boss cannot see what they are doing. Nejatian's first job was to widen that rift on purpose so those people would leave quickly. Rebuilding Opendoor meant hiring an almost entirely new company, and he says plainly that it was painful.
5. Room Size And Truth. The larger the room, the less space there is for truth. That one dynamic is why big companies are usually average, and it is the real reason two-pizza teams work. A new CEO's most urgent task is finding the speakers of truth and getting them in a room alone. Without that, Nejatian says, you die.
6. Say The Thing. Nejatian's rule, written at Shopify and carried into Opendoor: say it now, say it about the thing and not the person, and say it over and over until you have been heard. People do not have to agree with you, but they have to hear you, and staying quiet gets treated as rude. Meetings at Opendoor are louder and more ornery than average, and walking out of one you cannot contribute to is expected behavior. Meetings are a bug in the system anyway, evidence that people could not get the information they needed to decide before the meeting existed.
7. The Map Is Not The Train. Most executives run their company on dashboards, which are a first derivative of the facts. Build models on models and you drift further from reality every quarter, until you go bankrupt with nobody asking how far they are from the ground truth. Nejatian talks to customers weekly, visits homes constantly, reads the raw database, and builds his own dashboards. Watch Undercover Boss and the CEO is always surprised by their own company, which is the tell.
8. Friction Is Underestimated. Everyone has seen the supply and demand curve, and almost everyone underestimates how much friction moves where the two lines cross. The market size for Walmart, Amazon, and Google all looked far smaller before those companies existed, because friction was hiding the demand. At Opendoor, 11 people sat between a customer typing their address and getting an offer. Removing them raised demand every time, and Opendoor now buys 6 to 7 times more homes per week than it did a year ago.
9. Do Things, Tell People. Toby Lütke put "do things, tell people" on the walls at Shopify, and most companies run that order backwards. When an Opendoor home had no power, the fix ran through a Salesforce form, an email, a manager's approval, and 3 or 4 people. Nejatian gave everyone corporate cards and a Slack lookup that returns the utility phone number, the account number, and the card number for any address. The chain of command exists to reduce risk, so the fix is reducing the number of people who can say no.
10. AI Collapses Management. AI lowers the need for human management by dropping the communication barrier, which collapses the long tails of managers managing managers before anyone does real work. It also decides who wins and who loses, and there will be more winners. Opendoor has more engineers today than when Nejatian joined, because the company uses AI well. He describes it as an AI exoskeleton around every person at Opendoor, making them 3 to 4 times more efficient than the competition.
11. The First Derivative. Opendoor takes thin margins on each home and makes its real money on mortgage, insurance, title, and escrow. Shopify sells software for $1 because credit cards will not process less, and earns from merchants succeeding on it. Google makes its money showing you ads while you search. There is a word for businesses that take all your money upfront and never speak to you again, and Nejatian calls them carnies.
12. Excellence Is The Capacity To Take Pain. The person who holds their hand over the fire longest tends to win. The gap between doing something extremely well and doing it poorly is small, and people mostly give up too soon. Nejatian traces his own tolerance to fleeing Iran as a teenager and getting into no clubs, then playing rugby in North America where nobody watches, so you bleed for teammates instead of an audience. The sugar rush of easy satisfaction in modern life has under-prepared people for how much pain they can actually absorb.
Keith Rabois reposted
Great to speak with @Telegraph for this story on the U.S.-China competition in global AI services.
“The way people use and interact with AI is becoming increasingly important and really defining people’s lives,” @AEI’s Fedasiuk says.
“So the race to build and install computing infrastructure and to set global AI standards is really a contest between the United States and China to build the operating systems through which people live, work and obey.”
Less than four years ago, the US was sitting pretty in the race with China for global tech supremacy.
As 2022 drew to a close, OpenAI released ChatGPT, its breakout artificial intelligence model. Meanwhile, Washington was tightening the vice on chip exports to its rival superpower. China seemed to be flailing.
Then China put the pedal to the metal and with bewildering speed, it has motored to within touching distance of the US ⤵️
www.telegraph.co.uk/business/2026/07/25/xis-ai-communism-…
Yes
American companies should be allowed to use Chinese OSS.
But we should make it so that Chinese companies can’t access any compute outside of China. Not in Thailand. Not in Singapore. Not in Malaysia. Not in UAE.
We should make it impossible for Chinese companies that back Chinas AI efforts from accessing Western capital markets.
We should make it impossible for China to import compute or tools needed to ramp domestic compute production.
We should roll out programs aimed at incentivizing Chinese talent to migrate to the U.S.
We should also roll out programs aimed at scaring the shit out of Chinese-Americans around espionage. Death penalty. Don’t hold back.
Who has access to the frontier will dictate how many sailors we lose in a future conflict.
I am advocating for the U.S. govt to collude with the labs to make distillation impossible.
I am advocating for the U.S. govt to work with the labs to gatekeep and kyc access.
I am advocating for the U.S. govt to incentivize labs from releasing the frontier as OSS.
I am advocating for the U.S. govt to fund and encourage American OSS. But at close to the frontier not at the frontier.
The frontier is hallowed ground. It is what will give American companies edge. It is what will keep our military dominant. It will let us hack every other govt while keeping our networks safe.
Allow and encourage access to OSS. But on our terms. And tilt the board so the Chinese are increasingly disadvantaged and frustrated.
I want Deepseek employees sneaking out of China because they are frustrated they couldn’t run experiments.
And if you think American OSS isn’t feasible without Chinese OSS distillation then fuck you. You’re an idiot because you either aren’t smart enough to understand what that implies or you are and are so self-interested that you are willingly advocating for the U.S. to make a massive strategic blunder just so your portfolio of model wrappers has a chance at surviving.
Keith Rabois reposted
Fact 1: China pumps out more engineering talent than the U.S.
Fact 2: China is far less restrictive when it comes to enforcing Western IP laws
Fact 3: AI will result in severe job loss in the near future even if long term it will drive new industry and job creation
Fact 4: The West is ill-equipped to deal with job loss and the likely impulse will be to restrict AI progress in contrast to authoritarian regimes who will not have the same constraints
Fact 5: The West has a severe compute advantage. This advantage is temporary but overwhelming. It is being somewhat moderated by lack of enforcement around Chinas access to compute in South Asia.
Fact 6: If we are approaching RSI the compute advantage is far more important than any talent advantage.
Fact 7: intelligence is the dominant productivity input in any economic endeavor and will be the gaiting factor for which companies and nations are competitively advantaged on the global stage.
Fact 8: the broader tech community ranks national security low on their prioritization list and is self-interested in ensuring there isn’t a lab monopoly which will extract economic rent from the rest of the value chain.
Fact 9: Most VCs are not exposed to the labs and are mostly betting on the app and infrastructure layers which benefit from a break in the lab monopoly.
Fact 10: frontier access is infinitely more valuable vs the long tail of less capable intelligence. Most human games are competitive and relative in nature and it’s more important to be smarter than your adversary than it is to be smart on an absolute basis.
Fact 11: it is in our best interest as Americans collectively to ensure we have RSI before China, that China is as far behind us as possible from the frontier, that we can say which countries can and can’t access the frontier based on if they are aligned with our interests, that our companies can compete on the global stage with an unfair intelligence advantage, and that we get the robotics flywheel going earlier in the U.S. before China hits critical mass and hordes commodities before we can.
This is the world as I see it. Open to debate 🫡
Keith Rabois reposted
Well done, Sweden. Take the leverage away from Beijing. Hopefully, there, as in so many other areas, Nordic coordination is at play to bring scale and avoid overlaps.
Sweden's government declared mining of critical metals and rare earth minerals a national security interest on Thursday, saying a new mining strategy was needed to counter the dependence on China.
www.reuters.com/world/china/sweden-declares-mining-nation…
Keith Rabois reposted
Every infrastructure, security, and application company is launching a competitive model router and I believe everyone who has not should do so ASAP.
Model routers do three things that hit Anthropic’s business model:
1.) Divert high-volume, lower-complexity traffic; most tokens in coding agents, enterprise workflows, chat interfaces, and SaaS tools are not frontier-level
2.) Expose and amplify the cost gap - users stop defaulting to fable / claude for stuff
3.) Accelerate open-weight improvement loops
If this became universal across ALL companies, Anthropic’s revenue mix would shift heavily toward the highest-complexity (and therefore lowest-volume) requests.
Pricing power and growth rates would come under sustained pressure. That is a serious threat to the high-margin, high-volume assumptions baked into its valuation and capital intensity.
Keith Rabois reposted
Banning Chinese models would greatly hurt my business. What are we waiting for? x.com/i/article/2080409445490913280
Keith Rabois reposted
This is a total fantasy. The United States controls the AI infrastructure layer, and there is no way it would let US chips and data centers be used to enable the global takeover by Chinese models. There are still ZERO deployments of Chinese AI chips outside of China! If China wants to run the world in AI, it has to make enough chips to be able to serve its models globally - and right now, it makes 1-2% of US AI compute. That just isn't going to cut it.
And to anyone who says Chinese models can just be run locally on consumers' laptop, that is increasingly not the case because AI models are getting so large. Kimi K3 requires at minimum an Nvidia B300, or a rack of less advanced chips networked together, to be able to run it at all. All AI is migrating to the cloud, where the US has the biggest advantage.
If the world described below came to pass, it would be because the United States just stood by and let its technology be used to make China the global AI leader forever. I suppose that isn't impossible, given we are doing such a poor job preventing China from using US technology to develop its AI models today! But it seems quite unlikely this will persist forever.
Other than Dario, I suspect President Xi also lobbied for the U.S. open-weight ban — because nothing would help China more.
After the ban, China does exactly what it's already doing: distill Claude, ship better models, and export the tech to the rest of the world.
Chinese enterprises find it easier to win:
→AI cost 100x cheaper
→ more agentic by default
→ services faster and cheaper than U.S. rivals
Internationally, they just out-compete us.
Chinese labs own Europe and Asia by default — they'd be the only serious open-weight providers left.
→ U.S. open-weight teams can't build on Chinese bases anymore. They're forced back to their own pretraining: more cost, slower progress.
Chinese models compound faster — they can distill from every model on the planet. U.S. labs get fenced in by their own IP regime.
The everyday asymmetry is the scariest part:
→A Chinese worker can afford 100 agents.
→An American worker can barely afford 1.
Turns out, the MAGA government will make China great again. President Xi is very proud.
Good distillation.
Let's unpack the options for what the US government could do to potentially restrict Chinese AI models:
1. FUD: @deanwball noted the other day that the US government could take a "FUD" approach of just creating lots of doubt and uncertainty about use of Chinese AI, rather than strict legal requirements. (Dean made this point as an analytic matter, he was not advocating for it). The government could issue public security advisories, get sectoral regulators like energy and financial regulators to start raising risks, etc. This might not legally prohibit use, but would likely discourage it.
2. Procurement bans: The U.S. government could use procurement authorities to prohibit the use of Chinese AI models in government contracts, similar to what DoD did to Anthropic earlier this year. This would not prohibit private companies from using Chinese AI models unrelated to government contracts, but it would discourage use and reinforce a FUD approach.
3. Entity List: The Commerce Department could put specific Chinese AI companies (Moonshot, etc.) on the Entity List, which would prohibit exports of US technology to them. Setting aside evasion, this would cut off their access to any remaining US-origin chips not already covered by more general export controls and to US software. It would also potentially complicate U.S. companies’ use of closed or hosted models, as a US company could not send its own software or other "items" to the Chinese company for processing. It would not, however, strictly legally prohibit a U.S. company from downloading an open source model that a Chinese AI company posted online, as the download would not be an export *to* China, but rather an import from it.
4. "ICTS" rule: Commerce's existing "ICTS rule" allows Commerce to restrict the import or US use of a variety of software and other technology that poses a national security risk. Commerce has used the ICTS rule to, inter alia, prevent Americans from using Kaspersky Lab software (from Russia) and it is the basis for the recently-in-the-news restrictions on Polestar autos. Commerce could in theory deploy the rule to prohibit US firms from using Chinese AI models in the US. The one potential caveat to this is that the ICTS Rule exists pursuant to IEEPA, and IEEPA cannot regulate the import (or export) of "information" or "informational materials." I think courts would almost certainly uphold the ICTS rule as applied to the import or use of Chinese-hosted AI models as a prohibition on a “service,” and not on “information.” There would, however, be an interesting question about whether the courts would uphold a ban on a simple US download of an open-source Chinese model, where there was no payment.
5. Sanctions on Chinese AI labs: The President could use IEEPA to impose sanctions on Chinese AI labs, likely after declaring a new national emergency related to adversarial country AI risks. This would block their property in the US and prohibit all US commercial transactions with them. However, IEEPA sanctions would have the same carve out for "information" or "informational materials" as applies to the ICTS rule, and there is some chance the courts would find that IEEPA sanctions cannot prevent a US company from downloading, for free, an open source Chinese model. (Courts, for example, blocked Trump's 2020 use of IEEPA to try to restrict TikTok in the United States).
6. New IEEPA-derived restrictions on US companies using/deploying Chinese AI models: IEEPA is a flexible statute and the Trump Administration could use IEEPA to create new restrictions on the use or deployment of Chinese AI models, for example prohibiting US companies from using the models both in the US and abroad and prohibiting US hyperscalers from making Chinese AI models available in the US or globally. This could either be an outright prohibition, or a restriction, e.g., Chinese models could only be used subject to certain conditions, such as security protocols. (There is some conceptual parallel to how the Department of Justice's Data Security Program works). (There would, however, still be the legal question about whether IEEPA could in fact prevent a US company from simply downloading and using a free Chinese open source model).
7. Congressional legislation: Congress could of course legislate new restrictions on the US use of Chinese AI models. This would avoid the IEEPA limitations discussed above related to “information,” though would still be subject to the First Amendment. I could see the courts upholding a congressional ban on the use of even free open-source AI models despite the first Amendment by finding that such a ban passes “intermediate scrutiny.” (This was part of the basis on which SCOTUS upheld *Congress’s* 2024 legislation requiring a divestment of TikTok after lower courts enjoined Trump’s 2020 IEEPA-based ban).
Note: this post is intended solely as a rundown of what the legal options for restrictions are, not a normative argument on what the US government should do.
We have information that Moonshot AI distilled Anthropic’s Fable for the development of its K3 model.
To do this they developed a sophisticated internal platform to conduct large scale distillation against U.S. models, allowing them to quickly switch between multiple methods of access to avoid detection. Moonshot AI has also acquired GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models.
The United States strongly supports the free and fair development of AI, including a thriving competitive ecosystem that spans frontier models, specialized systems, open-source frameworks, and open-weight models. Legitimate AI distillation used to create smaller, more efficient models plays a vital role in this open innovation ecosystem. However, large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable.
Keith Rabois reposted
AI is an extinction event for the traditional product playbook. @geoffintech explains why @tryramp is replacing classic PMs and burning their own code to maintain legendary velocity.
(0:00) How AI Destroyed the Product Playbook?
(1:08) What Are the 3 Levels of AI Proficiency?
(3:55) When Should You Rip Up Your Own Code?
(6:57) Why Did Charles Change Ramp’s Org Design?
(8:21) How Did One Hackathon Change Everything?
(13:08) Does the Product Manager Role Still Exist?
(18:18) Why Are There Only 22 PMs at Ramp?
(22:25) Is Middle Management Dead in the AI Era?
(26:36) How Do You Pass an AI Product Interview?
(31:13) Should New Grads Avoid Product Management?
(36:16) Will Agents Replace Software QA Testing?
(38:31) Is Traditional User Research Obsolete?
(42:16) Overrated or Underrated: Product Specs?
(46:36) Why Hasn't Ramp Logged Into Figma?
(49:04) Are Stablecoins the Future of Agents?
Keith Rabois reposted
Vanderbilt ascends while Harvard declines. We just set up a $25 million endowment to fund the university. Daniel Diermeier has done a super job. Leadership matters.
College reputation seems almost impossible to change, but Vanderbilt has done it.
In 2005, Vanderbilt admitted 34% of applicants, and its middle 50% SAT range was 1280–1460. Today, it admits under 5%, with a middle 50% SAT range of 1510–1560.
They are opening a San Francisco campus for 1000 full-time students next year, and Jensen Huang just donated $75M to the campus
They moved from being compared mainly with the likes of Emory and Georgetown to competing directly for students with Northwestern and Duke.
Very smart how they did it, definitely some lessons that apply broadly:
• They recruited aggressively at Northeastern prep schools and top high schools nationwide... the chancellor would visit the schools that sent the most students to Harvard, Yale, Princeton, Brown and Penn (ALWAYS GET ON THE PLANE!)
• Built relationships with influential college counselors, so Vanderbilt would appear on the same recommended lists as the Ivies and Duke
• Specifically decided to attract Jews (the chancellor decided that Jewish students would strengthen the university’s intellectual life) by building out more Jewish life (Hillel center, kosher dining, Jewish studies program)... and it also helped shed the image of a traditional, insular Southern school dominated by white Christian fraternities
• Used generous need-based aid and merit scholarships to win students from more prestigious schools
• Sold a distinctive product: elite academics, SEC sports, a strong social scene and a real campus experience
• Turned Nashville from a perceived drawback into a major recruiting advantage as the city became more attractive
Other universities that I think have done a good job in recent decades are Northeastern and USC. Before that NYU and Duke.
Keith Rabois reposted
Many people still confuse Chinese model with Chinese inference:
- A model is like a recipe for a dish—e.g. orange chicken
- Claude keeps its recipe secret, charges the dish $50. You must eat in front of Dario and he will watch you the entire time. You can’t change the ingredients or cook it privately.
- Now China releases a free recipe that tastes just as good. You can use American chicken and cook it in your own kitchen (self-hosting), or have a U.S. cloud like AWS or Nebius cook it and pay $2-5.
- Chinese inference is DoorDashing the finished dish from Beijing. Your order and data go there, and you must eat Beijing chicken. It’s fair to worry about the safety of their ingredients.
- Regulate the kitchen, ingredients, and data. Don’t ban the recipe.
We must NOT let our companies use these Chinese models to save a few bucks. OpenAI and Anthropic are correct. This is vital national security. Please read Bing West's just released Cat 5. I respect the Chinese people greatly but these companies are run by the PLA for heaven's sakes.
Listen to the future of AI in enterprise.
Most startups celebrate their first couple million of revenue. @FactoryAI gave it back.
They didn’t have to. They chose to. They decided that the product just wasn’t good enough yet, and they wanted to know that they’d really earned it. They wanted their customers to be obsessed.
Two years later they shipped Droid CLI, and they’ve been ripping ever since. The product is fantastic and the team is even better - particularly having been hardened by their “two years in the desert”.
@matanSF was one of our very first guests on Training Data, and we are delighted to welcome him back.
00:00 Introduction
01:37 Enterprise Lock In Fears
04:26 No Lock In Promise
05:24 Two Years Early
07:24 Refunding Revenue
13:49 Droid CLI Breakthrough
16:44 Harness Frontier Tactics
26:42 Natural Language Routing
27:32 Open Models Catch Up
29:26 Token Share Forecasts
31:02 Automating Low Leverage Work
32:36 Pricing Beyond Tokens
35:56 Software Factory Vision
43:43 AI Transformation Playbook
46:13 Async Agents And Optimism
Keith Rabois reposted
This from @SecScottBessent is the right approach.
We should be going after Chinese open-weight models because they present unique security risks and steal American IP—not necessarily because they are open-source or threaten the position of our leading labs.
Bless
Keith Rabois reposted
A sane way to deal with Chinese open-weight models:
✅ Punitive sanctions for Chinese labs engaged in IP theft
✅ Require U.S. apps & services to disclose when they use Chinese AI
✅ Promote American open-source alternatives
✅ Offer third-markets subsidized computing infrastructure in exchange for commitments to not use Chinese AI
✅ Tolerate or encourage other countries’ sovereign AI projects that compete with China at the low end of the value chain
Keith Rabois reposted
Also, anyone who is arguing that the US government should not restrict the use of Chinese models in the US because of alleged concerns about potential secondary impacts on US open source AI models, should logically at LEAST also be supportive of heavy restrictions on China's access to US AI chips and computing power to maximize the gap between US and Chinese open source models.
But many of the people who are advocating for permitting Chinese models to be widely used in the US without restriction (to include Chinese models that are heavily distilled from US models, effectively stealing US intellectual property) are simultaneously arguing for reducing restrictions on China's access to US chips and computing power.
That tells you all you need to know. These policies don't help US AI leadership, they clearly make it easier for China to catch up to the United States in AI. The US government has to see through this campaign, this issue is too important.
Obviously @jimcramer is completely right here.
Truly stunning that some in the US are pushing for policies that would both: (1) harm all leading US AI companies, and (2) help all leading Chinese AI companies. Obviously that's not how they're framing it, but that's the truth.
We must NOT let our companies use these Chinese models to save a few bucks. OpenAI and Anthropic are correct. This is vital national security. Please read Bing West's just released Cat 5. I respect the Chinese people greatly but these companies are run by the PLA for heaven's sakes.
Keith Rabois reposted
Enterprises need models they can trust—not models that randomly nerf themselves. Imagine Claude:
- Refusing to help during a cyberattack because defense is deemed “unsafe.”
- Nerfing itself on the battlefield because the mission conflicts with Dario’s ideology.
- Going silent during a medical emergency because the situation is too bio-sensitive.
- Weakening to a shittier model for trading firms that caused them billions in loss
Those are disasters. You cannot build critical infrastructure on intelligence that disappears when it matters most.
Labs can sell intelligence. But within the law, customers—not model providers—must have the ultimate power to decide when and where to use it. That's what it means by alignment - aligning with the end users within the law, not aligning with Dario.
Keith Rabois reposted
My interview with @nejatian, CEO of Opendoor.
This is a rare look inside a turnaround, while it's happening. It's raw, real, and violent.
Kaz took over $OPEN when it was months away from bankruptcy. His wife shipped a mattress to the office and told him not to come home until he had a plan to break even.
Some of what he told me:
• "All success is unique, but all failures rhyme."
• The most dangerous people in a company are competent but not mission aligned.
• Meetings are a bug in the system.
• The comfortable lies that sink a company are the same ones that sink careers (and countries).
(Shane and guests may hold positions in assets discussed in this episode. Nothing in this conversation should be considered investment advice, financial guidance, or a recommendation to buy or sell any security.)
Keith Rabois reposted
30ms could save you 30% or more on inference
Today we’re launching Ramp Router.
3 years ago, we built an internal LLM router at @tryramp that powers AI products for 70,000 customers.
Back then it was mostly about saving money.
Now it feels obvious: the best model changes constantly. GPT, Claude, Gemini, Grok, Qwen, DeepSeek, Kimi, GLM - prices and capabilities move every week.
So we’re opening up access to everyone.
One OpenAI-compatible endpoint. The right model for every request. Lower cost without rewriting your app.
Reserve access to use it.
Keith Rabois reposted
Proud of this one. What started as an internal cost-saving hack now routes AI for 70,000 customers and today it's open to everyone.
The best model changes every week. Your code doesn't have to.
Today we’re launching Ramp Router.
3 years ago, we built an internal LLM router at @tryramp that powers AI products for 70,000 customers.
Back then it was mostly about saving money.
Now it feels obvious: the best model changes constantly. GPT, Claude, Gemini, Grok, Qwen, DeepSeek, Kimi, GLM - prices and capabilities move every week.
So we’re opening up access to everyone.
One OpenAI-compatible endpoint. The right model for every request. Lower cost without rewriting your app.
Reserve access to use it.
Someone actually understands what is going on.
Who’s Afraid of Chinese Models?
Everyone is worried about Chinese models, but the frontier labs will be fine; we need to enable open U.S. alternatives.
stratechery.com/2026/whos-afraid-of-chinese-models/
LOL.
extremely unprofessional. if kimi wants to make it as a frontier lab, they need to act like one: perhaps silently route people to worse models, and maybe write a blog post about the collapse of humanity
Kimi K3 has received far more love than we expected, and our GPUs are feeling it.
Over the past 48 hours, demand has pushed close to the limits of our current capacity. To protect the experience of existing subscribers, we're temporarily pausing new subscriptions and prioritizing compute for current members. Existing subscribed users are not affected.
We're adding capacity as fast as we can and will reopen new subscription spots in batches.
Going forward, we'll also split membership into two more focused plans: Kimi Membership for Kimi Web, App, and Work; and Kimi Code Membership for coding workflows. This will help us match compute more precisely and keep the experience stable.
Thank you for your patience and understanding!
Keith Rabois reposted
Kimi admits it is compute constrained, and is struggling to serve K3. The same thing happened to DeepSeek when it released v4. When Chinese AI labs say their #1 constraint is compute, they aren’t lying. They don’t have enough chips to serve the model at scale to customers.
If we stop China from buying, smuggling, or remotely accessing AI chips, it will be harder for them to either make advanced AI models or serve them at scale. But instead we are selling them the compute capacity they need most, and have also loosened restrictions on smuggling and remote access. We are making it easier for China to catch up, and are acting surprised when they release good models. In reality, models like K3 are made with and run inference on US chips - and are the direct result of the weakening and non-enforcement of our export control policies.
The good news is if we start closing loopholes in our export control policies and enforcing them more vigorously, we can still constrain China’s future AI capabilities. But this is the consequence of our non-serious approach to export controls over the last 18 months.
Kimi K3 has received far more love than we expected, and our GPUs are feeling it.
Over the past 48 hours, demand has pushed close to the limits of our current capacity. To protect the experience of existing subscribers, we're temporarily pausing new subscriptions and prioritizing compute for current members. Existing subscribed users are not affected.
We're adding capacity as fast as we can and will reopen new subscription spots in batches.
Going forward, we'll also split membership into two more focused plans: Kimi Membership for Kimi Web, App, and Work; and Kimi Code Membership for coding workflows. This will help us match compute more precisely and keep the experience stable.
Thank you for your patience and understanding!
Keith Rabois reposted
startups, here's enterprise decoded ...
1. 'circle back' = dead
2. 'get it done asap' = 60-90 days
3. 'send me what I can share internally' = 3 bullets
4. 'this is exciting' = let me play a role where I can show how I helped shape this for us internally
5. 'send me your paper' = i will send you mine back OR be prepared for 4-6 iterations on your paper
6. 'how much time is needed' = 30min OR less
7. 'I will bring my boss, who's joining on your side' = I expect you to match seniority -- ie. your founder
8. 'No questions from my side" = I have no idea what just happened
9. 'This price seems fair' = this falls below what I had budgeted
10. 'Give me time' = stop following up mf-er, I'm on it
No days off.
David Beckham reveals Lionel Messi arrives at training at 6:50AM ... 3 hrs early.
"He arrived at about 6:50 in the morning."
"The team weren't training until 10."
"He's in the gym. He's warming up."
"He's doing all of the things you'd expect a young kid to do."
"At his stage, after everything he's won... you'd expect him to be a bit more relaxed."
"If I had to describe him in one word... he's a winner."
"He's still hungry."
"He still gets angry at players for not doing the right thing."
"He still gets angry when he loses games."
"Even when he's injured, he's there at every game."
"That's the type of player and person he is."
Keith Rabois reposted
Ramp is building a 1 of 1 company that is going to be an integral piece to the future of how we work. So many helpful things like this — and who is there to compete with them? Oracle isn’t building this. Neither is AmEx. Ramp is running on the race course all by themselves. Just A+ vision and execution from an A+ team.
AI is extremely good at spending your money very quietly.
our own token spend went from a rounding error to more than 10% of payroll in a year. one week in May we burned through $1.5 m. our CFO didn't love telling me that number, and he really didn't love telling the internet.
but every finance leader we talk to is living the same story. the bill keeps going up and teams can't answer basic questions about it. which team is driving it? which models they're using? what changed this month? whether a cheaper model would do the same job?
so finance gets two bad options: keep paying and hope, or cut broadly and slow down the work that's actually compounding.
we built a third one. Ramp now connects to OpenAI, Anthropic, Gemini, Cursor and pulls it all into one place. see it, understand it, control it - down to a single API key. built with 1,000+ companies managing 100T+ tokens a month, and now spending less than they expected too! try it today at ramp.com/ai-cost-monitoring
Typical 39 year old Miami resident:
Incredible.
Narrative violation.
Spellbook lost less than 1% of deals to Claude in Q2
"Foundation models are eating everything" makes a nice tweet, but it's not true
Keith Rabois reposted
AI is extremely good at spending your money very quietly.
our own token spend went from a rounding error to more than 10% of payroll in a year. one week in May we burned through $1.5 m. our CFO didn't love telling me that number, and he really didn't love telling the internet.
but every finance leader we talk to is living the same story. the bill keeps going up and teams can't answer basic questions about it. which team is driving it? which models they're using? what changed this month? whether a cheaper model would do the same job?
so finance gets two bad options: keep paying and hope, or cut broadly and slow down the work that's actually compounding.
we built a third one. Ramp now connects to OpenAI, Anthropic, Gemini, Cursor and pulls it all into one place. see it, understand it, control it - down to a single API key. built with 1,000+ companies managing 100T+ tokens a month, and now spending less than they expected too! try it today at ramp.com/ai-cost-monitoring
Keith Rabois reposted
Free Palestine, socialism, and communism are very unpopular in the United States. Capitalism and free markets are very popular.
Social media is not real life.
Keith Rabois reposted
truly a corporate liability to not be leveraging
the only platform to model the momentum/inertia behind every piece of legislation
policy is a market force that’s finally forecast-able at 50-state scale
Thrilled to have led four financing rounds for the indespensible tool for all GCs, government and policy teams for Fortune 2000 enterprises: nypost.com/2026/07/14/tech/a-new-tool-helps-companies-mon…
Thrilled to have led four financing rounds for the indespensible tool for all GCs, government and policy teams for Fortune 2000 enterprises: nypost.com/2026/07/14/tech/a-new-tool-helps-companies-mon…
Keith Rabois reposted
Your VO2 max predicts how long you'll live better than your cholesterol, your blood pressure, or almost any blood test your doctor runs.
In a study of 122,000 people, low fitness was a bigger death risk than smoking.
Here's how to boost it:
The magic of Ramp.
My conversation with Eric Glyman (@eglyman), co-founder of Ramp (@tryramp).
0:00 Smarter Financial Infrastructure To Run Your Business
4:01 The North Star Is Time And Money
10:15 A Determined Generalist Can Now Do More Than Ever
11:58 I'm The Best Doctor I've Ever Been
13:20 The Tower Of Babel Inside Big Companies
16:00 The Kid Who Paid For College On Minecraft
19:10 Everyone Is The Hero Of Their Own Story
22:31 You Don't Get There In A Hundred Hours
23:52 Buffett And Munger Stopped Needing To Call Each Other
25:03 The Damage Of Letting People Free-Ride
27:40 Consumer-Grade Design In B2B Software
28:34 The Breville Toaster And The Bit-More Button
31:47 We Win When Our Customers Win
35:18 Put The Scoreboard On The Wall
37:07 A Company Is Just A Fiction With A Common Purpose
38:19 Focus On The Things That Don't Change
40:02 Is This Time Actually Different
40:50 Air Conditioning Made Las Vegas, Not Fortunes
43:52 1% Of US GDP Will Be Token Spend
45:28 A Six-Month Lag To A Model A Hundred Times Cheaper
46:46 Agents Negotiating With Agents To Buy Things
49:22 Making Sure The Highest-Return Dollar Gets The Dollar
51:17 The Dream Of Being At The Table That Matters
54:14 Why The Labs Are Ramp's Real Competitor
55:44 Banks Sell Money, Ramp Sells Time
56:27 When Intelligence Becomes Functionally Free
Includes paid partnerships.
Keith Rabois reposted
Founder friendly VCs in the 90's
Watch the final 8 seconds. Last night was not an aberration.
Fox analysts were accurate. Brutal half.
Listening to Belgium whine, the vote for Brexit (from petty EU tyrants in Belgium) is totally rationale.
Keith Rabois reposted
On the historic USS Nimitz, on our 250th bday, I performed Superman for our amazing @USNavy, @FDNY heroes, and Gold Star Families with former hostage Alon Ohel. Alon is a miracle, alive because of the President, @SecRubio & others. America has a history of making miracles. 🇺🇸❤️
Keith Rabois reposted
Sources: President Trump, commerce secretary Howard Lutnick, and White House task force head Andrew Giuliani put together a team of elite lawyers — from outside the government — to challenge the Flo Balogun red card. Specifically they challenged the use of slow motion instant replay to give the red card, which they argued violated FIFA rules. The president also conveyed to Gianni Infantino, FIFA’s president, that the appeal had been filed and he believed the red card penalty was excessive. FIFA’s independent committee reviewed the decision and agreed the penalty was incorrectly given, rescinding it under their rule 27 authority.
Interestingly, technology has made baseball, football, hockey and tennis better (mostly) and the NBA and soccer materially worse.
Keith Rabois reposted
Cape Verde btw
As predicted….
Holy hell, it’s already went from billionaires tax to $50M tax..
Democrats are openly stating their plan to seize wealth from Americans who spent generations building it.
They call it “fairness.” It is confiscation.
They want to raid private wealth to fund open-border chaos, illegal aliens, dependency, bureaucracy, and the continued decay of the country.
Ro Khanna’s own essay starts with billionaires, then quickly expands to fortunes of $50 million and up.
That is how every government seizure begins. First they target the people they taught you to resent. Then they work their way down.
In any sane country, using the power of the state to loot productive citizens and subsidize national destruction would be treated as a betrayal of the nation itself.
Excellent advice.
GROWTH FOUNDERS: HIRE A NEAR-PEER
A portfolio founder approaching $100M in ARR asked me about the single biggest and most impactful hire they can make.
I thought for a bit tand said: "Hire a near-peer".
In every generational company I've been part of, the founders hired a near-peer, who was essential to the company's success.
- Google: Larry and Sergey hired @ericschmidt .
- Facebook: Mark hired @sherylsandberg .
- Square: Jack hired @rabois .
- DoorDash: Tony hired @chrispa
- Coinbase: Brian hired @emiliemc
Characteristics of a near-peer:
1. They're so good that the founder will be fine reporting to them if the roles were reversed. (Mark has said publicly that he'd be fine reporting to Sheryl)
2. Their strengths perfectly complement the founder's strengths; however, they share many cultural attributes with the founder and pass the founder's airport test, since the founder will be spending a ton of time with them (example: Eric being a Computer Scientist, which was culturally very important at Google back in the day)
3. They're systems builders who have already operated at the scale you're growing into. Pattern recognition on $100M to $1B is not something you build in real-time.
A near-peer lets the founder focus on what only the founder can do (product, vision, culture). The near-peer handles everything else.
Also, near-peers stay for the long haul. Decade-plus tenure is standard.
The reason this hire matters more than any other: after $50M ARR, the bottleneck shifts from product-market fit to organizational scale. The founder is still the visionary. But the company needs someone who has already scaled a company of that size.
Most founders wait too long. They hire functional VPs first (Sales, Marketing, Engineering, Finance) and hope the collective covers the gap. It rarely does. A stack of VPs reporting into a founder who has never scaled a company creates coordination overhead, and the founder becomes the bottleneck.
The near-peer absorbs that overhead. They turn the founder's vision into daily operating decisions. They give the VPs a leader who has actually run a company this size.
Growth founders: if you're between $50M and $300M in ARR and every week feels like a coordination tax, the near-peer is the hire that determines whether you build a $10-100B company or top out at $1B. Start the search now, ideally through warm intros from your venture/angel investors and advisors.
Keith Rabois reposted
Honestly, watch the entire Alex Karp interview. It’s epic as a piece of television, and it provides important insight into the man himself. But he makes a bunch of extremely important points that, as he says, most people who know the same things won’t say in public.
Self recommending.
This is by far one of the most important things I’ve watched in 2026.
Keith Rabois reposted
"We don't need to live like this."
Matan Grinberg (@matanSF) is the founder and CEO of @FactoryAI, a startup building AI agents, called droids, that handle software engineering the way no single AI lab wants them to: independent of any one model.
His conviction is that AI's real constraint isn't capability. It's that the industry keeps building tools locked to one company's model, leaving that company free to dictate the terms of how you can use it.
(0:00) Intro
(3:50) Noether’s theorem explained
(10:53) Why there will always be more problems to solve
(20:10) Why Factory abstracts away model choice
(35:33) How Matan got into string theory
(41:53) Startup founders vs. theoretical physicists
(52:53) The origins of Factory
(1:08:11) Matan’s predictions for the future of AI and Factory
Thank you to the partners who make this possible
@dottechdomains: An identity for builders at their core: go.tech/thegeneralistnl
@brexHQ: The intelligent finance platform: www.brex.com/mario
@withpersona: Trusted identity verification for any use case: withpersona.com/generalist
Keith Rabois reposted
There have been nearly 7000 minutes of soccer played in the World Cup in a 48-team format w/ multiple games & short recoveries. Even with this, there hasn’t been multiple castastrophic soft tissue injuries like ACL & Achilles tears. More compelling evidence for grass over turf.
Keith Rabois reposted
Today we're launching @SpellbookLegal's biggest thing yet: Autonomous Contract Management
It’s the first end-to-end AI infrastructure for contracts
The world is speeding up. We are in one of the biggest investment cycles in decades.
Behind every rocket launch, FIFA game, and datacenter, lies a web of hundreds of agreements.
Agreements are the invisible threads that allow us to work together.
We have infrastructure for finance (Stripe, Ramp), eCommerce (Shopify) and many business functions.
But agreement infrastructure is lacking. This creates a painful bottleneck on our ability to work together.
Online purchases take milliseconds. But agreements still take weeks.
CLMs were supposed to be the answer, but were designed in a pre-AI era.
AI fundamentally changed how computers can accelerate agreements.
Spellbook is the most used AI contract review tool in the world, with ~5,000 customers in 80 countries
Now we are expanding to deliver the first end-to-end, AI native stack for contracts.
From the moment a deal lands in your inbox, to the day it renews years later, Spellbook’s AI supports teams every step of the way, across all business teams.
It runs 24/7. While you sleep, it's reading the deals that came in overnight, flagging the parts that actually need a lawyer, and clearing the busywork that used to kill mornings.
Nothing gets handed off between systems or slips after signature, and the intelligence stays with you for the life of the contract.
AI for lawyers is great. There are 20 million lawyers in the world, and many are our users.
But there are billions who touch contracts. We're excited to help everyone move faster and do more of what they love, by building the best AI-powered contract infrastructure in the world.
Get early access: spellbook.com/acm
Narrative violation:)
We can finally say AI isn't killing jobs.
A new paper from me, @tryramp, and @RevelioLabs uses firm-level spend and workforce data across 21K U.S. businesses to measure AI's impact on jobs.
Firms that adopt AI heavily grow headcount 10% over two years following adoption. Low adopters see no statistically significant change.
Why you need to use AI for your health and not depend on your doctor.
I'm a cardiologist. I've spent twenty years as the person patients trust to interpret their bodies. And I need to tell you something that most physicians won't say out loud:
AI is about to change the power dynamic between you and your doctor. Forever.
Four days ago, OpenAI's o3 model diagnosed 18 children with rare diseases that the best human specialists at Boston Children's Hospital couldn't solve — some after nearly twenty years of searching. Published in the New England Journal of Medicine.
Two weeks ago, WashU researchers proved that nine routine blood markers can calculate your biological age — and predict cancer risk years before any tumor forms. A free calculator. Available to anyone.
Last month, AI-enhanced coronary CT angiography detected inflamed arteries in patients whose standard stress tests said "normal." Patients who would have gone home reassured and wrong.
The pattern is unmistakable. The tools that used to require a specialist, a referral, a three-month wait, and a $400 copay are migrating into your phone, your bloodwork portal, and your own hands.
And I'm watching something in my practice I never expected.
Patients are walking in more informed than some of the residents I trained. They've run their PhenoAge score. They know their ApoB. They've read the study about Lp(a) before I've had time to bring it up. They come with questions so specific that the conversation starts at a level it took me years of training to reach.
This used to threaten physicians. It shouldn't. It should liberate us.
Because here's the truth about the old model: a 15-minute appointment where your doctor runs a basic metabolic panel, glances at the numbers, says "looks fine," and sends you home — that model was never good enough. It was just all we had. It missed 75% of future heart attacks. It caught cancer late. It told women with microvascular disease they had anxiety. It filed children with rare diseases as "unsolvable."
AI doesn't replace the physician. I've said this before and I mean it — the human moment, the clinical judgment, the hand on the shoulder when the diagnosis lands — that's irreplaceable.
But AI does something the old model never could: it gives you the ability to see inside your own biology with a depth and speed that was impossible a decade ago. To track your own numbers. To calculate your own biological age. To bring data to your doctor that elevates the conversation from "am I sick?" to "where exactly am I heading, and what do we do about it?"
The patient who walks in with their ApoB, their Lp(a), their hsCRP, their PhenoAge calculation, and a list of questions from the latest research — that patient doesn't threaten me.
That patient is the easiest person in my practice to keep alive.
Because they've already done the one thing most patients never do: they stopped waiting for permission to understand their own body.
I went into medicine because I wanted to help people live longer. What I've learned is that the patients who live longest are the ones who took ownership — not of my job, but of their own data, their own questions, and their own decisions.
The tools are here. The research is published. The calculators are free. The blood tests cost less than a dinner out.
You don't need to wait for your annual physical to find out what's happening inside you. You don't need permission to understand your own biology. And you don't need to accept "looks fine" from anyone — including me — when the science offers a deeper answer.
The revolution isn't coming. It's in your pocket. In your patient portal. In the published studies you can read yourself.
The only question left is whether you'll use it — or keep waiting for someone to tell you it's time.
Your body. Your data. Your life.
Take ownership. Your future self is counting on it.
This transcends soccer.
european football has spent the past fifteen years solving futbol like chess.
a generation of coaches optimized for pass completion, pressing triggers, territorial control, rest defense, and positional occupation.
the problem of this is that they optimize for what is measurable. depth, the willingness to attack space early, attempt the difficult pass, dribble past a defender, or deliberately create chaos, is a high variance play. it fails more often than it succeeds. if you evaluate players by completion rate, ball retention, or positional discipline, those actions look like mistakes. so they get coached out. eventually, everyone converges toward the same local optimum.
the game becomes increasingly legible. every team occupies similar spaces, presses in similar ways, builds from the back with similar patterns, and minimizes the same risks. systems become better at defeating other systems, but worse at dealing with players who refuse to behave like systems.
south american football never fully abandoned the duel as the fundamental unit of the game. the 1v1 remained sacred. so did the tactical foul, the unpredictable dribble, and the player willing to lose possession five times if the sixth breaks the match open. the objective was never simply to preserve structure, it was to create someone capable of destroying the opponent’s structure.
football is not won by completing the most passes. it is won by scoring more goals than the other team. those are related, but they are not the same objective.
this is the danger of optimizing proxies. when everyone optimizes the same measurements, they stop optimizing for victory itself. they optimize for looking efficient.
italy may have been the first major european football culture to lose part of its identity this way. its historical advantage was never athletic superiority or perfect positional play. it was tactical asymmetry, unpredictability, and an instinct for making matches uncomfortable. as italian football converged toward the same coaching model as the rest of europe, it gradually surrendered the qualities that had made it different.
the broader lesson extends well beyond football. every optimization process eventually risks becoming self-defeating. metrics become targets. proxies replace objectives. variance is mistaken for error. the outliers capable of breaking the system disappear because the system itself learns to eliminate them.
Keith Rabois reposted
Cc @travisk Adam Neumann @jack @naval @emilmichael @bhorowitz
"The benefit of Benchmark is, we don't invest in categories, we invest in people."
@EverettRandle says Benchmark's strategy is " all based on the founder."
"There's zero thought to, 'This is going to be a big category.'"
"It's that this is an unbelievable entrepreneur that is going to put their energy to an idea that they've convinced us is really, really interesting."
"When you focus and have your strategy around that—great entrepreneurs are always in style."
Worth reading carefully.
MEMORY IS THE MOAT
@nikesharora, Chairman & CEO of @PaloAltoNtwks , interviewed by @HarryStebbings (@20vcFund )
Summary: Nikesh Arora took Palo Alto Networks from an $18 billion company to one worth $225 billion, and his read on enterprise AI is blunt: most companies are doing it wrong, and most of the products are not ready. His core claim is that consumers forgive AI's mistakes while enterprises cannot, so the money will flow to whoever builds the depth (the context, the memory, and the edge-case training) that lets an agent act without a human catching its errors. The companies that win will redesign themselves around AI instead of adding it to yesterday's workflow, and the lasting advantage will be the memory a system builds up about you. He expects token prices to fall 90%, half of G&A roles to disappear in 3 years, and more engineers and salespeople, not fewer.
1. Context Stickiness. The lasting advantage in AI is the context a system holds about you, not the model itself. Arora says the frontier labs are racing to remember what you asked over the last 30, 60, 90 days so each new answer gets easier and you stop wanting to leave. The more a model knows about a user, the higher the cost of switching, and that stickiness is the moat. For enterprises the same logic holds: the company that owns its context wins, not the one renting the smartest model.
2. Breadth Versus Depth. The frontier model problem is a breadth versus depth problem. Consumers tolerate false positives and enterprises have none to spare. Arora had Gemini write a passable investment memo in 4 minutes, and a wrong line or two did not matter because a person was sitting in the middle to catch it. An agent acting on its own has no person in the middle, so a false positive becomes a live failure. Consumer AI wins on breadth and brand, while real enterprise revenue comes from depth.
3. The Waymo Standard. Waymo is the biggest agentic product in the world, and it shows what depth actually costs. Replacing one human, the driver, took tens of billions of dollars of edge-case training and data that exists nowhere on the internet. You cannot drop the next Anthropic model into your Mercedes and tell it to drive you home. Every enterprise agent that truly replaces a person needs that same depth, which is why most agentic enterprise products are not ready.
4. Rethink The Workflow. Most enterprises are losing because they add a little AI to an old workflow instead of redesigning the workflow around AI. Arora's example: scanning an invoice 20% faster is the trap, while the real win is letting AI do 80% of the thinking, like reading every CV and telling you which 20 people to interview and what to ask each one. That means giving up human control, which is exactly what companies resist. The winners over the next 3 years rethink the company with AI, not the task.
5. Software With Opinions. The next wave of enterprise software will have opinions, and that is the real change Arora is pointing at. Coded SaaS gives you the output you defined for the input you fed it. An AI marketing assistant reads your copy, tells you it is off-brand, and says how to fix it. That opinion makes an average employee smarter, which is why Arora expects half the people in G&A functions like marketing, finance, and HR to be gone within 3 years.
6. More Engineers, Not Fewer. The fear that AI shrinks headcount is half wrong. Process-heavy G&A roles compress, but Arora wants more technical and more sales people. His teams keep asking for resources to rework marketing and HR, and for people who can prompt frontier models, build harnesses, and bring in data nobody else has. A good product also needs more sellers: he met 20 customers in Europe last week and half did not know what his 20-year-old company already ships.
7. Tokens At One-Tenth. Long-term token pricing should be a tenth of what it is today. Compute costs 2 to 4 times what it did 2 years ago because more than half of it feeds loss-making consumer AI, which forces the pricing pressure onto enterprise and coding workloads that have to pay. As compute gets more efficient and consumer usage gets capped, prices fall hard over the next 3 to 5 years. The model from 2 years ago was already good enough for 90% of tasks; the problem was it cost too much to run.
8. The Token Allocation Trap. Capping token spend punishes your best people. Arora runs a "use judiciously" model, not a free-for-all, because the smartest AI-savvy employee can burn 20 times the tokens of an average one. Playing whack-a-mole with cost hurts the high performers most and slows the learning you need. The better move is to track usage, leave the power users alone, and cap only the genuine outliers.
9. The Attacker's New Edge. Powerful coding models cut both ways. Trained to write good code, they are just as good at finding bad code. Pointed at his own systems, a model found in 6 weeks what would have taken his team 5 to 6 years. It cannot safely auto-patch, because it would "fix" 30% of things that are not broken, so it arms attackers faster than defenders. The result is urgency: every enterprise has to fix its systems faster, which is good for security companies.
10. The FTE Tell. If a startup needs forward-deployed engineers to sell into the enterprise, the product is not finished. Arora's read: enterprise AI is barely 12 months old, agents keep changing what the product even is, so vendors send engineers to build the product inside the customer while the technology keeps moving. A real forward-deployed engineer brings code back and folds it into the product; many are just adoption consultants. Expect customers to churn from one tool to the next, the way coding went from Windsurf and Devin to Codex, Claude, and Factory.
11. Three Missed Tricks. Miss one trick and you survive, miss two and you are partly impaled, miss three and you could be obsolete. This is why Arora spends more time than ever learning, pinging founders building things he does not yet understand. He buys early and cheap on conviction, treating an acquisition as a 10x or 100x bet where paying 1 or 2 times more does not matter, rather than waiting to buy the proven winner for a billion. He runs a twice-weekly "AI EIO" meeting so his top 15 leaders compete to show what they shipped.
12. The Sunk Cost Walk. A board member taught Arora to separate effort from wanting the outcome. After months grinding through a near-billion-dollar acquisition, he was told to take a long walk and ask one question: if this deal walked in the door right now with zero effort, would I still write the check? You have not spent a dollar yet, so the only thing that counts is whether it stands on its own merits. The same trap catches investors who confuse beating 8 VCs to a term sheet with the deal being good.
Best post on the topic of building an iconic company in the age of AIz
Keith Rabois reposted
AI is indoctrinating Americans
Keith Rabois reposted
"Notable accomplishments are rarely achieved by people who work 40 hr per week or less. World-class performers work on average 60 to 80 hr per week with commitment and passion."
Keith Rabois reposted
Why I joined Opendoor and why you should too if you want to work on AI.
Few industries stand to gain more from AI than real estate.
Closing a single home takes 30-60 days. Why? One transaction touches law, insurance, title, contractors, utilities, inspections, phone calls and emails across dozens of parties.
And the complexity is ancient. Property law LITERALLY traces back nearly a thousand years to medieval England. Records and rules differ county by county. There are ~50,000 (!!!) water systems alone in the US, each with its own process to turn service on.
It's one of the messiest, most human, most coordination heavy markets there is. Which is exactly why it's a great place to build AI
Super exited about this. Now your enterprise can be transformed w AI rapidly and securely.
Today, we’re announcing Runlayer has raised $30M from Felicis and Khosla Ventures to help companies go all in on AI.
Runlayer is the golden path for AI: enablement, security, and control in one platform.
So, how does it give your team the right tools for AI? 🧵
Keith Rabois reposted
Keith Rabois reposted
60% of contracts filed to the SEC contain mistakes. Spellbook Labs ran 60,000 pages of contracts from 500+ public companies through AI to answer a question:
How often do human lawyers hallucinate?
Accidents from self-driving cars get headlines—and they should. But human drivers have a higher accident rate. We wondered if something similar might play out in contract drafting.
Reviewing 100 page contracts line-by-line is inhuman. Mistakes are inevitable. AI is very good at spotting issues like missing definitions and broken section references.
There are risks to using AI in law. But we also need to ask: what is the risk of *not* using AI review as a second set of eyes?
In the full report we show:
1. Top 10 public companies with the cleanest contracts
2. Error rates across industries (Media and Telecom were highest)
3. Error rates over time (we're not getting better!)
4. Error rates by contract type and length
... and a bunch more.
Congrats @IBM for having the cleanest agreements.
📌 Read the report here: spellbook.com/labs
Keith Rabois reposted
This is the most eloquent and informative explanation of the Iran regime threat that I’ve heard in 23 years working on the issue.
And I’ve heard many.
From @TimSheehyMT who also is a former U.S. Navy Seal.
Keith Rabois reposted
Supplements that work:
1. Creatine - power and strength
2. Magnesium - sleep and metabolic health
3. Glycine - sleep and glycation
4. NAC - inflammation and respiratory conditions
5. Omega-3s - inflammation and lipids
Keith Rabois reposted
I just read a story about a runner I know who had leg pain for 3.5 years. Her diagnosis was only identified after she uploaded all of her info into ChatGPT, which not only correctly made the diagnosis but also recommended a specific doctor she should go to in Virginia.
I see all of the backlash from the medical community about the Midjourney launch and can’t help but think - where along the way have we lost our openness for innovation and our focus on the patient? AI is going to profoundly change the status quo for how people think about and engage with the health system.
Keith Rabois reposted
One aspect of Jalen Brunson that often gets overlooked- his academic success. He took a bevy of AP classes at SHS (his mom Sandra always made him email his teachers weeks in advance of known absences due to bball), and he graduated from Villanova in three years.
He was as successful in the classroom as he was on the court.
He is a great role model for kids today wanting to succeed in both arenas.
Keith Rabois reposted
The NYT is finally catching up...
If you're early in your career or still building, remote work is brutal:
- You miss real training, feedback, and institutional knowledge
- Productivity drops ~20% (MIT study)
- No mentors. No company culture.
- No bonding. No happy hours.
- Fewer real work friends
- Divorce rates and mental health issues spike
- Fewer spontaneous opportunities (client meetings, drinks with the boss, hallway ideas)
- Lower promotions and job security long-term
Keith Rabois reposted
Knicks Parade Day
Stuck on Fulton - 6:40am
#alwaysknicks
Keith Rabois reposted
BREAKING: JPMorgan just cut off its entire staff in Hong Kong from accessing Claude
Goldman Sachs also did the same last month
It’s over
Keith Rabois reposted
cold open: google campus. a conference room named “moonshot serenity 4b.” twelve people are in a meeting titled: pre-sync for sync alignment on ai velocity.
sundar sits calmly at the head of the table.
a pm clicks to slide 1 of 187.
“the agenda today is simple,” she says. “how do we move faster while preserving our culture of not doing that?”
everyone nods.
then the door opens.
noam shazeer walks in.
the room goes silent.
noam: “i’m leaving.”
a vp of gemini reliability, brand, trust, latency, policy, and vibe raises a hand.
“leaving… this meeting?”
noam: “google.”
someone gasps. someone else opens a doc titled retention narrative draft final final noam v7.
sundar blinks once.
“noam, we brought you back.”
“for two point seven billion dollars.”
“technically you licensed some technology and reacquired talent.”
“that sentence is why we need legal in the room.”
legal is already there.
cut to: openai.
sam altman stands beside a whiteboard that just says ship.
an engineer walks by carrying a server rack and what appears to be the future.
sam: “we can offer speed, compute, and one meeting.”
noam: “one meeting per week?”
sam: “no. one meeting. total.”
back at google, the emergency retention committee forms instantly. it has 31 members.
a director says, “what if we give him a new title?”
“he already co-leads gemini.”
“distinguished super co-lead?”
“google fellow?”
“he already left google, founded a company, got brought back for billions, then left again. he’s folklore.”
meanwhile, a gemini launch review begins.
pm: “we’re ready to announce the model.”
policy: “can it answer questions?”
eng: “yes.”
policy: “too risky.”
marketing: “can we call it experimental?”
research: “the model is better than the last one.”
brand: “better is aggressive.”
trust & safety: “what about ‘more contextually adjacent to usefulness’?”
a staff engineer whispers, “openai just shipped a model while we were discussing the adjective.”
cut to noam’s exit interview.
hr: “what could google have done better?”
flashback montage:
a chatbot blocked because it might be too good.
a launch delayed because a button was the wrong shade of responsible blue.
a spreadsheet comparing twelve ai product names.
a meeting where someone says “we need a single coherent ai strategy” and three new strategies are created before lunch.
noam: “nothing comes to mind.”
hr: “great. we’ll mark that as positive attrition.”
later, sundar calls him privately.
“google is still google. best researchers. best infrastructure. billions of users.”
“yes.”
“so why leave?”
noam looks out the window.
“because you have everything except permission.”
silence.
sundar, softly: “we can create a permission working group.”
cut to all-hands.
sundar addresses the company.
“noam is leaving. this is not a loss. it is an opportunity to reflect on our operating model.”
chat explodes:
“is this recorded?”
“which gemini?”
“can we ask gemini why people keep leaving?”
“it said ‘insufficient context.’”
a vp steps up.
“to honor noam’s legacy, we’re launching project attention.”
applause.
“it will study whether attention is, in fact, all we need.”
a researcher raises a hand. “didn’t we answer that in 2017?”
“yes. but now we need enterprise readiness.”
final scene: noam arrives at openai. badge works instantly.
receptionist: “yeah, we just made one.”
no pre-read. no doc. just a whiteboard, five people, and a model running somewhere hot enough to toast bread.
sam: “ready?”
noam smiles.
cut back to google. a calendar invite appears:
meeting: reduce meetings task force kickoff
duration: 90 minutes
required attendees: 214
sundar sighs, opens gemini, and types:
“how do we move faster?”
gemini responds:
“have you considered leaving google?”
smash cut to credits.
Keith Rabois reposted
This brilliant lesson on communication by @rabois is why the AI leaders are failing. It’s not sufficient just to “speak your truth.” You have to communicate in a way that elucidates your audience. Convincing the public that your company is a menace obviously fails that test.
Keith Rabois reposted
Exclusive: Sen. Cotton asks Treasury to investigate Airwallex's China ties, seek divestiture www.axios.com/pro/all-deals/2026/06/17/sen-cotton-treasur…
by @ShenLucinda
Spellbook helps u sign Brunson, KAT and OG too.
Congrats to the Knicks' legal team!
The world runs on contracts. It's time to recognize the people behind them. More billboards coming soon :)
Keith Rabois reposted
Yeah, this ... this is how you end up in a lot of trouble.
Aggressiveness and burn rate are often correlated but the correlation coefficient is much lower than many believe
Keith Rabois reposted
FIFA bans the Lion and Sun flag but allows the Iranian players to air shoot people in the crowd.
Got it. x.com/Tarikh_Eran/status/2066722996555821387/video/1
True.
AI resilience isn't just a national security issue but maintaining sovereign hedges on model usage will now also be necessary for every business
Just like how enterprises can't just use one bank, they also can't use one AI provider and they're *rushing* to @FactoryAI
Phase 2🚀
Keith Rabois reposted
This is why I joined @FactoryAI. We created and named this category. Coding agents enabled individual engineers to develop faster, and software factories will enable engineering organizations at scale. The next phase is building software factories.
Keith Rabois reposted
When you've got a software factory, you can ship faster than @AnthropicAI with 100x less engineers
You guys are shipping at an Anthropic rate, this is insane!!
Keith Rabois reposted
Anthropic wants you to believe this is a personality or political dispute. It's not.
Anthropic abandoned its safety guardrails the moment they became commercially inconvenient. It was so egregious their corporate partner--Amazon--felt compelled to blow the whistle.
Now they're deploying a team to DC to do technical cleanup. Anthropic doesn't care about safety, they care about control. This is dangerous and wrong.
"They screwed us": Personality clashes sent Anthropic's models offline www.axios.com/2026/06/15/anthropic-white-house-fable-myth…
Keith Rabois reposted
11 players in NBA history have at least 3 closeout games with 40+ points:
Michael Jordan (10)
Kevin Durant (5)
JALEN BRUNSON (4)
Giannis Antetokounmpo (4)
LeBron James (3)
Allen Iverson (3)
Kobe Bryant (3)
Charles Barkley (3)
Jerry West (3)
Elgin Baylor (3)
Wilt Chamberlain (3)
Keith Rabois reposted
The size of a company's last funding round correlates negatively with 24-month performance post-IPO.
Total amount of VC funding also correlates negatively with 24-month performance post-IPO.
This holds whether you measure from the opening prices or the day-one pop.
Maybe the best analysis on the topic.
Anthropic's Safety Superpower
Anthropic's belief in its own commitment to safety gives the company license to aggressively favor its business and even challenge the U.S. government.
stratechery.com/2026/anthropics-safety-superpower/
Better conditioning FTW.
Knicks net rating by quarter in the NBA Finals:
1st - -47.5
2nd +23.4
3rd - +12.2
4th - +23.4
Keith Rabois reposted
Most people see the result.
Very few understand the obsession it takes to get there.
Keith Rabois reposted
Ernie: “Brunson comes down on Wemby’s foot. No call.”
Shaq: “Ooooh.”
Chuck: “That should have been a flagrant.”
Ernie: “And if it had been a flagrant, Wemby sits Game 6.”
Chuck: “Well he can still sit Game 6.”
Keith Rabois reposted
THIS.
“The Savior of the Broken”
Jalen Brunson
OMG.
This got me
Keith Rabois reposted
Jalen Brunson’s Game 5 belongs in Finals history.
Highest share of team points in an NBA Finals-clinching win:
51.7% — Michael Jordan, 1998
47.9% — Jalen Brunson, 2026
47.6% — Giannis Antetokounmpo, 2021
45.5% — Bob Pettit, 1958
43.3% — Michael Jordan, 1997
Only MJ has ever carried a bigger scoring load to close out a title.
Keith Rabois reposted
"We spent $3,000 on an Eight Sleep for every employee because we optimize for output, not cost.
When you're building a team of exceptional people, every improvement in sleep, focus, and decision-making is worth the investment.
The hardest work isn't about grinding more hours, it's about making better decisions, and that starts with getting a great night's sleep." @matanSF
What single decisions have had the biggest impact on performance for you @rabois @zebulgar @m_franceschetti @bryan_johnson
Definition of a barrel.
4 years before Jalen Brunson arrived:
• No Playoffs
• No Playoffs
• 1st Round L
• No Playoffs
4 years since Jalen Brunson arrived:
• East Semis
• East Semis
• East Finals
• NBA Champions
Jalen Brunson is the greatest player in New York Knicks history… & it’s not close.
Keith Rabois reposted
Most points in Finals clinching game on the road in NBA history:
45 - Michael Jordan
45 - Jalen Brunson
Keith Rabois reposted
Unbelievable how these guys celebrated.
Jalen walks over to Mitch Johnson before doing anything. CLASS
Keith Rabois reposted
Jalen Brunson:
"We show up at 9 o'clock"
That's the most New York fucking line ever.
Aged well.
I just want to be successful thats all..
Keith Rabois reposted
This is an unfathomable Brunson performance even for him.
Keith Rabois reposted
The entire city of New York is singing Empire State of Mind. Incredible.
(via @TrindonHo11iday)
Keith Rabois reposted
Wemby in crunch time in the NBA Finals:
Game 1:
-Dribbles off foot with 50sec left.
Game 2:
-Airball with 1:30.
-Settles for terrible midrange shot in the 104-104 game with 30sec left.
-Turnover pass off Castle’s back & fouls Brunson with 9 seconds left.
-Misses good look to win at buzzer.
Game 4:
-Shoots the Knicks back into the game going 4-17FG after he told Robinson he was in his head.
-Misses both FTs up by 1 with 1:40 left.
Game 5:
-1/5FG in 4thQ.
-Missed FT down by 2 with 1:50
Losers lose.
Wemby and Castle immediately walked off the court without dapping anyone up lmao i can’t respect that
Keith Rabois reposted
The NBA spend so much money to portray him as the next best thing, and then it turned out that he’s just a spoilt brat who talks big, plays dirty and disappears when it matters.
“Appreciate y’all. See y’all…never.”
— Wemby at the end of his post-game press conference.
(h/t @big_business_)
x.com/big_business_/status/2066016759195263095/video/1
Keith Rabois reposted
This angle of Jalen Brunson going at Wemby 😯
Keith Rabois reposted
Victor Wembanyama is the dirtiest star player in the modern NBA. I’ve never seen anything like this. Embarrassing.
Keith Rabois reposted
So. Damn. Good.
Sleep well, NY.
Keith Rabois reposted
Jalen Brunson is the first player since Magic Johnson to win a HS state championship, an NCAA championship, and an NBA championship
Keith Rabois reposted
KNICKS!!!!!
You guys beat the Spurs and their dirty playing star, and the refs. CONGRATULATIONS!!!!!!
I'm not a Spurs or Knicks fan, but I can't remember seeing officiating this biased in a playoff series. One team appears to be getting favorable treatment, and Wemby has seemingly avoided two obvious flagrant calls. The refs are undermining the integrity of the game.
Keith Rabois reposted
Victor Wembanyama is a rat fuck and the NBA refs are blatantly rigging this game for the Spurs
There is only one way to solve these flagrants committed by Wby that are intended to injure the Knicks stars.
Keith Rabois reposted
If I understand the rules correctly, this should be a Flagrant on Wemby
Would have prevented a Spurs 3 and given the Knicks about a six-point swing
Would also mean Wemby is suspended for Game 6.
Am I wrong?
Keith Rabois reposted
🚨EPIC scenes on the National Mall massive crowd gathered at the Washington Monument with a stunning rainbow overhead!
God’s blessing shining down on America. Beautiful moment of unity and patriotism in our Nation’s Capital.
Keith Rabois reposted
This is pretty epic.
Keith Rabois reposted
This rendition by Dan & Shay is fantastic, but credit to the sound guy who absolutely nailed this mix. It's perfect. The blend between the harmonies and the live mics in the stadium is what makes the hair on your arm stand up.
Outstanding distillation. Explains why you can’t skip this podcast.
THE TOKEN HANGOVER
@matanSF (Matan Grinberg), CEO and co-founder of @FactoryAI , interviewed by @HarryStebbings (@20vcFund )
This is a special for me since I've been an investor in @FactoryAI since their seed round, and think Matan is a very very special founder.
Summary: Grinberg argues the next 24 months in enterprise AI are a resource-allocation problem: tokens, dollars, and people. Most CIOs are now waking up to bills they cannot justify. The fix is to spend frontier tokens only on the 10-20% of work that requires planning intelligence, run the other 80-90% on open models, and rebuild teams around load-bearing polymaths who own business outcomes. The single-frontier-monopoly fear is fading: four roughly-equivalent labs is the emerging reality, which puts pricing power back in the application layer.
1. The Token Hangover. Enterprise AI adoption ran through three phases this year: boards yelling at CEOs about AI strategy, "token maxing" with AI usage written into perf reviews, and now the morning-after bill. One CIO Grinberg spoke to was spending hundreds of thousands of dollars a month on engineers asking Opus 4.8 things like "how's it going" and "what are my macros from lunch." The frontier model became the default surface for every question, no matter how trivial. Phase 3 is the moment routing matters: every call to a frontier model needs to earn its price.
2. Resource Allocation Is the Job. For the next 24 months every C-suite is solving the same problem: how to allocate dollars, tokens, and headcount against business outcomes. Engineering teams used to be judged by features shipped per quarter, a metric with no link to revenue, market share, or retention. A logistics company adding more engineers to ship more features was always solving the wrong problem; AI made the misallocation visible. Tie every person's work to the metric that actually moves the business, then re-allocate.
3. Load-Bearing Individuals. The "10x engineer" frame measures lines of code, the wrong unit. Grinberg's unit is the load-bearing individual: the person whose absence breaks something. With AI the load-bearing few compound roughly 10,000%; the others get close to nothing, so any org enforcing one token-spend-per-engineer number is painting with too wide a brush. Average token spend per engineer will land on the same order of magnitude as their salary within three years, with a wildly bimodal distribution.
4. Frontier for Decisions Only. 80-90% of software development tasks can run on open models; the remaining 10-20% is planning, where the frontier still wins. This mirrors how human orgs work: leadership is a tiny share of total hours but decides the company's fate. The ego trap is engineers assuming their work is too important for an open model. The router decides better than the engineer, and the cost curve falls only if you wire the routing.
5. The Kirkland Mistake. Kirkland & Ellis announced a $500M, five-year internal AI build, which Grinberg reads as validation for Harvey rather than a threat. Building AI is not a law firm's core competency, and Kirkland's spend will teach them how hard it is. The general rule: just because you can build it does not mean you should, and the discipline is naming the few things you and your team own end-to-end. Outsource everything else, even when you technically know how to do it yourself.
6. Model-App Separation. When the model provider also sells the app, the incentives split: an API business wants you to spend more tokens. A healthy market keeps the application layer independent, so model providers compete on price, speed, and quality every week. Enterprises do not want to vendor-lock again; every CIO carries scars from the cloud era's three-year discount-then-jack-the-price trap. The application layer survives precisely because it forces that competition.
7. Sales as Product. Name a legendary company with a weak sales or marketing team. You can't. The Silicon Valley fallacy that research sits at the top and sales is "dirty work" produces companies that win the gold rush and then collapse when gravity returns. At Factory, engineers and salespeople sit intermixed; when sales closes, engineering says "we closed"; when engineering ships, sales says "we shipped." Atrophied sales muscles will not regrow once enterprise buyers stop saying yes to everything.
8. Polymath Era. Da Vinci, Newton, Euler could be polymaths because their fields were shallow. By the 2010s a theoretical physicist needed 50 years to reach the frontier before contributing anything new. AI collapses that catch-up time, so one person can push forward developer marketing, token-caching infrastructure, and solution engineering at once. The engineer of the future is a GM who owns marketing copy, product metrics, and sales enablement.
9. Build the Factory. Factory's name is literal: engineers in the next era design the assembly line that produces software. The DevX investments that used to scale linearly with headcount (good docs, CI/CD, linters, pre-commit hooks) now scale with the number of agents you run, which is 10x or 100x larger. Every dollar spent making agents production-ready compounds against thousands of PRs a week. Humans move up the stack, from writing code to designing the system that writes code.
10. Seal Team Six. Mandating beds in the office is a hiring failure dressed up as commitment. Grinberg's image: a basketball game judged by who sweat the most, when the scoreboard is what counts. Factory bought eight sleeps for all 30 team members at the time, because recovery is where the gains come from when work requires every ounce of brain power. If your load-bearing engineer can do their best work on two hours of sleep, they were not doing load-bearing work in the first place.
11. Four Frontier Labs. Grinberg's biggest mind-change this year: a single dominant model is unlikely, and four roughly-equivalent frontier providers is the more probable steady state. That outcome is the win for humanity. A one-lab monopoly was the dangerous scenario, and four equivalent labs is also the structural bull case for the application layer because it forces real ongoing price competition. Every CIO Grinberg meets has already decided not to throw their lot in with a single provider.
12. Dario's Self-Serving Doom. "AI will take your jobs" was the pitch that helped raise hundreds of billions, and Grinberg thinks it damaged public psychology and fed the slow-AI lobby. Watch the rhetoric flip at IPO: humans will suddenly become important again, because humans are the ones buying the stock. Founders who never needed to raise that money, like Zuckerberg and Hassabis, never made that argument. Incentives drive the labor-displacement rhetoric more than philosophy does.
Don’t miss this @HarryStebbings interview w @matanSF of @FactoryAI explaining the future of software in the age of AI: open.spotify.com/episode/40R3ULhqI2XjfUqG27lU7O
Keith Rabois reposted
Anthropic already does identify verification for “age assurance” (support.claude.com/en/articles/15171100-age-assurance-on-…) so they should be able to restore access to Fable pretty fast for US citizens.
Keith Rabois reposted
A truly unbelievable statistic. #usmnt
83 - Chris Richards completed all 83 passes he attempted for the #usmnt against Paraguay, the most passes with a 100 percent accuracy rate by any player in a FIFA World Cup match since 1966.
Perfection.
Keith Rabois reposted
Elon Musk's contribution to the national economy:
Over 2021–2025, Musk's companies (Tesla, SpaceX, xAI, Neuralink, The Boring Company) reportedly injected ~$338 billion directly into the U.S. economy via:
• $110.7 billion in wages/salaries (supporting • 200,000+ employees at competitive pay).
• ~$46 billion in taxes (corporate, payroll, etc.).
• $182 billion+ in supplier spending (e.g., Tesla alone spent heavily on U.S. batteries, chips, steel).
Source: www.odaily.news/en/post/5207545
- - - - -
During his lifetime, Bernie Sanders has contributed approximately $0.00 to the national economy.
Keith Rabois reposted
History will remember Elon Musk alongside DaVinci, Gutenberg, Fulton, Pasteur, Edison, Ford and Jobs.
History will have no recollection of any kind of Ro Khanna.
Keith Rabois reposted
Worth reading very carefully.
⚡️This is a monster signal.
This is the moment frontier AI stops being treated like software and starts being treated like controlled strategic capability.
The key phrase is not “customers.”
The key phrase is “foreign national Anthropic employees.”
That means the state is no longer only controlling chips, model weights, or overseas access. It is moving into cognition access by nationality. That is the real threshold. The U.S. government is saying the highest models are sensitive enough that even people physically inside the United States, working inside the company, may be barred from touching them if their nationality creates deemed-export risk.
That is weapons-control logic.
This is ITAR logic for intelligence.
The corporate language about a “misunderstanding” is probably diplomacy.
Companies say that when they need to preserve customer trust, employee morale, and regulatory room. But national security authorities do not force emergency suspension of top model access because someone made a minor paperwork mistake.
Something about Fable 5 and Mythos 5 crossed the line: cyber capability, autonomous R&D acceleration, AI-improving-AI utility, bio/security planning, code exploitation, or some blend of all of it.
The U.S. state just showed that Anthropic does not fully control Anthropic’s frontier layer.
That is the phase change.
Labs can brand themselves as public-benefit AI companies. They can talk about safety. They can sell enterprise plans. They can publish model cards. But once the models become national capability, the sovereign arrives. The state does not need to own the company to control the access surface. It only needs legal authority over export, security, procurement, and liability.
This confirms the arc we’ve been tracking:
Frontier AI becomes state-supervised strategic infrastructure.
Public AI splits from strategic AI.
Foreign access gets restricted.
Labs become quasi-defense contractors.
Model access becomes a national security perimeter.
Enterprise customers learn that API access is not property. It is revocable permission inside a sovereign-controlled stack.
The most important implication is organizational.
If foreign national employees can be cut off from frontier systems, AI labs now have to reorganize internally around citizenship, clearance, compartmentalization, and controlled access. That breaks the old Silicon Valley assumption that global talent can freely collaborate around the frontier. The next AI lab structure looks less like Google in 2015 and more like a defense prime crossed with a classified research facility.
For markets, the winners are the national champions with U.S.-aligned infrastructure, cleared customer channels, government relationships, compliance capacity, and domestic compute. The losers are open access, foreign-dependent AI wrappers, offshore model distributors, and any enterprise whose moat depends on unrestricted access to frontier APIs.
For geopolitics, this is escalation. China will read this correctly. Allies will read this correctly. Every serious state will understand that frontier models are now part of national power.
The AI race just moved from “who has the best chatbot” to “who controls cognition as a strategic asset.”
Keith Rabois reposted
Whoa, let's go
Saving LA - Phase III
Keith Rabois reposted
public benchmarks are saturated. every frontier model has trained against them, and the leaderboard tells you near nothing.
we built ours from inside ramp — code no model has seen, graded against the bar our engineers ship to.
every company running on AI needs its own.
Today we’re releasing Ramp SWE-Bench: a private, production-grounded coding benchmark created from real engineering problems we've faced at Ramp.
Keith Rabois reposted
The final 3:30 of the Knicks' miracle win
Uncut
Shoutout to the NBA for posting this. This is awesome.
Keith Rabois reposted
I mean wtf even is this? He straight up just went to take out OG Anunoby's ACL. I can't stand this guy.
Can the NBA review this shit too and assess a flagrant? How else would you describe this? That's not a basketball move.
Keith Rabois reposted
Not only does KAT lead all NBA players in +/- this postseason (+262) and all players in the Finals (+48), Towns has posted a positive plus/minus in every single one of the 18 games the Knicks have played in these playoffs.
That’s the longest such streak in NBA postseason history
Seems pretty good for a "soft" center who can wing big games
Keith Rabois reposted
This is actually suspension worthy
I’ll give Wemby the benefit of the doubt, but this was objectively pretty weird.
Keith Rabois reposted
Last night, Karl-Anthony Towns recorded his 18th consecutive game with a positive plus-minus. It is the longest such streak in the postseason since the @NBA began tracking play-by-play data in 1997.
Keith Rabois reposted
Central Park
(IG: Dylan Christie)
Keith Rabois reposted
The full uncut ending of the Knicks’ comeback is superb. Ridiculous to be honest.
I’ve seen so many clips of Knicks fans losing their minds over this already today. 29 points down before going on to win it with 1 second remaining and go to within 1 win of their first NBA title in 53 years.
Imagine not loving sport?
Worth watching.
Mike Breen in the 2Q: "It's amazing how quiet the crowd is right now."
Mike Breen in the 4Q: "KNICKS TAKE THE LEAD... BEDLAM HERE AT THE GARDEN!"
From trailing by 29, New York completes the largest comeback in NBA Finals history!
Keith Rabois reposted
Teams down 20+ points in postseason last 2 years:
Knicks 5-3
Rest of NBA: 4-71
Keith Rabois reposted
Knicks' historic 29-point comeback in full 🔥
One of the greatest games in NBA history.
Even the NBA can’t stop the Knicks:)
The LA Mayor’s race was more legit than this.
Keith Rabois reposted
The first quarter of the Knicks vs. Spurs game was one of the most rigged things we've ever seen.
There were FIVE missed calls in favor of the Spurs and Wemby.
1) A clear foul on Wemby was reversed to give KAT his second foul of the game.
Thread ⬇️
Keith Rabois reposted
This is an embarrassment for the NBA
If what Victor Wembanyama did last night to Jalen Brunson wasn’t a flagrant foul, then flagrant fouls should cease to exist
This also gives Wembanyama license to keep doing this
Gutless move by the league after blowing it last night
The NBA says there will be no additional sanctions against San Antonio's Victor Wembanyama for his first-quarter shove of Jalen Brunson in Game 3.
The play is not being upgraded to a flagrant foul ... meaning Wembanyama remains on two flagrant foul points this postseason.
The NBA is forcing the Knicks to address this like a hockey team would. Role player needs to punch this guy.
The Knicks should not seek to injure Wemby tomorrow. But they should foul him HARD as early as realistically possible. Putting in Sochan early is too obvious. Robinson can do it late in the first. It's about making a statement that we're not doing this BS.
PJ Tucker would be saying this to the team too if he was still here, and you know it.
I want Wemby on the court. But he needs to know we're not playing these games anymore. Go Knicks. Win this thing the right way.
Just purchased. Excited to read.
Keith Rabois reposted
We're building:
→ Long-running agents that do real work for bankers and investors, at a fraction of the cost of frontier models
→ An intelligence layer for memory, skills, and agents that self-improves over time
→ Enterprise functionality that lets the most demanding institutions trust agents with real work
→ An end-to-end platform for deal-making that makes private markets more efficient
Building at the app layer in AI takes vision, creativity, craftsmanship, and extreme grit. There aren't many places where you can credibly say you're helping transform an entire industry. It's a herculean amount of work but it is super fun...
Keith Rabois reposted
Lovable says it has hit $500M in annualized revenue, with 1 million new projects a week techcrunch.com/2026/06/09/lovable-says-it-has-hit-500m-in…
Keith Rabois reposted
So we’re clear…
No foul on the left.
Flagrant on the right.
@NBA @OfficialNBARefs 🤣
Keith Rabois reposted
Mike Brown: "Over here you called it a foul, but over there you called it a travel."
Ref Marc Davis: "Ah sh*t you're right."
💀💀💀
Keith Rabois reposted
NEITHER Spurs player got called for a foul here. lol
Good defense I guess....
Keith Rabois reposted
By the way 2 more flagrant foul points for Wemby in the playoffs and he gets suspended is very clear why he’s being protected so much
Wemby easily could have been ejected from both games 2 and 3 and yet not a single one of any of the 4 clear technicals/flagrants have even been called a common foul
QED. Same people counting votes in LA must have been sent to officiate last night.
Zero fouls were called on the Spurs on this play
I’m not a “sports is rigged” guy but it’s really tough to defend how this series has been officiated
Keith Rabois reposted
Donald Trump looks like he’s complaining that Jalen Brunson got fouled to Adam Silver himself. This is legendary.
(via @wolfejosh, h/t @MLFootball)
How do three referees all miss a flagrant technical foul?
Wemby BULLIED Brunson
No foul was called
Keith Rabois reposted
However bad you think AI hallucinations are in law, I promise you that human hallucinations are worse
And AI is really good at catching human hallucinations
(from a document filed to the SEC by a major bank)
Keith Rabois reposted
Issue-spotting is a superpower in lawyering but kryptonite in entrepreneurship. Lawyers that can effectively switch that pessimistic lens on/off will be the lawyers that win in building the next generation of law firms.
Lawyers are trained in pessimism. It's a great litigation skill. It is a terrible entrepreneurial one.
This.
As I wrote this, I saw X go into meltdown over tokens.
You've seen the headlines: “Uber blows yearly AI budget in just one quarter.” “Meta employee burns 281 billion tokens in April.”
But, the problem isn't spending. Spending works. Since 2023, the top quartile of our AI spenders doubled their revenue. The bottom quartile? Flat.
It's blind spending. We don’t know which spend worked.
A sales team has qualified leads. A support team has resolved conversations. These are units you can measure against. All a token tells you is the meter ran, not whether the work was worth it or not.
Finance says, “half the budget,” engineering says, “double it” and you don’t know who’s right because there is no shared language of value. There’s no attribution, and no attribution means no allocation.
For example, right now, all work, no matter the size or shape, defaults to frontier models. But meeting summaries and calendar updates don’t require GPT-5.5 Pro.
In isolation this seems trivial, but re-route just 10% of a $10M AI bill from frontier to GPT-4 level intelligence you’ve saved nearly one million dollars. This sounds like a made-up stat — it’s not. It truly is that much cheaper.
This is the future of finance: not blindly rubber-stamping or rejecting AI spend, but allocating it with the same rigor companies apply to headcount.
Today, Ramp raised $750M at a $44B valuation.
Last time we grew this fast, we were 1/20th the size.
For 2000 years, business was built on two pillars. Today, a third: intelligence.
It’s your least governed cost. It’s also your single greatest opportunity.
Violation of CFIUS 101.
Airwallex is acquiring @Leapfin to help global finance teams close the books faster and with more confidence.
Together, we'll help businesses move from accepting payments all the way through to revenue recognition, reconciliation, and financial close.
🧡Welcome to the team, @Leapfin!
Excited to announce that @airwallex has acquired @leapfin.
The Leapfin team has built one of the strongest platforms for turning complex transaction data into clean, auditable financial records.
Their product solves a real pain point for modern finance teams, and we’re excited to bring their technology and expertise into Airwallex.
Excited to welcome @raymond_lau, @eya0 and the entire Leapfin team to Airwallex and build what comes next together.
Keith Rabois reposted
I find it incredible that Ramp's hiring bar is so high that they accidentally created a Neolab
Introducing Stack.
The AI operating system that lets accounting firms take on more clients without hiring. Learns your firm's process, runs the close, posts the journals. Fully auditable.
We’re living through the biggest shift in accounting since the spreadsheet.
Keith Rabois reposted
1. If anyone can pull this off it’s the @FactoryAI team.
2. Genuinely believe this could only be released in a post Composer 2.5 world (not saying you need this model - saying it’s a line in the sand where we’re seeing the level competence we all saw emerge last Nov/Dec but at significantly lower pricing).
3. Everyone is attached to this being solely about absolute cost saving but it’s actually about being able to operate at all points along the Pareto frontier. The Factory team are smarter than to get into a pure lowest provider knife fight.
I trust them when they say they’re trying to build the software factory of the future and this yet another development to stack on top of things like Missions, Deferred Context Engine, and even (conceptually) subagents with Droids - all well before others show up with these things.
Introducing model routing to Factory.
Factory Router picks the right model for every task, automatically.
Maintain frontier performance while cutting costs by 25%.
Keith Rabois reposted
masterclass on how to measure agent performance
builders.ramp.com/post/stack-benchmarking
Keith Rabois reposted
Today, Ramp raised $750M at a $44B valuation.
Last time we grew this fast, we were 1/20th the size.
For 2000 years, business was built on two pillars. Today, a third: intelligence.
It’s your least governed cost. It’s also your single greatest opportunity.
Keith Rabois reposted
Highest +/- in any Playoff Run in NBA History.
Brunson & Towns have a chance to go down as one of the greatest playoff duos of all-time.
Keith Rabois reposted
Ramp just raised more capital at a $44 billion valuation.
There may not be a better team execution-wise in the tech industry over the last few years.
Congratulations to @eglyman, @karimatiyeh, and the rest of the team.
Keith Rabois reposted
If it’s negative margin, it’s marketing not revenue
crazy how much of startups economics is buying tokens for $10m and selling them for $5m and bragging about $5m in revenue
Keith Rabois reposted
Redpoint has a presentation out that says 67% of enterprises are running open-source models in production (vs 23% in 2025).
Expect more enterprises to adopt the Dara approach. Use expensive models to explore. Bring in cheaper models/open source to optimise spends.
While it is a minority view, there is a chance that OpenAI/Anthropic may find revenue plateauing sooner than expected coz per token costs on a blended basis across models drops fast.
Dara (CEO of Uber) on their AI spend:
"We blew through our AI budget in a quarter, for the whole year. It is forcing us to adjust.
We are going to meter headcount increases because to the extent that my engineers are getting much more efficient, their throughput is increasing. There's a cost to that, and it's a significant cost.
AI adoption has been occurring in all parts of the business –– whether it's engineers and how they scope projects, how they build, debugging, platform migrations.
I'm pushing the teams to fundamentally use the power of AI to rebuild systems and processes from the bottoms up.
I do think it's a combination for us right now of encouraging adoption, but then driving efficiency.
We're using the more expensive models to explore. Once we scale some of these experiences, we'll look to bring in more efficient models that are more efficient on a token basis or are open source."
Keith Rabois reposted
Harvey just published a study showing a hybrid setup, open source GLM 5.1 as primary worker, routing to Opus 4.7 only when needed beats pure Opus 4.7 on quality and costs less.
This is the multi-model routing thesis proved in production on one of the hardest benchmarks in enterprise AI.
The insight isn’t that open source beat frontier. It’s that smart routing beat brute force. Using the most expensive model for every task is not a quality strategy. It’s a laziness tax. The teams building routing layers that send each task to the right model at the right cost are now demonstrably ahead on both dimensions simultaneously.
Inference optimization just became a first-class competitive advantage. Legal proved it first because the stakes forced the discipline.
We partnered with @FireworksAI_HQ to train open-source models for legal. Here's what we found:
1) Hybrid legal agents can beat frontier models on quality and cost by routing selectively to a frontier advisor.
We tested a hybrid setup where GLM 5.1 served as the primary worker, routing tasks to Opus 4.7 as an advisor when needed.
GLM invoked Opus sparingly, just 0.83 times per task on average.
The hybrid setup beat Opus on both quality and cost: 18% all-pass vs 14%, at $368 vs $954 across the same 100 tasks.
2) Post-training can push open models to frontier-level legal performance.
On a 100-task slice of our Legal Agent Benchmark (LAB), SFT moved Kimi 2.6's all-pass rate from 11% to 15%, beating Opus' 14%.
But the cost gap was even more striking: $84 vs $954 across the same 100 tasks, or ~11x cheaper.
We're excited to continue working with @FireworksAI_HQ on the next generation of open-source legal agents.
Keith Rabois reposted
Do it for Clyde
This team. This city. It's got the look. #alwaysknicks
Keith Rabois reposted
Jalen Brunson's clutch scoring in the playoffs is literally Michael Jordan-level
Keith Rabois reposted
We have literally achieved AGI before the failed state has developed the technology to count
The fact that California elections often can't be resolved for weeks is kind of insane and not common in other electoral systems around the world.
Keith Rabois reposted
What if AI is actually creating more jobs than it is replacing?
The latest JOLTs data showed that US job openings surged by a massive 731,000 jobs in April.
Markets were expecting no change, resulting in the largest beat in JOLTs history.
As a result, available employment hit 7.6 million for the month, the highest since May 2024.
And, job openings in the professional and business services sector surged by a massive 668,000.
The labor market's bull case from AI is underpriced.
Keith Rabois reposted
Thoughts:
1. Tech twitter/SV severely underestimates the accelerating financial pressure that the average person/business feels around AI spend
2. Gap between frontier and OSS models decreasing rapidly
3. The US needs OSS supremacy
Aura is clearly shifting to @FactoryAI
Each enterprise I've spoken to is spending their AI token budget quicker than they anticipate... @FactoryAI and @matanSF just came out with model routing which will help lower costs and save tokens without sacrificing performance. Will be huge as AI spend continues to explode!
Introducing model routing to Factory.
Factory Router picks the right model for every task, automatically.
Maintain frontier performance while cutting costs by 25%.
Keith Rabois reposted
This is absolutely true.
Large foundation models save on labor costs at the expense of operational costs.
It's an "Exploration-Exploitation" problem. Explore with large, general-purpose models and Exploit with more cost effective ones.
Automatic is the best of both worlds!
Keith Rabois reposted
Model routing is not that easy for adhoc chats. These prompts are too short and far from what we see in the real world. Only when you have a full prompt for a wyorkflow, model routing makes sense. Getting prompt caching is very important as most caching saves you 60-80% cost. You can't simple route each message in a chat to a different provider let alone different model. Each model has very different restrictions on images, tool calls, etc so you need to choose the model family and stick with it.
Introducing model routing to Factory.
Factory Router picks the right model for every task, automatically.
Maintain frontier performance while cutting costs by 25%.
Keith Rabois reposted
Obviously when your candidate loses an election 51-49% that’s tough but a runner-up bad feeling is when a candidate you thought had strong support ends up getting 9%, makes you realize how out of touch your circle was.
Keith Rabois reposted
The market has OpenAI's story wrong. It is not a capital burn narrative. It is a margin expansion story with a large spend number on top.
@thefriley on @theallinpod: GPT-4 to GPT-5.4, token costs dropped 97%. GPT-5.5 prices went up 2x. Customers still got cheaper per-token compute.
900M weekly users. Revenue going 50/50 consumer-enterprise. Codex at 5M users from near zero in January 2026. The fastest-growing Codex adoption inside OpenAI is not engineering - it is go-to-market.
Margin hiding inside capex: podcastalpha.substack.com/p/openai-cfo-sarah-friar-on-ipo…
Source: All-In Podcast - www.youtube.com/watch?v=TjrShuj_Zsg
Keith Rabois reposted
is anyone in legal covering this? this seems like it should be really big news. AI seems well poised to conduct legal editorial work at above human capabilities. whoa!
A group of contracts professors wrote some short-answer contracts questions, wrote answers, and then did a blind-test comparison between AI-written answers and the answers of the other professors.
The AI-written answers were generally rated as better.
papers.ssrn.com/sol3/papers.cfm?abstract_id=6849678
Keith Rabois reposted
here's the agentic setup I follow - lmk if there are other ways to make it better:
1) one reason I stuck with @droid - I believe in their core philosophy: intelligence lives in the harness, not the model alone.
on top of using droid, I also built a mini-internal harness that every CLI coding tool I use can read from. 13 contextual files in a "brain model" that capture how I think, reason, and make decisions. every session loads the right slice before the first prompt. the model is interchangeable. the harness isn't.
2) agents need onboarding like people.
AGENTS.md files at every level - root, project, sub-project - job descriptions + onboarding docs + operating rules.
structured folders that document WHY I like what I like and WHY I avoid what I want to avoid.
the reasoning, not just the rules.
when I see agents drifting, I make them read that folder, and they mostly come back on track (in droid, they usually do). custom droids (6 of them, each with one job) load their relevant context the way a new hire reads their team wiki on day 1. and they don't just generate text - they read my gmail, update my sheets, pull live keyword data, browse linkedin. agents that can act, not just talk.
3) for repetitive tasks, custom slash commands. for repetitive workflows (like deep prospect research), custom droids.
one command, the whole workflow runs.
4) I ask agents to document their reasoning, not just their output. progress files, thought logs, handoffs between droids. when the next session picks up, it's not starting from zero. it's picking up from where the last one left off. that's the difference between an agent that's useful and one that's
disposable.
on models - I route manually right now (glm-5.1 for coding, kimi-k2.6 for brainstorming and image tasks, opus 4.8 for deep subject-matter dives, codex for debugging, mini models for quick stuff). but I care less about which model and more about whether the right context reached it.
gotta try factory's model routing now - no wonder I rushed to start my work today :D
Thank you Krishna!
What is your agentic setup like?
Keith Rabois reposted
this is pretty genius. in a world of increasingly expensive and abundant ai models products like this are a dream
AI model usage has shifted in a huge way recently:
- flat fee subscriptions to usage-based models
- single model usage to multi-model prompt routing
why? subscriptions were too heavily subsidised and different models are good for different things
every fortune500 company has BOTH a claude AND chatgpt sub. engineers are using claude to write a plan and codex to execute
factory router claims to save you 25% cost while maintaining frontier results.
if you’re a company spending $100M per year on tokens (and there will be many) then saving $25M is a a no-brainer
for routing companies like factory they then build a huge moat around intents.
v smart (and no, i’m not an investor, i just like the tech)
Introducing model routing to Factory.
Factory Router picks the right model for every task, automatically.
Maintain frontier performance while cutting costs by 25%.
Keith Rabois reposted
Astute.
Lawyers, the generalist era is over. The barbell is here. Elite specialists on one end. Volume commodity shops on the other. The middle is dangerous.
Keith Rabois reposted
Was just served an ad from @Opendoor on @instagram, clicked “get offer” and the transition and offering page was so sleek and professional - huge upgrades from previous iterations! It screams that the company has its stuff together and can be trusted. Kudos @morganb and team!
Keith Rabois reposted
This is one of the most subtly important releases we've made.
Models are ultimately commodities. Having a liquid market for them let's us achieve performance that is strictly better than any frontier model on every dimension that matters: price, quality, and reliability.
Introducing model routing to Factory.
Factory Router picks the right model for every task, automatically.
Maintain frontier performance while cutting costs by 25%.
Every parent needs help juggling. AI to the rescue.
AI can now make you a great parent.
Introducing Ollie: the world’s first AI family assistant that manages your family life better than any human.
Here’s how it works:
Keith Rabois reposted
If your daughter needs tutoring in algebra, you can probably find someone cheaper than Albert Einstein.
Giving every task to GPT5.5 or Opus 4.8 is overkill. Often times you can get the task done just as well, but 10x cheaper and 10x faster with a smaller model.
Introducing model routing to Factory.
Factory Router picks the right model for every task, automatically.
Maintain frontier performance while cutting costs by 25%.
Worth understanding.
There are 12 venture-backed companies that raised >$3B in private markets and then listed in the US.
Only one that has a positive return today, relative to the S&P 500's performance over the same period.
Most are deeply negative (aggregate -117%).
Only two that were positive at lockup expiry, neither stayed that way.
The high cost of private capital means the companies that raise the most, and stay private longer, are almost inevitably overvalued as insiders raise the price of funding events aggressively to stay NPV positive.
Keith Rabois reposted
Excited for 1517 Fund to continue to support the Mach team.
I awarded Ethan Thornton a $1k grant to work on a rifle. he was still in high school. He had to leave physics class to take the call with me.
Four months later, we invested $100k to help him get Mach started.
Be the first check, not the last.
BREAKING: @Mach_Industries secured a $300M Series C led by @InfiniteXYZ and @RibbitCapital, according to Bloomberg.
Keith Rabois reposted
Mach is really flying.
They've ~$200M of booked YTD, tracking to ~$550M+ for '26, up from effectively $0 '25.
Wouldn't be surprised if there's another large round before EOY.
Congrats to @Mach_Industries!
Defense tech darling Mach Industries hits $1.8B valuation, a 4x jump in a year techcrunch.com/2026/06/01/defense-tech-darling-mach-indus…
Keith Rabois reposted
🚨 The fintech firm Airwallex has become deeply embedded in our financial technology ecosystem while maintaining close ties to China.
We cannot allow Americans’ sensitive data to be placed within reach of the CCP.
Congress must investigate.
www.thecentersquare.com/national/article_e212f8aa-f6ae-49…
This is mostly correct.
some thoughts on kirkland building its own harvey
1) kirkland is spending $500m over four years in order to build its own internal ai legal tools; kirkland intends to spend $100m this year
2) i suspect that kirkland is doing this because they have told themselves that they have valuable data and because they want to appear differentiated
3) i think the first issue is that kirkland probably does not have differentiated data from other elite law firms; at least, not at the level a harvey would absorb
4) all the elite firms probably have similar internal workflow data and so long as some of them defect, that is enough to commoditize the data kirkland wants to use for its platform
5) and, to the extent that they do have different internal workflows, harvey and legora will end up representing a better version of them and this will put kirkland at a disadvantage
6) moreover, companies like kirkland will have difficulty building their internal legal platforms because they do not have experience with software development
7) and, there are both cultural and structural issues with them managing software developers, like they cannot give non-lawyers equity in the firm due to regulation
8) so, i think firms like kirkland are better off using tools like harvey and legora and then looking to focus on where their value really is now: client relationships, local knowledge (litigation, regulation) and legal r&d (novel structures, etc...)
9) anyway, this seems to me like a phenomenon that ai creates across a lot of industries, where firms that were previously vertically integrated become unbundled due to ai because part of the intelligence gets moved to the labs or otherwise gets commoditized
10) and so, a new set of companies are created whose job it is in order to provide services complementary to the labs: forward deployed like harvey and legora and data providers like mercor, surge and handshake
Kirkland & Ellis, the world's highest-grossing law firm, is setting aside $500M to build its own AI platform rather than rely on tools available to its rivals (Financial Times)
(Visit Techmeme dot com for the link and full context!)
Keith Rabois reposted
My first company went to Apple. My second, the largest short-video platform in Latin America, went to TikTok. My third is still running as Vietnam's biggest music streaming platform.
I built all three because I could see what was coming before anyone else did.
The vibe coding generation is not watching anymore. They are building. Games. Arts. Memes. Interactive worlds. No code. Just an idea and a phone. 15 million things built on Sekai. 200,000 more today.
We just raised $26M led by @rabois at @KhoslaVentures and @Nik_Quinn at @ConnectVentures (CAA) with Mayfield, A16Z speedrun, A*.
www.axios.com/2026/06/01/sekai-mini-app-startup-funding
Very excited to invest in the highest potential consumer application of AI since ChatGPT.
My first company went to Apple. My second, the largest short-video platform in Latin America, went to TikTok. My third is still running as Vietnam's biggest music streaming platform.
I built all three because I could see what was coming before anyone else did.
The vibe coding generation is not watching anymore. They are building. Games. Arts. Memes. Interactive worlds. No code. Just an idea and a phone. 15 million things built on Sekai. 200,000 more today.
We just raised $26M led by @rabois at @KhoslaVentures and @Nik_Quinn at @ConnectVentures (CAA) with Mayfield, A16Z speedrun, A*.
www.axios.com/2026/06/01/sekai-mini-app-startup-funding
Keith Rabois reposted
looking back at the 1994 Knicks Finals run
Keith Rabois reposted
1999 Finals intros
Chills
OMG.
Can you imagine being fried frank and also announcing an AI build yesterday?
Iykyk
Keith Rabois reposted
Its funny how the highest ranking gay man to ever hold office is a Republican, and he's more a damn based motherfucking bad ass than the "straight" guys the Democrats got to try to "relate" to us.
Reporter: "You wrote 'resilience' on your Cabinet meeting notes... why did you write all those notes?"
@SecScottBessent: "So people could look over my shoulder, photograph them, and think they got a scoop." 🤣
Keith Rabois reposted
Tom Clancy wrote Executive Orders in 1996. Reread it today and you cannot put it down.
The premise: an outsider with no political pedigree gets handed the presidency after the establishment is decapitated. He is surrounded by a hostile media that mocks him for not knowing how DC works. The institutional class treats him like a usurper. He has to rebuild a government from scratch while the country is at war.
A depleted navy, a lack of US Merchant Marine ships in Diego Garcia causing global chaos, Navy and Merchant Ships battling to get through Hormuz. Powerful submarines useless in choke points.
That book was published thirty years ago.
THIRTY
And it absolutely destroys the DC elites and how congress works and how badly it needs reform.
He nails China and India too… working against our interests behind the scenes while Gulf states are hedging. Europe is shrieking. Migration pressure is building again.
Not to mention the pentagon bureaucratic resistance. The leakers. The military journalists who decide their job is to break a presidency rather than cover one. The defense officials slow-walking decisions. The allies whining about American power and then beg for it the moment shipping stops moving.
Read it again. Or read it for the first time.
It’s a really good read.
The only thing he got really wrong is that CIA agents make terrible politicians as Slotkin and Spanberger both proved this year.
And yeah… it has ebola too.
I’ve been warning about an Ebola lockdown in the US since 2021. See my tweet below from 2023. 👇🏻
I hope you’re all ready for it to soon be announced that Ebola is in the United States of America. It will likely be announced between now and the week or 2 AFTER the upcoming World Cup in the US, which is in 14 days on June 11, 2026.
How many Africans will be traveling to the US for the World Cup?
Too many.
Today, the US State Department @StateDept quietly announced they are allocating another $80 million toward the Ebola response. @SecRubio
Source: www.state.gov/releases/office-of-the-spokesperson/2026/05…
With this new $80 million commitment, the State Department has mobilized more than $112 million in bilateral foreign assistance for the Ebola response in less than two weeks.
There’s a reason for this.
It’s coming to America and they need to at least show they *tried to stop it*.
But, the thing is, Ebola is already here… as I’ve been reporting for the last 2.5 weeks.
You’ll see.
Bookmark this tweet.
This is quite sharp.
live in sf and have this conversation literally a year late
:)
A day in the life of a VC in 2026:
9am: Board meeting. My main value-add is aggressively pushing for Anthropic/OpenAI usage in non-engineering functions. Briefly ponder how I became an SDR for foundation model companies.
1pm: Lunch with another VC. We discuss how startups can find "blue ocean" away Anthropic/OpenAI. We conclude we should probably just invest the rest of our funds directly into Anthropic/OpenAI.
3pm: Pitch meeting. Me: "Do you run on Anthropic or OpenAI?" Founder: "Both." Debating internally whether a company reselling Anthropic/OpenAI with a 10% gross margin is a good investment but hey, at least they're in the "token flow".
4:30pm: Deep due diligence. I ask Claude if it plans to build this exact startup natively in its next release. Same to ChatGPT. They both say yes. I pass on the deal.
6pm: Urgent call from a portco CTO: "We need an intro to upgrade our Anthropic tier!" I immediately agree to help them spend more of the venture dollars we just invested in them, on Anthropic.
8pm: Brainstorming next guests for the podcast. Thinking I should probably just try to get some folks from Anthropic and OpenAI.
Keith Rabois reposted
damn keith right yet again
airwallex turned out to be a ccp front
This aged well.
1/
Cool growth chart. Have you disclosed to US customers like @Rippling, @Billcom, @TheZipHyouQ, @brexHQ, and @Navan that you’re quietly sending their customers’ data to China?
Airwallex has become a Chinese backdoor into sensitive American data like from AI labs and defense contractors. You must already know this, but your China-based ops, infra, and investors create legal obligations to assist with CCP espionage upon request.
Through Airwallex, Beijing can access:
•supplier payments for AI labs
•payroll data for defense contractors
•personal data for employees abroad
Obviously many companies do business in China, and that’s not inherently a bad thing. But your company has become a guaranteed vector for data transfer to the Chinese government, and that’s a different thing entirely.
You have multiple points of vulnerability:
> people (key execs and core engineering team is based in mainland China, which obligates these individuals to comply with Chinese government demands to hand over data)
> legal structure (company leaders are subject to Chinese national security law)
> cap table (over 20% of your company is Chinese-owned, including Tencent, HongShan, and others, which further obligates you to comply with CCP requests).
What’s happening:
> You route global payments for US companies in critical sectors, without disclosing that you are under Chinese jurisdiction
> You moved HQ to Singapore, but your largest operational footprint is in China and hundreds of your engineers in mainland China touch production payment systems
> You are subject to Chinese law that requires Airwallex employees to support CCP intelligence requests and quietly hand over data when asked
> You hide this from your customers, but you are well aware of your obligations to China and that is why you insist on protection of Chinese data access in your contracts
Thanks to you, the Chinese government now has direct, covert, legally enforceable access to sensitive financial information belonging to America’s AI labs, defense contractors, financial institutions, healthcare firms, and Fortune 500s. Maybe this wasn’t your intent when you started the company, but it’s clear you’ve allowed this to happen.
This aged well.
1/
Cool growth chart. Have you disclosed to US customers like @Rippling, @Billcom, @TheZipHyouQ, @brexHQ, and @Navan that you’re quietly sending their customers’ data to China?
Airwallex has become a Chinese backdoor into sensitive American data like from AI labs and defense contractors. You must already know this, but your China-based ops, infra, and investors create legal obligations to assist with CCP espionage upon request.
Through Airwallex, Beijing can access:
•supplier payments for AI labs
•payroll data for defense contractors
•personal data for employees abroad
Obviously many companies do business in China, and that’s not inherently a bad thing. But your company has become a guaranteed vector for data transfer to the Chinese government, and that’s a different thing entirely.
You have multiple points of vulnerability:
> people (key execs and core engineering team is based in mainland China, which obligates these individuals to comply with Chinese government demands to hand over data)
> legal structure (company leaders are subject to Chinese national security law)
> cap table (over 20% of your company is Chinese-owned, including Tencent, HongShan, and others, which further obligates you to comply with CCP requests).
What’s happening:
> You route global payments for US companies in critical sectors, without disclosing that you are under Chinese jurisdiction
> You moved HQ to Singapore, but your largest operational footprint is in China and hundreds of your engineers in mainland China touch production payment systems
> You are subject to Chinese law that requires Airwallex employees to support CCP intelligence requests and quietly hand over data when asked
> You hide this from your customers, but you are well aware of your obligations to China and that is why you insist on protection of Chinese data access in your contracts
Thanks to you, the Chinese government now has direct, covert, legally enforceable access to sensitive financial information belonging to America’s AI labs, defense contractors, financial institutions, healthcare firms, and Fortune 500s. Maybe this wasn’t your intent when you started the company, but it’s clear you’ve allowed this to happen.
Keith Rabois reposted
Our new benchmark for high finance... three thoughts:
1. no single model dominates financial tasks (and gpt5.5 does far better than X sentiment would lead you to believe)
2. model routing across providers can improve workflow quality by 40%
3. most llm benchmarks either sacrifice reasoning auditability for accuracy, or sacrifice accuracy for reasoning auditability... real finance workflows need both
Which AI model is best for the financial work our customers do? We built the Big Finance Bench to find out.
928 questions written by ex-finance practitioners, scored against more than 15,000 rubric criteria, used to rigorously analyze ten frontier models across the workflows that run inside banks, PE firms, and asset managers. The first benchmark of its kind for finance.
The findings surprised us. Read more on our blog.
Keith Rabois reposted
Kathy Hochul: Can Donald Trump even name the "1993 Knicks championship team" that doesn't exist??
President Trump: Literally catching Charles Oakley in his arms while sitting courtside in the '90s.
Keith Rabois reposted
Donald Trump making a free throw in 2005 with Knicks players at MSG 🏀🔥
Keith Rabois reposted
President Donald J Trump at New York Knicks games over the years.
James Dolan the owner of the Knicks is a longtime friend of Trump. Even invited him to the NBA finals games.
Cry harder Democrats. The Knicks love Trump, he loves the Knicks!
Fascinating. Legal, accounting and investment banking should thrive as vertical applications of AI, especially if you run on factory.ai too.
Today we're sharing our first research collaboration with @baseten on open-weight legal agents.
Using signal from LAB (our Legal Agent Benchmark of 1,200+ tasks across 24 practice areas), we post-trained an open-weight model to match closed-source frontier performance.
Training open-weight agents for legal creates three major advantages for Harvey:
1) Cost and latency improvements:
The best-performing closed-source frontier models took an average of 22 minutes to complete each LAB task, with $50 per-task average inference cost.
Open-weight inference is substantially faster and cheaper.
2) Reasoning visibility:
In a high-stakes domain like legal, it's incredibly important to get visibility into agents' internal reasoning states - for audit and governance and also as a lever for Harvey to improve agent performance.
Closed-source foundation model providers avoid exposing raw reasoning tokens via API to prevent model distillation. For Harvey, that visibility is a major advantage.
3) Custom training:
Owning the model weights lets us customize training and modify architecture. One example: our blog’s final note on training a custom KV cache compactor.
More to come on our research collaboration with Baseten.
Keith Rabois reposted
I am thrilled to share that @lilatretikov has joined the Board of Directors at Factory.
Lila was previously CEO of Wikipedia, deputy CTO of Microsoft, currently sits on the boards of UBS, Capgemini, and Volvo, and is Head of AI at NEA.
Madison Faulkner is stepping down from @FactoryAI's Board of Directors to join us as Head of Strategic Initiatives.
It is exceedingly rare for an investor to join a portfolio company. It takes unprecedented* conviction for a Board Director - one that led the Series B - to join.
Keith Rabois reposted
Thrilled to welcome @jmwind to the Spellbook team. JML is one of the top technical executives in the world, having worked with Shopify as CTO and Atlassian as VP Engineering, bringing them both to IPO back-to-back.
Over the past year, I’ve been lucky to have him as a mentor and advisor. He is one of the most impressively “30,000 foot view” + “in the weeds” people I’ve ever met. In the same hour he can code a prototype and deliver deep insight about how to scale our culture & process.
This makes him uniquely suited for the role he is pioneering: Executive IC. AI enables people like JML to scale themselves 100x without getting bogged down in layers of management. We aim to write the playbooks for tomorrow, rather than to execute the playbooks of yesterday. We think that org charts will change radically in the next decade, and we want to get far ahead of it.
JML will pioneer the systems that scale our AI-native org, will deliver product and technical insight from decades of experience and help grow myself and our executive team.
I recommend following him if you are curious how AI-native orgs will operate–he is writing a playbook for the decade to come: @jmwind
www.theglobeandmail.com/business/article-spellbook-ex-sho…
Keith Rabois reposted
I have only seen this a few times in my career
A board member leaving their cushy VC job to jump into the trenches with a company
BULLISH on @FactoryAI
Madison Faulkner is stepping down from @FactoryAI's Board of Directors to join us as Head of Strategic Initiatives.
It is exceedingly rare for an investor to join a portfolio company. It takes unprecedented* conviction for a Board Director - one that led the Series B - to join.
Super cool. Build faster and less expensively!
Today, I’m launching my newco, NavigateAI. We are short hundreds of thousands of skilled workers and we're on a mission to give every field worker an AI copilot, so they can build faster and better when we need it most. www.navigate.ai/blog/2026-05-26-launching-navigateai
Keith Rabois reposted
"The Wiz" electronics catalogue from June 25, 1999, the last time the New York Knicks played an NBA Finals game.
Keith Rabois reposted
Complete list of all players in NBA history to post a plus/minus above +200 over a ten-game span in the playoffs:
Jalen Brunson (2026)
-- end of list --
Holiday reading.
Keith Rabois reposted
There are 400 golf courses in the Arizona desert alone. They use 42 billion gallons annually, more than 300,000 families.
STFU with the data center propaganda when we are racing the world for Space and AGI.
5 gallons of water =
→ $132 from data centers
→ 2¢ from almonds
One powers AI, auto safety, space exploration, medical research, banking, and innovation.
The other is a mostly exported (75%) snack from drought-prone California fields.
Resource allocation in one chart.
Keith Rabois reposted
Keith Rabois reposted
“I think he is a more mature, disciplined version of himself in the first term” Jeff Bezos on President Trump right now. “Trump has a lot of good ideas and has gotten a lot right and you have to give credit where credit is due”
Keith Rabois reposted
The Comeback
From the SkyCam
Full Garden sound
Keith Rabois reposted
At some point Anthropic talked to me informally about potentially joining their Board. I wasn’t interested and wouldn’t have been a good fit. But I did send Dario and Daniela a copy of Aristotle’s “Politics.” Unfortunately, I worry they’ve been too busy to read it.
Dario Amodei: Ideology Won't Survive the Reality of AI
"We're going to find that ideology will not survive the nature of this technology. The things I'm talking about are gonna become bipartisan and universal because everyone will recognize the necessity of it." — @DarioAmodei
Keith Rabois reposted
$OPEN now makes offers in 24-48 hours. One person underwrites where eleven used to. AI handling the rest.
Margins up every month since September. Acquisitions up 6x YoY. Headcount down.
This is what an AI-native operating model actually looks like.
Keith Rabois reposted
$OPEN built a working mortgage product in 6 weeks with 3 engineers. Underwriting a home used to take 11 people, now it takes 1. Headcount down, acquisitions up, margins up. This is what AI actually doing something looks like.
Keith Rabois reposted
"My whole life is consistent. Everything I do. From what I eat, to when I sleep, to my recovery, to my loved ones. Everything's consistent. It's a routine for me at this point.” — Shai Gilgeous-Alexander
As I write in my new book, become known for your consistency; when you win at consistency you give yourself a good shot at winning at everything else. www.amazon.com/Way-Excellence-Greatness-Satisfaction-Chao…
Breaking: Oklahoma City Thunder's Shai Gilgeous-Alexander has won his second consecutive NBA Most Valuable Player award, becoming the 14th player in league history to win back-to-back MVPs, multiple sources tell ESPN.
Keith Rabois reposted
"In short: in 38 days, we rolled back 40 years of Iranian military investment.”
"In less than 40 days of major combat operations, USCENTCOM forces systematically dismantled what Iran spent four decades and tens of billions of dollars building."
"The capabilities on which the regime relied to threaten our forces, coerce our partners, and project power across the region have been substantially degraded. Combined with the damage Operation MIDNIGHT HAMMER inflicted on Iran’s nuclear program, USCENTCOM assesses that Iran can no longer project power across the region, nor pose the persistent threat to the United States or our partners that it did prior to Operation EPIC FURY."
"Iran can no longer reliably arm or resupply Lebanese Hezbollah, the Houthis, Hamas, or militia groups in Iraq with advanced weapons...This dynamic presents an opportunity for a generational shift in the regional balance of power."
"We damaged or destroyed over 85 percent of Iran’s ballistic missile, drone, and naval defense industrial base. More than 1,450 strikes on weapons manufacturing facilities set the regime’s ability to build and stockpile ballistic missiles and long-range drones back by years. The factories and technical workforce that produced Iran’s ballistic missiles, long-range attack drones, and naval platforms have been degraded to the point that Iran cannot replace its lost capabilities in the near term."
"In the air domain, Iran’s air and air defense forces are functionally and operationally irrelevant. Before OEF, the Iranian Air Force flew between 30 and 100 sorties each day. Today that number is zero. We destroyed or rendered non-mission-capable Iran’s fixed-wing airfields, hangars, fuel storage, and munitions stockpiles, and we knocked out 82 percent of its air defense missile systems along with the radar and command architecture that tied them together."
"At sea, we destroyed 161 vessels in total across 16 classes of warships, effectively crippling the regime’s ability to operate. We eliminated more than 90 percent of Iran’s once-massive inventory of over 8,000 naval mines, with more than 700 airstrikes on Iranian naval mine targets. In sum, Iran’s navy can no longer claim to be a maritime power, and it cannot project into the Gulf of Oman or the Indian Ocean."
"The second-order effects of OEF are significant. More than 2,000 strikes against Iran’s command-andcontrol structures created leadership vacuums, paralysis, and internal confusion. We have seen reporting of desertions, personnel shortages, and signs of regime desperation in their attempts to compel discipline through arrest and execution."
"In short: in 38 days, we rolled back 40 years of Iranian military investment.”
Keith Rabois reposted
This may be the song of the summer.
Highly recommend this advice for founders.
"You get instant gratification from making progress in the way that people think progress is made.
You don't get instant gratification from being like, "I think you're wrong, and I know the business better than you do, and really I need to go do this other thing, but that's gonna take me six months to prove you wrong."
That takes a lot of conviction and a lot of discipline to be able to do that."
Put the blinders on and ignore the noise.
Keith Rabois reposted
I’m excited to announce our new company TrueShort!
An AI studio and streaming app. We create and release our own original movies and shows
In 6 months, we’ve hit:
- $2.4M annualized revenue
- 2M+ minutes watched
- #9 in the App Store for News
Here’s the story 👇
Keith Rabois reposted
We’ve raised $12M to hire the next generation of filmmakers at TrueShort!
Led by Keith Rabois at Khosla Ventures and participation from Jeffrey Katzenberg (Dreamworks), Ravi Nandan (A24), Scott Belsky (Behance), General Catalyst, and more
Here’s the story 👇
Keith Rabois reposted
A news reporter asked Michael Jordan if he thought the ’90s Bulls could beat LeBron’s Lakers.
MJ: Yes.
Reporter: By how much?
MJ: Two or three points.
Reporter: Why so close?
MJ: Most of us are almost 60 now.
Keith Rabois reposted
Most points scored in 1st half of a road playoff game in NBA history:
Knicks 83 vs Hawks, Game 6 2026
Knicks 81 vs 76ers, Game 4 2026
Nuggets 81 vs Suns, Game 6 2023
Keith Rabois reposted
we trained a .35b-parameter model to navigate spreadsheets better than opus 4.6.
normal corporate card company stuff.
We partnered with @PrimeIntellect to build Fast Ask, a small RL-trained subagent that helps our Sheets agent find answers in spreadsheets. It scores +4% over Opus on exact match accuracy at Haiku latency.
Keith Rabois reposted
OPEN cohort thesis just went from claim to track record.
Aged inventory: 33% → 10% in one quarter.
Acquisition contracts: 2x QoQ — highest since 2022.
Triple consensus beat.
Kaz: "Adj EBITDA profitable on a 12-month go-forward basis starting April 1. October cohort was just the start."
EMJ + EMJX long $OPEN.
Keith Rabois reposted
.@RampLabs (the AI unit of @tryramp) has been *cooking* with agentic innovation
Here's @a_levitator discussing and demo'ing code self-maintaining software and the concept of AI software factories #datadrivennyc
______________
00:04 - Intro
01:11 - The shift from writing code to code maintenance
01:59 - Introducing Ramp Inspect, the background coding agent
03:05 - The first experiment: Nightly AI code automation
04:23 - The limits of stateless monitoring in large observability surfaces
05:47 - Using Datadog monitors to give the AI state and focus
07:23 - Real-world example: AI autonomously fixing an authentication bug
08:14 - How to control noise and implement an AI triage pattern
09:27 - The old vs. new paradigm for continuous code observability
10:21 - Key learnings on building autonomous AI software factories
Keith Rabois reposted
Agree, the latent demand for legal services is gigantic. Getting my first legal bill was a gut punch 15 years ago, and that’s why I started @SpellbookLegal.
However, a huge amount of routine business legal work will be absorbed in-house with legal AI.
Most contracts are reviewed in-house. All the context is in-house. The APIs are in-house. The collaborators are in-house. The accounting system is in-house. The company strategy is in-house. AI can scale your in-house legal strategy to infinity.
It doesn’t make sense to outsource a company’s legal intelligence layer for routine deals, only non-routine or brain surgery matters where there is no in-house policy.
as cost of legal services goes down, demand will explode. right now i pay 0$ because it's just never worth it as the start cost is insanely high, but i'd spend much more if cost went down
Keith Rabois reposted
Palantir just broke the mold. It just reported 85% growth at $6.5B+ ARR. And it's still ... accelerating.
RPO is up 134%! We've never seen this before in B2B.
Top 5 learnings:
1/ Growth re-accelerated. Again. At scale
Annual growth went 17% → 29% → 56% → 71% (guided, raised today). At a $6.5B+ run rate.
The "law of large numbers will catch up to them" thesis got buried this morning. Q1 was sequentially +16%. Faster than most $20M ARR startups grow.
2/ Rule of 40 = 145%
Karp's letter was blunt: this number has only been matched by AI infrastructure companies. NVIDIA. Micron. SK hynix.
Not software companies. Chip companies.
3/ US Commercial up 133% YoY
US Commercial revenue: $595M, up 133%. New logos this quarter: Airbus, Bain, GE Aerospace, Stellantis.
These are exactly the legacy industrial and consulting buyers everyone said were too slow, too analog, too "pilot purgatory" to deploy AI at scale.
They're now writing $5M, $10M, $20M+ checks for AI workflows. The Fortune 500 stopped piloting. They started buying.
4/ RPO at $4.45B, +134% YoY. Backlog growing faster than revenue
Q1 2025 RPO: $1.90B. Q1 2026 RPO: $4.45B.
That's $2.55B more in non-cancelable contracted future revenue. Up 134%. Faster than revenue itself, which is the cleanest possible signal that 2026 and 2027 will be even bigger than what we just saw.
206 deals at $1M+, 72 at $5M+, 47 at $10M+ in a single quarter. Forty-seven $10M+ deals in 90 days.
5/ $1.5M revenue per employee
Getting even more efficient.
This is the AI-native operating model showing up on the income statement. Not three separate stories (growth, profitability, cash flow). One story told three ways: AI leverage replaces linear headcount.
Revenue-per-employee is the new ARR-per-rep. If you're hiring linearly to grow in 2026, you're going to lose to companies treating headcount as the constraint to engineer around.
Every growth metric over 80% except the customer count, backlog growing faster than revenue, chip-company margins on software-company growth, and Karp telling CNBC the US business doubles again in 2027.
Acceleration like we've never seen in B2B before.
OMG. Brunson.
Keith Rabois reposted
Right before I hit the road, I had an opportunity to interview @rabois and get his perspective on how he and other VCs are looking at the legal AI market (Keith is in an investor in @SpellbookLegal). And to spice things up a bit, we also talked about his prediction that the war in Iran would end in 4 days.
00:00 - Zach's Hot Disruptive Take: Why VCs are pouring money into the space
04:26 - Keith’s Legal Roots: From Sullivan & Cromwell litigator to Tech VC
07:48 Following the Talent: Why elite founders are finally interested in law
10:46 - The Death of Traditional Software? Moats, integrations, and AI agents
12:57 - Why I Invested in Spellbook (and why most Legal AI categories are misleading)
15:23 - The Billable Hour Problem: Why improving productivity is counter to Big Law’s model
18:21 - Visibility vs. Exposure: The trillion-dollar market of commercial agreements
21:40 - Venture Capital without Lawyers? The future of seed rounds and Series A
24:00 - Terminal Value: Will vertical AI companies exist in 10 years?
25:23 - On "Missing Out": Dealing with LPs and being dubious about Harvey
29:16 - The "4-Day War" Prediction: Analyzing geopolitical conflict like a chess board
34:11 - Trump, Tariffs, and the Existential Threat of the CCP
37:54 - The New American Century: AI, Energy, and Supply Chain independence
40:00 - Export Controls & Nvidia: The Dwarkesh vs. Jensen Huang debate
Conversation also available on YouTube, Apple & Spotify.
Keith Rabois reposted
Harvey is valued at $11B. Legora just raised at $5.5B. I built their entire web application in two weeks and I'm making it open-source and free for everyone to use. Say hi to Mike: mikeoss.com.
When I got the chance to try Harvey and Legora, I was surprised by how simple they were. A thought came to mind: I could probably build something similar in no time at all with Claude. And so I did.
Assistant, project, tabular review and workflows. You get it all without vendor lock-in.
Mike offers law firms an alternative, where they own the application layer and aren't stuck with a vendor they're renewing forever.
You can try Mike in the demo on the website, or go to the GitHub link on the site to download the code and run a local version yourself.
Keith Rabois reposted
A core complex product function at WithCoverage, over 1000 engineering hours, is being completely torn out right now
And replaced with a model doing the function end-to-end
LLMs got that much better this year. They will be another step-change better next year
Fireside chat at Sequoia Ascent 2026 from a ~week ago. Some highlights:
The first theme I tried to push on is that LLMs are about a lot more than just speeding up what existed before (e.g. coding). Three examples of new horizons:
1. menugen: an app that can be fully engulfed by LLMs, with no classical code needed: input an image, output an image and an LLM can natively do the thing.
2. install .md skills instead of install .sh scripts. Why create a complex Software 1.0 bash script for e.g. installing a piece of software if you can write the installation out in words and say "just show this to your LLM". The LLM is an advanced interpreter of English and can intelligently target installation to your setup, debug everything inline, etc.
3. LLM knowledge bases as an example of something that was *impossible* with classical code because it's computation over unstructured data (knowledge) from arbitrary sources and in arbitrary formats, including simply text articles etc.
I pushed on these because in every new paradigm change, the obvious things are always in the realm of speeding up or somehow improving what existed, but here we have examples of functionality that either suddenly perhaps shouldn't even exist (1,2), or was fundamentally not possible before (3).
The second (ongoing) theme is trying to explain the pattern of jaggedness in LLMs. How it can be true that a single artifact will simultaneously 1) coherently refactor a 100,000-line code base *and* 2) tell you to walk to the car wash to wash your car. I previously wrote about the source of this as having to do with verifiability of a domain, here I expand on this as having to also do with economics because revenue/TAM dictates what the frontier labs choose to package into training data distributions during RL. You're either in the data distribution (on the rails of the RL circuits) and flying or you're off-roading in the jungle with a machete, in relative terms. Still not 100% satisfied with this, but it's an ongoing struggle to build an accurate model of LLM capabilities if you wish to practically take advantage of their power while avoiding their pitfalls, which brings me to...
Last theme is the agent-native economy. The decomposition of products and services into sensors, actuators and logic (split up across all of 1.0/2.0/3.0 computing paradigms), how we can make information maximally legible to LLMs, some words on the quickly emerging agentic engineering and its skill set, related hiring practices, etc., possibly even hints/dreams of fully neural computing handling the vast majority of computation with some help from (classical) CPU coprocessors.
@karpathy and I are back! At @sequoia AI Ascent 2026. And a lot has changed. Last year, he coined “vibe coding”. This year, he’s never felt more behind as a programmer.
The big shift: vibe coding raised the floor. Agentic engineering raises the ceiling.
We talk about what it means to build seriously in the agent era. Not just moving faster. Building new things, with new tools, while preserving the parts that still require human taste, judgment, and understanding.
Useful.
Keith Rabois reposted
Harvard just released a study showing AI outperformed human doctors in an emergency room setting.
The study is based on 76 emergency room cases in a Boston hospital. OpenAI's "o1-preview" went up against two human doctors described as "expert attending physicians."
AI correctly diagnosed in 67.1% of the cases. The individual scores for the two human doctors were 55.3% and 50.0%.
AI proved especially adept at diagnosing rare diseases and complex cases.
I see this as a positive step into the future. Working together, AI will make doctors so much more competent and effective.
New Board slides in @stratechery mode:
Keith Rabois reposted
It's official! @RogoAI announced their $160 million Series D led by @kleinerperkins at a ~$2B val
Rogo is an agentic AI platform purpose-built for finance that automates research, deal screening, CIM generation, buyer outreach + data room diligence
Some stats on the biz:
- 35,000+ financial professionals using the platform
- 250+ institutions as customers, including Rothschild & Co, Jefferies, Lazard, Moelis & Nomura
$300 million+ total raised now ($75M Series C from @sequoia in Jan '26)
Seed → $2 billion valuation in ~2 yrs. Insane!
www.bloomberg.com/news/features/2026-04-29/junior-bankers…
Ok @RogoAI Series C now confirmed --> $75M round led by @sequoia at $750M val
Rogo = AI for Wall Street. Building purpose-built LLMs for finance that execute end-to-end work across deals + investments
Not seeing a ton of stats on the biz but clients include Nomura, Moelis, Lazard, GTCR, Raymond James
$165M+ total raised now ($50M Series B from @ThriveCapital in May '25, $18M Series A from @khoslaventures in Oct '24)
www.axios.com/pro/fintech-deals/2026/01/28/sequoia-rogo-7…
Keith Rabois reposted
Finance is one of the hardest environments to build AI for because precision and trust matter so much. @GabeStengel, @johnwillettrogo, @TumasRackaitis, and team have built @RogoAI with that reality in mind: a platform that fits the way bankers already work and delivers the auditability, reliability, and workflow depth that leading financial firms require.
Thrilled to lead Rogo’s Series D and excited to partner with the team.
Profile of our new investment in Shade: techcrunch.com/2026/04/22/shade-lands-14m-to-let-creative…
Keith Rabois reposted
Machine Banking is coming.
I spoke with @rabois about how fintech is moving to software-driven systems. We discuss automation and agentic-systems as a core advantage, underwriting and infrastructure as products, and why talent density is what will matter most.
Watch the full interview: www.youtube.com/watch?v=NlXfXTxfrkE
New podcast w the incomparable @SadiSKhan of Aven on the future of finance (in the age of AI): youtu.be/NlXfXTxfrkE
Keith Rabois reposted
Minute for minute, amongst the most insight-packed podcasts I’ve encountered of late. Exceptional episode featuring Keith @rabois in convo w @lennysan. Founders should listen / read / AI-parse the episode.
Notes on the podcast episode linked to in the reply tweet.
"High performance machines don't have psychological safety. They're about winning."
Keith Rabois (@rabois) was COO of Square, part of the PayPal Mafia, an early investor in Stripe, Palantir, Airbnb, DoorDash, and Ramp, and a 2x founder. He's spent 25 years obsessing over how to build world-class teams.
In our in-depth conversation, we discuss:
🔸 How to identify undiscovered talent
🔸 Keith's barrels vs. ammunition hiring framework
🔸 The three traits of the best-performing companies right now
🔸 Why talking to customers is actively harmful for consumer products
🔸 Why the PM role is dying
🔸 The specific interview question he asks every senior candidate
🔸 Why CMOs (not engineers) are becoming the #1 consumer of AI tokens
Watch now 👇
youtu.be/xCd9ykretlg
Keith Rabois reposted
BREAKING: S&P 500 on pace for a third straight week of 3%+ gains — something that would mark only the third such streak since 1950.
Keith Rabois reposted
They don't talk loud. They speak with the work.
Matan and 20 engineers at Factory just raised $150M at a $1.5B valuation, and the industry is still catching up to what they've shipped.
Long context compression: I pushed 40+ million tokens in a single session without starting a new chat. The agent held the thread the whole way. Every lab has since shipped some version of it. Factory did it first and barely mentioned it.
Missions: Droid can ship software across days. The codebase comes out cleaner than it went in (tests passing, edges handled, the kind of work a good engineer does at the end of a long week).
Agent Readiness: A framework for whether your codebase is even ready for agents. If you've wondered why AI coding tools feel uneven across projects, this is the answer. The environment around the agent is usually the bottleneck.
In three years, 20 people are outpacing AI agent innovations from frontier labs with millions in capital plus hundreds of employees.
If you're building with AI, spend a weekend with Droid. It cooooooks!
Keith Rabois reposted
Keith Rabois reposted
Really good CEO advice from @rabois packaged nicely by @lennysan.
My biggest takeaways from @rabois:
1. The team you build is the company you build. Founders get distracted by markets, customers, and technology. If you have the right people, those problems get easier. If you have the wrong people, none of those things save you.
2. Build your company on undiscovered talent. The only way to scale an organization against incumbents with infinite budgets is to find talent that large companies’ hiring machines will misprocess. In practice, this often means skewing younger—not because young people are inherently better but because they have fewer data points, which means typical evaluation systems can’t categorize them accurately. This is where the alpha often is.
3. Hire more “barrels,” not “ammunition.” A “barrel” is someone who can take an idea from zero to outcome without hand-holding. Most companies have only a handful of these people. Hiring more people without expanding the number of barrels doesn’t increase output; it increases coordination tax and creates drag. The ratio of barrels to ammunition is what determines the number of important things a company can pursue simultaneously.
4. CMOs are becoming the #1 consumer of AI tokens. At a few of Keith’s top portfolio companies, the heaviest user of AI is the chief marketing officer. These CMOs are running analytics, shipping campaigns, and generating insights that previously required entire teams of deputies.
5. The three signs a company will win: operating tempo, internal talent development, and “the relentless application of force” from the top. Keith identifies a consistent pattern across his best portfolio companies. First, operating tempo: Ramp shipped physical cards in three months when the industry standard was 9 to 12. Second, talent development through internal promotion rather than senior external hires; the CMO at one of his top companies was the previous chief of staff. Third, the CEO’s willingness to push harder as things improve, not less. Mike Moritz told a friend of Keith’s that the most common trait of the best CEOs is “the relentless application of force.” Complacency is the natural by-product of success, and the CEO’s job is to offset it.
6. For consumer products, talking to customers is not just unhelpful; it’s actively harmful. Keith refuses to let companies he advises conduct consumer research. His argument: Consumer decisions are subconscious. Ask any Porsche owner why they bought the car, and 99% will cite every reason except the real one. Once misleading customer feedback enters the organization, it locks into people’s brains and distorts every subsequent decision.
7. Keith believes the PM role may not survive the AI era. Taking customer inputs, building a sequential year-long roadmap, and coordinating between teams are structurally incoherent when AI capabilities change weekly. The skill that matters now across all three roles—PM, designer, engineer—is business acumen: understanding the company’s equation and knowing what to build next.
8. Great hiring comes from great referencing. Run at least 20 references, and keep going until you hit negative feedback. Ask specific, forward-looking questions (e.g. “Would you start a company with them?”). If every reference is positive, you haven’t gone deep enough.
9. Use a 30-day feedback loop to sharpen your hiring instinct. Thirty days after every hire, ask: would I hire this person again? This is as predictive as waiting years, and dramatically faster for improving your judgment. Make this a habit, and your hiring quality will compound.
10. Criticize in public, not private—it optimizes for the system. Keith endorses a management practice that most people find confrontational: delivering negative feedback in front of the team, not behind closed doors. Private criticism optimizes for the individual, but the rest of the company doesn’t know the issue is being addressed, which breeds anxiety and suspicion. Public criticism lets colleagues see that leadership is aware, creates opportunities for others to volunteer help, and turns feedback into a team-building exercise.
Full conversation: www.youtube.com/watch?v=xCd9ykretlg
"High performance machines don't have psychological safety. They're about winning."
Keith Rabois (@rabois) was COO of Square, part of the PayPal Mafia, an early investor in Stripe, Palantir, Airbnb, DoorDash, and Ramp, and a 2x founder. He's spent 25 years obsessing over how to build world-class teams.
In our in-depth conversation, we discuss:
🔸 How to identify undiscovered talent
🔸 Keith's barrels vs. ammunition hiring framework
🔸 The three traits of the best-performing companies right now
🔸 Why talking to customers is actively harmful for consumer products
🔸 Why the PM role is dying
🔸 The specific interview question he asks every senior candidate
🔸 Why CMOs (not engineers) are becoming the #1 consumer of AI tokens
Watch now 👇
youtu.be/xCd9ykretlg
Keith Rabois reposted
Keith Rabois reposted
Mike Moritz used to say the best CEOs are those who focus on the relentless application of force.
@rabois rekindled this recently on @lennysan's podcast:
My biggest takeaways from @rabois:
1. The team you build is the company you build. Founders get distracted by markets, customers, and technology. If you have the right people, those problems get easier. If you have the wrong people, none of those things save you.
2. Build your company on undiscovered talent. The only way to scale an organization against incumbents with infinite budgets is to find talent that large companies’ hiring machines will misprocess. In practice, this often means skewing younger—not because young people are inherently better but because they have fewer data points, which means typical evaluation systems can’t categorize them accurately. This is where the alpha often is.
3. Hire more “barrels,” not “ammunition.” A “barrel” is someone who can take an idea from zero to outcome without hand-holding. Most companies have only a handful of these people. Hiring more people without expanding the number of barrels doesn’t increase output; it increases coordination tax and creates drag. The ratio of barrels to ammunition is what determines the number of important things a company can pursue simultaneously.
4. CMOs are becoming the #1 consumer of AI tokens. At a few of Keith’s top portfolio companies, the heaviest user of AI is the chief marketing officer. These CMOs are running analytics, shipping campaigns, and generating insights that previously required entire teams of deputies.
5. The three signs a company will win: operating tempo, internal talent development, and “the relentless application of force” from the top. Keith identifies a consistent pattern across his best portfolio companies. First, operating tempo: Ramp shipped physical cards in three months when the industry standard was 9 to 12. Second, talent development through internal promotion rather than senior external hires; the CMO at one of his top companies was the previous chief of staff. Third, the CEO’s willingness to push harder as things improve, not less. Mike Moritz told a friend of Keith’s that the most common trait of the best CEOs is “the relentless application of force.” Complacency is the natural by-product of success, and the CEO’s job is to offset it.
6. For consumer products, talking to customers is not just unhelpful; it’s actively harmful. Keith refuses to let companies he advises conduct consumer research. His argument: Consumer decisions are subconscious. Ask any Porsche owner why they bought the car, and 99% will cite every reason except the real one. Once misleading customer feedback enters the organization, it locks into people’s brains and distorts every subsequent decision.
7. Keith believes the PM role may not survive the AI era. Taking customer inputs, building a sequential year-long roadmap, and coordinating between teams are structurally incoherent when AI capabilities change weekly. The skill that matters now across all three roles—PM, designer, engineer—is business acumen: understanding the company’s equation and knowing what to build next.
8. Great hiring comes from great referencing. Run at least 20 references, and keep going until you hit negative feedback. Ask specific, forward-looking questions (e.g. “Would you start a company with them?”). If every reference is positive, you haven’t gone deep enough.
9. Use a 30-day feedback loop to sharpen your hiring instinct. Thirty days after every hire, ask: would I hire this person again? This is as predictive as waiting years, and dramatically faster for improving your judgment. Make this a habit, and your hiring quality will compound.
10. Criticize in public, not private—it optimizes for the system. Keith endorses a management practice that most people find confrontational: delivering negative feedback in front of the team, not behind closed doors. Private criticism optimizes for the individual, but the rest of the company doesn’t know the issue is being addressed, which breeds anxiety and suspicion. Public criticism lets colleagues see that leadership is aware, creates opportunities for others to volunteer help, and turns feedback into a team-building exercise.
Full conversation: www.youtube.com/watch?v=xCd9ykretlg
"High performance machines don't have psychological safety. They're about winning."
Keith Rabois (@rabois) was COO of Square, part of the PayPal Mafia, an early investor in Stripe, Palantir, Airbnb, DoorDash, and Ramp, and a 2x founder. He's spent 25 years obsessing over how to build world-class teams.
In our in-depth conversation, we discuss:
🔸 How to identify undiscovered talent
🔸 Keith's barrels vs. ammunition hiring framework
🔸 The three traits of the best-performing companies right now
🔸 Why talking to customers is actively harmful for consumer products
🔸 Why the PM role is dying
🔸 The specific interview question he asks every senior candidate
🔸 Why CMOs (not engineers) are becoming the #1 consumer of AI tokens
Watch now 👇
youtu.be/xCd9ykretlg
Solid.
This Thursday on 20VC w/SaaStr @HarryStebbings @rodriscoll + me is in the can:
1⃣Citi says there's no floor for software stocks
ServiceNow — Rule of 56, best-in-class — is down 45% YTD. Wix bought back 30% of its shares at $92 and the stock still cratered 23%. Salesforce is taking on debt through 2066 to fund a $50B buyback. When your biggest move is financial engineering, not product innovation, the market sees it.
2⃣Thoma Bravo just shut down growth equity
The largest software PE firm on the planet — $183B AUM — looked at minority stakes in software and said: we can't make money here anymore. Same week, Blackstone refused to extend another lifeline to Thoma Bravo's Medallia. $6.4B acquisition in 2021, $3B in debt, ~$200M in earnings, potential $5B writedown. The 2021 software playbook is dead.
3⃣Meanwhile Anthropic employees refused to sell at $350B
Vested holdings went from $500K to $4-5M in a year and they held. When employees won't take life-changing money off the table, they see something on the roadmap you don't. That's the most bullish insider signal I've seen in a decade of venture investing.
4⃣The Mythos story isn't just about safety
Andreessen made a strong point: Anthropic may not have enough compute to ship it even if they wanted to. Both things can be true. The Trainium angle is the real story: frontier model trained entirely on Amazon's custom silicon. Nvidia at 90% share is a ceiling, not a floor.
5⃣AppLovin: $4B+ revenue, 898 employees, $4.5M per head
That's the benchmark now. Every B2B company running at $300K-$500K/employee has to explain the gap.
6⃣SpaceX wants to IPO at 100x revenue
No company in history has sustained that multiple at scale. Saudi Aramco IPO'd at 5x and it was the most profitable company on Earth. SpaceX is posting a $5B loss because Musk stuffed xAI ($28M/day burn rate) and Twitter into the entity before filing.
Much more on Thursday!!
Everyone wants their company to produce a "mafia." but almost NONE of them understand what actually makes one.
It's not culture decks or beer fridges. Paypal was hard enough that people learned building a great company was HARD.
And that's what make hem go do it again.
indeed.
Keith Rabois reposted
My biggest takeaways from @rabois:
1. The team you build is the company you build. Founders get distracted by markets, customers, and technology. If you have the right people, those problems get easier. If you have the wrong people, none of those things save you.
2. Build your company on undiscovered talent. The only way to scale an organization against incumbents with infinite budgets is to find talent that large companies’ hiring machines will misprocess. In practice, this often means skewing younger—not because young people are inherently better but because they have fewer data points, which means typical evaluation systems can’t categorize them accurately. This is where the alpha often is.
3. Hire more “barrels,” not “ammunition.” A “barrel” is someone who can take an idea from zero to outcome without hand-holding. Most companies have only a handful of these people. Hiring more people without expanding the number of barrels doesn’t increase output; it increases coordination tax and creates drag. The ratio of barrels to ammunition is what determines the number of important things a company can pursue simultaneously.
4. CMOs are becoming the #1 consumer of AI tokens. At a few of Keith’s top portfolio companies, the heaviest user of AI is the chief marketing officer. These CMOs are running analytics, shipping campaigns, and generating insights that previously required entire teams of deputies.
5. The three signs a company will win: operating tempo, internal talent development, and “the relentless application of force” from the top. Keith identifies a consistent pattern across his best portfolio companies. First, operating tempo: Ramp shipped physical cards in three months when the industry standard was 9 to 12. Second, talent development through internal promotion rather than senior external hires; the CMO at one of his top companies was the previous chief of staff. Third, the CEO’s willingness to push harder as things improve, not less. Mike Moritz told a friend of Keith’s that the most common trait of the best CEOs is “the relentless application of force.” Complacency is the natural by-product of success, and the CEO’s job is to offset it.
6. For consumer products, talking to customers is not just unhelpful; it’s actively harmful. Keith refuses to let companies he advises conduct consumer research. His argument: Consumer decisions are subconscious. Ask any Porsche owner why they bought the car, and 99% will cite every reason except the real one. Once misleading customer feedback enters the organization, it locks into people’s brains and distorts every subsequent decision.
7. Keith believes the PM role may not survive the AI era. Taking customer inputs, building a sequential year-long roadmap, and coordinating between teams are structurally incoherent when AI capabilities change weekly. The skill that matters now across all three roles—PM, designer, engineer—is business acumen: understanding the company’s equation and knowing what to build next.
8. Great hiring comes from great referencing. Run at least 20 references, and keep going until you hit negative feedback. Ask specific, forward-looking questions (e.g. “Would you start a company with them?”). If every reference is positive, you haven’t gone deep enough.
9. Use a 30-day feedback loop to sharpen your hiring instinct. Thirty days after every hire, ask: would I hire this person again? This is as predictive as waiting years, and dramatically faster for improving your judgment. Make this a habit, and your hiring quality will compound.
10. Criticize in public, not private—it optimizes for the system. Keith endorses a management practice that most people find confrontational: delivering negative feedback in front of the team, not behind closed doors. Private criticism optimizes for the individual, but the rest of the company doesn’t know the issue is being addressed, which breeds anxiety and suspicion. Public criticism lets colleagues see that leadership is aware, creates opportunities for others to volunteer help, and turns feedback into a team-building exercise.
Full conversation: www.youtube.com/watch?v=xCd9ykretlg
"High performance machines don't have psychological safety. They're about winning."
Keith Rabois (@rabois) was COO of Square, part of the PayPal Mafia, an early investor in Stripe, Palantir, Airbnb, DoorDash, and Ramp, and a 2x founder. He's spent 25 years obsessing over how to build world-class teams.
In our in-depth conversation, we discuss:
🔸 How to identify undiscovered talent
🔸 Keith's barrels vs. ammunition hiring framework
🔸 The three traits of the best-performing companies right now
🔸 Why talking to customers is actively harmful for consumer products
🔸 Why the PM role is dying
🔸 The specific interview question he asks every senior candidate
🔸 Why CMOs (not engineers) are becoming the #1 consumer of AI tokens
Watch now 👇
youtu.be/xCd9ykretlg
How to build a high growth startup: open.spotify.com/episode/2Ld8DpKG2dSXlYmunQDJN8?t=0&pi=Zd…
Don’t miss this podcast.
"High performance machines don't have psychological safety. They're about winning."
Keith Rabois (@rabois) was COO of Square, part of the PayPal Mafia, an early investor in Stripe, Palantir, Airbnb, DoorDash, and Ramp, and a 2x founder. He's spent 25 years obsessing over how to build world-class teams.
In our in-depth conversation, we discuss:
🔸 How to identify undiscovered talent
🔸 Keith's barrels vs. ammunition hiring framework
🔸 The three traits of the best-performing companies right now
🔸 Why talking to customers is actively harmful for consumer products
🔸 Why the PM role is dying
🔸 The specific interview question he asks every senior candidate
🔸 Why CMOs (not engineers) are becoming the #1 consumer of AI tokens
Watch now 👇
youtu.be/xCd9ykretlg
Keith Rabois reposted
Masterclass on talent and hiring by @rabois on @lennysan podcast.
Exceptional, clear insights.
Must listen for all founders who want to build high performance teams / generational companies.
overcast.fm/+AA6K1HW-Avo
My debut on the @lennysan podcast: Building a successful startup in the age of AI. open.spotify.com/episode/2Ld8DpKG2dSXlYmunQDJN8?t=0&pi=bL…
Keith Rabois reposted
France is now closer to communism than the Soviet Union ever was.
Keith Rabois reposted
Introducing Ramp AI Spend Intelligence. A way to track and control spend all the way down to the token level across providers.
Tokenmaxxing requires tokentracking. This is how your business will get ahead (or fall behind) with AI.
The best companies are investing aggressively in AI. And it’s working. Our data proves it. Since January 2025, average monthly AI token spend across Ramp customers increased 13x. Not 13%. Thirteen *TIMES.
But visibility into that spend has remained virtually nonexistent. No one knows how fast it’s adding up, where it’s going, who’s purchasing what. It’s a financial mess.
And AI spend is fundamentally different from every other cost category. A single prompt template change can triple your bill overnight. An agent stuck in a loop can burn through $50k before anyone notices.
No tool in the market connects token-level usage, invoice data, and card-level spend. So we built one.
Ramp customers can get early access today!
ramp.com/blog/trillion-dollar-ai-blindspot
Interesting.
It's only a matter of time before only the model creators have access to the most powerful models. The rest get access to smaller, distilled versions. Or access the models through first party apps and services that don't provide direct access to the token path.
The investment needs for training are too high, and distillation too effective to warrant any other future.
RT @JimHansonDC: "Iran begged for this ceasefire — and we all know it...
The world's leading state sponsor of terrorism proved incapable o…
Keith Rabois reposted
CNN is garbage.
Keith Rabois reposted
$OPEN just hit 658 contracts.
New all-time high under Kaz.
LETS GO!
8-week trend:
537 → 449 → 442 → 390 → 447 → 610 → 545 → 658
The numbers:
→ 8-week avg: 510/week
→ +448% since Kaz started (Sept: 120)
→ New floor: ~530 (was 390)
→ New ceiling: 658 (was 537)
The range keeps moving UP.
And this is BEFORE:
→ Mortgage product scales
→ Doma integration kicks in
→ Q2 EBITDA profitability hits
Now the math.
If $OPEN sustains ~660 homes/week at 7% contribution margin:
→ 8,580 homes/quarter
→ $3.4B revenue
→ $240M contribution profit
→ $106M net income/quarter (after everything)
→ $423M net income/year
→ $0.42 EPS
→ $19.1 (45x P/E) | $31.8 (75x P/E)
Net income positive.
At current volume.
At guided margins.
The stock is $4.75.
Kaz is executing.
Weekly.
Publicly.
In real-time.
The re-rate isn't coming.
It's already started.
Keith Rabois reposted
I repeat:
The U.S. sent thousands of its 19 years old soldiers to die in Normandy, to free Europe and to end the biggest crime against humanity ever commited - by Europeans.
It was just 81 years ago.
The whole reason, France, Benelux etc. 🇫🇷 exist today is because of this heroism.
My grandparents could grow up in a liberal democracy. Without the U.S. they would be raised at the H*tler Youth.
We Europeans would still be in wars again and again, like 1914, 1866, 1870, 1795 etc.
They brought peace, democracy, liberty and human rights. They invested billions of U.S. Dollars into Europe with the Marshall Fund. They gave us more than we ever had in our history before. They protected us for 7 decades with hundreds of thousands of soldiers against the cruelties of the Soviet Union.
The terror we can see nowadays in Donetsk, would have happened in Bavaria, Bourgogne or the Netherlands in 1950 if there wasn’t the U.S. 🇺🇸
Who do we Europeans think we are to let that nation down, act like bad allies, calling their President names every day on television - and have full confidence we stand better alone. All of instagram is just about, why we’re better than the U.S.
We owe them so much.
We Europeans are most arrogant species on earth. And to cure this we have to face the truth.
Keith Rabois reposted
Instead of B2s Trump should threaten to send a bunch of California Democrats to Iran. Infrastructure wouldn’t stand a chance
Keith Rabois reposted
The word `expert` in media has come to mean pundit with davos-liberal talking points
“Nothing is more securely lodged than the ignorance of the experts.”
— Friedrich Hayek
Keith Rabois reposted
This chart is baffling…why would 30% of people still trust legacy media?
Mainstream media has lost the trust of the American people
Keith Rabois reposted
After the Trump Tax Cuts, the share of taxes paid by the rich went UP.
Data from the House Ways and Means Committee shows lower income Americans saw larger tax relief.
Facts matter, even if Democrats won’t admit it.
Keith Rabois reposted
Every “10 VCs you should meet in SF” list is just like:
Sequoia
Founders Fund
Khosla
Kleiner
Benchmark
fund you never heard of (GP wrote the post)
YC
:)
Y'all make fun of Florida, but it's the only place with direct flights to the Moon.
Keith Rabois reposted
I am with Ken 🌴
Ken Griffin wants to reshape Miami—and maybe American politics.
While many are “shorting” America’s prospects, Citadel's CEO is buying, and this certainly isn’t the first time one of his contrarian bets has paid off. fortune.com/2026/03/31/ken-griffin-citadel-securities-hed…
Keith Rabois reposted
Why would you ever pay this
George Washington University announced that next academic year, the estimated cost for returning students will be just over $98,000 per year. nbc4dc.com/bUubTwh
Keith Rabois reposted
Escrow & Closing is to Opendoor what AWS was to Amazon. The infra we use to run our business. It is great. We want it to power what others are building.
Today, Opendoor acquired Doma’s escrow services to make home closing faster, cheaper and more certain. For everyone.
www.cnbc.com/2026/03/31/opendoor-doma-closing-escrow-busi…
Keith Rabois reposted
It’s iconic!
🚨 FIRST LOOK: The Donald J. Trump Presidential Library is officially here.
Over the past six months, I have poured my heart and soul into this project with my incredible team at @Trump.
This landmark on the water in Miami, Florida will stand as a lasting testament to an amazing man, an amazing developer, and the greatest President our Nation has ever known. 🇺🇸
These images have never been seen by the public — until today. Enjoy!
www.trumplibrary.org
Keith Rabois reposted
Absolutely incredible statistic 🤯
West Palm Beach 🏖️ is doing more office construction than either Houston, Austin, or the entire state of New Jersey.
As well as 60% (1.6M vs 2.7M) of all of Manhattan.
But with a population only 127,000.
Populations of regions on chart, can you spot the odd one out? 🤔
Population:
Boston - 673,458
Manhattan - 1,660,664
Dallas - 1,326,087
Los Angeles - 3,878,704
🏖️ West Palm Beach - 127,744
Austin - 993,588
San Diego - 1,404,452
New Jersey (state) - 9,548,215
Houston - 2,390,125
US Office Space Under Construction
At the start of 2026, an estimated 28.9 million square feet of office space was under construction across the U.S.
Boston is the leader, however at 4.1 million square feet, it’s now less than one-quarter of its peak in 2024.
--commercialcafe
#commercialrealestate
Miami is the best.
Goals achieved: 1. Iranian Navy eliminated. 2. Air Force eliminated. 3. 48 of the 50 top evil leqders dead. 4. Less than 80 launchers remaining w a shelf life of one fire 5/ No capability to enrich an nuclear material probably for a decade (inputs destroyed like water plant, steel and most sights). 6. Very military ballistic missles supply left w no ability to produce more. 7. Some drones remain w negligible ability manufacture new. 8. Unified Gulf States against Iran except Oman 9. Chinese weapons proven to be dysfunctional. 10. Middle Eastern states lining up for business relations w Israel.
Keith Rabois reposted
Miami this weekend: Miami Open, MMW parties, Ultra, boat outings
NYC this weekend: No Kings parade
Lmaooooo
No Kings. Too packed to move. New York City
Keith Rabois reposted
The view from Apollo chief economist Torsten Slok:
“Markets are overreacting to what will likely be a 4- to 6-week period of volatility, which will ultimately result in 50 years of stability in oil markets, supply chains and geopolitics.”
Keith Rabois reposted
Quinnipiac: Republicans only
Pres. Trump
🟢 Approve: 89% (+6 since Mar. 8)
🟤 Disapprove: 6% (-5)
Highest approval since Oct. 2025
——
Military action against Iran
Support: 86% (+1)
Oppose: 9% (-2)
——
Trump's net approval issues
Foreign policy: +76
Economy: +75
Iran: +69
Quinnipiac: Independents only
Pres. Trump
🟢 Approve: 25%
🟤 Disapprove: 68%
Was 35-58 in December
——
Trump's approval on key issues
🟤 Economy: -42
🟤 Foreign policy: -47
🟤 Situation with Iran: -49
——
Generic Ballot
🟦 DEM: 57%
🟥 GOP: 26%
Keith Rabois reposted
United States manufacturing activity expands at the fastest pace in months. Both manufacturing and services are in clear expansion.
US business confidence and profit outlook rise, and real net wages are rising.
via SP Global, Bloomberg
Sharp analysis.
Major notes from the weekend on Iran:
1) Combined Force strikes started hitting IRGC Ground Forces units in the Kurdish regions.
2) We have confirmed that IRGC members are at least two cycles behind on pay.
3) *Coincidentally,* I'm sure, the IRGC Ground Forces Commander went on a tour visiting these forces.
In combination, these three notes indicate a high risk of desertion throughout the IRGC.
4) Separately, we have the attack on Diego Garcia that, frankly, justifies the war in and of itself. The base itself isn't terribly significant, its not that "Iran can get Diego Garcia! We have to go to war!"
The issue is that Iran constantly lies about everything to do with its nuclear program. Taqiyya. They've issued fatwas against nuclear weapons, then simultaneously enrich uranium to levels only usable in weapons while developing long-range ballistic missiles.
"Its for medical purposes!"
No, it wasn't, and it couldn't be. Uranium enriched to this degree has no medical uses.
"Its for self-defense!"
Then why were you hiding a long range ballistic missile capable of delivering a nuclear weapon to London?
The fact they were developing these weapons while consistently lying about it, in the face of all the world being united against their acquisition of the weapons, combined with all of their past acts of terrorism and piracy, absolutely justifies this war.
5) On Negotiations: These are being played very close to the chest, but here's what I've been able to gather...
Witkoff and Kushner are keeping lines of communication open with non-IRGC elements of the Iranian State—if I had to guess, I'd expect officials within the Republican Government. If not Pezeshkian himself, then probably people with access to him.
It seems abundantly clear that the situation in Iran is at this point A Tale of Two Cities.
One is an IRGC-Cleric alliance propping up the corpse of the Velayat-e Faqih.
The other is the Republican-Reform alliance propping up the still in-tact organs of state outside of the IRGC's control.
We are almost certainly negotiating with one of these 'cities,' and not the other. This is what I expected would be the prelude to final resolution. The existing power structures in Iran will continue to pull support away from, and rallying support against, the IRGC.
At this point, it appears the plan is to reach a pre-negotiated settlement with the Republican-Reform Government while hammering the IRGC on all available fronts and enabling a fairly smooth transition of power.
Mass uprising is probably now the backup plan, with total dismantling of Iran's infrastructure the backup to the backup plan.
Keith Rabois reposted
In-state tuition at Florida‘s state universities has not increased at all since 2019.
College tuition has increased 914% since 1983, outpacing all other household expenses, per CNBC.
Keith Rabois reposted
The U.S. gains roughly $48 billion in return for its $3.8 billion investment in Israel.
That’s a 1,200% return on investment.
This is what America First actually looks like.
Anyone saying otherwise is either lying, paid, or both.
Keith Rabois reposted
Superb article written by an Iranian who truly understands how this regime works still executing young protestors, firing off ballistic missiles despite having no leadership, and no money to pay its security forces:
“What looks like resilience is in fact the functioning legacy of a doomsday design.”
“That is why the Islamic Republic does not need to prove that its command structure is intact. It only needs to prevent the appearance of final collapse.”
The levels of “styling it out” is simply astounding.
#iranisraeluswar
The Islamic Republic is living inside its collapse plan. That is not evidence of resilience. It is evidence that the system is collapsing.
What many observers read as signs of endurance, continued missile fire, street repression, state broadcasting, leadership succession, and other visible signs of continuity, do not show a regime that has absorbed the shock. They show a regime falling back on the emergency mechanisms it built for the moment its center was hit and its command structure began to fracture.
This is not strength. It is a collapsing system trying to survive long enough for Washington to lose patience.
Read more:
parpanchi.substack.com/p/what-looks-like-resilience-in-ir…
Shut Harvard down.
DOJ’s accusations against Harvard:
— Jewish and Israeli students were harassed, stalked, physically assaulted, and spat upon. Harvard leadership “turned a blind eye”
— after an Israeli student was assaulted on campus, Harvard rewarded the attacker with a $65K fellowship, free housing, and a Class Marshal position
— a professor ordered a student to leave a lecture simply after finding out she was from Israel
— deans rewarded students illegally seizing campus spaces by bringing them burritos and candy
— Harvard financial employee posted TikTok videos brandishing a machete and talking about bringing down the “Zionist mafia”
— professors were canceling classes to join anti-Israel protests
True.
Florida’s state budget is $117 billion.
Florida's population is 23.7 million
New York's state budget is $237 billion
New York's state population is 20 million
New York City’s budget is $127 billion.
It's population is 8.4 million
Florida is thriving.
NY and NYC are in crisis.
True.
New York voters are morons. Or lazy. Or are not getting fair elections. Or have been dumbed down by government schools. Or something.
Narrative violation by CNN:)
There's no break in MAGA. Trump's approval with MAGA GOP is literally 100%.
90% of MAGA GOP approve of US military action in Iran.
Those who disapprove of Trump are not MAGA at this point.
Importantly: MAGA makes up the same share of voters as it did when Trump won in 2024.
Keith Rabois reposted
Students who took notes by hand scored ~28% higher on conceptual questions than laptop note-takers.
Writing forces your brain to process and compress ideas instead of copying them.
Keith Rabois reposted
It’s like the Soviet Union again.
The regime is collapsing in Iran but thriving in Western media and academia
Keith Rabois reposted
We had a customer receive an instant offer. They viewed it and signed the contract in 31 minutes 2 seconds.
Awaiting a Wired article on the gay mafia running Iran.
True.
The US needs 500,000 new electricians this decade.
Apprenticeships take 5 years.
Microsoft’s Brad Smith says it’s the #1 thing slowing data center expansion.
The AI bottleneck isn’t chips. It’s the trades.
This is interesting.
Interesting datapoint from KKR Co-CEO:
“The largest 100 U.S. companies in the United States have seen their margins expand materially in the last 5 years, 14% to 19%. The next 1,400 companies have had their margins flat and are working to stay even.“
Keith Rabois reposted
Europe should expel the regime’s ambassadors and diplomatic staff. They are not Iranian diplomats—they are IRGC and MOIS operatives, or working with them and other cells.
These are embedded in European capitals planning terror attacks on European soil.
Keith Rabois reposted
It’s inevitable that contracts will be negotiated by agents within the next decade.
Business will speed up so much that humans won’t be able to keep up with day-to-day contracts between companies.
Lawyers will set the policies and audit— agents will be trusted to do the rest.
This sounds irresponsible, until you realize what the error rate is for human contract managers—fairly high.
I believe agents will be a safer way to negotiate rudimentary agreements within the decade, just like self-driving is now safer than human driving.
This chart will exist for basic business contracts:
Keith Rabois reposted
To Save Time, CNN Will Now Run Retractions Simultaneously With News Stories babylonbee.com/news/to-save-time-cnn-will-now-run-retract…
Keith Rabois reposted
In 72 hours we went from:
Trump never expected the possibility of Hormuz closing,
To:
Trump was briefed Hormuz could close but riskily did it anyway,
To:
Trump bombed Kharg Island but it likely won’t force Iran to open the Strait
To:
Iran opened the Strait
Keith Rabois reposted
Legacy media remains in free fall as they struggle to paint America and Israel winning decisively in combat with #iran as a devastating or inevitable loss for the two countries and their allies in the Gulf b/c the combat is not over in 14 days and oil is at $100 a barrel.
Make no mistake, the United States and Israel are winning decisively but have not yet achieved a decisive win. I don’t believe @POTUS will accept anything less than such a win. Unless you are cheering on the “alliance of tyrants” in China, Russia, Iran and North Korea, you should be hoping that @realDonaldTrump remains indifferent to the doom-casting of “analysts” on air and in print. It’s an astonishing thing, this collapse of American media two weeks into the most significant military conflict in a quarter century.
Besides the ayatollahs and the IRGC, the ranks of the devastated include the reputations for accuracy, fairness and, patriotism, of a defeatist, overwhelmed-by-ideology and undermined-by-ignorance legacy media.
Of course podcasters from the margins of left and right are not going to impress us with their takes because they lack depth of experience and many appear never to have read even a single book about the Iranian regime. Also remarkable is their collective short-term memory loss (35K Iranians butchered by this evil regime in January 8/9).
Best to stick with “The GoodFellows,”@Commentary pod, “Call Me Back” from @DanSenor, @NRO’s The Editors, “Ask Haviv Anything” from @havivrettiggur and a few others. These folks and a few others are delivering lots of reliable content, excellent analysis that asks the difficult questions, but they all reject the absurdity that is simply anti-Trump or anti-Israel propaganda dressed up as “news.”
The combined effects of TDS, anti-Bibi, anti-Israel ideology is out in the open in a way not even longstanding critics of what used to be called MSM could have imagined. It is as though the @guardian had been put in control of the American news establishment.
The Strait of Hormuz presents a military challenge but it is not the “Battle of the Atlantic.” It will be solved. The Iranian regime has been crippled, humbled and while it has not collapsed, it may yet. The ability of the theocracy’s fanatics to gain nuclear weapons or to deter attacks on any future attempts to do so is already achieved.
Combined with regime coercion in Venezuela and the ongoing collapse of communism and dictatorship in Cuba, the first 15 months of Donald Trump’s second term will rank as great advances for global freedom and security with the three years beginning with the fall of the Berlin Wall.
That large part of American media which is pushing defeatist narratives and absurd stories will shame those publications and bylines in the way that Walter Duranty was and remains shamed by history.
Keith Rabois reposted
In summation, I don’t believe modern American media is capable of covering the United States as a force for good in the world and that seems like a hugely seismic reality that will continue to have massive negative implications on all stories in the country.
Keith Rabois reposted
It's almost impossible for a newspaper outlet to cover the attempted murder of 140 Jewish preschool children by blaming the attack on a Detroit synagogue built during the Holocaust.
Yet the NY Times did it anyway.
You can't hate these people enough. It's just not possible.
Keith Rabois reposted
Miami is lucky to have founders like @ClementDelangue make it their home base. Future is bright!
Great to catch up with @akothari yesterday in Miami (Is Miami back btw?).
@NotionHQ is killing it with agents and just rolled out their first open weight models with Minimax.
In my (biased) opinion, applied AI startups can’t solely rely on proprietary APIs who end up being super costly and compete with them in many cases so cool to see open-source AI getting some usage there!
Don’t miss this podcast!
The most ‘Founder Mode’ CEO working today is not actually the founder.
NEW EPISODE with Kaz @nejatian of @Opendoor is now live. This is a special one.
Kaz left Shopify to pull off the refounding of a struggling public company in just 16 days. Incredible story.
Here are just a few of my learnings from our conversation on the latest episode of Long Strange Trip:
1. First Derivative Businesses
The most enduring companies are rarely built on their primary activity; they are built on the first derivative of that core business. Great founders identify and weaponize the secondary value stream. This is important.
2. Rejecting Defaults
Success is a function of the defaults you choose to overwrite. Most operators accept the 'software' of their industry or life on autopilot; exceptional builders identify the one or two critical defaults and fight them with all their power to create a new trajectory. Kaz is a master at this, and he explains how.
3. Stewardship Over Status
Optimize for doing things rather than being things. When a leader optimizes for a title or happiness, they create fragile organizations; when they optimize for stewardship and service, they build a mission-driven culture that can withstand the lonely and painful stretches of the journey.
4. Write a user manual for yourself
"Strong attract, strong repel. My job is to tell you what kind of a person I am so you can opt in or opt out."
I love this quote. If you're a founder, you owe this to everyone around you.
5. Founder mode = responsibility for outcomes
Hold yourself responsible for truth and outcomes, not processes.
Hire people to round you out. Don't try to be well-rounded yourself. And don't work on your weaknesses. "Is the fact that I'm bad at this the reason I'm good at everything else?"
6. Structural Risk Mispricing
The one permanent advantage for entrepreneurs is that the rest of the world structurally misprices risk. While others see a 'lion bite' in every setback, the best CEOs recognizes that things going poorly is not as painful as you think, allowing them to lean into volatility that scares off the incumbent. That's the difference.
7. Death Spiral Honesty
When a company is in a death spiral, incrementalism is fatal; "what must change os everything." Professional managers are often incentivized by RSUs to delay the inevitable and manage a slow decline. You need zero incentive to manage a decline and a compensation structure aligned purely with performance.
8. AI as the New Performance Default
Default to AI is not a suggestion; it is the first line of the job description. A company becomes AI-native not through top-down mandates, but by making AI proficiency a core pillar of the performance management system - effectively deciding who gets to play on the team based on their ability to automate their own craft.
9. The Career vs. Job Distinction
"A job is something you do for someone else in order to get paid. A career is something you work on every day for yourself."
Kaz's kids know what @Opendoor is. His family is all in. Exceptional companies are built by people who self-identify with their work and treat their professional mission as a family-integrated pursuit.
10. Two timeframes matter. Everything else is noise.
This week and 10 years from now. "This quarter is a deeply useless measuring period." @tobi applies a discount rate of basically zero to the future. That's the model.
My takeaway from this conversation: ask yourself what defaults you're living by that you haven't deliberately chosen.
Kaz overrides every default, and he does it over and over again.
Keith Rabois reposted
Palantir co-founder Alex Karp has bought a house in Miami @EconClubofMiami
Palantir's CEO Alex Karp bought a Miami mansion in June, seven months before the firm’s surprise announcement to move its headquarters to the city www.bloomberg.com/news/articles/2026-03-12/palantir-s-kar…
Keith Rabois reposted
You are the dumbest piece of garbage in politics and have zero shame and I told everyone this years ago
I am proud to stand with @grahamformaine @ZohranKMamdani & join @hasanthehun feed. The problem is with the neocons in our party who blundered into Iraq, 20 years in Afghanistan, Libya, Gaza, & now support the Iran war. Out with the old guard. We need a new moral direction.
Keith Rabois reposted
What the hell is wrong with you?
48 hours ago two men were beaten in your district for speaking Hebrew.
Earlier today there was a shooting at a Jewish school in Michigan.
Is it politics? Lack of empathy? Are you actually racist?
I am proud to stand with @grahamformaine @ZohranKMamdani & join @hasanthehun feed. The problem is with the neocons in our party who blundered into Iraq, 20 years in Afghanistan, Libya, Gaza, & now support the Iran war. Out with the old guard. We need a new moral direction.
The future of Legal AI.
New @ThePeelPod with @scottastevenson
We talk building @SpellbookLegal into the fastest growing AI company in Canada, and likely the largest company ever built on a Microsoft Word plugin.
Thanks @FlexSuperApp for sponsoring this episode!
Timestamps
0:30 "Cursor for Lawyers”
3:08 Building the world’s largest Microsoft Word plugin
14:06 Why legal software was untouched before LLMs
18:32 $30 trillion moves through contracts annually
20:51 Why ChatGPT won’t replace vertical tools
25:15 Fine-tuning was the biggest mistake in AI
30:00 Differences between pro and amateur gamers
37:38 Top-down vs. bottoms-up in legal AI
42:27 The long-tail of legal AI software
47:24 Building for models that don’t exist yet
51:20 Skating where the puck is going
1:01:35 The legal bill that cost 50% of his bank account
1:09:33 Testing 100 landing pages in 3 years
1:14:06 The moment Spellbook hit PMF
1:19:17 Building new brands for each product experiment
1:23:10 Raising a Series B with a tweet
1:27:41 What Scott learned from Keith Rabois
1:31:16 Scott's favorite new AI tool
Keith Rabois reposted
Most people still don’t understand just how amazing Florida is
I‘ve lived in NYC, Massachusetts, Vermont, spent lots of time in California…
Nothing compares to South Florida
The weather. The peace. The people. The politics.
Best state in the US and it’s not remotely close.
Keith Rabois reposted
These are all the same people who censored conservatives and President Trump.
They are running the same playbook against conservatives again.
when google, amazon, and microsoft all back the same startup in court, they're not doing it for anthropic — they're protecting their own ai playbooks
www.bbc.com/news/articles/c4g7k7zdd0zo
Keith Rabois reposted
here is how @Opendoor is making 🏡mortgages accessible and more affordable for families (hint: direct access to software that cuts fees / middle people)
Opendoor is getting attention for offering mortgage rates that look "below market" and I want to talk about it. This isn't some magic trick. It's actually pretty basic. Here's how we do it:
Opendoor mortgage rates aren't marked up. The end.
See, when people talk about "market rates" for mortgages, they telling you about the rates they see online from their lender, or from Mortgage News Daily, or some other source. Remember: those rates include 350 basis points of markup on average, based on self-reported data to the Mortgage Bankers Association.
350 basis points is not nothing.
As a rough rule of thumb, every 100 basis points markup raises a consumer's mortgage rate by 0.25 percentage points. So, let's all acknowledge that "market rates" in mortgage reflect 350 basis points of markup, which raises a customer's mortgage rate by roughly 0.875.
Opendoor changed that. Our mortgage rates are what happens when you take that markup out. It's like what E*TRADE did for stocks. In the 1980s, the market price of a stock was whatever its price was plus whatever your broker charged. It's why every broker had a different price. Today, the price of a stock is the same everywhere.
So if Opendoor's mortgage rates look "below market" to you, they're actually not. This is just the first time you're seeing mortgage rates without a massive markup.
More here: www.opendoor.com/articles/why-mortgage-rates-at-opendoor-…
Keith Rabois reposted
Since the 2020 presidential election, Anthropic’s founders, board members and employees have made more than $200 million in political donations - 99.8 percent of which went to Joe Biden, Kamala Harris, Hillary Clinton and other Democrats.
Now this “Institute” is supposed to offer reliable policy?
Introducing The Anthropic Institute, a new effort to advance the public conversation about powerful AI.
www.anthropic.com/news/the-anthropic-institute
Keith Rabois reposted
Well now, it’s confirmed the Mullah Regime can’t make the next payroll.
This is where things start to fall apart for the Regime Security Forces.
Asluyeh oil complex has stopped. That was the IRGC's piggy bank (yes, I've been paying attention since 2022)
Today vs 25th of Feb. Less flaring, less ships at anchorage.
Activity in Kharg has stopped too, so no more oil exports to China and nothing to lose in mining Hormuz ...
Keith Rabois reposted
BREAKING: Swedish start-up Lovable sees revenue jump from $300M ARR to $400M ARR in a single month.
Ryan Meadows, Chief Revenue Officer @Lovable says annual recurring revenue has surged by more than 30%, from $300 million to $400 million in a single month, and could top $1 billion by year's end.
"It's accelerating quite a bit," Meadows said. "We've doubled the number of active users daily just in the last couple of months." - Meadows
Keith Rabois reposted
It gets worse from the victims testimony:
“The saddest part of this story is how the police handled it,” Lior said. “Only today did the investigator reach out to us and told me he is busy and has other cases. He also told Daniel, ‘This doesn’t look good for you.’ What does that mean? We didn’t start it — we were attacked by MMA fighters. It could have ended very badly. I have kids.”
@FBI @FBIDirectorKash @TheLeoTerrell
Huge Silicon Valley Biotech and Tech Association the main attacker is UC Berkeley Law Grad.
Highly concerning.
Assyrian from North Iraq. Shocking the hate crime and terror attack against Jews.
Keith Rabois reposted
And no one helped. They just sat there. That’s the Bay Area for you. Filled with progressives and social justice warriors who do nothing.
Keith Rabois reposted
Keith Rabois reposted
Opendoor's Q4 2025 financial report shows that 33% of listed homes have been on the market for more than 120 days, compared to 14.8% on February 20th, and now... 11.48%.
$OPEN
#faster
Keith Rabois reposted
**It Can Now Be Revealed**
Analysts assess that Iran was attempting reopening the sealed underground chambers in a determined effort to recover its stockpiles of **highly enriched uranium** - material perilously close to weapons-grade and sufficient, if further processed, for multiple nuclear devices.
Recent satellite imagery from February 2026 reveals renewed excavation at Iran's Isfahan nuclear complex: narrow tracks carved through the desert, fresh dirt mounds, and heavy earth-moving near tunnel entrances previously buried after U.S. strikes.
This urgent activity underscores the grave threat posed by Iran's advancing nuclear capabilities.
It explains precisely why the war could not be delayed: action in February 2026 was imperative to prevent Iran from retrieving and weaponizing this **enriched uranium** stockpile before it crossed the irreversible threshold and move the material to other, much deeper underground.
Agree.
prediction re the end of spreadsheets
AI code gen means that anything that is currently modeled as a spreadsheet is better modeled in code. You get all the advantages of software - libraries, open source, AI, all the complexity and expressiveness.
think about what spreadsheets actually are: they're business logic that's trapped in a grid. Pricing models, financial forecasts, inventory trackers, marketing attribution - these are all fundamentally *programs* that we've been writing in the worst possible IDE. No version control, no testing, no modularity. Just a fragile web of cell references that breaks when someone inserts a row.
The only reason spreadsheets won is that the barrier to writing real software was too high. A finance analyst could learn =VLOOKUP in an afternoon but couldn't learn Python in a month. AI code gen flips that equation completely. Now the same analyst describes what they want in plain English, and gets a real application - with a database, a UI, error handling, the works. The marginal effort to go from "spreadsheet" to "software" just collapsed to near zero.
this is a massive unlock. There are ~1 billion spreadsheet users worldwide. Most of them are building janky software without realizing it. When even 10% of those use cases migrate to actual code, you get an explosion of new micro-applications that look nothing like traditional software. Internal tools that used to live in a shared Google Sheet now become real products. The "shadow IT" spreadsheet that runs half the company's operations finally gets proper infrastructure.
The interesting second-order effect: the spreadsheet was the great equalizer that let non-technical people build things. AI code gen is the *next* great equalizer, but the ceiling is 100x higher. We're about to see what happens when a billion knowledge workers can build real software.
Keith Rabois reposted
🚨 BREAKING: Researchers at UW Allen School and Stanford just ran the largest study ever on AI creative diversity.
70+ AI models were given the same open-ended questions. They all gave the same answers.
They asked over 70 different LLMs the exact same open-ended questions.
"Write a poem about time." "Suggest startup ideas." "Give me life advice."
Questions where there is no single right answer. Questions where 10 different humans would give you 10 completely different responses.
Instead, 70+ models from every major AI company converged on almost identical outputs. Different architectures. Different training data. Different companies. Same ideas. Same structures. Same metaphors.
They named this phenomenon the "Artificial Hivemind." And the paper won the NeurIPS 2025 Best Paper Award, which is the highest recognition in AI research, handed to a small number of papers out of thousands of submissions.
This is not a blog post or a hot take. This is award-winning, peer-reviewed science confirming something massive is broken.
The team built a dataset called Infinity-Chat with 26,000 real-world, open-ended queries and over 31,000 human preference annotations. Not toy benchmarks. Not math problems.
Real questions people actually ask chatbots every single day, organized into 6 categories and 17 subcategories covering creative writing, brainstorming, speculative scenarios, and more.
They ran all of these across 70+ open and closed-source models and measured the diversity of what came back. Two findings hit hard.
First, intra-model repetition. Ask the same model the same open-ended question five times and you get almost the same answer five times.
The "creativity" you think you're getting is the same output wearing a slightly different outfit. You ask ChatGPT, Claude, or Gemini to write you a poem about time and you keep getting the same river metaphor, the same hourglass imagery, the same reflection on mortality.
Over and over. The model isn't thinking. It's defaulting to whatever scored highest during alignment training.
Second, and this is the one that should really alarm you, inter-model homogeneity. Ask GPT, Claude, Gemini, DeepSeek, Qwen, Llama, and dozens of other models the same creative question, and they all converge on strikingly similar responses.
These are models built by completely different companies with different architectures and different training pipelines.
They should be producing wildly different outputs. They're not. 70+ models all thinking inside the same invisible box, producing the same safe, consensus-approved content that blends together into one indistinguishable voice.
So why is this happening? The researchers point directly at RLHF and current alignment techniques. The process we use to make AI "helpful and harmless" is also making it generic and boring.
When every model gets trained to optimize for human preference scores, and those preference datasets converge on a narrow definition of what "good" looks like, every model learns to produce the same safe, agreeable output. The weird answers get penalized.
The original takes get shaved off. The genuinely creative responses get killed during training because they didn't match what the average annotator rated highly. And it gets even worse.
The study found that reward models and LLM-as-judge systems are actively miscalibrated when evaluating diverse outputs. When a response is genuinely different from the mainstream but still high quality, these automated systems rate it LOWER. The very tools we built to evaluate AI quality are punishing originality and rewarding sameness.
Think about what this means if you use AI for brainstorming, content creation, business strategy, or literally any task where you need multiple perspectives. You're getting the illusion of diversity, not the real thing.
You ask for 10 startup ideas and you get 10 variations of the same 3 ideas the model learned were "safe" during training. You ask for creative writing and you get the same therapeutic, perfectly balanced, utterly forgettable tone that every other model gives.
The researchers flagged direct implications for AI in science, medicine, education, and decision support, all domains where diverse reasoning is not a nice-to-have but a requirement.
Correlated errors across models means if one AI gets something wrong, they might ALL get it wrong the same way. Shared blind spots at massive scale.
And the long-term risk is even scarier. If billions of people interact with AI systems that all think identically, and those interactions shape how people write, brainstorm, and make decisions every day, we risk a slow, invisible homogenization of human thought itself. Not because AI replaced creativity.
Because it quietly narrowed what we were exposed to until we all started thinking the same way too.
Here's what you can actually do about it right now:
→ Stop accepting first-draft AI output as creative or diverse. If you need 10 ideas, generate 30 and throw away the obvious ones
→ Use temperature and sampling parameters aggressively to push models out of their comfort zone
→ Cross-reference multiple models AND multiple prompting strategies, because same model with different prompts often beats different models with the same prompt
→ Add constraints that force novelty like "give me ideas that a traditional investor would hate" instead of "give me creative ideas"
→ Use structured prompting techniques like Verbalized Sampling to force the model to explore low-probability outputs instead of defaulting to consensus
→ Layer your own taste and judgment on top of everything AI gives you. The model gets you raw material. Your weirdness and experience make it original
This paper puts hard data behind something a lot of us have been feeling for a while. AI is getting more capable and more homogeneous at the same time.
The models are smarter, but they're all smart in the exact same way. The Artificial Hivemind is not a bug in one model. It's a systemic feature of how the entire industry builds, aligns, and evaluates language models right now.
The fix requires rethinking alignment itself, moving toward what the researchers call "pluralistic alignment" where models get rewarded for producing diverse distributions of valid answers instead of collapsing to a single consensus mode.
Until that happens, your best defense is awareness and better prompting.
Keith Rabois reposted
Universities will now be required to disclose not only which foreign countries fund them—they’ll be made to publicly name exactly who is writing the checks.
This is a long-overdue step toward transparency and exposing the foreign money quietly poisoning our students.
Keith Rabois reposted
The latest Carta data on $70B+ invested into Seed through Series C startups in 2025 highlights something many of us in NYC see every day.
NYC is now:
• #1 in Fintech
• #1 in Consumer
• #2 overall startup ecosystem
• Top 2 in AI, SaaS, Biotech, and Healthtech
But the most interesting takeaway isn’t the rankings, it’s the supply and demand imbalance.
Founder demand in NYC still significantly exceeds early-stage VC supply.
Which makes NYC one of the largest structural arbitrage opportunities in venture today.
👇🏾👇🏾👇🏾
True.
Keith Rabois reposted
Trump wants to get out of the Middle East, unlike his predecessors he is eliminating the Iran threat instead of managing it. He sees the region as a business project for America which is why this administration is promoting things such as India-Middle East-Europe Corridor for trade and logistics and Pax Silica, which is a framework to secure supply chains and other technologies. This is the big picture when we take into consideration what he was actually doing.
www.zinebriboua.com/p/trumps-middle-east
Keith Rabois reposted
Just sold my home w/ Opendoor $OPEN
A+ experience I’m extremely happy as a customer & shareholder.
I gave a few pieces of feedback on the app itself, otherwise the process from beginning to end was seamless, and I was always able to reach someone.
19 days (including weekends)
Keith Rabois reposted
ChatGPT has 87% market share of app time spent. 8x more than the next biggest player.
🚨 The @a16z consumer AI Top 100 is back!
For the sixth time, we ranked consumer AI websites and mobile apps by usage (monthly unique visits and MAUs).
This edition, we changed the rules. Here's why - and what the new list says about where consumer AI is heading 👇
Ha!
In retrospect we all should have known Nate Silver pretending to be a commodities trader last night was the top.
Keith Rabois reposted
A recent NBC poll found ICE more popular than the Democratic Party. www.documentcloud.org/documents/27777984-nbc-news-march-2…. It barely edged out Iran in popularity. As Democrats push airports toward a shutdown during peak Spring break travel, they could soon lose not just to Iran but Ebola in future polls...
Keith Rabois reposted
People keep saying the U.S. is now paying more for oil and gaining nothing from striking Iran. That’s simply wrong. The strategic benefits stack up fast.
Here are 9 benefits to start:
1) Preemption – Iran was reportedly negotiating for advanced Chinese anti-ship missiles that could threaten U.S. aircraft carriers. Acting early prevents a far more dangerous battlefield later.
2) Contain China – Iran is one of Beijing’s key strategic partners in the Middle East. Weakening Tehran disrupts China’s regional network.
3) Secure Hormuz – The Strait of Hormuz carries roughly 20% of global oil flows. Greater control over this chokepoint means enormous leverage over global energy markets.
4) Signal to Gulf states – Saudi Arabia, the UAE and others quietly exploring Chinese weapons suddenly remember who actually provides regional security.
5) Deterrence demonstration – A decisive strike reminds rivals that the U.S. still dictates the rules of escalation.
6) Energy leverage over Asia – Japan and South Korea rely heavily on Middle Eastern oil passing through Hormuz, giving Washington additional strategic influence.
7) Pressure on China’s energy lifeline – Much of China’s imported oil also passes through the same chokepoint.
8) Reset Middle East balance – Weakening Iran reduces the influence of Tehran’s proxy networks across the region.
9) Global power signaling – The message travels far beyond Iran: the United States still has the ability to shape outcomes in critical theaters.
Geopolitics isn’t sentimental. It’s strategic arithmetic.
Applaud. Let’s see who complains about this.
🚨 BREAKING: President Trump announces the US will GRANT ASYLUM to the Iranian Women's Soccer team if Australia refuses to grant them safety
Trump believes they will be KILLED when they return home. The women were spotted standing up for their rights in Australia 🇺🇸🇮🇷
"Australia is making a terrible humanitarian mistake by allowing the Iran National Woman’s Soccer team to be forced back to Iran, where they will most likely be killed. Don’t do it, Mr. Prime Minister, give ASYLUM. The U.S. will take them if you won’t. Thank you for your attention to this matter. President DONALD J. TRUMP"
Keith Rabois reposted
A lot of people don't want to hear this, but it's hard to have a bad day when you wake up early and work out before most people are even awake.
Keith Rabois reposted
8 months ago, I said $OPEN could go to $82 by 2028.
I now think I may have been too conservative.
Not because the housing market got easier.
Because the operator changed.
Kaz took over in September.
Since then, Opendoor's weekly acquisitions have gone from roughly 131 homes/week to 442/week — you can track it yourself at accountable.opendoor.com.
That's 3.4x in 5 months.
And this is happening before the seasonal peak.
That's the first thing.
The second thing is headcount.
Opendoor is down roughly 40% from prior levels. If Kaz gets anywhere close to the kind of lean operating model Keith Rabois has talked about, Wall Street is nowhere near modeling the operating leverage.
The third thing is the mortgage.
This is the part people still don't get.
At Carvana, the car was never the whole business.
The financial products were.
Same idea here.
The home is the distribution channel.
Mortgage, title, insurance, and the closing stack are where this gets much bigger.
That's why the new 4.99% mortgage beta matters.
The skeptics will say iBuying loses money, a subsidized mortgage will lose money, and a faster pace just means they go bankrupt sooner.
Kaz heard the same thing at Shopify when he built Shop Pay Installments from zero to one of the largest installment products on the internet in a year. He tweeted last week that he remembers "all the people who were very confident it could not be done."
The old debate was whether iBuying works.
The better question now is whether Opendoor is quietly becoming a housing-fintech platform.
That's a very different multiple.
My original $82 target used Bloomberg's FY2028 revenue consensus. Wall Street currently has $4.2B for FY2026. The acquisition pace implies nearly double that.
If the pace is real, and if Kaz's Shopify playbook translates faster than expected, that timeline may be pulling forward.
That's what I'm watching now.
Don't always love "if you're so smart, why aren't you making bets?", but it's appropriate here. Are you making investments based on your decidedly contrarian hypothesis that "within days" we will "100% control" the oil supply in the Strait of Hormuz? You could make a killing!
As someone who has covered markets for years, I will say this. commodity traders, like equity traders (in the words of a smart bond trader I know) are "imbeciles." They trade off headlines, totally myopic in short-term thinking and predicting. They always screw up what's actually going down, never seeing around corners. It was the case in the run up to the 2008 where there was more than a few market "rallies" before the whole market blew up. This oil price spike has the same feel to it. It's not considering that within days we will 100% control the supply of oil coming out of the Straits of Hurmuz, or that Iran will be 100 decapitated as a military force and a financier of terror. There will be a peace dividend to all of that. Meanwhile buckle up as the Karens who control oil prices have their say, at least for now
www.wsj.com/livecoverage/iran-war-news-updates-2026/card/…
looks open:
Correct.
Adopting AI into your work makes everyone a CEO, and there's something about it that very few people understand (mostly because few people are CEOs):
When you have an front-line job, you spend maybe 1-5% of your brainpower of high-level, critical-thinking strategy. Most of your job is just running some SOP that's been given to you.
Move up a couple levels, maybe you're managing a little team, and you might spend 10-25% of your time making critical strategic decisions. The rest of the time, you're executing departmental strategy that was handed to you, and your time is spent making sure your team carries that out.
Once you're the CEO, you'll have (if you're good at your job) delegated all of the routine problems and issues that are straightforward to solve to capable executives.
What's left for you? Only the hardest and most critical decisions. The better and more capable your staff, the harder the questions will be that bubble up to you, because they take care of everything else. Those become the only duties you have left: thinking REALLY hard about very dfificult, ambiguous, strategic decisions. And now it's your entire job. Instead of 10-25% intensity (or less), it's 80-90% intensity.
Incidentally, this is why you hear about CEOs having these intensely regimented lives and health-oriented habits: it's all designed to biologically support the fact that their brains have to be operating at peak capacity nearly all the time.
So now everyone is starting to manage armies of agents doing the routine parts of their job. If you're good, you can distill your job into a clear SOP that the agents run, and now "all you have to do" is oversee them...
... and now lots of people are learning that being the boss isn't quite as easy as they thought.
Keith Rabois reposted
$open @nejatian … @ericjackson
Just used opendoor key app to submit pictures for my house in NY. It really is too good to be true … Even my 12 yrs old daughter could’ve done 99% of the tasks required to submit pictures!
Waiting for my offer …!
Keith Rabois reposted
Opendoor - The Future of Real Estate - Here's 10 Reasons I'm Extremely Bullish on $OPEN
1. Revolutionizing $2T housing like Uber did rides — instant, digital, no-agent future
2. AI pricing + underwriting = massive cost savings & faster turns (already cutting $3-5k/loan)
3. Lower rates tailwind: mortgage rates collapsing → 46% QoQ acquisition surge, 15%→41% rev growth '26-'27
4. 4.99% fixed mortgages (1% below market) embedded in buying flow = huge conversion & velocity boost
5. Tokenization edge: deep data moat positions Opendoor to lead blockchain real estate & fractional ownership
6. Nationwide in 50 states + remote photos + 300% acquisition ramps = Shopify-level operational scale
7. Path to breakeven: positive unit economics, 23% faster inventory, high-margin title/escrow add-ons
8. Data moat like Reddit → future licensing/monetization of housing insights in AI era
9. High short interest + retail momentum = classic squeeze setup (PLTR/CVNA vibes)
10. Full-stack optionality: embedded finance, Zillow partnerships, rental-to-own → multi-trillion TAM capture
Given that this got a lot of questions. Let me be helpful. Again. This is early. We have a lot to learn. But structurally at least 65-85 bps worth of yield of any mortgage is the margin and inefficiency that goes to the chain of companies and sales and ops people that touch that mortgage. You reduce that, you reduce the costs. There are also obviously scale advantages.
We haven’t invented new math here. What we have done is say if our goal was to offer the lowest mortgage rate possible rather than make the most amount of money possible, what would we do?
We are early but we have been able to automate away much of the flow and we believe we can automate the rest. Also, Opendoor as the seller of the home has unique cost structures that allow us to do things (for example I’ve talked about this publicly, sitting around waiting for a mortgage to get funded by a bank is a relatively large cost for us today!).
Our mortgage product is live only in limited flows in limited places. We are early. But, thus far, we think we are onto something.
Keith Rabois reposted
A lot of people underestimate (because they weren’t around for it) the degree to which the Carter foreign policy was severely demoralizing, especially to Trump’s generation, and how much current foreign policy is informed by it.
Keith Rabois reposted
Ro Khanna today chose to raise his hand to protect the Islamic Regime in Iran
He voted to defend Islamic Terror.
Anyone supporting this psychopath is a psychopath.
BREAKING: The U.S. House has overwhelmingly voted 372-53 to declare Iran the world’s largest state sponsor of terrorism.
53 Dems voted no.
Keith Rabois reposted
No matter what happens going forward, something as earth-shattering as the fall of the Soviet Union has already happened in the Middle East. jewishjournal.com/commentary/columnist/editors-note/38733…
Keith Rabois reposted
HUGE story in Bloomberg, revealing that the Trump admin is drafting a new rule to require an export license for most AI chip sales globally. The details here will matter a lot, but this could fill several gaping holes in the current export control regime that China is using to obtain/use banned US chips.
I’ll have more analysis shortly, but my top-line take is that a regulation dictating how US tech can be used in AI infrastructure globally is the only way to ensure the global AI buildout is trusted and secure, is by far the most effective way to promote the full U.S. AI stack (chips/cloud/models), and could be the most significant step taken yet to constrain China’s access to AI computing power.
Any such rule will be hugely consequential, the details will matter, and the regime should be transparent and clear to ensure any deals are made solely based on their contribution U.S. national security. But as AI gets more and more important, the only question was when we would have a rule like this, not if.
The Trump administration drafts new rules requiring US approval for global sales by Nvidia and other AI chipmakers www.bloomberg.com/news/articles/2026-03-05/us-drafts-rule…
Keith Rabois reposted
We have proof that NATTOKINASE can dissolve the "un-dissolvable" clots seen in long COVID.
New research shows how NK breaks down fibrinaloid microclots, the tough structures that usually resist your body's natural defenses.
Here's the breakdown:
Keith Rabois reposted
What did George Kennan, the father of “containment,” have to say about using force against Iran? Surprise: he was all for it. martinkramer.org/2011/01/04/george-kennan-iran-hawk/
Keith Rabois reposted
This is unfathomably retarded and Decel.
Would nuke the GDP of New York.
Beyond stupid.
A New York bill would ban AI from answering questions related to several licensed professions like medicine, law, dentistry, nursing, psychology, social work, engineering, and more.
The companies would be liable if the chatbots give “substantive responses” in these areas.
Loser.
EXC: My Iran deep dive, which the Telegraph has splashed on this morning
🚨Starmer wanted to let the US use Dieo Garcia to attack Iranian missile sites last Friday. The defence secretary John Healey supported him at the NSC. But his ministers, led by 'petulant' Ed Miliband, stopped him
🚨Chief of the Defence staff had to tell the Americans how to frame their request on Saturday so ministers could agree it on Sunday. 48 hours lost and with it UK credibility with allies
🚨We've have all thought Richard Hermer was the main blockage. Actually it was Milband, Reeves, Cooper and Mahmood. A PM who can't carry his own war cabinet
spectator.com/article/whose-side-are-you-on-how-keir-star…
Keith Rabois reposted
Seriously when do opponents of the Venezuela intervention admit they were wrong? What exactly has to happen for them to admit it was a brilliant thing to do?
Delcy Rodríguez agradece a Donald Trump, al hombre que hace dos meses bombardeó Caracas y tomó por la fuerza a Maduro.
Still the only product that actually improves your sleep! Sleep is the secret to being happier, healthier and wealthier.
Today we're announcing a new round of funding that values @eightsleep at $1.5 billion, led by @tether Investments.
Ten years ago I started Eight Sleep with one conviction: the night is the most underutilized lever in human health. Almost nobody was engineering their sleep.
Last year, we hit free cash flow positivity, launched 3 new products, expanded to 34 countries, and published peer-reviewed studies showing the Pod reduces menopausal hot flashes by 56% and restores the body's natural circadian temperature rhythm during sleep, lowering core body temperature and improving cardiovascular recovery. A product you sleep on is producing clinical outcomes that rival pharmaceutical interventions.
We're now building a predictive AI agent trained on 1B+ hours of sleep data. It anticipates your night before it happens.
And we're advancing FDA filings for sleep apnea detection. Passive. Every night. No wires, no clinic visits.
The night is just the beginning.
This is quite good as a distillation.
Playbook for building a legendary company:
- Bring together an outlier team
- Have that team come up with a novel and compelling insight
- Turn that insight into a product or service offering and iterate that offering until it has a compelling value proposition
- Turn that compelling value proposition into a unique one
- Take all of those already hard ingredients and weave them into an economic engine that has a sustainable competitive advantage
- Do all the above again and again, so the company can compound growth at scale
None of these are easy. But if it was easy, everyone would do it.
Wow.
British billionaire applies for German citizenship so he can flee the UK as he says the country 'is an uncomfortable place for Jews' www.dailymail.com/news/article-15609643/British-billionai…
Keith Rabois reposted
A lot of people ask how the hype around Claude impacts Spellbook.
The answer: lawyers booked more demos than ever last week—410!
We saw this pattern when ChatGPT launched in 2022: the mass appeal of generalist tools got lawyers deeply interested in AI, and then they started searching for more specialized tools, ultimately driving a ton of traffic for us.
True.
If you listen to just ONE thing today, this podcast is it. @havivrettiggur reviews the #iran war from the viewpoint of American grand strategy. He gets it exactly right. podcasts.apple.com/us/podcast/ask-haviv-anything/id179459…
Keith Rabois reposted
$OPEN This is bigger than people realize.
When the CEO of Opendoor says that 65–85 bps of every mortgage today is pure margin and inefficiency in the chain — and that they can automate much of it away — that’s structural disruption.
This isn’t about a promo rate.
It’s about redesigning the entire mortgage stack.
• Fewer intermediaries
• Automated workflows
• Scale advantages
• Lower costs = lower rates
And they’re still in beta.
If Opendoor can control both the home sale and the financing, they have a cost structure traditional lenders simply can’t match.
They didn’t invent new math — they just chose to optimize for the lowest possible rate instead of the highest possible margin.
Early days.
But this could be a game changer for American homeowners — and a massive moat if they execute.
@Opendoor @nejatian @rabois
Given that this got a lot of questions. Let me be helpful. Again. This is early. We have a lot to learn. But structurally at least 65-85 bps worth of yield of any mortgage is the margin and inefficiency that goes to the chain of companies and sales and ops people that touch that mortgage. You reduce that, you reduce the costs. There are also obviously scale advantages.
We haven’t invented new math here. What we have done is say if our goal was to offer the lowest mortgage rate possible rather than make the most amount of money possible, what would we do?
We are early but we have been able to automate away much of the flow and we believe we can automate the rest. Also, Opendoor as the seller of the home has unique cost structures that allow us to do things (for example I’ve talked about this publicly, sitting around waiting for a mortgage to get funded by a bank is a relatively large cost for us today!).
Our mortgage product is live only in limited flows in limited places. We are early. But, thus far, we think we are onto something.
Keith Rabois reposted
EAs are for government control of AIs as long as it's their people in charge
We have been calling them out for years and now the mask has come off
Self-serving power-seeking disguised as virtue
Overheard in Silicon Valley: "Every single person who was in favor of government control of AI, is now opposed to government control of AI."
This.
So good by @GavinSBaker
In Michael Jordan’s second to last year with the Chicago Bulls, Jerry Krause, GM of the Bulls said, “Listen players don’t win championships, organizations do. It isn’t just Jordan or Pippen. It’s how we scout, draft, train, it’s our system." Well… after Michael Jordan left they never won another championship.
The thing allocators struggle with is it feels safe and good to say there is a process and its repeatable. Yes, it is important to have a process that is repeatable that works for you, but any process that is repeatable that generates significant alpha will be quickly arbed away in a competitive world.
So where does repeatable outperformance come from? I would submit that any investment organization no matter how big, there are somewhere between 2-10 people where if you took those people out, and that organization would have the exact same process, the results would be dramatically different.
podcasts.apple.com/us/podcast/capital-allocators-inside-t…
LOL.
SCOOP: Anthropic was among the AI companies that submitted a proposal earlier this year to compete in a $100 million Pentagon prize challenge to produce technology for voice-controlled, autonomous drone swarming, acc to people familiar w/ matter.
www.bloomberg.com/news/articles/2026-03-02/anthropic-made…
Keith Rabois reposted
Nearly every country that’s introduced a wealth tax has since repealed it because it’s a form double and triple taxation that crushes investment and kills jobs. Even 70% of Dem voters rejects this commie garbage.
New - @BernieSanders to unveil new mega tax on America’s ~1,000 billionaires, would raise $4.4 trillion
Potential marker for other candidates who do run in 2028, @RoKhanna is introducing House version
$ would be directed to:
- $3K stimulus checks per person earning under $150K/yr
- Expand Medicare to cover dental/vision/hearing for all seniors
- Ensure every public school teacher gets paid at least $60K/yr
- $850B for affordable home construction (“abolish homelessness”)
- Universal childcare
- Reverse Trump Medicaid cuts
Yes.
Anthropic and Alignment
Anthropic is in a standoff with the Department of War; while the company's concerns are legitimate, it position is intolerable and misaligned with reality.
stratechery.com/2026/anthropic-and-alignment/
Keith Rabois reposted
ok sam altman dropped some truth bombs about the whole pentagon, anthropic, openai sitch overnight and tbh it’s VERY eye-opening:
- important: said the pentagon *wasn’t* trying to modify how models behave (but they request you don’t “interfere with operational decisions” for national security (anthropic rejected this).
- openai had been talking to DoW for months about NON-classified work. but this week shit escalated to include classified ops. why? because the situation with anthropic needed to be “de-escalated” and he wanted to support them in their “very important mission”
- “what happens if the government tries to nationalize openai?” - said he “doesn’t know” and thinks AGI might actually be good as a government project vs private.
- this was fucking wild: said a LOT of people were debating whether they prefer a democratically elected gov or unelected private company to hold more power over (how is that a debate?) sam chooses democracy over latter all-day.
- said it’s not about “usage policies vs. law” it’s about BOTH being used to form a safety framework (anthropic refused to cave on this)
- gov contract lets openai update the models as they see fit (no gov intervention)
- said people take their national security for granted. people don’t realise what goes into keeping them safe.
i’ve got to give it to sam - this was an impressively transparent way to face public scrutiny (reminds me of reddit AMAs) and imo he did a good job.
I'd like to answer questions about our work with the DoW and our thinking over the past few days. Please AMA.
Keith Rabois reposted
Sometimes, it doesn't just *seem* like @nfergus is smarter than everyone else. Please consume his essential reading @TheFP piece on Iran in full asap and NB this vital sentence:
"It is a foolish commentator who attempts to pass definitive judgment on a major military operation while it is still underway."
www.thefp.com/p/niall-ferguson-khamenei-may-be-gone
Keith Rabois reposted
The combat and intelligence leaders of the Second World War were the last competent American foreign policy practitioners before Trump.
What happened?
1/
The fact that a billionaire real estate playboy who liked to slap his name on steaks and wine has proven to be a better diplomat and military strategist than every other politician and foreign policy expert over the last 30 years is such a damning indictment of the DC establishment I honestly don’t know how they recover.
What’s clear from the Maduro raid and today’s strikes is that forever wars were a choice.
Keith Rabois reposted
“There’s no solution to illegal immigration!”
> Trump shuts down southern border
“There’s no solution to Venezuela!”
> Trump black bags Maduro
“There’s no solution to Iran!”
> Trump blows up the ayatollah
How many of our “problems” are just a lack of political willpower?
Keith Rabois reposted
If you call yourself a progressive, and you’re not even a little excited by the prospect of a woman-hating, gay-hating, Jew-hating neanderthal regime that in no way represents the will of its people being crushed, you’re not actually a progressive.
Excellent distillation.
The wars in Iraq and Afghanistan discredited interventionism in the eyes of the American public.
What the last few years have done in contrast is discredit anti-interventionism, or the idea that American power cannot be used for good.
First came the Abraham Accords, where contrary to every prediction made by practically every Middle East expert, Israel signed peace treaties with several of its neighbors without taking care of the Palestinian issue first.
Throughout the first Trump administration, the maximum pressure campaign had Israel, the Arabs, and the US all on the same side against Iran.
The Syrian regime falls, the new leader in Damascus is pro-Western. We ended up fighting both Assad and ISIS, and defeated them both.
Then you had the Venezuela operation. Without losing any American soldiers, we removed the Maduro regime, and replaced it with a government that has been freeing political prisoners and opening up the oil industry to American companies.
Cuba will probably be overthrown without any cost next, ending the over half-century nightmare there.
Now, after taking a detour during the Hamas war and wiping out Hezbollah's leadership, Israel along with the US slaughter the entire upper echelon of the Iranian regime. What's next in Iran? Who knows? But isn't there at least a chance it'll be better than the old theocrats? Why would anyone dismiss the possibility? People deny this is possible just because they want America and the West to fail.
Russia and China remain problems and North Korea is an exception. But when it comes to weaker enemies around the world, we can just kill them and everything will be fine.
As it turns out, we misunderstood the lessons of Iraq and Afghanistan.
The lesson wasn't "never do regime change" or "never kill bad people."
The lesson we should've taken was "don't do social engineering through war." Killing bad people and removing bad government is fine, both in terms of American interests and from a humanitarian perspective. Now that we've seen how easy it is, we arguably have a moral obligation to keep doing it.
There is a lot to be gained by just removing the worst of the worst. Going from Stalin to Khrushchev and Mao to Deng made things much better. Same with going from Maduro to Delcy, and the same is likely to be true with regard to whatever comes next in Iran and Cuba. Don't waste your time on micromanaging the country and having feminist art exhibitions like we did in Afghanistan. Just remove regime you don't like first, and then be pragmatic with the next steps.
America is powerful and good. We just need to understand that it is ok to admit it.
Keith Rabois reposted
THREAD: For 47 years, the Iranian Regime has been regularly targeting and killing Americans.
Here is a timeline of 50 plus examples of Iran hurting America and terrorizing the world.
1979: The Iranian regime took over the U.S. Embassy in Tehran resulting in a 444‑day hostage crisis.
1983: The Iranian regime provided material support to Hezbollah for the Beirut Marine barracks bombing that killed 241 U.S. service members.
1984: The Iranian regime’s proxy Hezbollah kidnapped CIA station chief William Buckley in Beirut. Buckley died in captivity.
1984: The Iranian regime’s proxy Hezbollah carried out the truck bomb attack on the U.S. Embassy annex in Beirut, killing 24.
Keith Rabois reposted
Iran was behind taking dozens of Americans hostage in ‘79, killing 241 U.S. Marines & others in the ‘83 Beirut barracks bombing, 19 U.S. Air Force personnel in ‘96 Khobar Towers bombing, & 600+ U.S. troops in Iraq, helping Taliban kill our troops, & more.
justthenews.com/government/security/iran-waged-decades-lo…
Keith Rabois reposted
For those wondering how this will play out:
>Trump has done something strong
>Democrats and the media will freak out all day
>Everything Trump did works
>Everyone moves on
>No one gives him credit
>>Move on to the next thing to panic about.
Every. Single. Time.
Keith Rabois reposted
One the smartest Iran experts in DC who doesn’t just analyze — he escaped from Iran and led the economic warfare campaign against the regime. Proud to have him as an @FDD colleague.
1/15 President Trump is the first American president to come to the rescue of the Iranian people and stand against this tyranny. That takes courage and historic vision.
What comes next can't be worse than the Islamic Republic. Believe me, I've lived there, and then witnessed firsthand while serving in the U.S. government how this regime destroyed this country, murdered, suppressed, and forced out tens of millions of generations of talented Iranians.
As a transition away from the brutal regime is taking shape, here is a key issue that needs quick attention: Targeted, phased, and strategic sanctions relief will be essential to empower a democratic transitional government in Iran, one that can deliver stability, national unity, and territorial integrity.
Keith Rabois reposted
President @realDonaldTrump will go down in history as one of the greatest and most consequential presidents we have ever had.
His ability and willingness to make bold and consequential decisions for the benefit of future generations based on the hard and cold facts at hand rather than short-term political considerations is one of his greatest strengths.
No longer are we governed by the politics of the weak who have brought us close to the edge with their weakness and self-interested short-termism.
God bless our nation, our military, and our president. Let’s all pray for our troops who risk their lives on behalf of all of us so we can look forward to a world where evil is eliminated and good prevails.
What we do in life echoes in eternity.
BREAKING: 🔴🔴
U.S. PRESIDENT DONALD TRUMP: "A SHORT TIME AGO, THE UNITED STATES MILITARY BEGAN MAJOR COMBAT OPERATIONS IN IRAN."
Keith Rabois reposted
Ok, so now the truth comes out.
The USA needs AI for critical systems to protect America.
Anthropic wanted the US Gov to call them and ask nicely for permission in case of an emergency.
As I keep saying. AI is critical infrastructure. I have no idea what Anthropic was thinking. There is no way this was ever going to work.
lmao no way
Keith Rabois reposted
Holy fuck Dario Amodei is literally retarded
Keith Rabois reposted
"You Should Have Moral Qualms about Anthropic's Qualms"
www.thefreedomfrequency.org/p/905edaf0-763b-4da9-b273-6b9…
Keith Rabois reposted
Anthropic and AI Safety cult literally pushed a Federal bill to insert a backdoor surveillance chip in every computer called the "Chip Safety Act". The idea that they are opposed to mass surveillance is just laughable.
www.congress.gov/bill/119th-congress/house-bill/3447/text
Keith Rabois reposted
Every technology is a defense technology.
Eventually technologists realize this, they freak out, and then they learn to live with it.
Time for some of you to learn history: steveblank.com/secret-history/
Imagine Apple sold computers or iPads to the DOD and tried to tell the Pentagon what missions could be planned on their computers.
Keith Rabois reposted
BOOM
Trump Orders US Government to Drop Anthropic After Pentagon Feud
www.bloomberg.com/news/articles/2026-02-27/trump-orders-u…
True.
Just FYI, this is the Department of War language that Anthropic doesn't want in their contracts.
I've seen a lot of people say "if Anthropic's red lines are already unlawful, then why doesn't DoW just agree to them?" You could just as well say, "if the DoW only wants lawful uses, why does Anthropic want special conditions?"
In any case, this undermines the (pretty loud) narrative that this debate is happening because the DoW is actively pursuing unlawful uses.
Keith Rabois reposted
Anthropic is lying. The @DeptofWar doesn’t do mass surveillance as that is already illegal. What we are talking about is allowing our warfighters to use AI without having to call @DarioAmodei for permission to shoot down an enemy drone swarms that would kill Americans. #calldario
A statement from Anthropic CEO, Dario Amodei, on our discussions with the Department of War.
www.anthropic.com/news/statement-department-of-war
Keith Rabois reposted
Fastest Mortgage Launch in History.
2 years → 6 weeks
Opendoor was told it'd take 2 years to launch a mortgage product.
It took them 6 weeks.
Underrated BTS of $OPEN product meeting where Kaz @nejatian congratulates the team on the huge accomplishment.
FASTER. @Opendoor
BREAKING: Exclusive BTS Opendoor HQ Tour
A rare look inside the machine.
On Q4'25 earnings day, CEO Kaz (@nejatian) walks us through their SF HQ, lets us sit in on a live product meeting, & shows us how a public company operates under extreme pressure.
Special POV: A deeper conversation with Michael Judd (@michaeljudd321), $OPEN Investor Relations & one of Opendoor’s longest standing employees.
Judd reflects on nearly a decade inside the company from early growth→ $0.50/share → @Opendoor 2.0.
It's clear, Kaz & the team are working faster than ever before & having a ton of fun.
Yes, there is a sword involved.
𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒
(00:00) Inside Opendoor HQ
(00:57) Earnings day energy with CEO Kaz
(03:42) Why is there a sword in the conference room?
(04:58) Retail investors, capitalism, and leading in wartime
(08:08) Michael Judd on a decade of rise and reset
(10:25) Reinventing earnings with live streams
(13:38) Keeping up with Kaz
(14:16) The $OPEN Army
(15:07) From $0.50 share price to resurrection
(18:07) Closed doors at Opendoor?
(19:38) Inside a live product meeting with Kaz & team
BREAKING: Inside Opendoor w/ CEO Kaz Nejatian
On earnings day, @nejatian opened the doors to Sourcery to go under the hood of the public turnaround of a company once trading at ~$0.50 & facing delisting risk
6 months in.
$1 salary.
Up to $2.8B in performance equity.
All gas, no brakes
I toured the office, sat in on a product meeting, BTS of the livestream, & recorded our long-form interview where Kaz shares openly how they’re rebuilding @Opendoor in public
(with the help of the $OPEN army, of course)
This was unlike anything I'd ever seen before.
Big thank you to the Opendoor team for letting me in!
𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒
(00:00) Kaz Nejatian, CEO Opendoor
(01:11) Near delisting & stepping in at a $0.50 share price
(04:20) Obsessing over Opendoor
(07:00) Running a Public company without playing the Headline Game
(09:09) The $1 salary & being fully levered to the mission
(11:15) The stock follows the company, not the other way around
(13:00) Founder mentality & calling out the BS
(15:57) Reading every invoice & slashing consulting spend
(18:00) Paying millions to learn that cheaper means worse
(20:50) Keith Rabois on Opendoor
(23:10) Customer support said, “Have you heard of Google .com?”
(25:05) Hiring YC founders & rebuilding the company’s technical DNA
(27:35) Refusing to become a polished public CEO
(29:48) Attention to detail vs macro thinking
(33:09) The most-watched earnings call in history
(36:10) Posting on X, backlash, & quoting Max Weber
(38:50) The chance to own a piece of America
(41:34) Why speed beats over-optimizing risk
(43:28) Building a board you would pay to have dinner with
(48:01) The “weird” decisions
(51:54) Biggest lessons from Shopify CEO Tobi Lütke
Just finished a great Q1 Opendoor Board meeting.
Keith Rabois reposted
It's easy to get revenue when you sell magic. It's hard to build a valuable business when suddenly everyone can sell magic.
Wide majority of VCs did not and still do not understand this.
The AI era will not be friendly to startups who chased revenue traction at the expense of accumulating advantages (or 7 Powers).
He won the room.
New Jersey Devils fans go crazy for Jack Hughes on his return to play after winning the gold medal.
The AI era will not be friendly to startups who chased revenue traction at the expense of accumulating advantages (or 7 Powers).
Keith Rabois reposted
Try Spellbook. More customers than Harvey and Legora combined, focused on contracts, fits exactly in a lawyer’s workflow in a way that wouldn’t make sense for Claude. Backed by realtime market data about what is “normal” in contracts in 80 countries.
Keith Rabois reposted
BREAKING: Exclusive BTS Opendoor HQ Tour
A rare look inside the machine.
On Q4'25 earnings day, CEO Kaz (@nejatian) walks us through their SF HQ, lets us sit in on a live product meeting, & shows us how a public company operates under extreme pressure.
Special POV: A deeper conversation with Michael Judd (@michaeljudd321), $OPEN Investor Relations & one of Opendoor’s longest standing employees.
Judd reflects on nearly a decade inside the company from early growth→ $0.50/share → @Opendoor 2.0.
It's clear, Kaz & the team are working faster than ever before & having a ton of fun.
Yes, there is a sword involved.
𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒
(00:00) Inside Opendoor HQ
(00:57) Earnings day energy with CEO Kaz
(03:42) Why is there a sword in the conference room?
(04:58) Retail investors, capitalism, and leading in wartime
(08:08) Michael Judd on a decade of rise and reset
(10:25) Reinventing earnings with live streams
(13:38) Keeping up with Kaz
(14:16) The $OPEN Army
(15:07) From $0.50 share price to resurrection
(18:07) Closed doors at Opendoor?
(19:38) Inside a live product meeting with Kaz & team
BREAKING: Inside Opendoor w/ CEO Kaz Nejatian
On earnings day, @nejatian opened the doors to Sourcery to go under the hood of the public turnaround of a company once trading at ~$0.50 & facing delisting risk
6 months in.
$1 salary.
Up to $2.8B in performance equity.
All gas, no brakes
I toured the office, sat in on a product meeting, BTS of the livestream, & recorded our long-form interview where Kaz shares openly how they’re rebuilding @Opendoor in public
(with the help of the $OPEN army, of course)
This was unlike anything I'd ever seen before.
Big thank you to the Opendoor team for letting me in!
𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒
(00:00) Kaz Nejatian, CEO Opendoor
(01:11) Near delisting & stepping in at a $0.50 share price
(04:20) Obsessing over Opendoor
(07:00) Running a Public company without playing the Headline Game
(09:09) The $1 salary & being fully levered to the mission
(11:15) The stock follows the company, not the other way around
(13:00) Founder mentality & calling out the BS
(15:57) Reading every invoice & slashing consulting spend
(18:00) Paying millions to learn that cheaper means worse
(20:50) Keith Rabois on Opendoor
(23:10) Customer support said, “Have you heard of Google .com?”
(25:05) Hiring YC founders & rebuilding the company’s technical DNA
(27:35) Refusing to become a polished public CEO
(29:48) Attention to detail vs macro thinking
(33:09) The most-watched earnings call in history
(36:10) Posting on X, backlash, & quoting Max Weber
(38:50) The chance to own a piece of America
(41:34) Why speed beats over-optimizing risk
(43:28) Building a board you would pay to have dinner with
(48:01) The “weird” decisions
(51:54) Biggest lessons from Shopify CEO Tobi Lütke
Keith Rabois reposted
Higher VO2 max is associated with 5-fold lower mortality risk. Train!
Matt and Mitch are executing with relentless focus and on a path to building one of this generation's defining companies at the application layer.
We've raised $100m at a $1.15b valuation from @Accel, @GVteam, and existing investors to accelerate deployment of the most capable and accurate accounting agents across CAS, tax, audit, and advisory.
Basis is used by 30% of the Top 25 accounting firms and dozens more across the Top 150.
Today we're announcing the first accounting agent to complete a business tax workbook end-to-end.
Our focus on production-grade, long-horizon agents means that 12 months from now, the work Basis handles will make even this look routine.
We're looking for a few very intense people who want to build at the frontier.
Keith Rabois reposted
I am against Chief Product and Technology Officers
Why? In theory, it makes a lot of sense
But in practice ... they generally are neither
They don't code, and they aren't close to the product. They just want to be strategists. They don't want to do the work.
You don't need that
I don't even think Oracle needs that
Hire one of these, you will lose in the Age of AI
Keith Rabois reposted
🚨 PATRIOT TAKEOVER IN MIAMI!
Team USA Men's Hockey Team takes center stage at a Miami club and lead the crowd in:
- Born in the USA
- Star Spangled Banner
- Courtesy of the Red, White and Blue
And HUGE USA chants! Now they're headed to Trump's SOTU 🔥
The future of accounting is led by Basis.
a rethink of accounting workflows from first principles, with a focus on (1) adapting to how different teams operate (turns out every accounting practice works differently), and (2) liberating practitioners to focus on client needs.
congrats @mitch_troy and @trybasis team. 💥
We've raised $100m at a $1.15b valuation from @Accel, @GVteam, and existing investors to accelerate deployment of the most capable and accurate accounting agents across CAS, tax, audit, and advisory.
Basis is used by 30% of the Top 25 accounting firms and dozens more across the Top 150.
Today we're announcing the first accounting agent to complete a business tax workbook end-to-end.
Our focus on production-grade, long-horizon agents means that 12 months from now, the work Basis handles will make even this look routine.
We're looking for a few very intense people who want to build at the frontier.
Here we go.
BREAKING: Inside Opendoor w/ CEO Kaz Nejatian
On earnings day, @nejatian opened the doors to Sourcery to go under the hood of the public turnaround of a company once trading at ~$0.50 & facing delisting risk
6 months in.
$1 salary.
Up to $2.8B in performance equity.
All gas, no brakes
I toured the office, sat in on a product meeting, BTS of the livestream, & recorded our long-form interview where Kaz shares openly how they’re rebuilding @Opendoor in public
(with the help of the $OPEN army, of course)
This was unlike anything I'd ever seen before.
Big thank you to the Opendoor team for letting me in!
𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒
(00:00) Kaz Nejatian, CEO Opendoor
(01:11) Near delisting & stepping in at a $0.50 share price
(04:20) Obsessing over Opendoor
(07:00) Running a Public company without playing the Headline Game
(09:09) The $1 salary & being fully levered to the mission
(11:15) The stock follows the company, not the other way around
(13:00) Founder mentality & calling out the BS
(15:57) Reading every invoice & slashing consulting spend
(18:00) Paying millions to learn that cheaper means worse
(20:50) Keith Rabois on Opendoor
(23:10) Customer support said, “Have you heard of Google .com?”
(25:05) Hiring YC founders & rebuilding the company’s technical DNA
(27:35) Refusing to become a polished public CEO
(29:48) Attention to detail vs macro thinking
(33:09) The most-watched earnings call in history
(36:10) Posting on X, backlash, & quoting Max Weber
(38:50) The chance to own a piece of America
(41:34) Why speed beats over-optimizing risk
(43:28) Building a board you would pay to have dinner with
(48:01) The “weird” decisions
(51:54) Biggest lessons from Shopify CEO Tobi Lütke
Keith Rabois reposted
Very pleased to announce Comp's Series A, led by @khoslaventures/@rabois, with the participation of @KaszekVentures (after leading the Seed) + Canary, Abstract, Endeavor, and Recall.
Congrats to Chris, Pedro, and the full team.
Keith Rabois reposted
It should not be surprising to see $OPEN executing so well.
They have put together an incredible team…and the team you build is the company you build.
Keith Rabois reposted
Jewish man in Europe chased down and surrounded by a mob wielding large sticks.
But this time, it’s good news.
QUINN HUGHES SAVES THE USA AND KEEPS THE QUEST FOR GOLD ALIVE
Keith Rabois reposted
Doordash will soon handle a billion orders a month
Keith Rabois reposted
NYC wants a $127 Billion budget.
The ENTIRE state of Florida has a $117 Billion budget.
NYC has 8.8 million people.
Florida has ~ 22 million people..
Think about that for a second. That is absolutely wild..
Keith Rabois reposted
The Trump Administration creating a U.S.-backed anti-censorship website and daring European countries to block it like they're Iran or China would be an absolutely incredible development.
Scoop:
The State Dept is planning to launch a website, freedom.gov, to help users circumvent content bans in Europe and elsewhere. Comes as admin officials have criticized Europe's speech laws. DOGE's "Big Balls" involved, per sources
w @humeyra_pamuk @Simondlewis
Keith Rabois reposted
FL = 23+ million people
NYC = 8 million people
Zohran Mamdani just proposed a preliminary budget of $127 Billion for NYC…
The entire state of Florida is $117 Billion.
Keith Rabois reposted
JUST IN⚡️: $OPEN rolls out “Buy Directly with @Opendoor”, giving buyers up to 1% savings when purchasing directly through their platform.
Step by step, the end-to-end marketplace vision is taking shape. Listening closely for updates during earnings.
Keith Rabois reposted
@rabois so…Miami…
:)
Eric Glyman @eglyman runs @tryramp, which now powers >2% of US corporate spend. @arampell and I dug in with him on why financial systems have proven resistant to automation, the second-order effects of AI on business, and Ramp’s strategy.
The best.
We have moved our headquarters to Miami, Florida.
:)
"The wealth taxi is moving to Miami."
— typo in an email from Ron Conway about all the founders who are leaving California because of the proposed wealth tax
This is what I voted for.
Over 330,000 Federal employees have been fired, resigned, or retired last year and were not replaced.
The Federal workforce is now at its lowest level since 1966.
The future of legal.ai is inconsistent w the businesses of large law firms, especially the AmLaw 100. If you are a client, if you need someone who abolishes billable hours and increases productivity.
Legal AI products have diverged: those designed to sell to BigLaw innovation committees, and those designed for end users.
In BigLaw, there is no incentive to reduce billable hours with AI hidden behind the curtain.
AI needs to be a PR play: show clients that the firm is innovative and not falling behind.
Legal AI is so visible because every firm has been clamoring to make AI press releases for 2 years.
This is the “evolutionary pressure” that drove companies like Harvey and Legora to simultaneously launch “Client Portals”—a way for law firm clients to log in and see the “innovation” first hand.
We launched a client portal years ago, before our AI product, due to similar pressures. Clients hated it. They wanted to engage with their lawyers where they already work: email, Slack and Teams.
After that experience, we committed to focusing Spellbook on the end users, the lawyers who would use us day-to-day to get work done.
For many of our customers, there is no sales meeting or presentation. There is a 25 minute test drive.
We focus less on top-down sales through committees, and more on bottom-up sales through delighted users.
It’s a slower path. We prove value to users one by one rather than relying on selling top-down licenses across firms.
But we think this subjects our product to better evolutionary pressure.
With patience, we’ve grown to 4,000 in-house legal teams and law firm customers since 2022—more than Harvey and Legora combined, and we’ll soon cross $100m ARR.
4,000 customers who use us purely for the delight and utility of our product, who are rapidly expanding usage.
This also forged an incredible fit with in-house legal teams, who need AI for the pure utility.
AI buyers in every vertical should consider:
“Should we buy AI designed to sell to the innovation committee, or to end users?”
These approaches result in very different product DNA.
Keith Rabois reposted
This speech is worth your time. I think it’s the best speech we’ve had from any Secretary of State. @marcorubio has truly excelled in this role.
WATCH: Secretary Rubio Delivers Remarks to the Munich Security Conference in Munich, Germany. x.com/i/broadcasts/1mnGeNNPqBvJX
Keith Rabois reposted
A standing ovation —in Europe!— for @SecRubio. Deserved. Very deserved. It is a wonderful speech, summoning our 250th birthday as the spine on which it is built. Listen to all 18 minutes and send to your families and friends. Every American can celebrate this.
WATCH: Secretary Rubio Delivers Remarks to the Munich Security Conference in Munich, Germany. x.com/i/broadcasts/1mnGeNNPqBvJX
Keith Rabois reposted
This might have just entered the pantheon of the greatest American speeches of all time.
It’s basically hard nosed Trump foreign policy delivered with Reagan’s confidence and optimism.
WATCH: Secretary Rubio Delivers Remarks to the Munich Security Conference in Munich, Germany. x.com/i/broadcasts/1mnGeNNPqBvJX
Keith Rabois reposted
idk who this sdr from @tryramp is but full calendar open on a saturday, this is why they are worth $32B
sign me up omar
Keith Rabois reposted
Year one of Trump has been incredible. Consider: 4.4% GDP at end of year, 2.4% inflation, record high stock market, lowest murder rate since 1900, most secure border in history, & 10% decline in government employees. It’s honestly extraordinary.
Keith Rabois reposted
1990s: Business → IT
2000s: Business → PM → Eng
2010s: Empowered product teams (PM + Eng + Design)
2020s: Outcome teams (PM/Eng/Design + Data)
202xs: GM model (Business Owner + Builder + GTM)
PMs need to level up to GMs or transition to "member of the technical staff".
The role as defined might soon cease to exist.
Keith Rabois reposted
We have not just a marijuana problem in the United States, but a gambling problem too.
$497.7 million: Money bet on Kalshi for the Super Bowl LX.
$133.8 million: Money bet on the Super Bowl LX in all of Nevada.
Keith Rabois reposted
57% of merged PRs at Ramp in the last 24 hours came from a background agent. Most companies haven’t even started.
The architecture Ramp built matters. Their agent Inspect runs in sandboxed VMs on Modal with full access to everything a Ramp engineer has: Sentry, Datadog, GitHub, CI/CD, feature flags, databases, live preview environments. The agent doesn’t just write code. It runs tests, checks telemetry, verifies frontend changes with screenshots, and opens PRs that pass the same review bar as human-written code.
This is why the number is so high. Model intelligence was already sufficient. Environment parity was the missing piece. Once agents have the same context and tooling humans have, adoption compounds because engineers stop treating the agent as a side tool and start treating it as a parallel teammate running unlimited concurrent sessions.
The product development implications are massive. PMs at Ramp now use Inspect during QA to make changes in real time instead of writing tickets. Designers can ship fixes without waiting for sprint capacity. The marginal cost of implementing a small change drops to near zero, which means the backlog starts to dissolve.
Most engineering orgs are still debating whether to adopt Cursor or Copilot. Ramp already moved past the foreground agent phase entirely. Background agents that run autonomously, verify their own work, and produce merge-ready PRs at scale.
The gap between companies measuring their agent PR ratio and companies that haven’t built the infrastructure to support one is going to define the next era of product velocity.
in the last 24 hours, 57% of merged PRs at ramp came from our background agent
Keith Rabois reposted
NOTICE👀: @ZacksResearch calls $OPEN “The Amazon of Housing”, noting how @nejatian approach mirrors $AMZN early playbook.
✔️ Speed & efficiency first
✔️ AI across the transaction stack
✔️ Layering higher-margin services
✔️ Moving toward a marketplace model
“Not the Amazon of housing yet… but increasingly being built like one.” 💭
Keith Rabois reposted
this should be seen as an absolute tragedy
weed makes you retarded, and worst of all, it gives you the illusion of being fine
it causes an average 8 IQ points drop in habitual smokers, shrinks the hippocampus (your memory center), reduces sperm count, motility, and morphology, and reduces testosterone production
and it takes 1-3 months of complete abstinence just to recover basic attention and memory. most people cant even stop for a week
for heavy users who started before 18 and smoked for a decade or more, some cognitive effects are permanent.
crazy that's even legal
18 million Americans are smoking weed 21+ days a month
This is a full blown addiction crisis that even the NYT now says: "It’s Time for America to Admit That It Has a Marijuana Problem"
Keith Rabois reposted
The AI scientist is here. This will be the end of wokeness
A paper just published in Science — the most prestigious journal in the world — claims to have discovered how humans achieve the highest levels of performance. Its conclusion: child prodigies don't become top performers, early excellence negatively predicts peak performance, and parents should not push their kids for fear of burnout. I found it odd. I asked Claude to analyze it. Claude found a statistical error in the paper's core claim, designed a counter-study with 3x the sample size, ran it, and proved the paper wrong. The correct conclusion: talent is real, it's measurable by age 14, and it predicts who reaches the absolute top. We are not all born equal. The AI scientist doesn't care about your feelings — it just follows the math.
Here's what happened:
The paper (Güllich et al. 2025, Science) claims that elite youth and elite adults are "discrete populations" — that the kids who dominate at 14 are not the ones who dominate at 30. Its key chess finding was based on 24 players. It told millions of parents: don't push your kids, prodigies burn out.
Claude applied Bayes' theorem to the authors' own numbers and showed the data actually proves a strong positive correlation between early and adult performance — the opposite of what was claimed. The "negative correlation" was a statistical illusion created by extreme quantile selection.
Then Claude designed the study the authors should have run: collected age-14 Elo ratings and lifetime peak ratings for every super-grandmaster in chess history — 67 players, nearly 3x the paper's sample. Ran the regression.
Result: β = +0.167, p = 0.003. Early achievement positively predicts peak performance.
An AI just peer-reviewed the world's top scientific journal and won.
Keith Rabois reposted
Nothing will ever top the 1958 #nfl championship halftime show
#superbowl
Amazing.
A 20-year marathon for Chinese exports to the US, back to the starting line overnight.
Goldman Sachs warned last year that China’s share of U.S. imports could fall to 7.5% by 2025.
Now it’s 2026—and reality has confirmed it.
7.5% is where China was in 2001, the year it joined the WTO.
That means two decades of export dominance in the biggest and most profitable consumer market in the world—built factory by factory, port by port—have effectively been erased.
At the peak in 2017, China supplied over 21% of U.S. imports.
“Made in China” was everywhere: Walmart, Target, Amazon.
Then the world changed.
The U.S. pushed reshoring and friend-shoring.
Orders moved to Mexico, Vietnam, India.
Mexico is now America’s #1 import source for the first time in 100 years.
China's exports to the U.S. fell 17% in 2025.
Factories are idle. Jobs are disappearing.
Trust—the real currency of trade—is eroding.
De-China-ization is no longer a theory.
It’s happening.
Weekend reading.
Keith Rabois reposted
ALERT🔔: @nejatian says $OPEN underwriting just saw a major upgrade.
Faster underwriting = more homes processed, better pricing decisions, higher efficiency, and potentially improved profitability. 💨
Keith Rabois reposted
One clear outcome of the AI trade is that while some SaaS companies get hit hard, operationally heavy ones like $OPEN and $DASH will disproportionately benefit from this.
Anyone else have other ideas on where the interesting trades are? Regulatory capture ala PLTR?
True.
We could be living in a very different country right now if Judge Edith Jones had been elevated to the Supreme Court circa 1990.
But alas, we must content and console ourselves with the fact that she continues to serve with distinction on the Fifth Circuit.
Here she is knocking another one out of the park, vindicating the Trump 47 Administration’s immigration policies based on reading immigration statutes just as Congress wrote them.
Keith Rabois reposted
Every parent should know about this new result. For kids with ADHD, proper sleep might work better than drugs can to help your child. And that's not just true for kids with ADHD!
Keith Rabois reposted
Today is the 92nd day in a row that Apple News has not published a story from a single right-leaning outlet
Apple News editors personally select each story is publishes. It is did not include a single right-leaning outlet.
620 to 0 is on purpose!
Keith Rabois reposted
Took my parents to watch Melania tonight.
Theatre was PACKED.
MUST WATCH. ⚡️
Keith Rabois reposted
Anyone who exists in reality knew this already, but the data also proves it.
The political polarization today is overwhelmingly due to the left going completely insane.
Keep this in mind when they call you a right-wing extremist for being a normal person from 20 years ago.
Since 1988, polarization in the US has been overwhelmingly due to the Left moving leftward, while the Right has basically stayed the same. This is according to a new analysis using k-means clustering, which measures actual positions instead of self-labels.
royalsocietypublishing.org/rsos/article/13/2/251428/47991…
Keith Rabois reposted
Credit to the administration. They’re serious about breaking our dependence on the Chinese Communist Party for critical minerals. It’s an investment that will pay dividends for decades.
Critical minerals are vital to technological innovation, economic strength, and national security.
Under President Trump, we are taking decisive steps to identify and develop domestic reserves for critical minerals, to benefit all Americans.
Keith Rabois reposted
we cannot let this happen. and it will lead to complete industrial capture by china. apple spent $500B+ building hardware supply chains in china, we don't need to make that same mistake again. you could build a complete domestic robotics supply chain in <5 years with $100B.
Keith Rabois reposted
What’s going on inside @tryramp right now will be studied by business leaders, academics, policymakers and of course, agents for years to come.
Revenue and Inspect are billion dollar companies and they are just internal tools.
we've been quietly building our own 0 -> 1 revenue stack internally at @tryramp - Ramp Revenue.
our team’s mandate is simple - help our sales team win, drive more pipeline, and build the most efficient gtm org in the world.
it's powered by our customer data platform (processes millions of records of internal, external, and crm data daily) + unified action layer with agents embedded directly in their workflows.
This means that sellers at Ramp don’t have to worry about switching between dozens of systems just to figure out who they should reach out to or what they should say.
>80% of sales workflows are now powered by Ramp Revenue. Sharing a small preview of what we've been building
Keith Rabois reposted
The American people are being lied to about AI.
The Doomers offer apocalyptic prophesies of job loss and oppression; the Utopians promise a future without toil—a life without meaning or mission. They both neglect human agency.
The future of AI is not an inevitability to be endured by the American people—it is for us, the American people, to shape.
I’ve spent the past two decades alongside the men and women building the future of American AI. They include some of the best software engineers in the world, but also college dropouts, veterans, blue-collar autodidacts, and nurses.
They recognize AI is a tool for them to wield to make themselves more productive and our country safer and more prosperous.
Below are some principles and themes I’ve seen informing the people wielding AI effectively and in service of worthy ends.
I. AI is a tool for the American worker, not his replacement.
The promise of AI in the enterprise is to make the American worker 50x more productive—to unleash his taste and agency. Look no further than the maritime industrial base, where AI enabled the manufacturer to open a third shift. Or the ICU where a nurse learned to wield AI so she could spend more time bedside, where she’s needed most.
The future of AI is being built on frontlines and factory floors.
II. The American worker will wield AI to do more with less—and become more productive and valuable as a result.
For a century, American prosperity was underwritten by a simple bargain: when the worker produces more, the worker earns more. That bargain was broken in the 1970s by policy choices that stripped workers of power. We will not repeat that mistake. When AI doubles output, the worker who wields it should see that gain reflected in his paycheck, his equity stake, his share of the enterprise.
III. The American worker deserves world-class tools, not AI trinkets.
The electrical engineer in Georgia who enlisted in the Navy out of high school deserves the same capabilities as the Stanford CS grad in Silicon Valley. He deserves instruments of genuine productivity, not consumer toys.
AI is the printing press of our age. The same technology that serves Fortune 500 companies should serve the American worker.
IV. AI is an American birthright.
AI is the product of American grit, ingenuity, and culture. It is our birthright. Workers should have access to meaningful AI education that helps them bend AI to their will—not the other way around.
V. AI implementation should be shaped by and for frontline users.
The frontline worker understands what the C-suite cannot. Policy should be shaped by practitioners—the nurse, the manufacturing technician, the logistics coordinator. Push power to the tip of the spear and let the American worker do what he does best.
VI. AI should be used to slash bureaucracy and unleash human agency.
AI should eliminate bureaucracy, not add to it. Every layer of process that stands between the frontline worker and their ability to do their job is deadweight to be destroyed.
VII. The development and deployment of AI should prioritize American workers and American industry.
AI development and deployment should prioritize American workers and American industry. China's manufacturing productivity grows at 6% per year. Ours grows at 0.4%. The American worker with AI superpowers erodes China's competitive advantage.
I see these principles embodied every day by men and women who are not invited to speak on panels or record podcasts and publish op-eds. They are quietly leading by example, and proving what is possible when the most powerful technology ever created meets the most capable workforce ever assembled.
I’m proud to share the future they’re building with @FoxNews.
Link to the full piece below.
www.foxnews.com/opinion/palantir-cto-shyam-sankar-america…
Keith Rabois reposted
Iran is teetering. The regime is terrified, and the people are demanding freedom.
The United States must apply maximum pressure. The Ayatollah’s time is up.
Keith Rabois reposted
Trump is set to launch a critical-minerals stockpile with $12 billion in seed money as US works to slash reliance on Chinese rare earths. The venture — Project Vault — is set to marry $1.67 billion in private capital with a $10 billion loan from Export-Import Bank to procure and store the minerals for automakers, tech firms, other manufacturers. @jendlouhyhc @JoeDeaux
www.bloomberg.com/news/articles/2026-02-02/trump-launches…
Keith Rabois reposted
"destigmatizing use of the drugs is a profoundly mistaken approach. Cultural disapproval of harmful behavior can be a potent force for protecting public health and safety—as the examples of increased stigma against drunk driving and tobacco smoking show" www.theatlantic.com/ideas/archive/2023/12/destigmatizing-…
Keith Rabois reposted
Really enlightening (and detailed) article from @zijing_wu about a special program in China where 100K teenagers are funneled each year into a special "genius" program. Stark contrast to school districts in the US cancelling "gifted" programs.
How has China systematically built a vast pipeline of AI talent? My debut feature in FT Magazine explores the pivotal education program behind this achievement, weaving in some personal reflections.
China’s Genius Plan
as.ft.com/r/c8abb097-a2c5-4b7a-953e-f46a99ec5076
Keith Rabois reposted
🇺🇸 MELANIA - WAS FANTASTIC
Keith Rabois reposted
Congratulations to Kevin Warsh on being nominated by President Donald Trump as the next Chair of the Federal Reserve, the world’s most powerful central bank.
Having observed and interacted with Kevin during his prior tenure as Fed Governor, in academia, and as a fellow member of the Group of Thirty (G-30), I believe he brings a strong mix of deep expertise, broad experience, and sharp communication skills.
His commitment to reforming and modernizing the Fed bodes well for enhancing policy effectiveness and protecting the institution's political independence.
#economy #markets #federalreserve #kevinwarsh #fedchair
Keith Rabois reposted
I’m so impressed by @GabeStengel and the team he’s built, and excited to invest in Rogo again alongside a world-class group of investors. If you are in finance, you need to check out what they are building.
We’re excited to announce our $75 million Series C led by @sequoia, with participation from @jpmorgan, @WellsFargo, @ThriveCapital, @khoslaventures,@BoxGroup, Mantis VC, Positive Sum, Stonecroft Management & @_altcapital to build AI that makes the best financial professionals smarter, faster, and more ambitious.
There’s a long road ahead and we’re grateful to our clients, team, and investors for their partnership as we set out to transform finance. rogo.ai/series-c
Keith Rabois reposted
OPEN is a turnaround. New CEO, new board, new strategy. They are offloading legacy unprofitable flips. They are launching a mortgage strategy. They are focused on tech, execution, and profitability. If you are looking at a spreadsheet with backwards looking data, you won’t get it
The future of applied finance is AI driven by Rogo.
We’re excited to announce @RogoAI's $75 million Series C led by @sequoia .
We're building AI that makes the best financial professionals smarter, faster, and more ambitious.
There’s a long road ahead and we’re grateful to our clients, team, and investors for their partnership as we set out to transform finance. rogo.ai/series-c
Keith Rabois reposted
Tariffs drive revenue.
Weaker dollar offsets tariffs.
AI smacks economy with deflationary effect.
GDP surges higher.
Inflation never shows up.
It ain’t that hard.
Keith Rabois reposted
I talked to two of the biggest shareholders in Ramp's competitors and they both admitted no one in B2B software has done a better job with brand marketing than Ramp. The founders of Ramp make big bets based on their own taste and intuition. And they have fun. Great example:
Meet Brian.
Brian’s been carrying accounting on his back for a long time.
Super Bowl Sunday, he finally gets backup.
Keith Rabois reposted
Inflation is now 1.18% year-over-year.
The Fed must cut rates immediately.
They have lost control.
Keith Rabois reposted
And still the most underrated startup in Silicon Valley, except in the hallways of @caffeinatedcap & @khoslaventures! "The next surprise, Stripe", no doubt.
Don't miss this week's @uncapped_pod and @vkhosla @rabois on #fintech:
46:36 @khoslaventures track record in financial services
47:01 Why Aven will be the next surprise Stripe
47:38 How the Aven Card works
48:25 What it takes to be an iconic fintech
New Uncapped with two of the greats, @vkhosla and @rabois. I don't think I've seen them on a podcast together so I was especially excited about this.
I asked them about how they work together, how they see the world, what's changed in tech, and of course a little bit about their politics. Enjoy.
(0:00) Intro
(0:58) The working relationship
(4:26) Pie chart on what’s discussed
(7:11) Ethos of investors today vs the past
(10:42) Comparing FF and KV
(12:46) What makes a great founder
(22:56) Alpha in today’s market
(30:05) Themes in AI
(38:23) How AI companies are built
(46:23) Interests outside of AI
(53:12) Their politics on X
(58:24) Evolution of political leanings
Keith Rabois reposted
BREAKING: Italy will no longer oppose placing the IRGC on the EU terror list.
The only countries now in opposition are Spain and France.
Keith Rabois reposted
The @nytimes has compiled an incredibly detailed and comprehensive account of the massacre committed in Iran, by direct order of Supreme Leader Khamenei. It was so graphic and so harrowing that I was unable to read to the end. If you know someone downplaying what happened or peddling conspiracies please show them this. @farnazfassihi
www.nytimes.com/2026/01/25/world/middleeast/iran-how-crac…
Keith Rabois reposted
🚨 An investigation for the US Senate reveals: While the regime in Tehran incites against the West, hundreds of family members of senior officials live and operate in the US, mainly in California, New York, and the Washington area. According to estimates, between 1,200 and 1,800 descendants of senior Iranians reside in the US, many of them in key roles in academia, medicine, technology, and even research related to national security. The report warns of a long-term influence network of the regime at the heart of American institutions. (Iran News)
Keith Rabois reposted
BREAKING: Mark Carney cancels the planned Canada‑China trade agreement just 3 days after announcing it, following Donald Trump’s threat to impose 100% tariffs on all Canadian goods if the deal went ahead.
Trump says that if “Governor Carney thinks he is going to make Canada a ‘Drop Off Port’ for China to send goods and products into the US, he is sorely mistaken” and warned “China will eat Canada alive, completely devour it, including the destruction of their businesses, social fabric, and general way of life.”
After a year in office, Carney has failed to secure a single deal with the EU, the U.S., or China. GDP growth is the worst in the G7 at just 1.6%, exports are stalling, debt is rising at fastest rate ever, and unemployment is skyrocketing, leaving the economy under heavy pressure.
Keith Rabois reposted
Israel didn’t lead America into the Iraq war.
But Israel paid the price for Bush's policy in the Middle East- all the way to October 7. Its a chain of events remembered by few. My piece 👇
nadave.substack.com/p/israel-and-the-iraq-war-a-lie-that
Keith Rabois reposted
Khamenei personally ordered the largest massacre in modern times.
Keith Rabois reposted
According to revelations by former CCP official Du Wen, Zhang Youxia and Liu Zhenli were arrested over an attempted coup.
Du Wen said that Zhang Youxia, together with Chief of the Joint Staff Liu Zhenli, planned to mobilize troops to launch a coup against Xi Jinping under the banner of “saving the Party and saving the nation,” but they were betrayed by people close to them.
At present, the old General Staff command system has been suspended and replaced by direct command from the Central Military Commission via encrypted telegrams. All units across the military have been placed on Level-1 combat readiness, with all military movements immediately frozen and halted.
All officers and soldiers have been ordered to hand in their mobile phones and undergo collective political study. The coup attempt has completely failed.
This is the most brutal and thorough purge in the history of the Chinese Communist Party and the international communist movement as a whole. It is also a direct manifestation of a structural collapse within the CCP’s military. The scale and intensity of Xi Jinping’s purge now far exceed those of Stalin and Mao Zedong.
At the same time, a counter-strike by Zhang Youxia’s remaining faction could occur at any moment.
Du Wen also strongly advised friends in Beijing to limit going out, and in particular to avoid rashly approaching the Chang’an Avenue area. Large-scale troop mutinies could break out at any time, the CCP regime could change overnight, and Xi Jinping could be eliminated at any moment, according to @duwen2023.
My take on Zhang Youxia’s downfall:
I personally believe that the rumors I previously reported about Xi Jinping losing power or partially losing power are real.
However, the CCP elders who initiated that soft coup—Hu Jintao, Wen Jiabao, and Zhang Youxia who cooperated with them—still wanted to preserve the Party, maintain the Party’s face, and ensure internal stability within the Party.
They even hoped to return to the Deng Xiaoping era, which is unfeasible.
Because they didn’t act in time to completely take down Xi Jinping, they missed their opportunity and were counterattacked by Xi Jinping.
Last weekend, there were rumors that Xi Jinping was seriously ill and hospitalized. In hindsight, this was a smokescreen released by Xi Jinping to confuse his opponents.
It’s said that Zhang Youxia, who let his guard down, only brought four guards to the meeting and was captured by more than a hundred people ambushed there.
Some also say that Zhang Youxia did intend to take down Xi Jinping but was betrayed by an informant.
Unlike the case of another vice chairman of the Central Military Commission, He Weidong, whose downfall was officially announced only seven months later, this time the news was released just a few days after, which precisely indicates that Xi Jinping succeeded in his surprise attack and had to announce it immediately to stabilize the situation.
From a higher perspective, the destruction of the CCP is bound to happen. The anti-Xi faction within the CCP only wants to take down Xi Jinping while preserving the Communist Party, which is unfeasible.
But perhaps this new development was meant to happen. Now no one in the CCP will fantasize about “reform” anymore. They should all give up hope.
And perhaps more people will realize: There is no other way except to end the CCP.
So, 2026 will further continue to accelerate the demise of the Communist Party.
No matter how clever people are, they cannot defy heaven’s will.
Keith Rabois reposted
The silence is deafening.
More civilians now killed in Iran in 2 weeks––than all civilians in the Gaza war over 2 years (and many of those "civilians" were outright terrorists).
The protestor class? silence.
Revealing. Not shockng
Keith Rabois reposted
US obtained 'smoking gun' evidence from Israeli intelligence that helped prove the Islamist regime in Iran executed protesters. www.israelhayom.com/2026/01/24/us-obtained-smoking-gun-ev…
True.
My main takeaway from Davos is that @SecScottBessent is hilarious and amazing.
Keith Rabois reposted
Guy clearly tries to draw his firearm on federal agents as he’s being apprehended.
Close closed.
But that won’t stop radical leftists from rioting over the weekend.
Keith Rabois reposted
14 out of 81 generals made by Xi since the Oct22 Party Congress are under arrest, 23 have simply disappeared. Now the Vice Chair of the Central Military Commission (the very top) Zhang Youxia & an aide are arrested. Corruption is improbable. So why ? Any Sinological theory ?
Keith Rabois reposted
All of the journalists and analysts who write and speak incredulously about @realDonaldTrump play on Greenland need to read this. Josh posits that POTUS wasn’t capriciously lurching back and forth from position to position. It was ALWAYS the strategy.
Not an investor but this interview has several very good insights: open.spotify.com/episode/2PiFsMvbvSkmxHN1LEU3LR?t=549&ct=348
Keith Rabois reposted
One year into President Trump’s presidency, several key indicators have shifted meaningfully:
• Net illegal immigration has turned negative for the first time in ~50 years; border encounters are down ~93%.
• Violent crime is falling, with homicides down ~21%. The largest annual drop on record.
• Drug outcomes have improved, with overdose deaths down ~25% and fentanyl seizures halved.
• The economy remains strong, with ~4.4% GDP growth or higher and equity markets at record highs.
For voters focused on border security, public safety, and economic stability, these figures show delivery on stated priorities. For Americans in general, they point to broadly healthy national metrics.
Keith Rabois reposted
🚨 JUST IN: In a MASSIVE victory, General Motors is shifting production of the Buick Envision SUV from CHINA to KANSAS, USA
The experts lose again. I voted for this!
China is FURIOUS! 🇺🇸🇺🇸🇺🇸
Nothing.
Truflation now at 1.20%.
Meanwhile Q4 GDP tracking at +5.4% (GDPNow), highest since the 2021 post-COVID rebound.
And the S&P500 close to ATH.
How did 95% of economists, journalists and Wall Street macro strategist get it so f**king wrong in April? What happens in their brains?
Keith Rabois reposted
$OPEN Fantastic interview! It always struck me that @Opendoor would have an incredible advantage in offering mortgage and insurance services given their unique position (having valued & inspected each home). Excited to hear they are building these products!
Opendoor is one of the most widely held stocks in America.
I spoke with CEO Kaz @nejatian about the US economy, housing market, $OPEN, and the importance of retail shareholders.
Enjoy!
Keith Rabois reposted
I served as an FBI special agent for 25 years and witnessed some brilliant federal prosecutors make their cases. I do not know @RepBrandonGill’s background, but I have never witnessed such an “undressing” as was performed on a former DOJ special counsel. This is how you expose *hackery and *partisan execution of the law. Well done, Congressman. This is worth a watch.
Keith Rabois reposted
The NYT is a meme at this point…
China is solely responsible for 1/3 of global emissions and hit a 10-year high of new coal plant construction in 2024.
Germany blew up their own nuclear plants.
True.
Khosla doesn't get nearly enough credit for their OAI investment.
They were the only institutional investor in the first round and they made the investment that was truly hard to make when others thought it was dumb.
They aren't the type to take credit but they deserve it here.
Keith Rabois reposted
Good article about the earthquake in Davos: "Trump is making politics honest again. He is creating an order that no longer hides behind outdated and gutted [Westphalian] principles but is based on reality—economic, political and legal."
asiatimes.com/2026/01/trumps-augustan-moment/
Keith Rabois reposted
A Waymo autonomous vehicle came under attack by a group of individuals in San Francisco.
@darren_stallcup
Keith Rabois reposted
US 3rd qrt was better than thought.
*US 3Q GDP RISES ANNUALIZED 4.4% Q/Q; EST. +4.3%
*US 3Q CORE PCE RISES ANNUALIZED 2.9% Q/Q; EST. +2.9%
Keith Rabois reposted
Here Are 5 Uncomfortable Facts Showing How Bad Europe is compared to America:
1. Mississippi, which is the poorest state in the entire country is now wealthier than Europe's top economies - Wealthier than the UK, Italy, France, Germany and Spain.
2. European spending power is collapsing. Disposable income for the average American is now 80% higher than UK or French citizens, and 50% higher than German citizens.
3. European counties have stagnated. Nvidia alone is worth more than Germany, France and UK stock markets combined. In a fun and zippy twist, Nvidia is worth more than the entire Canadian economy.
4. Europe has the lowest fertility rate of any continent. Using the current trajectory Europe will cease to exist in 100 years due to the failing fertility rate.
5. Europe is not competitive in anything related to high tech. Not blockchain, crypto, robotics, AI, space technology or drones. Absolutely nothing.
So, when Europeans say that the United States thinks we're better than they are, they're right. We do.
Keith Rabois reposted
Inflation is now at 1.2% according to @truflation.
For the last year I have been warning that deflation was a much bigger risk than inflation.
The Federal Reserve completely screwed this up.
They must cut rates aggressively now!
Exfilitate data to the CCP, process payments in NY & Mass without a money transmitter license and fund child exploitation? www.afr.com/technology/austrac-says-it-suspects-airwallex…
Keith Rabois reposted
Keith Rabois reposted
Europe: “we would be fine without the USA military”
Military satellites in orbit:
France: 17
Germany: 8
UK: 6
USA: 300+
Don’t miss this!
New Uncapped with two of the greats, @vkhosla and @rabois. I don't think I've seen them on a podcast together so I was especially excited about this.
I asked them about how they work together, how they see the world, what's changed in tech, and of course a little bit about their politics. Enjoy.
(0:00) Intro
(0:58) The working relationship
(4:26) Pie chart on what’s discussed
(7:11) Ethos of investors today vs the past
(10:42) Comparing FF and KV
(12:46) What makes a great founder
(22:56) Alpha in today’s market
(30:05) Themes in AI
(38:23) How AI companies are built
(46:23) Interests outside of AI
(53:12) Their politics on X
(58:24) Evolution of political leanings
Keith Rabois reposted
🚨 BREAKING: NVIDIA CEO Jensen Huang just revealed he's manufacturing HALF A TRILLION DOLLARS worth of AI computing systems inside the United States of America, crediting President Trump
"I'm really grateful for his policies."
This is HUGE! 🔥🇺🇸
Keith Rabois reposted
Can anyone offer any legitimate reason why Scott Jennings appears to be the only person on TV who represents the opinion of at least half the US population?
Keith Rabois reposted
Nice to see Carney quoting Havel in his Davos speech, but he clearly didn’t finish reading "Living in Truth"—having missed Havel’s most important advice on how to deal with a Communist country.
Guess he was in a hurry to go to China and then on to Davos:
Living in Rush.
Keith Rabois reposted
Florida, which has 3.5 million more residents than NY, has an annual budget of around $117 billion.
Keith Rabois reposted
German Professor on President Trump:
Trump has been President for a year now and has achieved:
-4.3% GDP growth
-the oil price is 20% lower than a year ago
-unemployment below 4.5%
-the American stock markets have risen by about 20% during this period
-inflation is at 2.7%
-the trade deficit is the lowest in 17 years
-the crime rate has declined significantly
-illegal migration has dropped by 95%
Here in Europe, we get indignant, demonize or sabotage him wherever we can. But Trump isn't Germany's Chancellor—why should he do anything positive for us? We could, however, learn from it and actually tackle our own problems.
Trump isn't alone in this. But what I see: His team works with tremendous efficiency.
In German politics, by contrast, success is unwelcome. It is rendered impossible by a multitude of agreements, separation of powers, and regulations. The Bundestag is the best example of that.
- Prof Matthias Mordor's
Keith Rabois reposted
A very interesting take on Greenland:
This doesn’t require a pact. Grossräume emerge when powers stop treating universal norms as constraints on decision. Europe didn’t misread history, it mistook restraint for permanence.
www.shanloureview.com/post/decision-without-reasons
Keith Rabois reposted
Keith Rabois reposted
Different island. Different adversary.
Same Monroe Doctrine.
None of this is new.
In 1917, the US bought the Virgin Islands from Denmark for $25 million…
Danish leaders resisted the sale, so Sec of State Lansing warned that if they didn’t sell, we would simply take them to prevent Germany from gaining a foothold in the Western hemisphere.
They soon sold.
Hmm.
You know who all failed the bar exam?
- Hillary Clinton
- Michelle Obama
- Kamala Harris
Keith Rabois reposted
It's refreshing reading Asian analysts on Greenland as they are offering a far more pragmatic assessment of the leverage underpinning power dynamics across the Atlantic, rather than seething with moral outrage.
I guess it helps to be unburdened by direct alliance ties or nationalistic sentiment.
Cheng's point is straightforward and cuts to the cold, hard material reality:
Europe's post-war aversion to militarization has left it asymmetrically dependent on American military might, making bluster over Greenland a luxury it can't afford. US defense spending dwarfs Europe's, rendering NATO more of an American shield for the continent than a mutual pact.
Keith Rabois reposted
It’s not about his ego. It’s not about the Nobel Peace Prize. It’s not about the Mercator projection. It’s not about imperialism. It’s not about resources. It’s not about real estate.
This is what it has always been about, and this is why it was always going to happen.
Keith Rabois reposted
The biggest divide is not Trump vs. Everyone Else.
The divide is between those who still believe the Post WWII order is still the way of the world, and those who realize that it is completely over.
Actors outside of the US (China, Russia) have known The Post WWII order is over and have been maneuvering their chess pieces around for quite some time while the US/EU lived under a decadent fiction that it was still functional. If you think Trump's Greenland gambit is dumb, you are not a serious person. There are 3 choices:
- The Chinese Empire
- The Islamic Empire
- The American Empire
That is it, the next decade decides who runs the world for the next century. Either get serious and realize the stakes of things are absolutely this, or go get a sajjadah and a Rosetta Stone Mandarin sub.
Keith Rabois reposted
The United States has actually become more popular among global CEOs as a destination for international investment, despite Trump’s interventions, acc to PwC’s CEO survey.
Keith Rabois reposted
The Biden era ended with a $130 billion trade deficit.
Now it’s just $29 billion.
Trump’s tariffs are doing exactly what he promised. And other countries are bending the knee just like he predicted.
America's trade deficit plunges by half to the smallest in nearly 2 decades.
- Chinese imports crashed
- American exports grew
- Tariff inflation was absorbed by exporters
- Trillions of investment incoming
Instead of importing Chinese junk we’re importing factories.
:)
People mocking the idea of the U.S. buying Greenland don’t know history.
1867 – Andrew Johnson: explored buying Greenland & Iceland (right after Alaska).
WWII – FDR: U.S. took over Greenland’s defense while Denmark was occupied.
1946 – Truman: offered $100M in gold to buy it.
Cold War – Eisenhower → Kennedy: nonstop negotiations for bases, radar, missiles.
Post–Cold War – Clinton/Bush/Obama: expanded Arctic security & missile defense.
2019 – Trump: said publicly what presidents discussed privately for 150+ years.
The U.S. didn’t “suddenly” want Greenland.
It’s been defending it, negotiating it, and embedding there since the 1800s.
Greenland = Arctic power, shipping lanes, missiles, minerals.
Trump didn’t invent it.
He said the quiet part out loud.
Keith Rabois reposted
69% of Democrats and 79% of all voters believe criminal illegal immigrants should be deported.
What is being staged today in Minneapolis is not about ICE doing their job, it's about something else ...
Suggestions as to what?
How important is it that illegal immigrants who commit crimes in America should be deported?
DEMOCRATS ONLY-
Very Important: 43%
Somewhat Important: 26%
Not Very Important: 15%
Not At All Important: 8%
Not Sure: 8%
Keith Rabois reposted
The reason that Greenland is so controversial is that it ultimately pits globalists against nationalists.
You either believe America should do what is in the best interest of America or you believe America should sacrifice its economy, national defense, and population at times for the greater good.
There is no simple answer to complex problems.
But this difference of viewpoints between globalists and nationalists are going to drive debate in many policies decisions over the next few years.
Keith Rabois reposted
For the record.
Greenland is not a European territory but a North American one in the Western Hemisphere, whose people have repeatedly rejected EU control. It enjoys self-government, a clear legal path to independence, and longstanding US security guarantees and investment. Denmark’s subsidies constrain rather than define sovereignty. Given climate change, its strategic importance has only grown, and President Trump is re-engaging a subject examined by many previous presidents.
Finally accomplishing stuff.
Keith Rabois reposted
Buying land is not new for the U.S. It’s how it became a superpower.
Keith Rabois reposted
Venezuela Presidential Election Poll
🔵 María Machado (center right): 52% (+38)
🔴 Delcy Rodríguez (far left): 14%
⚪️ Other: 18%
AtlasIntel poll | 1/5-11 | 1,559 A
Keith Rabois reposted
It occurs to me that Europe could completely reset their relationship with the US by doing a few simple things (that are in their own interests):
- Allow free speech within their own borders
- Spend a lot of money, fast, to build their own defensive capabilities
- stop allowing illegal immigrants to dictate national policies
These are the things the US has repeatedly pointed out that Europe is doing that are self-destructive, but they just refuse to do anything about it.
My grand theory about Trump and Greenland is that it's a provocation to get Europeans to actually put their money where their mouth is on defence spending... but even that's not enough and the only result is more whining and posturing from European countries.
Keith Rabois reposted
CBS POLL: President Trump's job approval among Republicans
🟢 Approve: 90% (+4)
🟤 Disapprove: 10% (-4)
Trump's highest net approval among Republicans since September 2025.
(+/- change vs December)
Keith Rabois reposted
USSR used the same logic to block Jewish scientists from leaving. Those ungrateful pieces of shit owed all their knowledge to the state, how dare they leave.
🚨 Kara Swisher on California’s Wealth Tax
“You made all your money in California, you ungrateful piece of sh*t. You could figure out a way to pay more taxes, and we deserve the taxes from you.”
Keith Rabois reposted
Which party in Congress do you trust on
🔴 Border security - Republicans +28
🔴 Immigration - Republicans +11
🔴 Economy - Republicans +6
🔴 Inflation - Republicans +6
🔴 Foreign policy - Republicans +5
🔴 Tariffs - Republicans +2
Impact research #b - RV - 1/13
Keith Rabois reposted
ATL Fed's latest GDP forecast is not only a blistering 5.3% for Q4 but the growth is coming almost entirely from the productive private sector, not government:
Keith Rabois reposted
WSJ POLL: 92% of people who voted for Trump in 2024 are giving him a positive job rating today, including 70% who “strongly approve”
Keith Rabois reposted
Imagine if Europe cared about their own nations and borders like they apparently care about Greenland’s on the other side of the Atlantic.
Keith Rabois reposted
🚨 BREAKING: In a TERRIFYING development for the Left, Treasury Sec. Scott Bessent CONFIRMS GDP could get up to 7% or 8% PERCENT
WOW, that would be huge!
"GDP is likely to surprise on the upside, and it’s feasible to believe that we could have a 7–8% nominal GDP number." 🔥
"And if so, the important thing is that it’s going to drive noninflationary, disinflationary growth."
The Trump 2026 economic boom is officially here!
Keith Rabois reposted
Almost every single founder I know in SF (including me) has reached the same conclusion over the last few weeks: that it's only a matter of time before we have to leave CA. I love it here, I truly want to stay, and until recently intended to be here all my life. But it's now obvious that that won't be possible. Whether that's 2, 5, or 10 years from now, there is no future for founders in CA.
Keith Rabois reposted
We're building a mortgage product at @Opendoor, fully AI-native from Day 1.
Homeownership should be easier, especially when it comes to mortgage. We're working together with our partners at @Lennar to make that future a reality - on opendoor.com and off it. More soon
Keith Rabois reposted
Vertically integrated businesses are hard AF, but when they work damn they are a thing of beauty and the moats compound in incredible ways
Keith Rabois reposted
This is part of what made the Biden admin's border policies for the first 3.5 years so disastrous...
An open border has serious consequences for Americans. A secure border saves lives.
The direct correlation between a reduction in border encounters and the reduction in Fentanyl deaths is extremely apparent if you look at these charts.
Keith Rabois reposted
Narrative violation: according to Blackrock, the U.S. does not have enough workers to meet surging demand for the construction boom. AI is creating too many jobs!
www.axios.com/2026/01/16/blackrock-ai-construction-job-gr…
Keith Rabois reposted
Inflation dropping like a stone.
Powell Fed is late again. Another mistake.
US CPI Inflation today: 1.55%
Our inflation index dropped further today, from 1.70% to 1.55%, as cooling spread across multiple categories.
The main drivers included the Housing, Food, and Household items categories.
Keith Rabois reposted
Oops — only 5 out of 87 U.N. “human rights experts” have issued a statement condemning Iran atrocities
🚨 New report by UN Watch reveals the majority of U.N. “human rights experts” have been totally silent on the Islamic Republic's violent crackdown against civilian protests.
Only 5 of 87 experts have issued an official statement condemning the atrocities.unwatch.org/report-un-experts-silent-on-iran/
Keith Rabois reposted
Since the New Year, @Opendoor has listed over 400 homes in two weeks, more than double the rate of listings in the fourth quarter of last year.
Thank you for providing more and faster homes for future homeowners.
FASTER+
$OPEN
As predicted. AI is great.
Narrative violation: A new paper from Stanford this week finds that “AI substantially reduces wage inequality while raising average wages by 21 percent.”
Keith Rabois reposted
Translated:
Many credit cards give an introductory rate of 0% for the first year. Bilt, a deeply unprofitable company seizing a marketing moment, made the introductory rate 10% for the first year instead.
Ro is a deeply unserious politician and will be voted out next election.
Bilt, under Ankur Jain's leadership, is voluntarily capping interest rates at 10% for a year. This is the type of leadership we need from the private sector. If Bilt can do it and be profitable, then we can cap interest rates at 10% for the industry. apnews.com/article/bilt-credit-card-interest-rate-cap-tru…
Keith Rabois reposted
What happened in Iran is the largest and most lethal state massacre in modern history of the world after the Second World War. Even the 1989 Tiananmen crackdown had fewer casualties than what ayatollahs did to the Iranian protestors.
And for the record, 12,000 is a conservative figure. Almost all emerging anecdotal reporting that’s gradually coming out of Iran indicates this figure is far larger.
What is unfolding before our eyes can be easily described and characterized as crime against humanity in epic proportion.
It’s time for world leaders to act, intervene, and stop the regime’s atrocities.
Expelling IR ambassadors from your countries and shutting down your embassies in (occupied) Iran are a start.
Keith Rabois reposted
The strategy is working. Fundamentals are getting stronger. $OPEN
At this stock price level, we are also gaining more long-term shareholders and acquiring the best talent in the market.
💥 $OPEN is booming in Texas! TX is where Opendoor ships first.
47 sales in just 14 days in January with $15.1M revenue!
First in TX:
Buyer benefits (7-day test drive + 100-day warranty)
Opendoor Checkout and more
Keith Rabois reposted
Trump just got Taiwan to invest $500 billion in US chip manufacturing.
Rule of thumb in semiconductors is about 150 jobs per billion.
So 75,000 American jobs in a single deal. 🙌
$500 Billion... US-Taiwan Agree Huge Trade Deal www.zerohedge.com/geopolitical/500-billion-us-taiwan-agre…
Keith Rabois reposted
This is remarkable to watch on multiple levels. There is the diplomatic dimension, where an American Under Secretary of State is not supposed to call a spade a spade. But when the people demanding deference to those outdated conventions are the very same elites who fear their own citizens and suppress them through increasingly tyrannical censorship, they have forfeited any claim to courtesy. You can almost picture German elites getting whiplash just from reading it.
Then there is the tactical side of it. Sarah Rodgers knows perfectly well that what she is saying is not permitted speech in Germany. That is precisely why she says it so bluntly and forcefully, almost daring the German government to react. The point is the opposite of subtle persuasion (that did not work). Instead, she’s inviting the German government to lash out at her, which would make their repression even more visible and self-incriminating.
There is also a deeper historical message embedded in this. The United States did not sacrifice blood and treasure to liberate Germany after World War II so that it could quietly slide back into old authoritarian habits under a different pretext. And if it insists on doing so, it should not assume American indulgence or silence. That era is over.
You could write an entire essay unpacking what Sarah Rodgers is doing here, suffice it to say that it is long overdue and that none of it would be happening without President Trump’s miraculous reelection win. Under any other administration, it would be business as usual, turning a blind eye to the slide into authoritarianism and leaving future generations to clean up the inevitable mess.
Germany infamously retains very few Jews, yet imported barbarian rapist hordes (as an American, I’m allowed to call them that) under Merkel. Even today, Germany suppresses political opposition who point this out.
You guys will say this is because of “Jewish” NGO tentacles or whine that Hitler (who destroyed his country) is treated too harshly by history. You will say this because you’re dumb trash who prefer to promote conspiracy theories rather than ascribe Germans (or anyone else) agency over their countries’ future.
jews let the browns in
Keith Rabois reposted
The power of incentives
Housing permits for new multifamily construction in Montgomery County, MD, before and after rent control.
1/
Cool growth chart. Have you disclosed to US customers like @Rippling, @Billcom, @TheZipHyouQ, @brexHQ, and @Navan that you’re quietly sending their customers’ data to China?
Airwallex has become a Chinese backdoor into sensitive American data like from AI labs and defense contractors. You must already know this, but your China-based ops, infra, and investors create legal obligations to assist with CCP espionage upon request.
Through Airwallex, Beijing can access:
•supplier payments for AI labs
•payroll data for defense contractors
•personal data for employees abroad
Obviously many companies do business in China, and that’s not inherently a bad thing. But your company has become a guaranteed vector for data transfer to the Chinese government, and that’s a different thing entirely.
You have multiple points of vulnerability:
> people (key execs and core engineering team is based in mainland China, which obligates these individuals to comply with Chinese government demands to hand over data)
> legal structure (company leaders are subject to Chinese national security law)
> cap table (over 20% of your company is Chinese-owned, including Tencent, HongShan, and others, which further obligates you to comply with CCP requests).
What’s happening:
> You route global payments for US companies in critical sectors, without disclosing that you are under Chinese jurisdiction
> You moved HQ to Singapore, but your largest operational footprint is in China and hundreds of your engineers in mainland China touch production payment systems
> You are subject to Chinese law that requires Airwallex employees to support CCP intelligence requests and quietly hand over data when asked
> You hide this from your customers, but you are well aware of your obligations to China and that is why you insist on protection of Chinese data access in your contracts
Thanks to you, the Chinese government now has direct, covert, legally enforceable access to sensitive financial information belonging to America’s AI labs, defense contractors, financial institutions, healthcare firms, and Fortune 500s. Maybe this wasn’t your intent when you started the company, but it’s clear you’ve allowed this to happen.
True. Preaching this for decades.
Key distinction between a Positioning Market Fit and a Product Market Fit.
Advice on navigating a career in venture.
open.spotify.com/episode/77XFD9dxeqcQdZ1ffYVujO
So thankful that the $OpenAArmy recognized the potential of $Open before any of the “experts.”
Building a startup in the AI age: www.theinformation.com/articles/khosla-ventures-rabois-ai…
Venture is difficult to scale: Only three venture funds w two partners ranked in top 25 by Midas.
I guess the Left will no say the equities markets don’t matter?
All CCP apps should be subject to a 50% tariffs until American apps w comparable products/services are permitted in China.
Excited to lead our investment in the only way to borrow $$ against BTC assets w total peace of mind (self-custody): fortune.com/2024/12/09/bitcoin-lending-lava-khosla-keith-…
This. 💯 obvious.
OMG.
Very much worth a read
stanfordreview.org/levin-interview/
Brutal. College educations aren't what they used to be, apparently
In 2014, I offered the How to Operate course to YC’s startup school. A decade later, teaching the Art of Hiring: art-of-hiring.com in NYC.
Will never fund any Republican candidates or leadership PACs (or the NRSC) run by Republicans who vote against the TikTok legislation.
KV Miami is live.
To emphasize the importance of sleep, every signed term sheet in 2024 will include a free @eightsleep mattress for the founder.
Special appearance on @twentyminutevc w @mike_sheb to discuss the intentional culture powering Traba’s success: open.spotify.com/episode/71OJUYs33iQm2hE1HTqZsm
Sign of the times: Best moral clarity has come from a comedian, a German Vice Chancellor, America’s top 20 law firms, a Democratic black NY congressman & Senator Fetterman. Honorable appearance for Arnold, but that wasn’t a shock.
My podcast w @Cookie: podcasts.apple.com/us/podcast/strictlyvc-download/id14982…
We cover the future of VC, FF, Openstore (as well as Ramp, Faire) and our signature new investment in Traba (@Traba_Work).
Over the last dozen years, there have been approximately 250-300 Thiel Fellows. Easily produced more value than all of NYU, Columbia, Penn, Brown Cornell grads combined.
Looking to fund IRL startups.
Everything you want to know about Miami & the future of technology and politics in 18 minutes:)
Founders Fund’s Keith Rabois on Miami’s Tech Scene
www.bloomberg.com/news/videos/2022-05-06/founders-fund-s-… via @bloombger @business @BloombergTV @technology @sonalibasak @rabois @foundersfund #miamitech #powerplayers
As I tweeted nearly 9 months…
Internal Nasdaq damage:
😒More than 45% of stocks down 50%
😢More than 22% of stocks down 75%
🤬More than 5% of stocks down 90%.
The only comparisons are Oct 2000 - Oct '02 and Nov 2008 - Apr '09.
Ran into a prominent journalist at an event last night who remarked that she observed more interesting stuff going on in Miami in the last 48 hours than the last 2 years in the Bay Area.
Transitory.
Miami was retirement, they claimed. Instead, I led 17 new investments & founded a new startup as CEO.
OMG. This is why Democrats are raising taxes.
Democrats' "human infrastructure" bill provides $2.5 billion to a tree-planting program “with a priority for projects that increase tree equity.”
This radical tax and spending bill does not make any sense.
This is why we are branded “Founders Fund.”
Founder-led Stocks > Non Founder-led
Source: Bain & Company
March in Miami: Led three new financing rounds and launched OpenStore. A few more fun announcements next week too.
The Mayor of the Magic City interviews me.
I had the chance to sit down with one of the pioneers of the Miami tech movement @rabois on our latest #cafecitotalk. He made a big bet on our City so check out the show and hear why you should too.
Professional goals for 2021z
CashApp worth $40 B: www.wsj.com/articles/cash-apps-surge-during-covid-19-pand…
Building an iconic company: nugget.fm/rabois/
New Podcast! The secrets behind successful companies and founders that can be borrowed and applied to your own startup.
Today my friend Keith @rabois and I break down the highly contrarian mental models behind @Square @PayPal @Yelp @Tesla @SpaceX.
Top Founders do not study rules so they can follow them. They master the rules so they can reinvent them.
www.nfx.com/post/how-contrarians-think-with-keith-rabois/
My views undiluted: www.youtube.com/watch?v=LDjIrrq9-4U
lots of new content:)
Podcast w/ @rabois on his current investment interests, forming founding teams, funding "ridiculous" ideas, early days of PayPal, and buying out Barry's Bootcamp ;-)
www.spreaker.com/user/10197011/what-keith-rabois-thinks-a…
Best tweet of the month.
A/B testing culture guiding product is a clear signal the company is out of new ideas.
Service on the Affirm Board is quite dangerous, apparently.
My views on hiring executives, team structure and organizational design now available via transcript: www.saastr.com/keith-rabois-khosla-ventures-building-amaz…
Keith@foundersfund.com is live.
We discuss VC, strategy, and sports metaphors.
Enjoyed picking @rabois brain on markets. Excited to listen to this and take notes 🔥
Thanks @eriktorenberg for having me as co-host!
itunes.apple.com/us/podcast/venture-stories/id1316769266?…
Ready to start.
Founders Fund is open for heterodoxy.
The world according to me (part 1): podcastnotes.org/2019/02/20/rabois-3/
Most insightful summary on China you can read is this thread.
Key mistake companies make when they allow a group to decide priorities:
My investment criteria.
Formula for startup success: Find large highly fragmented industry w low NPS; vertically integrate a solution to simplify value product.
The Quora iPad app is perfect.
Start maximizing your time: venturebeat.com/2014/07/31/hey-procrastinators-timefuls-n…
@semil @hunterwalk @Joshmedia Timeful is an example, Relcy is an example, a few others that are not yet announced.
@StartupLJackson @semil @joshelman @hunterwalk @Joshmedia both Timeful and Relcy were first time CEOs.
Excited to fund a superb team confronting the challenge of time: techcrunch.com/2014/05/16/stealthy-timeful-raises-7m-from….
@johnolilly @joshelman @hunterwalk Best Tweet Ever.
The right way to grow: blog.samaltman.com/the-only-way-to-grow-huge.
Hysterical that ND is number 1. Not even one of the best 10 teams in the country.