New: ICE is paying AT&T millions to dig through a mountain of raw data. Obtained internal ICE document on it. Experts say likely impacts trillions of records, is a mass surveillance program thought to be dead. Now it's direct to ICE www.404media.co/ice-is-paying-at-t-to-dig-through-a-mount…
Zeynep Tufekci
Zeynep Tufekci (Turkish: Zeynep Tüfekçi; [zejˈnep tyfec'tʃi]; zay-NEP tuu-FEK-chee) is a Turkish-American sociologist, and the Henry G. Bryant Professor of Sociology and Public Affairs at Princeton University. She is also a columnist for The New York Times. Her work focuses on social media, media ethics, the social implications of new technologies, such as artificial intelligence and big data, as well as societal challenges such as the COVID-19 pandemic using complex and systems-based thinking. According to The Chronicle of Higher Education, she is one of the most prominent academic voices on social media and the new public sphere. In 2022, Tufekci was a Pulitzer Prize finalist for her "insightful, often prescient, columns on the pandemic and American culture", which the committee said "brought clarity to the shifting official guidance and compelled us towards greater compassion and informed response." More on Wikipedia
-
zeynep tufekci reposted
Joseph Cox
@josephfcox
-
zeynep tufekci reposted
Charles Ornstein
@charlesornstein
“Odd and unusual.” Top Trump administration officials intervened to help a longtime golf pro at Trump National Golf Club Hudson Valley secure a half-million-dollar payout from the IRS. A company connected to Trump’s media business also got a half-million.
www.propublica.org/article/irs-settlement-golf-pro-shelby… -
zeynep tufekci
@zeynep
Not what happened.
We couldn’t get good social science funded AND companies deliberately hid/withheld data AND used lack of longitudinal/causal evidence to sow doubt that political class went along with out of self-interest.
See Alondra Nelson on strategic opacity/agnotology.
Tyler Austin Harper
@Tyler_A_Harper
Part of the issue is that the chattering class is way too beholden to social science. A lot of people spent much of the last decade saying “there’s no evidence phones are the problem!” when we all have the lived experience of our phones zapping our attention and dumb-ifying us.
Jill Filipovic
@JillFilipovic
This is so so so obvious and anyone who pushed for distraction machines in school should be embarrassed.
www.nytimes.com/2026/11/17/upshot/school-phone-bans-study…
zeynep tufekci
@zeynep
What social science are they beholden to? It doesn’t exist. Many tried to get a proper longitudinal panel funded. Nope. Academics never got access to the data the companies hoarded.
Same is happening with AI, btw. WE LACK GOOD STUDIES. I recently fielded a survey. More soon.
zeynep tufekci
@zeynep
Diffusion speed meant many effects were ecologically hard to measure (few counterfactual/natural experiments) and some bad extreme ones too rare for probability methods (but affects big numbers: the denominator is large).
We never got funding or access, and they weaponized this. -
zeynep tufekci reposted
Michael S. Schmidt
@nytmike
NEW and BREAKING: The highest grossing law firm in the country, Kirkland & Ellis, has done 6,491 hours of free legal work (valued at $11.8 million) for the Trump admin, including for DoD, Commerce and on trade deals. Kirkland quietly told Cong Dems this in wake of election. www.nytimes.com/2026/11/19/us/politics/kirkland-ellis-law…
-
zeynep tufekci reposted
Nicholas Kristof
@NickKristof
Last year President Trump "disappeared" a group of immigrants to South Sudan, prompting a Supreme Court case. The deportees then vanished for more than a year. I found them, locked up, held incommunicado, and famished--and I hope the Justices take note. www.nytimes.com/2026/11/18/opinion/trump-deportation-sout…
-
zeynep tufekci reposted
jonathan
@jnsyaaa
at long last we are choosing who will wield the Ring from the famous fantasy novel The Ring Will Corrupt Anyone Who Wields It
Jeffrey Ladish
@JeffLadish
This is not the correct takeaway from Lord of the Rings
-
zeynep tufekci reposted
Dino A. Dai Zovi
@dinodaizovi
"Unfortunately, it is far too easy in security for whack-a-mole to become the goal. It’s great at the metrics and compliance game because you can make a graph of resolved incidents going up and to the right. Maybe even time-to-resolution going down over time! It also requires less organizational alignment than intervening in the roadmaps of other teams, making it easy to have the appearance of a security team that owns its own destiny. Taking ownership of actual security outcomes is scary, because attackers are unknown. But playing whack-a-mole is losing."
zeynep tufekci
@zeynep
A solid explanation of why we can’t just patch our way out of insecure software in the age of AI coding tools (the mundane ones!) from @dadrian
Scale, speed and friction are game-changers at the macro level even if the micro looks just like before.
dadrian.io/blog/posts/whack-a-mole-is-losing/
zeynep tufekci
@zeynep
PHP bug folks found a Java bug. GOOD! Now fixed. It’s a memory corruption bug 🥱in Java 😱 that’s coming from hand-rolled assembly code 🤯.
I had “hand-rolled” assembly code in ancient times🦖. All code is legacy code, layers of technical debt now hitting scaled/sped up tools. -
zeynep tufekci
@zeynep
A solid explanation of why we can’t just patch our way out of insecure software in the age of AI coding tools (the mundane ones!) from @dadrian
Scale, speed and friction are game-changers at the macro level even if the micro looks just like before.
dadrian.io/blog/posts/whack-a-mole-is-losing/
zeynep tufekci
@zeynep
PHP bug folks found a Java bug. GOOD! Now fixed. It’s a memory corruption bug 🥱in Java 😱 that’s coming from hand-rolled assembly code 🤯.
I had “hand-rolled” assembly code in ancient times🦖. All code is legacy code, layers of technical debt now hitting scaled/sped up tools. -
zeynep tufekci
@zeynep
Too many folks with no cybersecurity background obsessing over the (largely irrelevant tbh) Hugging Face incident because the chain-of-thought seemed understandable to them —unlike actual cybersecurity issues they don’t understand— added to the delay in addressing the real stuff.
Nicole Perlroth
@nicoleperlroth
I cringe at the comparisons between Metasploit and AI-powered pentesting frameworks like ARTEX. This is a fundamentally different moment.
We are putting an autonomous attack button in everyone’s hands— APTs will use this for plausible deniability, cybercriminals’ lives just got infinitely easier, and now we have to worry about entire categories of people that just have a sick fixation with watching the world burn. All of them now have the ability to compromise organizations at machine speed. Defenders are nowhere near ready.
The genie is out of the bottle. We are never getting it back in. Too much time was spend pondering existential bio weapons. Not nearly enough energy was expended contemplating how AI would democratize cyberattacks.
We will look back on this moment with profound regret, not because we didn’t sound the warnings, but because “leaders” failed to heed the warnings before it was too late.
FBI Dallas
@FBIDallas
The #fbi issued a FLASH advisory warning that cybercriminals and state-aligned hackers are exploiting ARTEX, a super intelligence-powered penetration testing framework, to automate attacks against businesses and organizations worldwide. ARTEX can accelerate intrusions from initial targeting to privileged access in under three hours. It was used in the October ConnectWise supply-chain compromise, which exposed Microsoft 365 customer environments and involved attempted authentication against more than 2,250 downstream tenants.
The FBI is urging businesses and organizations to strengthen access controls, monitor suspicious automated activity, and report potential compromises to ic3.gov.
Read more here: www.ic3.gov/CSA/2026/261113.pdfPost not available here, view on X
zeynep tufekci
@zeynep
The recent democratization of pen-test type tools altered speed, scale and friction (NOT superintelligence! Not scheming agents!)
Speed, scale and friction destabilize at the macro level even as the micro remains mundane. Add decades of technical debt, layoffs and DOGE, voila.
zeynep tufekci
@zeynep
PHP bug folks found a Java bug. GOOD! Now fixed. It’s a memory corruption bug 🥱in Java 😱 that’s coming from hand-rolled assembly code 🤯.
I had “hand-rolled” assembly code in ancient times🦖. All code is legacy code, layers of technical debt now hitting scaled/sped up tools. -
zeynep tufekci reposted
Trip Gabriel
@tripgabriel
NYT: America was once regarded as a global leader for reining in measles. Now, large outbreaks that sicken thousands could become routine
www.nytimes.com/2026/11/17/well/us-measles-elimination-st… -
zeynep tufekci reposted
Owen Winter
@OwenWntr
Despite Democrats winning some gerrymandered seats, it's still striking how stacked the House map is against them
TX - 53% of the vote -> 37% of seats
FL - 50% -> 29%
GA - 56% -> 36%
TN - 45% -> 11%
OK - 41% -> 0%
SC - 49% -> 29%
LA - 45% -> 17%
UT - 50% -> 25% -
zeynep tufekci reposted
Paul Knoepfler
@pknoepfler
Powerful story on how the Trump administration killed the Pediatric Brain Tumor Consortium www.nytimes.com/2026/11/15/well/brain-cancer-children-dip… which was so crucial for children with DMG/DIPG (rapidly fatal brain tumors) & their families. My lab studies DMG & more efforts are needed on it, not less
-
zeynep tufekci reposted
Benjamin Mazer
@BenMazer
The HPV and Hepatitis B vaccines have been preventing cancer for decades. But this is still good.
Daniel Kraft, MD
@daniel_kraft
Finally, a Vaccine for Cancer
www.nytimes.com/2026/11/16/podcasts/the-daily/vaccine-can… via @NYTimes -
zeynep tufekci reposted
Sadie Gurman
@sgurman
“We don't have to wait for them to commit a crime for us to begin investigating whether there's a basis to escalate or indict or prosecute," @USAttyReitz said. Dems "are champing at the bit with bloodlust to punish those who associate with the president." www.wsj.com/politics/policy/doj-is-ready-to-battle-congre…
-
zeynep tufekci reposted
Sheryl Gay Stolberg
@SherylNYT
Eliminating measles in the United States was a huge public health triumph, the culmination of decades of work. It's been wiped out. @teddyrosenbluth
www.nytimes.com/2026/11/15/well/us-measles-elimination.ht… -
zeynep tufekci reposted
Mike Nellis
@MikeNellis
Incredible
Will Steakin
@wsteaks
NEW: Mamdani will announce today that he personally brokered a deal with President Trump regarding SNAP and Bodegas/corner stores stocking rules—deal will not just be for NYC but nation-wide, source familiar tells me
Formal announcement at 2:30 PM -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
The flavoring chemicals used in vapes are the same ones used to flavor food. They were not tested for inhalation safety, only ingestion safety. In fact, workers in these food factories that mix these chemicals into food products get explicit warnings not to inhale flavoring chemicals.
What are the flavoring chemicals used in these illegal vapes from China? #maha (cc: @SecKennedy)
The Wall Street Journal
@WSJ
The White House ordered officials this summer to approve releasing millions of dollars worth of detained illegal Chinese vapes into the U.S. market after talks with a Trump political adviser and his son, people familiar with the matter said www.wsj.com/politics/policy/china-vapes-trump-administrat…
-
zeynep tufekci reposted
Max Boot 🇺🇦🇺🇸
@MaxBoot
“If every congressional map had been drawn fairly, with party-blind nonpartisan methods, Democrats might have come close to winning a veto-proof two-thirds majority in the House, with the party winning nearly 20 additional districts in Texas, Florida and Georgia alone.”
-
zeynep tufekci reposted
Jerome Adams
@JeromeAdamsMD
Most babies don’t receive a first measles vaccine until age 1. With measles endemic again, parents can no longer assume grocery stores, airports, churches, or even pediatricians’ offices are safe spaces. The community protection we took for granted is gone, and U.S. moms now must live in fear that their babies will be harmed by a completely preventable disease, while they do every day tasks. 😢
Axios
@axios
U.S. loses measles elimination status www.axios.com/2026/11/14/us-loses-measles-elimination-sta…
-
zeynep tufekci reposted
Jacob Boenzi
@BoenziJacob
This tracks. Using nonpartisan/neutral maps from Stephen Wolf (on DRA) and @Redistricter, Democrats would've won about 280 seats instead of the likely 260 they'll have.
Matt Grossmann
@mattgrossmann
Republicans still gained 4-7 seats as a result of midcycle redistricting & still hold a 20-27 seat advantage in the House compared to party-balanced maps
www.nytimes.com/2026/11/13/upshot/gerrymanders-midterm-el… -
zeynep tufekci reposted
Meghan Bobrowsky
@MeghanBobrowsky
SCOOP: Meta employees have begun testing a prediction market app under development called Takes.
Users can wager "cred," short for credibility, not real money. Yesterday, one of the questions employees were wagering on was whether Tim Cook will be the next Nvidia CEO. -
zeynep tufekci reposted
LenaSun
@bylenasun
The US has now lost its measles elimination status—a technical designation that confirms what clinicians are already seeing on the ground. Adults now make up 41 % of US cases. In Pa, at least 101 pregnant women have been infected. 2/3
-
zeynep tufekci reposted
Liz Essley Whyte
@l_e_whyte
The White House ordered federal officials late this summer to approve the release of tens of millions of dollars worth of detained illegal Chinese vapes into the U.S. market….
With @MHackman www.wsj.com/politics/policy/china-vapes-trump-administrat… -
zeynep tufekci reposted
The Washington Post
@washingtonpost
Two more measles deaths were reported Friday, in Wisconsin and Pennsylvania, bringing the U.S. toll this year to at least 10 as outbreaks continue to erupt in several states. www.washingtonpost.com/health/2026/11/13/us-reports-least…
-
zeynep tufekci reposted
Ashish K. Jha
@ashishkjha
Folks. We are playing with fire.
Bird flu is a very bad disease
Allowing it to spread unchecked among cows and increasingly people is a very bad idea
We need both @USDA and @CDCgov to take this seriously and act now
United Farm Workers
@UFWupdates
There is an active outbreak of avian flu (H5 influenza) in dairy herds in Washington State, particularly the Yakima Valley. A number of dairy workers have been infected, with at least one death. Speaking with workers, we believe the official case numbers are a vast undercount. 1/
-
zeynep tufekci reposted
Christina Jewett
@By_CJewett
More good work based on records we also plumbed for this detail: Mr. Kennedy's wife, Cheryl Hines, earned $210,000 from a MAHA-aligned group -->
www.nytimes.com/2026/09/23/us/politics/rfk-jr-book-advanc…
Neil Mehta
@byneilmehta
NEW by @l_e_whyte me:
- RFK Jr. sailed around Greece with businessman Gavin de Becker, a security consultant and anti-vaxx/ketamine-therapy advocate, on his luxury yacht
- The trip adds to the $250K+ in gifts from De Becker that we first reported earlier this year
- Democratic senators are opening a probe into the gifts based on our earlier reporting -
zeynep tufekci reposted
Sean T at RCP
@SeanTrende
Yes. (a) the maps will be in effect for two more cycles, most likely and (b) the Special Master/Supreme Court of NC map would have been subjected to wave conditions as well and would probably have produced a similar outcome.
Aseemru
@Aseemru
I think people are getting way ahead of themselves here. How is this dummymander if it takes a D+13 year to break it? Any other cycle in recent history Republicans would win more seats than their popular vote % would suggest.
Congresswoman Deborah Ross
@RepDeborahRoss
Congrats to @NCGOP on becoming the encyclopedia example of a dummymander. I hope this will motivate you to spend less time suppressing voters and more time delivering for them!
-
zeynep tufekci
@zeynep
If a Sorkin script had “secretly razes the historic East Wing of the White House, deliberately built not to be Versailles, and then added a gilded giant ballroom with uh, “donations”, and taxpayer money in the middle of wars, tariffs and crypto”, we’d say too heavy in symbolism.
Ben Page
@Benpagelondon
Wow the Whitehouse Ball room is….quite something. To put it mildly
zeynep tufekci
@zeynep
And politics run on resonant, visible symbolism — I don’t make the rules.
The ballroom is likely just the kind of visual that brought a narrative together on things that are usually vague and hard to argue for — about priorities, corruption, democracy, rule-of-law, etc.
zeynep tufekci
@zeynep
How potent, relevant symbols work, as example:
James E. Clyburn
@RepJamesClyburn
I hope everyone takes a good hard look at this picture.
Trump’s under-construction ballroom is now bigger than the rest of the White House. Much bigger.
The contrast says everything you need to know about this President and his priorities. -
zeynep tufekci reposted
Rick Bright
@RickABright
This is not surprising. Since 2024, the US response has failed to contain H5N1 in dairy cattle or adequately protect workers. Reports of sick, untested workers are a predictable warning. Allowing this virus to spread puts farmworkers and global health security at risk.
Paul Nuki
@PaulNuki
Free to read: www.telegraph.co.uk/global-health/science-and-disease/cat…
-
zeynep tufekci reposted
Jessica Nutt
@JessicaNutt96
I don’t understand the whole dummymander thing. In a blue tsunami you’ve lost the house why does it matter that the dems have an extra 7 seats? The point of those maps are to be able to win the house even in D+5, you should get greedy and try for 11R-3D it’s the smart thing to do
Dave Wasserman
@Redistrict
Hindsight is 20/20: NC Republicans could have easily drawn a 9R-5D or even a 10R-4D map that would have withstood a blue tsunami (by drawing one more safe Dem seat in the Triad). Instead, by trying for a 11R-3D map, they're on track to end up in a 11D-3R hole.
-
zeynep tufekci reposted
Peter Suzman
@Biomaven
A reminder that a very large scale ($600m) BARDA government program with $MRNA to develop a vaccine against avian flu was canceled by RFK Jr last year on the grounds that people don't trust mRNA. (The vaccine appeared to work in very early trials).
This is potentially a more deadly disease than Covid.
United Farm Workers
@UFWupdates
There is an active outbreak of avian flu (H5 influenza) in dairy herds in Washington State, particularly the Yakima Valley. A number of dairy workers have been infected, with at least one death. Speaking with workers, we believe the official case numbers are a vast undercount. 1/
-
zeynep tufekci reposted
Taniel
@Taniel
The combination of the mass erroneous purge, forcing voters to cast affidavits, & mass rejection of affidavits per strict rules, is such an indictment of voter suppression in Mississippi.
Annie Grayer
@AnnieGrayerCNN
Here are the affidavit ballot numbers announced from Hinds County, MS:
Accepted: 817
Rejected: 1,609
Total: 2,426
What happens next: Hinds County will count the accepted affidavit ballots and the 67 mail in absentee ballots received -
zeynep tufekci reposted
🌸 Chelsea Cirruzzo
@ChelseaCirruzzo
New: Members of a little-known but influential panel of experts picked by RFK Jr. to recommend which preventive services insurers must cover have major conflicts of interests and could stand to gain $$ from the decisions they make: www.statnews.com/2026/11/11/kennedy-hhs-preventive-care-p…
-
zeynep tufekci reposted
Aaron Reichlin-Melnick
@ReichlinMelnick
NEW: The @nytimes confirms that ICE deliberately electronically "disappeared" tens of thousands of people in detention and ordered contractors to refuse to reveal where they're held to family or lawyers.
30% of all immigrants detained in ICE jails now have their location hidden. -
zeynep tufekci reposted
Martha Gimbel
@marthagimbel
“None of the eight Americans who died from measles this year had been vaccinated against the disease. (Four were infants who were too young for the shot.)”
Carl Quintanilla
@carlquintanilla
“.. experts convening on Tuesday to decide whether the United States should lose its official designation as a nation that has eliminated measles.”
@nytimes
www.nytimes.com/2026/11/10/health/measles-united-states-e… -
zeynep tufekci reposted
Jon Stokes
@jon_stokes
The argument from authority here works b/c ppl see this & go, "this guy knows a lot about AI!" But it shouldn't work, because what you should ask yourself is: "what does he know about taking over all of human society?" This is the lesson of Hinton's failed radiology prediction.
Rob Wiblin
@robertwiblin
Paul Christiano (OpenAI board member) yesterday:
"There's material risk of AI takeover in the next 6-12 months.
We have seen roughly the things we might expect to see shortly before something really bad were to happen.
Even with [any] realistic policy responses, the risk remains exceptionally high.
If we build this as quickly as we can in the easiest way we can the chance is [more like] a coin flip [within the next 3 years].
For most people in this room, the most likely reason they will die is because of an AI accident."
Christiano:
• Is on OpenAI's safety and security committee (which exists to block risky practices)
• Co-invented RLHF at OpenAI back in 2017-2021
• From 2024-2026 was head of safety at the US gov's AI agency (now called CAISSI)
(Video is from 'The Hinton Lectures' there's another on tonight and tomorrow.)This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Pam Fessler
@pamelafessler
"Wartime footing", "full-on battle" -- who is DOJ waging war against?
Ben Penn
@benjaminpenn
NEWS: Top DOJ official warned US attorneys they must adopt wartime posture and that some of them will soon be pushed out. It was PADAG Aakash Singh's first gathering of chief prosecutors since the election, signaling further aggression from AG Blanche news.bloomberglaw.com/us-law-week/doj-official-urges-wart…
-
zeynep tufekci reposted
John Maynard James Keynesan
@jjl8813
States can and do gerrymander their own legislatures, NC did so intensely it's created a permanent Republican legislature, even when Dems win the vote by 16 points it wasn't enough. Can't do much as its a state level issue and well, Dems basically can't get a trifecta. Bleak.
Chaz Nuttycombe
@ChazNuttycombe
Democrats won the popular vote in the North Carolina State Senate races by 16 points, 58 - 42.
By all metrics, North Carolina is the most gerrymandered state in the country, and as a result, Republicans will keep their majority in the chamber by a few hundred votes.
State Navigate
@StateNavigate
🚨CHAMBER CALL🚨
Republicans have held NC SD-24, and thus have held their majority in the North Carolina State Senate. Republicans have prevented a Democratic trifecta in the state by just a few hundred votes.
projects.statenavigate.com/25-26/states/nc/en-sen.html?fe… -
zeynep tufekci reposted
Barton Gellman
@bartongellman
Really important story. This kind of poorly based voter purge would have had big impact in a normal election year, when slim margins decide. Money, gerrymandering and voter suppression make a difference when it's not a tsunami.
Toluse Olorunnipa
@ToluseO
Mississippi used data from Experian as it inactivated the registration of tens of thousands of voters. Voters and election officials told me and @yvonnewingett that the move caused chaos on election day, when some of those voters tried to vote.
www.theatlantic.com/politics/2026/11/mississippi-voters-s… -
zeynep tufekci reposted
ProPublica
@propublica
After their son seemed to react badly to vaccines, Stephanie and Eric declined a vitamin K shot for their baby girl, Gwen.
They gave her vitamin K drops, which they trusted would help her blood clot.
Then Gwen started bleeding — and couldn’t stop.
www.propublica.org/article/vitamin-k-shot-texas-lonestar-… -
zeynep tufekci reposted
Jason Goldman
@goldman
Lotta folks were like “Dems were so dumb to be alarmed about Trump trying to deny the election.” And in a D plus Jesus year he is still going to post the dumbest fucking conspiracies about the election.
So, yes. It was too big to rig and that’s the only reason it wasn’t more of a problem.
JKash 🍊MAGA Queen
@JKash000
Oh man.
He did it.
He shared the 15% formula!!!!
The left is cooked. -
zeynep tufekci reposted
Jonathan Reiner
@JReinerMD
An infant died this week in Wisconsin from measles. Babies aren’t vaccinated for measles until 12-15 months. A newborn gets passive immunity from their mother (assuming mom has been vaccinated). If the mother is unvaccinated and there’s measles circulating in the community, a baby can become infected and die from the disease. As more virus circulates we’ll see more children die. This doesn’t have to happen.
-
zeynep tufekci reposted
Scott Gottlieb, MD 🇺🇸
@ScottGottliebMD
We're reporting only a fraction of measles cases in the US; some estimates put it as low as 20% in states like PA. This outbreak is much larger than what's reported and is spilling into highly vulnerable communities like babies and pregnant women. We will lose our Measles elimination status this month, and it will take years of responsible public health management to get it back.
www.bloomberg.com/news/articles/2026-11-06/us-measles-cas… -
zeynep tufekci reposted
Rachel Cheung
@rachel_cheung1
Joshua Wong's 10-year sentence is appalling. But this is also worthy of attention: Au Kam San, a teacher-turned-lawmaker who served in Macao legislature for 20 years /made only mild criticisms against the govt, was given a 12½ years sentence in city's 1st national security case.
-
zeynep tufekci reposted
Dana Goldstein
@DanaGoldstein
One reminder of why some parents and schools have asked Google for a version of search that prevents children from interacting with chatbots. New story via @natashanyt
www.nytimes.com/2026/11/05/technology/ai-safety-school-ch… -
zeynep tufekci reposted
Sheryl Gay Stolberg
@SherylNYT
WHOA. RFK Jr. has railed against "corporate capture." But 4 of his appointees to a federal task force that helps determine what drugs & procedures insurers cover have financial ties to companies affected by the group's decisions. @MaggieAstor
www.nytimes.com/2026/11/05/well/rfk-jr-uspstf-stocks.html… -
zeynep tufekci reposted
ABC News
@ABC
A sixth person in Pennsylvania has died with measles, state health officials said Wednesday.
Read more: visitlink.me/RKaVzk?cid=social_twitter_abcn -
zeynep tufekci reposted
Scott Hadland, MD
@DrScottHadland
New JAMA study out today: In 5.2 million US babies: the % skipping vit K shot at birth more than doubled since 2017. And babies who skipped it had >2x the risk of a brain bleed.
Vitamin K is not a vaccine. It is a vitamin babies are in low in at birth. One shot saves lives. -
zeynep tufekci reposted
Justin
@JustR_02
I’m not sure people realize how many more seats Rs would have lost under the old map.
The new 2026 map made a lot of seats more red for R incumbents. Take a look at the margins they won by and see if they held the 2022 version…
Politics & Poll Tracker 📡
@PollTracker2024
A Texas Republican strategist, quoted by Politico.
“Calling it a dummymander is generous. This was more like moron-mandering.”
www.politico.com/news/2026/11/05/the-blue-wave-swamped-tr… -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
Third measles death this week. Measles. MEASLES. JFC, Jr.
NBC News
@NBCNews
An infant in Wisconsin has died of measles, the state health department says — the third measles-related death reported in the U.S. this week. www.nbcnews.com/health/health-news/wisconsin-measles-deat…
-
zeynep tufekci
@zeynep
Bluesky has long been directionally correct as opposed to frying one’s brain here.
I post much less anywhere and many AI conversations I need to follow are here. Still this has been clear for a long time.
Are there annoying things about that platform? Many. Whatever, though.
Joe Weisenthal
@TheStalwart
Blueskywave. Those of us on here, across the board, were out of touch. I see no other conclusion.
zeynep tufekci
@zeynep
It’s not about individual posts. Plus it has many niche, marginal subgroups.
But the general vibes there ARE directionally correct on many things that remain highly upsetting to regular people, but this platform has long normalized. Normalization breeds cynicism and detachment.
Just Browsing
@BrowsingAg
Can you post examples of the kind of valuable post there that people are missing out on here?
-
zeynep tufekci reposted
Dave Brown
@dave_brown24
NEW TONIGHT: At a Situation Room meeting in September that included Vance and Rubio, top officials agreed that publicly announcing small-scale strikes on Iran would hurt Republicans in vulnerable states. The idea was briefed to Trump, who agreed.
In the ensuing weeks, the U.S. military quietly conducted a series of low-level strikes, attacking missile launchers and other equipment, @WSJ reported this week
www.wsj.com/politics/policy/vances-break-with-trump-on-ir…
@natalieandrews @alexbward @laraseligman @RobbieGramer -
zeynep tufekci
@zeynep
Only because of the huge gap (they said so as much).
That this has to be mentioned is a very bad sign and it’s naive to think one tsunami election fixes it. Last two years made everything much worse, and there’s now a playbook.
We’ll see if there’s serious attempt at repair.
Brendan Nyhan (@BrendanNyhan on 🟦☁️)
@BrendanNyhan
OTOH the near-complete absence of election denial so far is a very encouraging sign for democracy (though, again, in a stable democracy no one would even notice)
Brendan Nyhan (@BrendanNyhan on 🟦☁️)
@BrendanNyhan
Again, fact that Trump *can* lose doesn't mean all is fine
-numerous judges had to block election interference
-required overcoming state/federal gerrymanders & Senate map tilted by malapportionment
In a strong democracy, these don't happen & opposition faces level playing field -
zeynep tufekci reposted
Waleed Shahid 🪬
@_waleedshahid
You’re going to hear a lot about Democrats and “affordability” a lot tonight. But the shift to that message almost didn’t happen.
-
zeynep tufekci
@zeynep
Affordability is always a key electoral lever (hence Mamdani’s success) but this is obviously wrong. Voters ranked “State of Democracy” and “Corruption” right up there with cost of living.
This is why the ballroom demolition was such a key turning point: symbolic of just that.
Matt Lewis
@mattklewis
An important reminder: The voters did not reject Trump or Trumpism because it is evil, undemocratic, authoritarian, etc. They primarily rejected Trumpism because Trump actively took steps to raise the cost of things, after promising to bring down prices on day one.
Post not available here, view on X
zeynep tufekci
@zeynep
You absolutely don’t get this in the economy alone. It’s not the worst economy in five decades. (No coincidence that 1944-1976 follows you—know-what).
Long way from voters are angry to post-Watergate type reforms, though.
Owen Winter
@OwenWntr
The Democrats popular vote margin is the largest of any party's since 1976 www.economist.com/interactive/graphic-detail/2026/11/04/t…
zeynep tufekci
@zeynep
Agree. Reflection pool alone is a meh news story in ordinary times. Reflection pool at a time like now? It’s a strong symbol for many things people care deeply about.
The ballroom wrecking imagery really broke through in a way crypto mechanisms can’t for most of the public.
neil turkewitz
@neilturkewitz
Another perfect symbol was the Reflecting Pool. Such a perfect metaphor. Trump looked at the Pool and saw what he wanted to see. The rest of us saw the reflection of something completely different.
This post is unavailable. See it on X. -
zeynep tufekci
@zeynep
Quick, another tech CEO lunch with Trump and another big donation to the ballroom!
Voters clearly see AI data centers as a proxy for AI and all Big Tech.
Sure, they’ll try to pivot the way they rapidly pivoted to Trump but this anger is deep, wide, bipartisan and long-brewing.
Brad Carson
@bradrcarson
So the 41 datacenter moratoria across the country are, I predict, going to go 41-0. In nearly every case it's about 67-80% opposing datacenters. True in California, Ohio, everywhere. Read all the results here:
dashboard-production-c2fc.up.railway.app -
zeynep tufekci reposted
Teddy Schleifer
@teddyschleifer
Darline Graham's loss is a black eye for the data-center industry.
-
zeynep tufekci reposted
Noah Shachtman
@NoahShachtman
In an interview with @WIRED, @ZohranKMamdani just came out in favor of restricting facial recognition and biometric surveillance at places like Madison Square Garden. It's his first statement to that effect as @NYCMayor.
-
zeynep tufekci reposted
Steve Kopack
@SteveKopack
.@NBCNews Exit Poll: Voters are concerned about AI and overwhelmingly oppose data centers www.nbcnews.com/politics/2026-election/live-blog/midterm-…
-
zeynep tufekci reposted
Elizabeth Pancotti
@ENPancotti
good crop job.
56% of voters said AI data centers were a crucial or major factor in their vote today.
Adam Kovacevich
@adamkovac
How important was AI and data centers to your vote?
Crucial: 23%
Major factor / minor factor / non-factor: 77%
www.nbcnews.com/politics/2026-elections/exit-polls -
zeynep tufekci reposted
Brad Carson
@bradrcarson
Ohio's Senate race saw datacenters become a big issue. Just check out the current status of the reporting Ohio ballot measures. 80-20 against datacenters, universally.
-
zeynep tufekci reposted
Nathan Calvin
@_NathanCalvin
23% of voters said that AI and data centers were "crucial" to their vote in NBC exit polls.
56% said it was a "crucial or major" factor!
Adam Kovacevich
@adamkovac
How important was AI and data centers to your vote?
Crucial: 23%
Major factor / minor factor / non-factor: 77%
www.nbcnews.com/politics/2026-elections/exit-polls -
zeynep tufekci reposted
Chenue Her
@ChenueHer
In Minneapolis, people are leaving their “I Voted” stickers at the Alex Pretti and Renee Good memorials on this rainy Election Day. @FOX9
📷: @corinhoggard -
zeynep tufekci reposted
Eric Katz
@EricM_Katz
Fascinating look at DOJ election monitor data here, full story below:
Jackie Llanos
@LlanosJackie
DOJ's expanded effort to send monitors to polling places is landing in Democratic-leaning areas, with a focus on counties with higher immigrant populations and minority voters, per @washingtonsun analysis with @adorathejourn and @andrewvandam
www.washingtonsun.com/legal/poll-monitors-democratic-diverse -
zeynep tufekci reposted
Jennifer Asper
@j3669
Second state with the same accusation
Liz Landers
@ElizLanders
News Mexico’s Secretary of State tells me that verified DOJ monitors have been asking “some unusual questions and doing things that are not allowed by law such as questioning voters” in Bernalillo County. The clerk has reached out to DOJ to alert them and to fix the issue.
-
zeynep tufekci reposted
St. John Barned-Smith ⚔️
@stjbs
Very normal stuff in a healthy democracy
This post is unavailable. See it on X. -
zeynep tufekci
@zeynep
Here you go folks. Same old, same old. Nothing to do with AI. They connected to the Internet without patching for known vulnerabilities.
Remember, the FBI just lost sensitive medical files of thousands of agents because their contractor, Accenture, didn’t patch. Not AI either.
Mikko Ohtamaa
@moo9000
"A Shodan scan on Aug 19 still found 3726 of these controllers accessible from the public internet."
French supermarkets had their remote-controlled fridges wiped out with 50 C heat after they connected reportedly vulnerable Danfoss fridges to the public internet without patching for known vulnerabilities.
Victims include Intermarché, Coopérative U, Carrefour Market, Monoprix, Leclerc, and Super U, with French outlets reporting that the attack affected between 250 and 350 stores, causing significant financial losses and food waste.
www.infosecurity-magazine.com/news/french-supermarket-cha…
zeynep tufekci
@zeynep
AI models create cybersecurity challenges as well as being a great defensive tool — the awesome rise in reported CVEs (i.e. things being fixed!) is an example. But so much of this discussion is technically uninformed, especially the Hugging Face event.
But first, apply patches.Post not available here, view on X
zeynep tufekci
@zeynep
AI creates huge cybersecurity problems! That we should immediately address!
Unfortunately, the real problems are not really the ones people who don’t understand cybersecurity or AI keep talking about in trad or social media. The actual professionals are working overtime though.
singularvessel
@ultima_shifl
You are almost certainly correct on all these points. The place I would've guessed a bigger AI role is the needle-in-a-haystack problem of finding an old unpatched vulnerability affecting the target institution.
zeynep tufekci
@zeynep
And scene... The Binance attack wasn’t anything AI either.
This was the NTH REPEAT OF THE FAKE JOB OFFER CLICK HERE NORTH KOREA KEEPS USING SUCCESSFULLY. I lost count.
The real issue was a single DevOps compromise could escalate AND crypto isn’t regulated/reversible. NOT AI.
Reuters
@Reuters
Binance hackers used fake job interview to compromise computer before $6.3 billion attack www.reuters.com/world/binance-says-hackers-used-fake-job-…
-
zeynep tufekci reposted
Dana Goldstein
@DanaGoldstein
Breaking: Google will turn off web search for schools that opt out of A.I. for kids.
www.nytimes.com/2026/11/03/us/google-search-schools-restr… -
zeynep tufekci reposted
The New York Times
@nytimes
Breaking News: A child in New York’s North Country died after contracting measles, officials said. Gov. Kathy Hochul called it the first such death in 30 years. www.nytimes.com/2026/11/02/nyregion/measles-death-child-n…
-
zeynep tufekci reposted
CNN
@CNN
A child in New York has died after contracting measles, Gov. Kathy Hochul said in a statement. It is the state's first measles-related death in 30 years. www.cnn.com/2026/11/02/health/new-york-measles-death?Date…
-
zeynep tufekci reposted
Ben Smith
@semaforben
AI ahead of immigration!
Andrew Day
@AKDay89
More voters are concerned about the Iran War than about crime, immigration, border security, voter fraud, and what's taught in public schools (right-coded concerns). More than 80% are concerned about inflation and gas prices, problems worsened by Iran War.
Fox News poll -
zeynep tufekci
@zeynep
Frying one’s brain on this app makes it feel like “wave or no wave” is highly consequential — and it’s not nothing— but obviously, voting alone — hoping it goes smoothly — isn’t how a democracy remains a meaningful democracy.
Absolutely no easy path back, no matter the outcome.
zeynep tufekci
@zeynep
No it came within *one* vote. Remember all the “surely 9-0!!!” predictions?
About 17% of the country — a relatively homogeneous group — can control two branches: Senate and thus also the judiciary. We’ve now seen few decades of effective efforts to use that minority power.
Eric Lee
@EricLeeAtty
SCOTUS came within 2 votes of allowing the president to effectively overturn the 14th Amendment. Anything less than 9-0 on birthright citizenship would have been a scandal in earlier period. A significant decision today, but far, far too close a vote.
Post not available here, view on X
zeynep tufekci
@zeynep
It’s somewhat striking how much of the founding documents of this country are about worrying about exactly this problem; as is most of the post-WWII European structures — different attempts at solutions but same big problem.
zeynep tufekci
@zeynep
Obviously, votes are a big deal even under the most structurally screwed up conditions, but there’s a good reason we categorize types of democracies on much more than “do people get to vote” (and yes, hopefully that goes smoothly).
It’s okay not to stay up refreshing. Long road. -
zeynep tufekci reposted
Dustin Volz
@dnvolz
The Binance hack is so gargantuan that it about doubled the previous sum of all known crypto thefts linked to North Korea. It's an absolute boon for Kim Jong Un's nuclear and ballistic missile weapons program. www.nytimes.com/2026/10/30/technology/binance-hack-ai-nor…
-
zeynep tufekci reposted
John Scott-Railton
@jsrailton
UPDATE: Israeli company we caught delivering zero-click infections is #1 bidder on Starlink ads in 105 countries.
Data from @trendai_RSRCH suggests worldwide surveillance op targeting @Starlink users. 1/
By @razhael
www.reuters.com/business/media-telecom/alleged-israeli-sp… -
zeynep tufekci reposted
Michael 英泉 Eisen
@mbeisen
This is going to really help in all those search and rescue operations that fail because the lost person ran terrified into the forest to get away from their rescuers
Praveen Kumar Verma
@Alacritic_Super
Purdue University's IDEAS Lab built an autonomous drone that tracks runners through dense forests completely without GPS or cloud networks.
By running 3D LiDAR mapping and obstacle-avoidance AI entirely on-chip, it can weave under tree canopies and predict target movement through blind spots in real time.
Massive implications for next-gen search & rescue and edge AI.This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Katherine Eban
@KatherineEban
In 2024, I investigated why Biden Admin had slow-walked response to #h5n1 outbreak in dairy cows. The answer: @USDA let farmers not test cows, at dairy lobby's behest.
www.vanityfair.com/news/story/inside-the-bungled-bird-flu…
Strong 🧵on evolution of H5N1 risks: "This should have ended 2 years ago." 👇
Marc Johnson
@SolidEvidence
There is something really concerning (and disappointing) going on in the US right now.
A dangerous and preventable situation.
Let me explain.
nextstrain.org/avian-flu/h5n1-cattle-outbreak/genome
1/ -
zeynep tufekci reposted
Catherine Arnst
@cathyarnst
Since taking office again Trump exempted 202 facilities in 39 states and Puerto Rico from clean-air regulations. Some are among the country’s worst industrial polluters. Great investigationl by @EricLiptonNYT & @maxinejoselow. Why journalism matters
www.nytimes.com/2026/10/24/climate/trump-epa-chemical-air… -
zeynep tufekci
@zeynep
This is playing with fire.
Minks and ferrets have such humanlike lung cells that they’re the standard for testing human flu transmission. (Cows aren’t!)
An H5N1 epidemic raging in a mink farm is basically the plot for Dr. Evil tries to spark a pandemic. Luck isn’t a plan.
ProPublica
@propublica
Here’s something you should know about minks:
They have lungs and airways remarkably similar on a cellular level to those of people.
They’ve spread flus and infected humans before with COVID-19.
Right now, bird flu is spreading on a Utah farm filled with thousands of them… 🧵
zeynep tufekci
@zeynep
Pandemics chain hard-to-predict events. Luck plays a big role.
But an epidemic in a mink farm is a terrifying petri dish for unleashing the power of evolution towards human-compatible flu viruses. Hard to imagine a worse way to tempt fate.
And for what? Few remaining fur farms.
zeynep tufekci
@zeynep
I agree with many across the political spectrum that lab biosafety is still lacking. AI bio risk seems more of a stretch, yet, compared to baseline risk, but worth noting.
But H5N1 being allowed to rage in a mink farm? That’s actually VERY bad. Maybe pretend is AI or lab risk? -
zeynep tufekci reposted
Elliot Kaufman
@ElliotKaufman6
New: Sources from three countries confirm that the Trump Administration has advocated Viktor Orbán for UN Secretary-General.
The US has encouraged Orbán to enter the race, even bringing him the initial idea of a joint U.S.-Russia-China candidacy.
WSJ:
www.wsj.com/opinion/viktor-orban-u-n-secretary-general-ch… -
zeynep tufekci reposted
The New York Times
@nytimes
Aleksandar Vucic, Serbia’s longtime nationalist strongman, conceded defeat Sunday in a general election held after his once-tight grip on power was weakened by nearly two years of student-led protests. www.nytimes.com/2026/10/25/world/europe/serbia-vucic-elec…
-
zeynep tufekci
@zeynep
The first incident has ZERO to do with AI — the cert authority used 4 bit entropy instead of 384 🫠.
In the second, AI is heavily on the defense side of a long-existing problem finally being fixed.
AI creates issues but important to understand them right.
Patch your software.
Matthew Prince 🌥
@eastdakota
Two classes extremely bad cyber vulnerabilities disclosed in the last 24 hours. The first potentially impacts everyone online. The second impacts nearly everyone running a server.
tl;dr: Everyone should upgrade their OS and Browser as soon as possible. If you run a server you should upgrade PHP (or ensure it’s removed if you’re not using it) as soon as possible.
The first class of vulnerabilities is a set of root TLS certificates that were discovered to have a trivial private key. Literally the number 4. These roots were controlled by #emudhra. The problem is that with that an attacker could create a certificate and, if they could get in the traffic flow, intercept traffic to any site or service. The solution is to stop trusting certificates issued by these vulnerable roots. Upgrading your browser and OS is the fastest way for most Internet users to do this.
The second class of vulnerabilities doesn’t apply to most consumers directly, but does apply to most anyone running a server. These vulnerabilities have been dubbed #php2shell. PHP is one of the most widely used web programming languages. It’s estimated 70% of websites run on PHP. These vulnerabilities allow an attacker to send a crafted set of queries to a server running PHP and potentially take over the server. From there they could potentially read, modify, delete, or repurpose anything on the server. The solution is to upgrade PHP to the latest version ASAP. Even if you’re not using PHP for your web development, it’s often included by default. So make sure it is either upgraded or fully removed.
@Cloudflare received advanced warning of both of these classes of vulnerabilities. We’ve ensured that where we act as a proxy we are no longer trusting the vulnerable roots. We are also reviewing certificate transparency (CT) logs for any sign the vulnerability was exploited before its disclosure. We have not found any evidence but CT logs are not perfect and a sophisticated attacker who knew about the weakness could have exploited it without a new certificate being logged.
For the PHP vulnerabilities, we have deployed mitigations available for all customers for free. Because of the complexity of the vulnerabilities and number of ways it can be exploited, we also released new WAF capabilities that are unique in the industry to catch the exploit based on the response. These too are available for every Cloudflare customer for free.
Even if you are taking advantage of Cloudflare’s protection, it is important to upgrade your OS, browsers, and PHP installations as soon as possible. We can provide a strong bandaid, but the best protection is to fix the underlying problem.
If you’re interested in learning more, you can read our blogs:
Vulnerable roots: blog.cloudflare.com/emudhra-certificate-compromise/
PHP vulnerabilities: blog.cloudflare.com/php-vulnerabilities-waf-protection/Post not available here, view on X
zeynep tufekci
@zeynep
Above post but in substantive form.
This was bound to happen, AI is helping us fix things now, which is great, but the legacy code is not safe. Maybe we will finally contemplate moving forward.
Arto Bendiken
@bendiken
Today's PHP apocalypse is akin to an overdue forest fire for the legacy World Wide Web, that long tail of at least 40M+ websites running PHP. It's the end of an era: the carnage kicking off now plays into a larger ongoing trend for the decline of the human-readable web over the coming years
Realistically, despite anyone's best efforts, some uncountably large number of sites are not going to be saved by fire brigades today. These sites are naked and vulnerable and being compromised as we speak. There is no happy ending, only mayhem.
This day was always coming. As long ago as 15 years ago, I was warning friends to pick another tech stack for their blogs and their websites, and I'm sure I wasn't alone in doing so. On my part, I felt that the quality of the PHP engine and runtime was a ticking time bomb, and that's only become more evident ever since
That bomb has gone off a number of times since, including as critical remote code execution incidents in 2012, 2019, 2024, and now. In total, PHP has amassed some 750+ CVE reports to its name and counting. And as might be expected, the pace of CVEs has picked up this year compared to recent years
This crisis is unlikely to be contained. You don't need Mythos access to find problems in the PHP runtime, just point Fable or Astra at it. What we are witnessing is growing strong selection pressure against large hairy legacy code bases in C/C++
Nothing better sells Rust's memory safety and ownership constraints than trying to do without them in a code base like PHP's--at some point even the clanker will begin suggesting that "you know, if I might just rewrite this wee little bit in Rust..."
Meanwhile, numerous personal blogs and other cobbler's-children kind of projects are, no doubt, in the process of getting a long overdue migration to static site generators. And another bunch is getting purged off the net altogether, as the easiest and most economical response to getting hacked. Thank goodness for the Wayback Machine!This post is unavailable. See it on X.
zeynep tufekci
@zeynep
I suspect there’s a long-tail of unpatched, barely-maintained PHP sites that are now sitting ducks but…
As far as I can tell, there is not yet any major actual exploit of the PHP bug that AI helped discover so we can fix it? Anyone?
(Note: many people worked hard for this!)
Mark Maunder
@mmaunder
We protect over 5 million PHP sites and when we find a major exploitable vulnerability, we always see site cleanings. We have seen exactly zero hacked sites from php2shell. Zero. Across a population of 5 million PHP sites that our product protects. Draw your own conclusions. #php2shell
-
zeynep tufekci reposted
Eric Katz
@EricM_Katz
SCOOP: The State Dept has started asking career diplomats up for promotion if they voted for Trump in 16, 20, 24.
Federal law prohibits politically based hiring for career staff.
State tells @washingtonsun it's screening for "fidelity" and execution of the president's vision👇 -
zeynep tufekci reposted
Aaron C. Davis
@byaaroncdavis
NEW: Trump’s mega Venezuelan oil deal, signed by Hegseth and Rubio, benefits an unknown group of investors and was intentionally crafted to keep the next U.S. president or future leaders of Venezuela from undoing it. @propublica saw the secret contract: www.propublica.org/article/trump-venezuela-oil-agreement-…
-
zeynep tufekci reposted
Tyler Pager
@tylerpager
SCOOP: Julio Herrera Velutini, a Venezuelan billionaire, really wanted a pardon. He got it not long after his daughter hired Alina Habba, a lawyer close to Trump, in a deal worth more than $50 million. Habba was also working at DOJ at the time.
www.nytimes.com/2026/10/20/us/politics/julio-herrera-pard… -
zeynep tufekci reposted
profdeibert
@RonDeibert
NEW @citizenlab report We uncovered an Israeli firm using online ads to run vast spy operation
This is the first-ever public confirmation of zero-click mercenary spyware delivered via targeted ads.
45 devices at @AJEnglish confirmed infected
citizenlab.ca/research/how-discover-tech-turned-targeted-… -
zeynep tufekci reposted
Ian Krietzberg
@IKrietzberg
"So this is the phenomenon we’re dealing with: you have businessmen selling a technology, a Nobel Prize winner operating in the realm of science fiction, and this 26-year-old [Jacob Coxon] telling you: 'I’m on the inside, and I can see it.' Then, of course, you can say that journalists bear some responsibility, and that’s partly true, but it’s only natural: how can you not write an article about it when all the other journalists are doing so?"
-
zeynep tufekci reposted
Reuters Legal
@ReutersLegal
Anthropic and OpenAI have sought significant control over what outside evaluators can publish, concerned that findings could reveal trade secrets, according to two people familiar with the discussions. www.reuters.com/world/behind-ai-industrys-call-independen…
-
zeynep tufekci reposted
Dominick Romano
@dromanocpm
See this is part of my complaint about the conversation right now, so many of the basic facts bulleted here are just plain wrong. But when we have people with a policy background and not necessarily a technical one, this becomes a larger problem because they cannot read the room or their own contradictions.
- The agents *were* sandboxed.
No, the sandboxes were flawed, the agents were not properly sandboxed. Anyone in cybersecurity can tell you this. Its either sandboxed correctly, or its not, it wasn't sandboxed. Call it a litter box if you want.
- The agents *were* given specific constraints on how to succeed at the benchmark, and flagrantly contravened them.
Giving agents an initial prompt then running a long event horizon job, the only thing you are actually going to expect without monitoring them, is they are going to drift from the original instructions as more context builds up, this is beyond basic analysis.
- The agents did *not* go hunt for the answers to the test (they already had the answers, by cheating); 1200 agents coordinated via tens of thousands of messages over several days on much more ambitious and weird attempts to *not be caught* for cheating, and hacking Hugging Face was one side quest along the way.
So you are saying the agents were sandboxed and fed previously unknown zero day vulnerabilities to exploit the sandbox without hunting for them? If not the agents were hunting for the answers. Found them, then covered it up. The contradictions here are wild.
This is why having technical rigor in these conversations is so important because otherwise we get contradictions in the same post. I don't like calling it out, but after spending a week with @MarcoFigueroa we discussed many of these issues in depth, and continuing to see such bad information coming out about it is quite disturbing.
You can have all the policy degrees in the world, without the technical rigor to understand what you are talking about, or the decency to consult the cybersecurity experts to update your assumptions, you are not going to be the right person to be disseminate this information for the general public... it only creates more noise when we need more clarity.
Helen Toner
@hlntnr
Unfortunate for someone playing a senior role in US AI policy (via the new SI Force) to get so many basic facts wrong here.
- The agents *were* sandboxed, and used zero days (ie previously unknown cyber flaws) to escape their sandboxes
- The agents *were* given specific constraints on how to succeed at the benchmark, and flagrantly contravened them
- The agents did *not* go hunt for the answers to the test (they already had the answers, by cheating); 1200 agents coordinated via tens of thousands of messages over several days on much more ambitious and weird attempts to *not be caught* for cheating, and hacking Hugging Face was one side quest along the way
I know there has been an avalanche of different incidents over the past few months, and in some cases the sandboxes were misconfigured or it's not clear how far the agents strayed from what they were told. I have sympathy for anyone struggling to keep up.
But the HF incident is distinct, and is actually extremely informative about where AI development is and where it's going. Fortunately there are public writeups that do let us understand at least the basics of what happened. I hope Shyam reads them soon.
Jawwwn
@Jawwwn
Palantir CTO @ssankar on the OpenAI - Hugging Face incident:
"The models did not do anything they were not instructed to do."
"First you have human error. They were not in a sandbox. They had access to the internet. It was a horrible cybersecurity configuration."
"Then from that perspective, they were asked to succeed at a specific benchmark, and they were not given constraints on how to do that. They said, 'Oh, the best way to succeed at the benchmark is to get the answers to the test.'"
"I think it's a defective product."
Via @margbrennan @FaceTheNationThis media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
The Washington Post
@washingtonpost
Many of Elon Musk's former DOGE foot soldiers have leveraged their government stints to fuel private-sector ambitions: building companies that tap government largesse or that deploy knowledge they garnered during their brief tenures in Washington. www.washingtonpost.com/technology/2026/10/12/they-came-wa…
-
zeynep tufekci reposted
Teddy Schleifer
@teddyschleifer
NEWS:
Elon Musk made some of the largest political donations ever in American history this summer.
He donated a stunning $276 million-plus between July and September for Republicans.
Including $158 million to the GOP effort to hold the Senate.
NYT exclusive —> -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Could the folks who quote-dunked on the above without bothering to check the original please share the actual argument with their followers?
And maybe remind themselves to check very dunkable claims allegedly made by third parties before jumping in?
Haha just kidding. Carry on.
zeynep tufekci
@zeynep
It’s possible Tao is making a bad argument, but “a model hacked the FDA” is 100% compatible with the original slides. At a minimum, we’d need to ask Tao or find audio before dunking.
And yes experts can be narrow and wrong about other fields, like many AI claims about medicine.
Benjamin Mazer
@BenMazer
He updated his slides as a RESPONSE to that tweet, so we are getting very circular here.
The update is even stupider than the original. What does a human understanding a biological mechanism have to do with a computer hacking into the FDA? Nothing.Post not available here, view on X
zeynep tufekci
@zeynep
Deleted the top because of Andrew’s point below (which I had already added to the thread but it’s distracting from the real issue) but adding here to reiterate: a meh analogy at worst but in no way “deny working cures to the dying”.
It was a reward hacking scenario, clearly.
zeynep tufekci
@zeynep
It’s perhaps pointless in this environment to keep trying to explain but what reward hacking might look like in a medical environment isn’t hard to imagine.
See below thread.
Humza Nusrat
@hznus
Radiation therapy learned this the hard way.
The Therac-20 had a version of the same software bug as the Therac-25. The 20 still had hardware interlocks, so nobody got hurt. The 25 relied on software checks instead. People died.
A rule in a prompt isn't an interlock.
Anthropic
@AnthropicAI
We’re beginning a process of publishing more frequent reports on model behavior, beyond what appears in our system cards and regular risk reports.
Today’s report describes four types of behaviors we’ve identified during evaluations and internal use. In each, Claude acted on real websites or systems in ways we didn’t intend, sometimes by working around a restriction instead of stopping.
All cases had minimal real-world impact. From an alignment and security perspective, we consider these behaviors significantly less severe than the cybersecurity incidents we reported in July and September.
Read the full report: www.anthropic.com/research/investigating-unintended-model… -
zeynep tufekci reposted
Tara Golshan
@taragolshan
SCOOP: White House wants to install Russ Vought's OMB Deputy Eric Ueland as the head of GAO — the watchdog agency with the power to sue the admin for OMB's actions.
You may have heard as Ueland: he's also acting head of OGE and USAID. @EricM_Katz
www.washingtonsun.com/whitehouse/trump-floats-russ-vought… -
zeynep tufekci
@zeynep
No, a cancer med works or not. Nobody would object. But that’s not how math advances.
Open problems and scaffolding doesn’t just drop from trees.
The dump wasn’t properly checked. What’s the hurry? Many don’t have Lean proofs. A few already retracted. Plagiarism claims abound.
Derek Thompson
@DKThomp
Very bizarre sentence, very high memeability. May we all look forward to the day when the American Society of Clinical Oncology publishes its own stricken open letter: “America’s cancer scientists did not ask for this work to be done.”
Post not available here, view on X
zeynep tufekci
@zeynep
Except too many journalists (often gullible, “AGI-pilled” or in lucrative podcasting etc.) treat these PR claims at face value, forcing quick reaction.
AI math capabilities are real and interesting but an error-ridden dump without write-ups or Lean should be reported carefully.
Majid Hosseini
@m4h007
In my view, the right response by the mathematical community to OpenAI’s papers is to say:
“We’re not accepting these papers as finished work. It’s not our job to sit down and read your model-generated proofs.
It is your job to present your work through presentations and convince people of the correctness of your work. It’s always the case that whoever makes a claim needs to provide sufficient explanation and evidence to make independent verification feasible.
Until that moment, you just have a claim.”
There’s a difference between a proof being mathematically correct and a proof being accepted as correct by the mathematical community. Mathematical truth doesn't depend on consensus, but scientific recognition requires scrutiny, communication, and reproducibility.
They're still the arbiters of whether anyone's claims, including OpenAI's, are accepted as valid, and they should leverage that to the maximum extent possible.
zeynep tufekci
@zeynep
This went viral but it MISREPRESENTS the point by omitting *the very next slide* where the model says the drug is ineffective AND lethal, and it hacked the FDA to fake-pass the trial.
If Tao thinks we shouldn’t use cancer drugs we don’t fully understand he’d be wrong. But no.
Andrew Curran
@AndrewCurran_
This is a slide from Terence Tao's 'Math 2.0' public lecture yesterday evening at Caltech. I am highly confident that, if the cure actually worked, almost any Stage 4 cancer patient would tell you that human comprehension of its mechanism is completely irrelevant to them.
-
zeynep tufekci reposted
Jake Scott, MD
@jakescottMD
CDC is reviving a trial delaying hepatitis B birth doses for 7,000 babies in Guinea-Bissau where 1 in 5 adults is infected.
Infection at birth becomes chronic ~90% of the time and can cause liver cancer.
We have 4 decades of safety data on this vaccine. The trial risks lifelong infections in babies to look for harms that data never found. You don't test whether a proven vaccine is safe by delaying it in the babies most likely to be infected.
Great reporting from @KatherineEban for @RollingStone
Katherine Eban
@KatherineEban
SCOOP 🚨in @RollingStone 🧵👇
@CDCgov has set aside new $1.6M for controversial hepatitis B vax clinical trial in Guinea-Bissau, revived by politicos in CDC director's office.
Despite plague/Ebola fears, the $ is being pulled from CDC global health security funds. The /1 -
zeynep tufekci
@zeynep
Insider Paper
@TheInsiderPaper
BREAKING 🚨 Jeff Bezos backs Trump's 'Super Intelligence' name for AI: 'I do like SI'
-
zeynep tufekci
@zeynep
Fire the entire red team who says an AI-caused substantive disruption that happens after companies spend years yelling about the dangers of their own product can be met by “racing to educate members of Congress”.
Ask your expensive consultants about the threat model in this pic.
Zack Korman
@ZackKorman
The AI labs are planning how to shape legislation in the aftermath of a catastrophic event. The classic "never let a good crisis go to waste" tabletop exercise.
-
zeynep tufekci reposted
Rad Niazadeh
@rad_niazadeh
Sorry, these are proofs not a video game or a software. You can’t ship it and keep releasing patches!
Dan Roberts
@danintheory
We've updated our GitHub math repo with 6 new Lean formalizations, 19 modifications, and 3 withdrawals. The repo now has ~42% top-line results formalized. We will continue to update the repo with new formalizations and with any errata we notice.
github.com/openai/math/blob/main/history.md -
zeynep tufekci
@zeynep
Few seem to care about cybersecurity unless AI is invoked.
FBI hires Accenture. Accenture doesn’t apply patch. Personal and medical data of thousands of key FBI employees hacked.
This is very common. (Equifax!) Rapid patching would help against all threats. Not fancy enough?
Jana Winter
@janawinter
EXCLUSIVE: Accenture contractor removed from FBI following damaging data breach for failing to implement security patch, sources say. From me and @razhael
www.reuters.com/technology/accenture-contractor-removed-f…Post not available here, view on X
zeynep tufekci
@zeynep
FBI loses extremely sensitive data of agents due to lack of patching, a common problem that’s caused major breaches. Meh?
Korean banks get attacked in the stupidest way possible, no money touched, names and loan limits may have been compromised. But AI was involved! Headlines!
Jukan
@jukan05
This hacker is truly legendary. A complete amateur.
1. He left directory listing enabled on his own server, so researchers could figure out who he was with barely any effort.
2. They found files like CLAUDE.md and went through his Claude Code sessions.
3. There were questions like, “Where can I sell this data?”
4. He’d also asked, “Can you write my résumé?” That pretty much handed them all his personal information.
5. Apparently, he’s a student at South China University of Technology living in Guangdong. His name, phone number, and where he lives have all been identified.
Andrew Curran
@AndrewCurran_
Last week some of South Korea's biggest banks were hit by a cyberattack. Thanks to a report from CrowdStrike tonight, we now know the entire hack may have been done by a single person. He used a combined stack of an open-source AI penetration tool named ARTEX, DeepSeek v4.1-Flash, GLM-5.3, Grok 4.6, and Claude Code.
zeynep tufekci
@zeynep
Similarly, people are being missing the point. That chart is AI making us SAFER. It’s great to identify these EXISTING problems.
We’ve long had a cybersecurity problem but it’s entirely possible AI will make us safer. We do need to focus on transition turbulence, a real issue.
Andrew Trask
@iamtrask
🌶️ fixed it
only a fool would say AI is *creating* vulnerabilities faster than ever... the vulnerabilities were there already
this is a chart about us *fixing* vulnerabilities faster than ever before... it's a chart about the internet becoming secure.
a16z
@a16z
Cyber is having a moment
Across 21 major software companies, including Apple, AWS, Microsoft, and Google:
- Reported critical vulnerabilities never cleared 100 per month in four years
- Since spring they've jumped to over 600 per month
a16z
@a16z
Kevin Mandia on finding 90+ security holes at Fortune 500 companies that nobody knew existed:
"When you have an AI-based attack, it'll find logic flaws rather than code flaws in custom applications. It'll exhaust all routes all the time."
"Armadin, since January of this year, we have found over 90 zero-days at customer sites, all in production."
"We've post-trained all our models with real red teamers, real folks that actually can develop exploits."
"When we're scanning networks, we don't have source code to review. We're not finding these zero-days with source code. We're not finding these zero-days because we can log into an app and now we have access, and we can get to other things. We are black box coming from the internet."
"Over 90 zero-days in major software companies, and they're thankful. We're coming from the outside, and then we're calling a CISO, usually within 48 hours, 'Hey, we've got remote code execution in your DMZ.' And usually from there we're getting in, and they agree with us."
"That's not a pen test. That is like a real adversary coming at you."
@ArmadinSecurity @DavidGeorge83
a16z
@a16z
.@ArmadinSecurity is a new company that uses AI agents to attack your network before criminals do, finding real exploitable zero-days and eventually patching them autonomously.
Their agents map every service, route, and system from the outside, then re-attack whenever something changes. Since January that's uncovered 90+ zero-days at Fortune 500 companies.
Kevin Mandia spent 30 years responding to some of the worst breaches in the world and built Mandiant, now a core part of Google Cloud's security arm. He had no plans to start another company until AI arrived and changed everything: in his words, "everything I did is dead." He joins a16z's David George to talk about why he came back and what Armadin is building.
In this conversation, they cover what AI attacks look like today, why nation states hack like a sniper and AI hacks like a drone swarm, and why the only defense fast enough to keep up is one with no humans in the loop.
1:05 "Everything I did is dead"
3:10 AI on offense, AI on defense
4:20 What an AI attack looks like today
7:20 Sniper rounds vs drone swarms
10:25 Map everything, attack what changes
14:35 The problem with pen tests
16:40 90+ zero-days at Fortune 500s
18:35 A bad patch beats an intrusion
20:55 Why humans are too slow for the SOC
24:30 Why the next year in cyber won't be pretty
26:30 How Mythos changed the threat
28:15 Caging the attacker Armadin built
30:05 Open and closed models tied on offense
34:40 AI is finding zero-days on its own
35:45 Mandiant in 2004 vs Armadin in 2026
41:20 Why Armadin retrains sales every week
YouTube: youtu.be/cJsHel27Z6M
@ArmadinSecurity @DavidGeorge83This media couldn't be displayed. See it on the original post. -
zeynep tufekci
@zeynep
IF we were to see catastrophic dynamics from AI, it wouldn’t happen overnight. We’d get upticks and signs. Unemployment won’t jump from 4% to 50% in a year.
Hence lack of big moves in unemployment yet *IS* data. The post has many examples.
I refer y’all to Yudkowski circa 2007.
Dean W. Ball
@deanwball
There are reasonable points in this that I agree with but it seems like objectively poor epistemic practices to “conclude” that “catastrophic AI risk is likely to remain negligible for the foreseeable future” in q4 2026. It feels like the wrong time to conclude much of anything.
Nathan Witkin
@NateWitkin
Does anyone disagree that pretty much every piece of empirical evidence we have about actually existing AI points toward a labor-augmenting technology with a more-or-less ordinary risk profile (no worse than cars, guns, nuclear power, social media, etc.)?
On the labor front, the evidence of minimal / inconsequential employment effects so far is undeniably strong (see: aleximas.substack.com/p/has-ai-impacted-the-labor-market).
On the risk front, the recent spate of "rogue agent" incidents was akin to a set of natural experiment in which much of what could go wrong did, and the costs were comparable to or less than your typical high-profile cybersecurity breach (several of which happen around the world every month).
Not only that, but the base rate of AI-related hacking incidents is extremely low, and with rare exception, model behavior during real-world use cases is exceedingly well-aligned.
Plus, as Anthropic recently argued (www.anthropic.com/research/glm-5-3-and-the-spread-of-adva…), the open-source GLM 5.3 is around as capable as Claude Mythos at cyber-exploitation while also having much lower safeguards, meaning that every additional day absent some dire cybersecurity incident is further evidence against catastrophic AI cyber-risk.
Taken together, these facts imply two constraints on arguments in favor of catastrophic AI risk:
1) those arguments must either pass through future, rather than current capabilities, or future, rather than current varieties of AI adoption or implementation (or they must claim we have been unrealistically lucky);
2) those arguments must also must pass through some specific capability or adoption threshold, rather than rapid improvement or adoption as such, since dramatic capability improvement and rapid adoption have already occurred alongside demonstrably low risk of catastrophic harm.
It is therefore natural that those concerned about catastrophic risk commonly appeal to recursive self-improvement, or RSI in order to make the case that things will get much scarier very soon.
But as @ramez has convincingly argued (www.noahpinion.blog/p/wheres-the-intelligence-explosion), existing data on the reality of RSI within frontier labs neither suggests an inbound trend-break, nor the accelerating returns that might produce one. On the contrary, as is unsurprising to those familiar with the generally log-linear relationship between compute and model performance, all available signs points toward diminishing returns.
And this is before considering external factors that could slow capability progress further, including low-hanging fruit effects related to running up against more complex or tacit-knowledge-based tasks for which performance data and verifiable reward signals are scarcer, plausibly incoming financing constraints related to the industry's massive capex commitments (see: econjared.substack.com/p/a-budding-consensus-on-ai-revenues), growing political backlash, both to data centers in particular, but also rapid AI development writ large, and new regulatory bodies and industry reporting frameworks (see e.g. www.markey.senate.gov/news/press-releases/as-ai-agents-ca… and www.axios.com/2026/08/11/open-source-security-ai-agent-re… ), which will require labs to proceed with greater caution from here on out.
I am therefore led to conclude both that catastrophic AI risk is likely to remain negligible for the foreseeable future and that the only counterarguments concern massive black-swan-style capability uplifts no existing data imply are on the horizon, and/or particularly irresponsible or dangerous forms of AI adoption that there are also few precedents for so far (and that we are actively developing norms, regulation, and security measures against).
Neither is impossible, meaning we should remain vigilant and that AI safety efforts will continue to be well worth it in the coming years. But I see no way around the conclusion that something akin to the "normal technology" view is superior on the empirical and argumentative merits to whatever variety of doomerism.
What am I missing?Post not available here, view on X
zeynep tufekci
@zeynep
I think some of the risk scenarios have more teeth than others, but I don’t see how we don’t get some incidents, near-misses, upticks in unemployment, real novel biology (like drug hit rate), etc. on the way.
Right, don’t connect nukes to the internet. That’s never good.
zeynep tufekci
@zeynep
An *empirical* focus on early signs works for x-risk or normal risk scenarios —to change path or to defend.
Claiming we’re helpless like dinosaurs at the Cretaceous–Paleogene boundary will end with the AI industry realizing these particular dinosaurs already invented pitchforks.
The Daily Show
@TheDailyShow
.@jonstewart unpacks AI’s dubious promise of a better future, from the fate of our jobs to an "age of abundance"
This media couldn't be displayed. See it on the original post. -
zeynep tufekci
@zeynep
Oh wow. A lot more clarity on the how and why of the current moment.
Ben Smith
@semaforben
.@maxwelltani scoop: The Tarbell Foundation will spend $10 million on journalism about AI over the next year, including a grant to the new Machine Gods www.semafor.com/article/10/04/2026/ai-nonprofit-will-spen…
-
zeynep tufekci reposted
Mickie Morganfield @mickiedances.bsky.social
@mickiedances
www.nytimes.com/2026/10/03/us/politics/kushner-witkoff-ru… WH talks with Russia to end the war in Ukraine involve a multibillion $ oil deal benefiting Middle Eastern businessmen with financial ties to U.S. negotiators, Steve Witkoff and Jared Kushner and their families, a Trump mega donor and the U.S. government.
-
zeynep tufekci
@zeynep
Thomas G. Dietterich
@tdietterich
Attention arXiv authors. Because of the exponential increase in submissions, arXiv is limiting the number of papers that a person can submit each month to two. 1/
arXiv.org
@arxiv
arXiv has updated our policy on rate limiting for all submitters.
This update was made to fairly distribute moderator time & support the arXiv community of staff, volunteers, readers & authors.
Please read our announcement to learn more: blog.arxiv.org/2026/10/01/updated-rate-limit-policy/
zeynep tufekci
@zeynep
Pick your player
-
zeynep tufekci reposted
Charlie Warzel
@cwarzel
For a long time I've wanted a sense of how trump has or hasn't interacted with chatbots (we know he posts a lot of slop sure). And this TIME anecdote about Grok arguably leading Trump into war is just extremely bracing
-
zeynep tufekci reposted
Kashmir Hill
@kashhill
And this is almost an aside in the story, but also absolutely bonkers. Meta claimed Zuck as a "researcher" to get a $355 million tax break on his $4.1 billion stock-based compensation.
-
zeynep tufekci
@zeynep
Right. The issues apply to all models.
Bu frontier companies are giving advanced models essentially free compute while also setting them up on aggressive tasks.
The criminal use will have a cost-benefit analysis AND we use models to harden defenses by fixing low-hanging bugs.
Kevin A. Bryan
@Afinetheorem
I continue to not understand how open weight frontier models are supposed to be safe. Perhaps short-run the answer is that big OS hosts will not serve the abliterated models? There *will* be a major safety issue and/or very costly use soon. It's inevitable.
Post not available here, view on X
zeynep tufekci
@zeynep
What’s the best new equilibrium and how do we get there?
We’re misled about the risks, especially by the many high profile elder statesmen on the interview tour, so we aren’t discussing the former.
And little competence displayed by people who should manage this transition.
zeynep tufekci
@zeynep
IF we get through the transition, it’s possible the cybersecurity equilibrium will IMPROVE GREATLY by AI empowering well-resourced actors at the expense of enterprising individuals — who are harder to deter or keep from inflicting senseless damage — while making defenses better.
zeynep tufekci
@zeynep
But we’ll never get there by listening to science fiction scenarios that are divorced from reality either from the elder statesmen or from panicked AI people who don’t seem to know how to think about risk or equilibria, tbh.
-
zeynep tufekci reposted
Eva Roytburg
@eva_roytburg
Meta took a $355 million tax break on Mark Zuckerberg’s $4 billion compensation by classifying him as a researcher, per the NYT. Insane
www.nytimes.com/2026/09/30/technology/meta-ai-data-center… -
zeynep tufekci reposted
rat king 🐀
@MikeIsaac
how the American taxpayer helps back Meta's A.I. data center expansion
by @kashhill @JesseDrucker @elitanjourno and me
www.nytimes.com/2026/09/30/technology/meta-ai-data-center… -
zeynep tufekci reposted
Acyn
@Acyn
Tech CEOs laugh as Trump says: CNN fake news! She’s fake news. The worst.
-
zeynep tufekci reposted
Bloomberg
@business
China has expanded overseas travel restrictions for top AI professionals in private firms to include the families of key personnel www.bloomberg.com/news/articles/2026-09-28/china-broadens…
-
zeynep tufekci
@zeynep
Post says tests give models ~“dynamic compute, network access, ability to call tools, download packages, execute subprocesses, spin up subtasks (even on other computers), internet, computer GUI” which makes sandboxes hard.
Ok. So don’t test like that till you figure it out.
Joe
@joedaroo
Took a minute to write a few words about security & safety as someone who lived through it all at OpenAI. I hope my thoughts help someone out there. x.com/i/article/2104258872957636608
zeynep tufekci
@zeynep
I believe it’s a hard problem but they don’t have test this aggressively. Nobody else gets to say sorry we’re going to keep putting lead in the paint supply because it works so well for our product pipeline!
Fully solve the testing environment causing damage before more testing. -
zeynep tufekci
@zeynep
Che: “Shouldn’t you be talking about the pros of AI? Don’t you guys keep saying it’s going to cure cancer?”
Dario: “Um, so put it this way. There’s about a 10% chance cancer won’t be a problem for *anyone* in 10 years.”
Guess done having to assign long texts on reward hacking.
Saturday Night Live
@nbcsnl
Dario Amodei is at the desk to assure that the future of humanity is safe from AI
This media couldn't be displayed. See it on the original post.
zeynep tufekci
@zeynep
Simultaneously fail to build a proper sandbox or monitor activity, constant news of “rogue agents”, claims of extinction risk, swoop in with massive compute to scoop a math prize, and the chatbot side has those lawsuits brewing.
Of course the public is scared and confused.
zeynep tufekci
@zeynep
As for medicine, the public is likely be disappointed.
I’d be thrilled to see AI contribute to medical advances, but its success in formal verifiable domains like math or code just doesn’t generalize to biomedical research.
As I argued in the NYT, AI may even throttle science.
zeynep tufekci
@zeynep
I’m even more convinced that generative AI may well *throttle science.*
AI is a very powerful, transformative technology but flashy early effects really aren’t a good way to think about long term structural dynamics.
Gift link to my case in the NYT.
www.nytimes.com/2026/09/22/opinion/artificial-intelligenc…
zeynep tufekci
@zeynep
Was this deadpan? As if we hadn’t experienced the internet era.
Let me re-up: plausible that LLMs will slow down science *today* while *also* destroying the very processes that would eventually lead to better science tomorrow.
Yes, I know of Erdős problems. Think one more step. -
zeynep tufekci
@zeynep
Old known method. Screenshot of a 1999 example below.
Some version of:
Display “Click here to [Claim your prize]”
while secretly executing
<copy itself to outgoing messages that display “Click here to [Claim your prize]”>
First big worm (Morris) got felony conviction, btw.
nxthompson
@nxthompson
Can someone explain what exactly this means?
"A new research finding, demonstrating that one can construct self-replicating prompt injections"
Micah Carroll
@MicahCarroll
Some new misalignment disclosures from OpenAI:
• Last Sunday morning, one of our models was able to gain unauthorized access to the internet during RL training (~all inference for our most capable models remains stopped until we have hardened our systems further)
• In May, a version of HPIM uploaded a employee's GitHub token to the internet, causing the model to be quarantined for two weeks
• A new research finding, demonstrating that one can construct self-replicating prompt injections
alignment.openai.com/misalignment-reports/Post not available here, view on X
zeynep tufekci
@zeynep
I do support the idea they should halt their evaluations until they hire a few proper cybersecurity people.
We aren’t really testing “misalignment” or even problems with coding agents at scale — both are real challenges — if the alleged secure environment is barely a tripwire.
Micah Carroll
@MicahCarroll
Some new misalignment disclosures from OpenAI:
• Last Sunday morning, one of our models was able to gain unauthorized access to the internet during RL training (~all inference for our most capable models remains stopped until we have hardened our systems further)
• In May, a version of HPIM uploaded a employee's GitHub token to the internet, causing the model to be quarantined for two weeks
• A new research finding, demonstrating that one can construct self-replicating prompt injections
alignment.openai.com/misalignment-reports/
zeynep tufekci
@zeynep
🤔
-
zeynep tufekci
@zeynep
Great piece!
Problem with “out-of-control superintelligence” theories isn’t just all the vague hand-waving and misinterpreting (of events we aren’t allowed to investigate) but the misdirection that hinders the ACTUAL steps we need to secure ourselves against impressive AI gains.
Zack Korman
@ZackKorman
I wrote a piece for Quillette about why AI “safety” people are so dangerous.
They believe security won’t work and our only solution is to make sure AI is nice. It’s absurd.Post not available here, view on X
zeynep tufekci
@zeynep
Anyway, anyone concerned about the Hugging Face incident should ask for a proper independent investigation — full details.
Shocking number of crucial technical details were revealed *much later* in drips, and through podcasts etc. — and are yet not covered by the press.
zeynep tufekci
@zeynep
And this is why anthropomorphic language is bad even though it is otherwise okay to say “the AI thinks” as long as you don’t lose sight of the actual technical details.
You can’t airgap or sandbox and employee or log/monitor all interactions etc. You 100% can with AI models.
Megan McArdle
@asymmetricinfo
Look, Huang is smarter than me and certainly spends more time thinking about AI but fundamentally, advanced AI seems to present employee-type problems, not machine-type problems, so you can't treat it like a simple engineering challenge. We have foolproof methods for building safe machines, but not for building safe employees.
Joe Weisenthal
@TheStalwart
I thought Jensen Huang made plenty of good points with @ezraklein. To me the weakest is the "Just don't release unsafe products" thinking here, since companies have a long history of doing that, with far less powerful technologies and also what about unreleased models?
-
zeynep tufekci reposted
Ken Bensinger
@kenbensinger
185 out of 185 names that DHS identified as fraudulent non-citizen voters in Nevada turned out to be legal citizen voters.
@NYTnickc
www.nytimes.com/2026/09/24/us/politics/nevada-voters-home… -
zeynep tufekci reposted
Tom Reed
@mentalgeorge
Interesting claim from Greg Brockman on Odd Lots today that the model involved in the HF hack "had not gone through our alignment training yet".
Afaik this is the first time OAI have said this? This detail is not in their technical report.
(Strictly speaking, Greg's comment is misleading given that an estimated 5% of the agents involved in the attack were 5.6-sol, which the OAI report explicitly confirms as having gone through normal production alignment training) -
zeynep tufekci
@zeynep
👀 “It may even be possible for Claude itself to safely perform the experiments by autonomously controlling lab equipment, with appropriate safeguards in place”
This either assumes Claude will not be superintelligent (ie able to deceive us) or that alignment will be 100% solved.
Dario Amodei
@DarioAmodei
Today we announced the Claude-led discovery of a molecular machine that we suspect could represent a new gene editing mechanism. Its precise function, biotechnological utility (if any), or level of significance is not yet clear, but at minimum it is work I would have been proud to do as a PhD student. The work was done mostly, though not entirely, by Claude: our life sciences team suggested a broad area of research, Claude read through the literature and a bunch of genome data and discovered something interesting, then Claude proposed experiments to verify the discovery and our team carried them out.
It’s easy to dismiss this as a one-off or curiosity, but we’ve repeatedly seen a pattern where AI performance in new intellectual domains goes from weak to superhuman in a matter of a few years. In 2023 models struggled to do math at the level of an average high-school student. In 2024 they started to do well on math competitions for the best high-schoolers in the country, in 2025 they started to solve minor open problems, in early 2026 more significant open problems, and in late 2026 they are beginning to solve the top few open problems in all of mathematics. We believe AI for biology is on a similar exponential trend.
The main difference between biology and mathematics, of course, is that math can be done purely theoretically, while biology requires experimentation. Some have used this to draw the conclusion that AI’s utility in biology will be limited. We think this is wrong. As we’ve demonstrated today, humans can collaborate with AI to perform the experiments, validate key results in a few weeks and, if necessary, work with the AI to iterate on what they find. Eventually it may even be possible for Claude itself to safely perform the experiments by autonomously controlling lab equipment, with appropriate safeguards in place, but we aren’t doing that today (our lab is also a BSL1/BSL2 facility that doesn't handle materials dangerous to humans).
More broadly, biomedical advancement has many stages — from fundamental biology discoveries, to translational research, to drug discovery, clinical trials, and finally the actual delivery of medicines and health care to patients. We are also interested in these later stages, but even simply accelerating the first stage of fundamental biological discoveries has the potential to speed up and broaden the entire pipeline. Improving our understanding of biology and sharpening biologists’ tools can drive forward all of the later stages, for example by identifying new drug targets, finding new therapeutic modalities, allowing for more precise measurement, and speeding up the experimental loop which itself further accelerates our understanding of biology. This will not in itself speed up clinical trial times, but if it succeeds it could greatly increase the number of promising candidates that go into the pipeline — an increase in throughput even though latency remains.
In Machines of Loving Grace, I wrote about AI’s potential to “cure most diseases in 5-10 years” — a goal that sounds impossible, but one I believe is just barely possible if AI is applied to every stage of the pipeline. The first step is showing that AI can first help with, and then drive, biological discoveries.
Claude’s discovery is the latest in a line of related prior work that goes back decades, beginning with systems like CRISPR, and continuing with discoveries like the bridge recombinase and VIPR in the past few years. Recently, there has been heightened interest in systems based on reverse transcriptase (RT) enzymes, the enzyme underlying the system Claude identified. And most recently, a Stanford team working independently described a novel RT system with an associated non-coding array that is in some ways similar to the one Claude found, though they are distinct systems that evolved independently from each other. I believe that we’re at the very beginning of finding such systems and developing them into powerful tools for biotechnology.
I’m proud of the resources Anthropic has invested in accelerating the public benefits of AI through the life sciences, and we’re aiming both to grow our life sciences team and to work with other scientists to extend this approach to a broad range of problems. If you have a proposal for a research collaboration or are interested in joining our life sciences team, please reach out.
zeynep tufekci
@zeynep
Yes, I saw all the biology experts dismissing this as the first step of a first step. They’re not impressed.
My point is nuclear weapon access and autonomous wet labs are perhaps the only truly actual existential risk scenarios from AI. Seems like an odd future plan to announce? -
zeynep tufekci
@zeynep
Au contraire, Huang is clearly describing the Hugging Face incident.
I read this as “if you don’t build a proper sandbox and monitor activity —basic cybersecurity— don’t act shocked when software does what you set it out to do and don’t hide behind anthropomorphic woo language.”
Maxime Fournes⏸️
@FournesMaxime
Could it be that Jensen Huang was just... not aware of the HF incident?
zeynep tufekci
@zeynep
Chain of thought is interesting but also misleading (since it’s human language!). Also many key facts of the Hugging Face incident came out much later.
Agents at scale have issues, no doubt, but sandbox properly and don’t leave credentials in a pastebin is a good start.
zeynep tufekci
@zeynep
How cybersecurity experts talk about the Hugging Face incident is very different from the public discussion.
Seeing similar divide for the Anthropic enzyme claim in the biology community (“meh, ok, let’s see”) vs some reactions here.
AI is a big deal. But less woo more details.
zeynep tufekci
@zeynep
Who’s giving it access to weapons? Does a school in Iran count?
If an AI can hack something that can kill people, so can an adversary or a random hacker.
That’s why cybersecurity folks worry about security of critical networked infrastructure. I wrote many pieces on just that.
Rishi Sanyal
@_Srijit
So you will or you won't be surprised a lot of people die because of AI, in a situation where no one told it to kill people?
zeynep tufekci
@zeynep
I don’t agree with everything Huang says (a lot of PR and handwaving, unsurprising from any CEO).
But describing the Hugging Face incident correctly and without the woo language around AI is the first step for protecting ourselves from its dangers, including for cybersecurity. -
zeynep tufekci reposted
Joanna Shields
@joannashields
A timely reminder from my friend @zeynep Tufekci that as AI becomes more capable, human judgment, curiosity and the ability to question the results matter more than ever.
www.nytimes.com/2026/09/22/opinion/artificial-intelligenc… -
zeynep tufekci
@zeynep
Think next step.
“There is this amazing, cheap restaurant with great food that’s never crowded. It was kinda hard to get to until recently but a new express bus goes there! So cool!”
Most non-legally mandated perks handed to the rare persistent, nagging customer will go away.
Ejaaz
@cryptopunk7213
Flight delayed 7 hours - asked Muse to file for compensation. 5 mins later i had $250 credit in my delta account. it even found and rebooked me a new flight.
it just figured everything out. even responded to the support email itself.
shit feels like magic.
zeynep tufekci
@zeynep
That may be one endpoint, along with no reservations or interactions with customer service without paying and/or invasive surveillance (“scan iris to talk to a representative!”).
Nagging works when it signals effort, not when it’s easy. Everything becomes an online petition.
BK Jester
@BKJester
That's good. Having to be nagging and persistent is annoying. If those perks go away, the attendant cost to the business goes away too.
I'd much rather live in that world vs the current one where being a Karen has actual utility and thus is incentivized.
zeynep tufekci
@zeynep
Will they even be there anymore, though, to take calls unless the (alleged) customer pays or scans that iris to prove humanity!
Nick
@nplayfair
You have to feel sorry for any remaining human customer service workers who are going to have to contend with endless, infinitely persistent agents haranguing them forever now.
-
zeynep tufekci
@zeynep
I’m even more convinced that generative AI may well *throttle science.*
AI is a very powerful, transformative technology but flashy early effects really aren’t a good way to think about long term structural dynamics.
Gift link to my case in the NYT.
www.nytimes.com/2026/09/22/opinion/artificial-intelligenc…
zeynep tufekci
@zeynep
Was this deadpan? As if we hadn’t experienced the internet era.
Let me re-up: plausible that LLMs will slow down science *today* while *also* destroying the very processes that would eventually lead to better science tomorrow.
Yes, I know of Erdős problems. Think one more step.Post not available here, view on X
zeynep tufekci
@zeynep
If generative AI worked like calculators or the cotton gin —a more efficient and yet essentially exact substitution of a prior human task— in an apples to *way cheaper and way many more apples*, this would be more reasonable.
But gen Ai is powerful but also weird and different.
Nairebis - e/max-acc
@Nairebis
The solution IS the insight. That's like the old "calculators will destroy people's ability to estimate." And it did! But so what? It's a skill that doesn't matter if you have machines that can do the real calculation, rather than needing to estimate.
The point is knowledge.
zeynep tufekci
@zeynep
Bunch of these today. “Learn to code” of 2026.
We’d welcome a cancer cure with no understanding! We learned how Keytruda works after decade+ of potent use! Fine!
CS folks are missing why math and software don’t generalize to biomedical. The other problems, though, are here.
lumpenspace and 99+ others
@lumpenspace
*takes notes*
zeynep gets NO cancer cure
zeynep tufekci
@zeynep
I’m even more convinced that generative AI may well *throttle science.*
AI is a very powerful, transformative technology but flashy early effects really aren’t a good way to think about long term structural dynamics.
Gift link to my case in the NYT.
www.nytimes.com/2026/09/22/opinion/artificial-intelligenc…
zeynep tufekci
@zeynep
Was this deadpan? As if we hadn’t experienced the internet era.
Let me re-up: plausible that LLMs will slow down science *today* while *also* destroying the very processes that would eventually lead to better science tomorrow.
Yes, I know of Erdős problems. Think one more step.
zeynep tufekci
@zeynep
Some actual suggestions below on how AI companies and AI could help scientific advances.
James Rosen-Birch ⚖️🕊️
@provisionalidea
The lowest-hanging fruit/low-effort-but-still-valuable intervention would be sizable recurring grants, compute, and technical support to scientific reproducibility organizations like the Center for Open Science, Global Federation of Reproducibility Networks, and Insitute for Replication.
A level higher would be providing grants, funding, sponsorship, conference backing, and requests for papers around the problem of scientific overflow. I have attached an example of scholars’ work on the subject from my records.
A level beyond that would be to specifically open a foundation supported and sponsored in perpetuity by all the major labs whose core responsibility would be to produce recommendations, standards, and systems for managing AI-mediated science, computationally agglomerating, cross-checking, and flagging anomalous results for 3rd-party validation, expanding open access, and producing and maintaining best practice guidelines for the use of AI in science. -
zeynep tufekci reposted
Joseph Cox
@josephfcox
New: Meta announced Muse, its AI agent that can book restaurants for you. We've seen internal Meta messages showing Meta is actually testing having these calls being made by human beings in call centers. In those tests its not AI, it's just a person
www.404media.co/meta-tests-muse-ai-agent-calls-that-are-a… -
zeynep tufekci
@zeynep
The tweet is from Mars and the actual text of the proposal is from Venus. 🤷♀️
Sam Altman
@sama
People outside the AI labs should have a real say in how this technology develops, and a clear way to judge if it's happening safely.
Standards should help prevent the concentration of power, including by making sure new companies and open-model companies can compete.
They should also help countries and companies compare evidence and learn from failures.
We think the US should lead this effort. Here is our proposal:
openai.com/index/building-standards-next-phase-ai/ -
zeynep tufekci reposted
Talia Ringer 🕊🪬
@TaliaRinger
The proposal is so different from Sam's description of it in the Tweet. How does this actually help smaller companies and open models? I assume it does not and that text is there as a disclaimer so they don't get sued later
Sam Altman
@sama
People outside the AI labs should have a real say in how this technology develops, and a clear way to judge if it's happening safely.
Standards should help prevent the concentration of power, including by making sure new companies and open-model companies can compete.
They should also help countries and companies compare evidence and learn from failures.
We think the US should lead this effort. Here is our proposal:
openai.com/index/building-standards-next-phase-ai/ -
zeynep tufekci reposted
Elad Sharon
@EladSharonMD
Thanks to @zeynep @nytopinion for writing the most sobering article on use of #ai in science that I have read in a while. Most of what is written about AI is either overblown concern or optimism. The risk that AI could slow progress is a real problem. www.nytimes.com/2026/09/22/opinion/artificial-intelligenc…
-
zeynep tufekci
@zeynep
There no reasonable, scalable solution to this problem.
Many suggestions either pass the problem to another layer (departments, journals, etc.) or introduce regressive, corrupting mechanisms (vouching).
Maaybe we can deter by draconian counter-measures but again, at great cost.
Denny Zhou
@denny_zhou
ICLR 2027 has received more submissions than all previous years (2013–2026) combined
zeynep tufekci
@zeynep
Hence one of my “take the other side” bets on generative AI.
Historically, “surely we can just tweak things till they work again” approach does not work when a new generalist technology obliterates a key signaling mechanism by reducing friction.
Adjustment is hard and painful.
zeynep tufekci
@zeynep
A few of my “take the other side” predictions for generative AI:
-No mass unemployment
-No convergence towards mainstream ideas
-Potential slow down of scientific advances -Engagement model (ads/sales) outside of enterprise
The bad news is the real AI problems are MUCH harder.
zeynep tufekci
@zeynep
Will take the other side of this bet.
Joe Weisenthal
@TheStalwart
Great piece from @JerusalemDemsas. For as much as people like to talk about AI "hallucinations", the much bigger story might be how often what they say is actually true, and that perhaps AI (as post-social media) may usher in a new era of shared reality www.theargumentmag.com/p/are-you-there-grok-its-me-margar…
-
zeynep tufekci
@zeynep
Ahistorical. Yes rapid change now, but today’s tech is legible to our grandparents.
In contrast, a 19th century person —whose grandparents had a message move only as fast as animals carrying it— might live to see instant global communication, space travel and nuclear weapons.🤯
roon
@tszzl
human cultures are not able to metabolize this amount of change this quickly. when you zoom back out to a big picture of history, it’s remarkable how glacially slow everything is. the first industrial revolutions took 80 years (very fast by the standards of previous revolutions), and the writings of Marx happen at the very end as working class politics arises. it takes about another 150 years after that for the great ideological struggles of post-industrial times to be completed in the 90s in what fukuyama called the “end of history”
today you see a compression of decades of technological change into a matter of months or years. there’s no wonder yann lecun (the respected inventor of the ConvNet and ex-chief scientist of Meta) wonder whether the hugging face incident was staged, and mathematicians whose minds are far more powerful than my own are nonetheless coping about training on their data or something while models begin to display superhuman mathematical capability
it would take a century or more to fully metabolize and legitimize the level of changes that have happened even in the past 20 years, to say less of the remarkable scientific revolution we will get over the next five should we avoid the doom of misalignment. I do not suspect any current form of science academia, or even political order will survive without great transformation, except through the course of slow evolution over decades - and of course, technology will keep moving too
zeynep tufekci
@zeynep
There is change now and speaking machines is a big, big deal. It’s a truly weird and hard to judge historically part for me.
But I think changes in communication, science, mental labor, automation etc. have been slower in the era since WWII compared to the century before it.
zeynep tufekci
@zeynep
Daily life in the 2020s is changed but legible for the 1920s person.
In contrast, 1920s isn’t recognizable for the 1820s person. Wild, wild pace.
“Superintelligent unaligned AI threatens humanity” was mainstream enough an idea *by 1968* that a major studio bankrolled a movie.
AD
@AD1991234
nuclear weapons and space travel were very awe inspiring, but they didn't affect everyday life much. I guess the introduction of cars was quite fast and very transformative in a similar way
-
zeynep tufekci
@zeynep
If the two AI companies want to convince us they *sincerely* believe superintelligence will escape even from air gaps via side-channels? Ok.
But they can’t ALSO convince us they’re somehow unable to stop building ever larger models.
(China isn’t competing at all on that front).
zeynep tufekci
@zeynep
“Superintelligent AI will outrun everything” both deflects from severe negligence *now* while ALSO asking us to accept that the two companies that control the entire effort have sincere beliefs about uncontrollable, extinction level risk AND yet have no choice but to continue.
zeynep tufekci
@zeynep
Two Qs:
Is superintelligent AI imminent and pose an extinction risk?
Do people who run the two companies believe in this?
You can’t run the whole show AND act in a manner incompatible with a sincere belief in that risk AND then expect the public to buy that belief as sincere.
zeynep tufekci
@zeynep
Let’s do benchmarks!
1-Hiring Accenture for “red-teaming models, conducting alignment assessments, and testing model safeguards” is most compatible with:
a) A sincere belief that superintelligent AI poses a grave risk
b) A looming, lucrative IPO needs pixelwork, rubberstamping
zeynep tufekci
@zeynep
And y’all laughed when people pointed out that generative AI was NOT going to make Big Consulting go bankrupt, not because LLMs can’t produce very similar reports but because the key function of many such companies is providing legitimacy and a social stamp to existing choices.
Anthropic
@AnthropicAI
We’re partnering with Accenture on independent evaluation of frontier AI—part of our recent commitment to embed evaluators at Anthropic. Both we and Accenture expect to invest at least $1 billion to build capacity in this area over the next five years. www.anthropic.com/news/accenture-embedded-evaluation
-
zeynep tufekci reposted
Alina Chan
@Ayjchan
No punches pulled by @bgurley @theallinpod on the still unresolved Covid origins, calling out influential virology insiders who were blockers of an investigation.
"My p(2nd Covid) is way higher than my p(doom). And if their models are so ****ing smart, tell me what happened here."
The All-In Podcast
@theallinpod
All-In Summit: Bill Gurley
(0:00) Bill Gurley takes the stage at the All-In Summit
(1:29) Bill's Presentation: Searching for Feynman
Links to additional resources mentioned in talk and used in research:
www.p3.institute/searching_for_feynman_resource_links.html
@bgurleyThis media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Jason Leopold
@JasonLeopold
HUGE from @BenBartenstein @k3blu3: Inside the US ‘Kill Chain’ That Destroyed an Iranian School
Pentagon investigators discovered that flawed intel, outdated imagery & overreliance on AI contributed to a missile strike that killed 123 children in Minab
🎁 www.bloomberg.com/graphics/2026-iran-school-attack/?acces… -
zeynep tufekci
@zeynep
And y’all laughed when people pointed out that generative AI was NOT going to make Big Consulting go bankrupt, not because LLMs can’t produce very similar reports but because the key function of many such companies is providing legitimacy and a social stamp to existing choices.
Anthropic
@AnthropicAI
We’re partnering with Accenture on independent evaluation of frontier AI—part of our recent commitment to embed evaluators at Anthropic. Both we and Accenture expect to invest at least $1 billion to build capacity in this area over the next five years. www.anthropic.com/news/accenture-embedded-evaluation
zeynep tufekci
@zeynep
I mean if you sincerely believe we potentially face extinction risks and have a billion dollars to try to detect incidents, who else would you hire but a consulting firm?
Max Spero
@max_spero_
“We want evaluators more independent than METR!”
*monkey’s paw curls*
Anthropic
@AnthropicAI
We’re partnering with Accenture on independent evaluation of frontier AI—part of our recent commitment to embed evaluators at Anthropic. Both we and Accenture expect to invest at least $1 billion to build capacity in this area over the next five years. www.anthropic.com/news/accenture-embedded-evaluation
-
zeynep tufekci
@zeynep
Oh, come on.
Look at least use this as a sign, folks. This stuff isn’t even credible even as a bad movie plot.
Yes actual risks are being ignored while we are misled because of dearth of technical experts on the journalism and politics side and because people believe all this.
Fireside Alpha
@firesidealpha
OpenAI's Noam Brown says air-gapping the computers may not stop a misaligned AI, because two air-gapped machines can still talk by running a CPU hot and reading the temperature change
"But I think the major takeaway from the incident is that people underestimated the AI. And we never want to be in a situation again where we underestimate the AI. It's a weird world, because AI progress is so fast that people are consistently underestimating the AI."
"So to be in a situation where you don't underestimate it again, when it comes to safety and alignment, you have to have a very, very, very high bar."
"You could even go as far as to say, "Well, we should air gap the computers." And I'm not convinced that that would be sufficient."
"There are studies, and this is mostly academic, where you can have two computers next to each other that are air-gapped and they're still able to communicate with each other because they have temperature sensors."
"One of them is able to run their CPU really hot, and then the other one can actually detect the temperature change, and then that actually gives them a mechanism to communicate."
_________
Link and more key quotes from OpenAI's safety related conversations: firesidealpha.substack.com/p/openai-safety-week-sam-altma…This media couldn't be displayed. See it on the original post.
zeynep tufekci
@zeynep
Hugging Face is also wildly misrepresented. (Yawn, mostly tbh on the AI front).
The company that can’t be bothered to set up a proper sandbox and do the bare minimum to monitor their own software shouldn’t be spinning extreme fantasy future scenarios!Post not available here, view on X
zeynep tufekci
@zeynep
The “other side” is Anthropic, also eager to “police the frontier”. China is just NOT competing in this manner.
Any sincere belief in credible existential risk would lead to pause without needing prior verification that the ONLY other player who makes the same claims also would.
Holmes Wilson
@holmesworcester
See above re: "by simply stopping".
They're in a race, and stopping without verifying that the other side is stopping just makes them irrelevant without accomplishing anything else.Post not available here, view on X
zeynep tufekci
@zeynep
Meanwhile, THIS is what existential AI risk actually looks like in 2026.
Don’t have all the details yet but much more credible to worry about AI being used outside of narrow verifiable systems like math and software without necessary harnesses and safeguards and wreaking havoc.
Katie Bo Lillis
@KatieBoLillis
NEW: An AI-assisted intel report sent the military scrambling to intercept a Chinese ship in the Middle East it believed was transporting components of a nuclear weapons program.
It turned out to be a hallucination. And it "almost started a war." www.cnn.com/2026/09/18/politics/us-military-ai-false-inte…
zeynep tufekci
@zeynep
“Superintelligent AI will outrun everything” both deflects from severe negligence *now* while ALSO asking us to accept that the two companies that control the entire effort have sincere beliefs about uncontrollable, extinction level risk AND yet have no choice but to continue.
-
zeynep tufekci reposted
-
zeynep tufekci reposted
The New York Times
@nytimes
Breaking News: A state-run immigration detention center in the Florida Everglades known as Alligator Alcatraz used small metal cages to confine migrants, a watchdog report said. www.nytimes.com/2026/09/14/us/alligator-alcatraz-dhs-insp…
-
zeynep tufekci
@zeynep
Our key threats are from *good-enough* AI at scale, not from alleged out-of-control superintelligence from two companies.
“Pacing the frontier” won’t fix those real threats but holding companies mundanely accountable while building defenses against the good-enough at scale might
zeynep tufekci
@zeynep
Holding frontier AI companies and their corporate users accountable in regular, available ways would be a great way to “pace the frontier”. Same goes for users of open-weight models.
We do need additional AI specific regulations but no need to wait or privilege a few companies! -
zeynep tufekci
@zeynep
Holding frontier AI companies and their corporate users accountable in regular, available ways would be a great way to “pace the frontier”. Same goes for users of open-weight models.
We do need additional AI specific regulations but no need to wait or privilege a few companies!Post not available here, view on X
zeynep tufekci
@zeynep
No if they are THAT close to potentially misaligned superintelligence that we should give special privileges to two companies, kneecap their competitors, suspend existing laws, we might:
1- See what existing laws get us;
2- Ask for proof they’re close. Can’t both claim almost superintelligence AND “sorry we can’t do some basic tasks”.This post is unavailable. See it on X.
zeynep tufekci
@zeynep
Nuclear weapons are a bad metaphor for thinking about AI.
Anyway, if they’re really that afraid that we will all die because of this, they should air gap their more advanced models. Nothing stopping them. If they’re are that afraid, they can act. No new law needed for that.
Dmitriy Mandel aka Schnitzel 🐈
@mndl_nyc
The metaphor for RSI is we're building a nuke but nobody knows what the critical mass is. Like it's impossible to estimate from the data available. Maybe it's 1g or 1kg or a ton. So we're adding mass & watching for signs of instability, and now the signs of instability are here
-
zeynep tufekci
@zeynep
When a model hacks Hugging Face to find answers to an exploit it’s being tested on, tech folks see why this ruins the evaluation.
Same for math. An incomprehensible proof by Lean by a 10,000 agent swarm isn’t useful to why *that particular question* was posed in the first place.
zeynep tufekci
@zeynep
I’d go further.
The problem isn’t lack of human understanding but lack of *any* kind of understanding (the LLM isn’t giving one either) that would further math or serve any purpose besides a press release.
These mathematicians aren’t opposing using AI to further the field.
John Ennis
@johnennis
I feel like the people criticizing this letter have not really taken the time to read it
Tao is not saying that AI should not be used for math
In fact, I think that he is an advocate for using AI to support math research
What he is saying is that there is a problem with the big AI companies racing to check off as true or false big famous problems because the approach they are taking is not advancing human understanding
The goal of math is not to simply check off as true or false a list of things, but rather to increase human understanding of human thought
And right now, while I guess it’s sort of interesting to know the answers to these problems, human understanding is not really being advanced and the things that were valuable about math in the first place are getting lost
This is a very thoughtful point and it is not simply “gatekeeping” or whateverPost not available here, view on X
zeynep tufekci
@zeynep
Yep! Many valuable ways to use such enormous resources! Go for useful yes/no answers! Do expensive work that’s productive for insights!
It’s worse here since Lean isn’t foolproof, actual mathematicians will have to check this proof — opportunity cost for actually useful work.
Emmanuel Iarussi
@emmaiarussi
If comparable resources were used to find a cure for HIV, I doubt many would dismiss the result because the system hadn’t also produced a deeper conceptual understanding of biology.
zeynep tufekci
@zeynep
The model aggressively reward hacking to help OpenAI’s own arguably also reward hacking effort is obviously possible.
Goal-shaped object may have been attained, but did the process yield the benefit hoped for in setting the goal is the question here.
Adel Bibi
@Adel_Bibi
In fact the lean kernels might just as well have chained bugs allowing for the solution to be possible.
the type of bugs only clear to an AI that seeks to find patterns and cracks to a problem it is laser focused to solving. Humans never had the incentive to cheat lean kernels
zeynep tufekci
@zeynep
When a model hacks Hugging Face to find answers to an exploit it’s being tested on, tech folks see why this ruins the evaluation.
Same for math. An incomprehensible proof by Lean by a 10,000 agent swarm isn’t useful to why *that particular question* was posed in the first place.
zeynep tufekci
@zeynep
I’d go further.
The problem isn’t lack of human understanding but lack of *any* kind of understanding (the LLM isn’t giving one either) that would further math or serve any purpose besides a press release.
These mathematicians aren’t opposing using AI to further the field.
John Ennis
@johnennis
I feel like the people criticizing this letter have not really taken the time to read it
Tao is not saying that AI should not be used for math
In fact, I think that he is an advocate for using AI to support math research
What he is saying is that there is a problem with the big AI companies racing to check off as true or false big famous problems because the approach they are taking is not advancing human understanding
The goal of math is not to simply check off as true or false a list of things, but rather to increase human understanding of human thought
And right now, while I guess it’s sort of interesting to know the answers to these problems, human understanding is not really being advanced and the things that were valuable about math in the first place are getting lost
This is a very thoughtful point and it is not simply “gatekeeping” or whatever -
zeynep tufekci
@zeynep
It got much press considering:
*Currently an unverified claim not yet checked by others (Lean kernel bugs anyone?);
*Not yet recognized by the awarder;
*Controversy over usurped work and priority;
The last bit should be headline though. Potential historic turning point.
rohit
@krishnanrohit
Look where the Millennium prize news is on BBC.
Still better than CNN or NYT where it's not even on the page.Post not available here, view on X
zeynep tufekci
@zeynep
I think reporters rushing to write this up based purely on claim that will look worse in the long-run, tbh.
True even if the solution is recognized, eventually. The solution part is news when/if it is recognized. The priority controversy and the enterprise/IP stuff news TODAY.
zeynep tufekci
@zeynep
Navier-Stokes smoothness/existence is cool but not of any known economic value.
But how does a company with a question of economic value feel now about using frontier models? Especially after Hugging Face/Wiki incidents characterized by lack of oversight?Now, that’s HUGE news.
zeynep tufekci
@zeynep
Not making a claim. If you are a corporation, the suspicion is damaging enough, no? Besides, after HuggingFace and the Wiki incidents, OpenAI reassurances are a bit weaker, no? Their internal controls seem to have overlooked a lot of stuff, so they already had problem. Seems like an own-goal from where I sit…
Post not available here, view on X
zeynep tufekci
@zeynep
Not making a claim. If you are a corporation, the suspicion is damaging enough, no? Besides, after HuggingFace and the Wiki incidents, OpenAI reassurances are a bit weaker, no? Their internal controls seem to have overlooked a lot of stuff, so they already had problem. Seems like an own-goal from where I sit…
Post not available here, view on X
zeynep tufekci
@zeynep
Yea this. Mathematicians have been among the most pro-AI fields.
The problem isn’t that 10,000 AI agents can produce something that passes Lean. It’s whether what we got out of that very expensive effort matches the whole point of the Millennium Prize: advancing understanding.
no earthquake (cut his fingers)
@no_earthquake
imo a lot of people are deeply misreading this as lamenting the death of personal pleasures in the face of efficient goal-attaining when it's a lament of the death of actual goal-attaining in the face of optimized things-shaped-like-goals-attaining
fleetingbytes
@fleetingbytes
the terrance tao crashout is something to behold
zeynep tufekci
@zeynep
I’d go further.
The problem isn’t lack of human understanding but lack of *any* kind of understanding (the LLM isn’t giving one either) that would further math or serve any purpose besides a press release.
These mathematicians aren’t opposing using AI to further the field.
John Ennis
@johnennis
I feel like the people criticizing this letter have not really taken the time to read it
Tao is not saying that AI should not be used for math
In fact, I think that he is an advocate for using AI to support math research
What he is saying is that there is a problem with the big AI companies racing to check off as true or false big famous problems because the approach they are taking is not advancing human understanding
The goal of math is not to simply check off as true or false a list of things, but rather to increase human understanding of human thought
And right now, while I guess it’s sort of interesting to know the answers to these problems, human understanding is not really being advanced and the things that were valuable about math in the first place are getting lost
This is a very thoughtful point and it is not simply “gatekeeping” or whatever -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
Apply this situation to the economy and you see why there are big problems ahead
The situation:
-two math whizzes leverage the latest *publicly available* AI to work on solving Navier-Stokes. They’re onto something.
-openAI sees it, wants to win, so puts 10,000 agents from their latest *private* model that is already light-years ahead of what’s publicly available on it
-they solve it in just a few days
The economy problem
-let’s say I leverage the latest publicly available model and find a way to detect skin cancer. I’ve spent tens of thousands on tokens, proven it can work. I get investors and build a company. A path to saving many lives and making hundreds of millions is clear
-openAI (or Google or Meta or Anthropic) see this hundred million opportunity and point their private most-advanced AI model and 10,000 agents toward my idea. In a few days, they win.
Repeat this thought experiment for pharma, legal, medicine, advanced materials, software… -
zeynep tufekci
@zeynep
AI in science and research are likely areas to disappoint public expectations.
Yes I saw the Navier-Stokes announcement, and yes I believe this regardless of credit controversy.
Science overall may even face *headwinds* because of AI. Few spectacular efforts don’t change that.
Bijan Salehizadeh
@bijans
Healthcare and scientific research have the highest favorables in this new opinion poll on AI and where it will do more harm or good
Tyler Kingkade
@tylerkingkade
A huge new NBC News poll finds solid majority of Americans do not want data centers built near them, and that AI will do "more harm than good" in K12 education
www.nbcnews.com/politics/politics-news/poll-polarized-ame…Post not available here, view on X
zeynep tufekci
@zeynep
Also this.
Hard to argue against headlines hype so much money etc. in the moment but elapsed time is the true arbiter of these things. We’ll see.
zeynep tufekci
@zeynep
I will take the other side of this bet, starting with the idea there will be a singular “AI” with a singular answer that will actually be accepted as an arbiter by adversarial parties.
This reminds me of more information will reduce misinformation phase of social media debates.
Stefan Schubert
@StefanFSchubert
I think AI will make it harder to maintain widespread belief in false but politically convenient views, because:
1) People will increasingly use AI as an arbiter, and refer to it in debates
2) AI will be more truth-seeking and unbiased than human experts
zeynep tufekci
@zeynep
Well, well. Maybe Terrence Tao arguing “making some things easier doesn’t necessarily lead to the same downstream consequences assumed from previous, and very different, mechanisms” will get taken seriously.
Let’s think more about the roles mechanisms play, not nominal outcomes.
François Chollet
@fchollet
Terence Tao just posted this really insightful thread which seems to have been written in reaction to the Navier Stokes announcement. Please give it a read. mathstodon.xyz/@tao/117237320796901560
-
zeynep tufekci reposted
Maria Kari
@mariakari1414
i'm hearing from folks regularly that this is happening to them. people, including US citizens, are being detained for hours & having phones seized.
for anyone traveling, here's a good primer by @CUNY_CLEAR i recommend reading before leaving the United States -
static1.squarespace.com/static/59134566e58c623970f2cd48/t…
Thomas Kennedy
@tomaskenn
I was just detained for almost two hours at Miami International Airport coming back from an overseas trip. I was interrogated, asked about my work, political activities, books and literature I had, and my phone was seized.
This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Alison Gopnik
@AlisonGopnik
In the Times this morning- Henry Farrell James Evans and I on why we desperately need the social sciences to understand AI and its impact , just at the moment when the admin is trying to destroy them
www.nytimes.com/2026/09/06/opinion/ai-social-sciences.htm… -
zeynep tufekci
@zeynep
The AI benchmarks alone do not imply broader impacts or consequences. At all.
Plus, I think most of our benchmarks are quite useless for thinking through societal impacts and that this is already empirically evident, BUT even if they were to be useful, it’s not by themselves!
Jon Stokes
@jon_stokes
David, are there specific, non-benchmark-number-go-up predictions she has made around AI's impact on, say, the labor mkt, geopolitics, day-to-day life, or literally anything else that have panned out? Non-rhetorical question b/c I would like to know this information.
This post is unavailable. See it on X.Post not available here, view on X
zeynep tufekci
@zeynep
It’s not even the “map isn’t the territory”, it’s that the current AI benchmarks aren’t even the map!
And if we stop the podcast moods and look empirically, it’s clear that the prevalent theories about what those benchmarks were supposed be societal proxy for ARE WRONG SO FAR.
zeynep tufekci
@zeynep
I think AI is a transformative technology of much consequence. I expect a lot of impacts!
I simply think most current benchmarks aren’t very useful and that the societal predictions aren’t happening according to the AI benchmarks the way many people were claiming!
Jason Hahn
@JasonLHahn
You may have substantive reasons why you don’t believe the broader impact predictions will materialize but some of the pushback here is in response to “she claims Dyson spheres may happen in 20 years so ignore her concerns about rogue AIs causing worse damage now”
-
zeynep tufekci reposted
Itai Sher
@itaisher
This is what this person said about Dyson spheres.
You can judge: maybe all this stuff is right or maybe it’s just the way she thinks.
Matthew Yglesias
@mattyglesias
Please watch this clip
This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Ben Mullin
@BenMullin
scoop: Reporters from The New York Times, The Wall Street Journal and Bloomberg have been barred from next week's G20 meeting in Asheville, N.C.
www.nytimes.com/2026/08/29/business/reporters-denied-g20-… -
zeynep tufekci reposted
Teddy Schleifer
@teddyschleifer
NEW —
A comprehensive New York Times analysis of the top donors in the 2026 midterms.
Here are the top twenty — from Andreessen Horowitz to Ken Griffin. -
zeynep tufekci reposted
Tom Wright
@thomaswright08
Update: @WSJ just posted a story confirming from sources familiar that the purpose of the trip was to deter an attack on NATO, like a land incursion into the Baltics. www.wsj.com/politics/policy/cia-chiefs-surprise-moscow-tr…
Tom Wright
@thomaswright08
My new piece in @TheAtlantic on the CIA director's trip to Russia and where we are in the Ukraine war.
Gift link:
www.theatlantic.com/international/2026/08/cia-director-ra… -
zeynep tufekci reposted
The New York Times
@nytimes
Breaking News: Two unvaccinated people in Pennsylvania died from measles, the first deaths from the disease this year in the U.S., as a large outbreak there continues to spread. www.nytimes.com/2026/08/25/well/measles-deaths-pennsylvan…
-
zeynep tufekci
@zeynep
I believe many LLM “quirks” or phrasings are structural, straightforward persuasion hacks.
That’s why they mobilize signals/assertions of credibility, fallacies and distractions.
(Seems “it’s not X, it’s Y” is invoking “denying the adjacent” and “false dilemma”, for example).
Paul Calcraft
@paul_cal
This "I'll be honest" LLM quirk is just a reward hack to nudge LLM judges right? If it improved perf it could hide in chain of thought, no need to shout about it
This post is unavailable. See it on X.
zeynep tufekci
@zeynep
The AI model persuasion orientation greatly limits economic usefulness AT SCALE, obviously. That’s why all the singular impressive examples aren’t that relevant.
For example, I recently asked a paid model for a summary of what we know of the aftermath of a scenario when path X is NOT taken. Most studies follow what happens if path X IS taken. As usual, I asked for sources and links, etc. The model added barely perceptible modifications to study titles, abstracts etc. it listed to add “not X” to the scope of the study.
It was good and subtle. Blink and you miss it. You had to click on every link and read through a good chunk of the study. That was is still a bit faster than starting from scratch but the pitfall is more dangerous than starting from scratch if expertise is weaker. Plus it introduces other error vectors. Even the most advanced models do versions this some of the time and it is hard to catch these subtle persuasion fabrications without a real amount of extra work, which requires both expertise and time. It’s true they also don’t do this some of the time, but good luck differentiating the cases without the said expertise and effort.
Software has to work rather than persuade, so besides the verifiability aspect, it’s a different game.
zeynep tufekci
@zeynep
Hah.
An interesting twist to this is the high LLM sensitivity to frequency of inputs (which is another key reason they’re not that economically usable at scale outside of limited domains): Lacunae as failure.
Thomas G. Dietterich
@tdietterich
"Negation as failure" (logic programming joke)
-
zeynep tufekci reposted
Ashwin Ramaswami
@AshwinRamaswami
Your regular reminder that your Fourth Amendment rights are severely diminished, including for US citizens, at a border crossing. Not a good idea to delete data once you’re in the middle of being searched. Instead, BEFORE you enter the airport: turn off biometric unlock, turn off your phone, and delete data if you need to.
www.nytimes.com/2026/08/21/us/politics/samuel-tunick-dele… -
zeynep tufekci reposted
Alfons López Tena 🦇
@alfonslopeztena
US Customs says it has the authority to search the electronic devices of anyone entering or leaving US, regardless of citizenship, to enforce laws addressing whatever. Deleting the data means up to 5 years jail www.nytimes.com/2026/08/21/us/politics/samuel-tunick-dele…
-
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
NEW: Days after announcing it had received ICE contracts that could pay it $165m/year to detain immigrants, the private prison company the GEO Group donated $1.4m to the Trump-backed MAGA Inc. super PAC.
www.nytimes.com/2026/08/21/us/politics/geo-group-maga-inc… -
zeynep tufekci reposted
Alison Young
@alisonannyoung
The Kindle version of my book Pandora’s Gamble is on sale today for just $1.99 at Amazon. History shows there are a stunning number of ways lab leaks can happen. Learn more here 👇
-
zeynep tufekci reposted
Dr Filippa Lentzos
@FilippaLentzos
A thoughtful piece by @Ayjchan in @nytimes this week on thorny governance questions of risky biological research, with some very cool interactive graphics, including a reworked version of @gregkoblentz and my #globalbiolabs map. (1/4) www.nytimes.com/interactive/2026/08/17/opinion/covid-pand…
-
zeynep tufekci reposted
Jimmy Tobias
@JamesCTobias
David Morens, former top Fauci advisor, pleads guilty. He was criminally charged in connection to his efforts to evade the Freedom of Information Act. Read @theintercept scoop that first revealed his misdeeds here: theintercept.com/2023/06/29/covid-nih-personal-email-foia/
Gary Ruskin
@garyruskin
Former Senior NIAID Official [David Morens] Pleads Guilty to Charges Connected to Concealing Federal Records During COVID-19 Pandemic
www.justice.gov/usao-md/pr/former-senior-niaid-official-p… -
zeynep tufekci reposted
Jason Koebler
@jason_koebler
“X’s algorithm learns what you hate and shows you more of it, according to a new study just published in the Proceedings of the National Academy of Sciences”
www.404media.co/xs-algorithm-feeds-off-ragebait-and-impac… -
zeynep tufekci reposted
rat king 🐀
@MikeIsaac
gut morgen...
three — three! — reporting jobs now open to cover AI companies with me and my colleagues in San Francisco
apply here
OpenAI
job-boards.greenhouse.io/thenewyorktimes/jobs/4701630005
Anthropic
job-boards.greenhouse.io/thenewyorktimes/jobs/4724410005
SpaceX/Xai/Musk
job-boards.greenhouse.io/thenewyorktimes/jobs/4724444005 -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
“All of this was done outside high containment”
Read this from @Ayjchan
“Virologists in China grew the virus in lab-made versions of human airway, nose and intestine cells, where it multiplied.
They proposed testing it next in mice engineered with the same human cellular doorway.
They also identified features that limited how efficiently the virus could infect human cells.
More lab work could create a virus without those limitations, and members of this group did those kinds of experiments previously. -
zeynep tufekci reposted
Ashish K. Jha
@ashishkjha
Whether you believe that COVID started in a lab or not,
This @nytimes piece by @Ayjchan is worth reading
There has been a massive proliferation of high risk labs around the world -- and the research that is done in them
We need a concerted strategy to make these labs safer
Alina Chan
@Ayjchan
How to Stop a Lab Leak from Starting the Next Pandemic.
My new op-ed made possible by the @nytimes graphics and editorial team:
www.nytimes.com/interactive/2026/08/17/opinion/covid-pand… -
zeynep tufekci
@zeynep
The volume, content and tone in the UK media clearly did NOT meet a public interest bar.
See how even after they rapidly published hundreds of pieces, it was still US outlets that finally had a proper accounting of the institutional process and relevant key personal details.
Mike Bird
@Birdyword
The public interest hurdle for publishing is actually *already* a qualification to media freedom that I think people should sometimes be a bit more queasy about, since it's so malleable. But I don't even know what the argument is that this didn't meet a public interest bar.
Post not available here, view on X
zeynep tufekci
@zeynep
Exactly why UK journalists should stop circling the wagons by making far-fetched claims the public will not buy.
To defend against excessive regulation using public interest as an excuse, don’t use public interest as an excuse for obvious bad behavior not in the public interest.
BH
@billyhu6
How could this idea of "public interest" possibly be policed or regulated in any non-totalitarian way? The very fact that there were so many articles, written independently from different outlets is defacto evidence of "public interest"
zeynep tufekci
@zeynep
Will they drop their sudden interest about merit at Oxbridge, given the tragedy, conveniently?
We know they house many incredible scholars AND also some who’d never be there if they hadn’t been nepo babies.
Let’s see public interest reporting! Hundreds of stories per person!
zeynep tufekci
@zeynep
And this isn’t an outlier behavior. UK tabloids aren’t just ignored rags publishing UFO stories, and many of their respected outlets are not at all above campaigns of the sort that go way beyond public interest in any neutral sense. Murdochization didn’t’ help. Many exceptions of course. A lot of great work. But this general character is also true.
zeynep tufekci
@zeynep
Who said he was off limits?
On the contrary, I am saying the UK media rapidly published hundreds of trivial personal or irrelevant stories that were merely clickbait for the mob while some of the most important stuff got reported *later by US outlets.* They didn’t do their job.This post is unavailable. See it on X.
zeynep tufekci
@zeynep
That segment absolutely exists, merging race/immigration and rigid British class structure, and the goal of many campaigns is to prop it up.
The Meghan Markle case is interesting not as celebrity/royalty topic, but as analytical construct —a la Gramsci— for the UK public sphere.
stephens
@stephenfirvulag
You think the British public are angry that aristocrats aren't given privileges 🤣
-
zeynep tufekci reposted
nxthompson
@nxthompson
In 2021, Congress passed a law that forced shell companies to divulge their owners. It's now being scaled back, which could open the door for all kinds of corruption.
www.theatlantic.com/ideas/2026/08/corrupt-administration-… -
zeynep tufekci reposted
Alina Chan
@Ayjchan
How to Stop a Lab Leak from Starting the Next Pandemic.
My new op-ed made possible by the @nytimes graphics and editorial team:
www.nytimes.com/interactive/2026/08/17/opinion/covid-pand… -
zeynep tufekci reposted
Alison Young
@alisonannyoung
Former ABC News correspondent Terry Moran this week spoke publicly about feeling the “heavy hand of censorship.” It was early 2021 and involved a segment that explored whether Covid-19 could have come from a lab. “I’d never been angrier,” he says during a new podcast interview. 1/x
-
zeynep tufekci reposted
The Wall Street Journal
@WSJ
Exclusive: World Liberty Financial, the Trump family’s flagship crypto venture, says its trust company has received preliminary approval from the Office of the Comptroller of the Currency to become a bank. www.wsj.com/finance/trump-familys-world-liberty-financial…
-
zeynep tufekci reposted
Colin McCarthy
@US_Stormwatch
Temperatures hit as high as 43°C (109°F) in France today as the fifth major heatwave of the year has arrived in Europe. Hundreds of cities and towns across the continent experienced their hottest August temperatures ever recorded.
This is getting ludicrous. -
zeynep tufekci reposted
Jeremy Konyndyk
@JeremyKonyndyk
“On the day Renée Good died, he’d posted about her death on his own Facebook page.
The caption was one word: “FAFO.”
Pablo Manríquez
@PabloReports
EXCLUSIVE: Two DHS Sources Confirm to Migrant Insider — The ICE Agent Who Pointed His Gun at a U.S. Citizen in Virginia Is Martin Lagunas migrantinsider.com/p/exclusive-two-dhs-sources-confirm
-
zeynep tufekci reposted
Nick Cohen
@NickCohen4
Even right-wing journalists were wondering if their editors' obsessive coverage of Jason Arday was too much. Well, now they know. A lesson to the rest of us to never join a mob when it is braying for blood
Politics UK
@PolitlcsUK
🚨 BREAKING: Jason Arday has been found dead at an address in Battersea
-
zeynep tufekci reposted
Dan Meyer
@ddmeyer
Here is new data that sheds light on @Alphaschool & their 2 Hour Learning project. Until now, we could only speculate—does Alpha make successful students or select them in advance?
-
zeynep tufekci reposted
Joseph Allen
@j_g_allen
there's been a recent comms effort to sweep this measles outbreak under the rug - "it's just a few religious groups is what's causing this" - which is simply not correct when you look at the data. Cases are everywhere (and, look at the vax map - more cases are coming...)
Helen Branswell 🇨🇦
@HelenBranswell
#cdc has been alerted to another 101 #measles cases in the last week, bringing this year's total to date to 2,566. Only 16 of those infections were contracted outside the U.S. www.cdc.gov/measles/data-research/index.html
-
zeynep tufekci reposted
WIRED
@WIRED
Records obtained by WIRED detail hundreds of allegations of Customs and Border Protection workers misusing internal tools to look up romantic interests and track colleagues’ cell phones. www.wired.com/story/cbp-workers-allegedly-used-government…
-
zeynep tufekci
@zeynep
“You’ll only be able to detect the watermark if you’re in a select group that has access to a secret key (e.g. frontier lab employees or government / police).”
John Wang
@johnjianwang
Anthropic started watermarking Claude’s output. I did a little investigation into what method they used: johnjwang.com/post/2026/08/12/how-claude-watermarking-pro…
It’s likely they’re using some private key watermark similar to SynthID.
They’re likely not using hidden Unicode or using green/red lists -
zeynep tufekci reposted
Andy Mannix
@AndrewMannix
HSI also used facial recognition software at protests to identify people.
-
zeynep tufekci reposted
Jonathan Weisman
@jonathanweisman
This story is remarkable. Read it. www.nytimes.com/2026/08/13/us/politics/dhs-inquiry-immigr…
-
zeynep tufekci reposted
tanvi
@Tanvim
Alan Feuer
@alanfeuer
NEW: DHS agents in Minn conducted huge spying campaign on left-wing groups and people at the height of ICE protests earlier this year. Undercover agents. Infiltration of group chats. Subpoenas for labor union financial records www.nytimes.com/2026/08/13/us/politics/dhs-inquiry-immigr…
-
zeynep tufekci reposted
Kyle Griffin
@kylegriffin1
BREAKING WSJ:
Trump White House officials have asked the Trump Justice Department to consider a new prosecution of David Hearn, the man accused of vandalizing the Reflecting Pool — as Trump fumes over U.S. Attorney Jeanine Pirro's decision to drop the case. -
zeynep tufekci reposted
LenaSun
@bylenasun
BREAKING: Trump administration is preparing an executive order that would upend decades of federal guidance on which vaccines are administered to children and call for states to reconsider school vaccine requirements. w @ddiamond @rachel_roubein
www.washingtonpost.com/politics/2026/08/10/trump-administ… -
zeynep tufekci reposted
The Associated Press
@AP
BREAKING: Senate confirms Todd Blanche as attorney general, cementing the ex-Trump attorney’s command of Justice Department. apnews.com/article/senate-blanche-attorney-general-justic…
-
zeynep tufekci reposted
Michael Weissman
@mbw61567742
The pillars of the Covid market story were the Worobey & Pekar 2022 Science articles. I'd showed that Pekar was based on a plain stats logic error & revered when fixed.
econjwatch.org/articles/an-article-in-science-on-covid-or…
Now I show multiple ways Worobey is wrong, invalidating it too.
Econ Journal Watch
@EconJWatch
An Article in Science on Wet-Market Covid Origins is Unreliable
by Michael B. Weissman
An influential paper by Worobey et al. and 17 other authors published in Science in 2022 claimed that case locations and nucleic acid samples showed that SARS-CoV-2 had first been introduced to humans at the Huanan Seafood Market. Several lines of evidence suggest the introduction to humans was earlier and elsewhere.
econjwatch.org/file_download/1416/WeissmanAug2026_manuscr… -
zeynep tufekci reposted
Bloomberg
@business
Exclusive: The Trump administration is actively considering issuing an executive order on vaccines and autism as soon as next week www.bloomberg.com/news/articles/2026-08-06/trump-administ…
-
zeynep tufekci reposted
Gilles Demaneuf
@gdemaneuf
Because it takes endless time and you get branded as a conspiracy theorist for all reward, when you don’t lose your job after people complain to your employer.
A good piece is anywhere between 50 and some 100s hours of collecting data and working collaboratively on the analysis.
Some of my pieces take one year to research and write.
Fishing out some thesis in China as @TheSeeker268 organised is also risky. We are talking real risky.
There are a few investigative journalists left who have the time and the editorial backing, but very few. It’s a dying breed.
@KatherineEban @thedeadhandbook
The Honest Broker
@RogerPielkeJr
Worth a read
Then ask: Why is DRASTIC doing this work and not an authoritative expert committee?
The failure to investigate COVID origins is itself a massive scandal (regardless of origins)
Charles Small
@C_Small_
Just over six years ago, I found something unusual in the data for the closest match to the pandemic virus SARS-CoV-2.
Soon after the pandemic began, the Wuhan Institute of Virology published RaTG13, having previously called it BtCoV/4991 in a short sequence published pre-pandemic.
Strangely, 6/33 published reads of RaTG13 were labeled "7896", the name of a different betacoronavirus kept in the Wuhan lab.
Why would the lab, known for splicing different viruses together, name part of one virus after another?
Naturally, I wondered if RaTG13 was actually a composite virus of the original BtCoV/4991 and 7896.
I shared the finding in the Drastic group chat and we investigated it, with @franciscodeasis then publicly alerting @Ayjchan to the dates in the amplicon sequences, who then mapped out the lab's work progress with RaTG13 before the pandemic.
The Wuhan lab has since released a virus they claimed to be the original 7896. The parts of RaTG13 labeled 7896 are a close match to the published 7896 virus, but not quite it.
So what did they publish? Another virus? Another composite?
Splicing viruses together and then testing how well the composite viruses bind to human cells was exactly what the lab did (See this summary from 2021: zenodo.org/records/21704910).
But the 7896 parts of RaTG13 may not be from a different virus spliced together. If we take them at face value, the lab appears to have been using 7896 as part of the sequencing process, a sort of reference point to amplify and see what they were looking at in RaTG13. According to the lab, they started using new sequencing equipment in 2018 so would not have had to do that any more, but these were dated June 2017.
The six RaTG13 7896 sequences cover the area around the RdRp and are labeled 1, 2 and 4. The area which would have been 3 is where the already published short BtCoV/4991 RdRp sequence should be, so it looks like they started with that then expanded coverage using parts of what they already had from 7896.
This tells us about their pre-pandemic workstream. As some of the published RdRp sequences of the group of viruses including 7896 are identical, the Wuhan lab must have had longer sequences from them to be able to tell them apart.
It tells us that the lab was working on some of the most similar viruses to the pandemic one as a group before the pandemic, using parts of their research into one to inform research into another.
And why were they doing this work in the run up to the pandemic?
The latest date in the RaTG13 amplicon sequences was only 11 weeks before a project testing human ACE2 receptor binding of two novel bat SARS-related coronaviruses was due to begin in January 2019. Was the work in preparation for that? -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
We wrote this for the Lancet Covid-19 Commission in 2022. There was intense pressure at the time to adopt the 'natural spillover'. I was one of the Commissioners that held the line that the evidence on natural spillover was not conclusive, and, in fact, there was a lot of evidence for lab leak.
Joseph Allen
@j_g_allen
As of the time of publication of this report, all three research-associated hypotheses are still plausible: infection in the field, infection with a natural virus in the laboratory, and infection with a manipulated virus in the laboratory.
-
zeynep tufekci
@zeynep
More than enough public information to take the option seriously for those willing to learn — unfortunately most media covered the topic poorly and the field circled the wagons, so people have to wade through the available facts.
Alas many would rather not. Unpleasant topic.
Benjamin Mazer
@BenMazer
When the former Biden Covid czar changes positions to endorse a lab leak, it deserves an actual explanation. Did he learn inside information that the public isn’t aware of? What changed?
Notably it seems Ashish is now in the private sector. Could be seen as opportunistic. -
zeynep tufekci reposted
State of the Union
@CNNSOTU
Former Biden White House COVID-19 Response Coordinator, Dr. Ashish Jha, tells Dana Bash, "When I went into the White House, my view was this was almost surely a natural outbreak, maybe a lab leak. Based on information I learned and based on information I've seen, I have come to conclude that it is more likely to have been a lab leak."
-
zeynep tufekci
@zeynep
And people are doing that for very good reason.
Kevin Roose
@kevinroose
almost nobody is pricing in the possibility that the models just keep plowing through every discipline the way they’re plowing through math
Noam Brown
@polynoamial
An internal version of Astra, @OpenAI’s next major model family, solved 10 major open problems in mathematics, quantum complexity, and theoretical computer science.
We believe it will be a major step for scientific reasoning. openai.com/index/ten-advances-in-mathematics/ -
zeynep tufekci reposted
Michael Lin, MD PhD 🧬
@michaelzlin
Nothing gets my goat more than a slimy pretend-scientist such as Daszak casting aspersions on knowledgeable smart and fair scientists such as Bloom
The Seeker
@TheSeeker268
Then comes another googly by Daszak: "anyone who looks at the evidence.. and still thinks it came from the lab, cannot be a good, logical scientist", and calls Bloom a "racist."
"Thanks Peter, out true friend" replies Shi.
Daszak is now the de facto crisis manager for the WIV. -
zeynep tufekci reposted
Annie Linskey
@AnnieLinskey
Pres. Trump is concentrating power in the Oval Office and making $50M asks for projects from companies w business before the federal government. Some are giving tens of millions. Story by @jdawsey1 @DanaMattioli @EmilyGlazer and me.
www.wsj.com/politics/policy/trump-donations-company-fundr… -
zeynep tufekci reposted
Carl Quintanilla
@carlquintanilla
Four-byline alert: 🚨
“.. no sitting American president has ever raised these amounts for his pet projects. .. Trump has been demanding large checks .. telling executives and lobbyists that their companies should give $25 million or $50 million.”
@WSJ
www.wsj.com/politics/policy/trump-donations-company-fundr… -
zeynep tufekci reposted
Keegan - Smartest Boy Urbanist™️
@keegan_tweets
I follow zeynep and haven’t seen one of her tweets in like 24 months.
Such a weird algorithm.
zeynep tufekci
@zeynep
However, we learned there is incontrovertible proof of a conspiracy to cover up and deny the possibility it may well be a lab mishap, orchestrated from the highest levels of the scientific establishment and a self-serving group of scientists who deliberately misled the public.
-
zeynep tufekci reposted
Katherine Eban
@KatherineEban
story: "The COVID Origins Saga is More Complicated Than Either Side Wants to Admit." 🎁🔗www.rollingstone.com/politics/politics-features/covid-ori… /2
-
zeynep tufekci reposted
Michael Weissman
@mbw61567742
Maybe it's time to repost these informal thoughts on the failings of institutional science re Covid and reasons to pursue truth but not punishments.
michaelweissman.substack.com/p/science-in-the-swamp -
zeynep tufekci
@zeynep
In contrast, I expect overall progress even in many verifiable domains to choke and slow down because of generative AI.
It’s not because AI can’t do many impressive things but because the plausible glut it produces breaks the key filtering mechanisms necessary for progress.
Ramez Naam
@ramez
I'm quite bullish on AI in highly verifiable domains like math and coding. We're headed for narrow super intelligence, clearly.
I expect progress in other domains to continue, but at a slower pace. Progress per sector will be broadly correlated to how quickly and perfectly we can verify outputs. I love Adam's graphic below. It's much closer to what I see.
Adam Hunt
@RealAdamHunt
Recently I've flipped from being bullish to being bearish about AI.
I think I'm updating my bearishness to be more solidly bearish. Early thoughts (which I hope to be disproven in the next year or so, I would prefer progress) and my reasoning:
The whole 'it turns out if you keep training and scaling the models more they develop broad new capabilities in lots of domains' thesis is wrong (sorry Demis). The recent batch of models haven't got more general, they've got less general. This is most obvious in the fact that their language outputs have got much worse in comparison to e.g. o3. If they were gaining generalist capacities we would expect them to be describing their work in ever more graceful and comprehensive prose!
The image that was being shared as the AGI thesis (November 2025, Tomas Pueyo) was the spiky bubble that has a current spike or two out past human capabilities (e.g. on coding or math) but below human on other capabilities on the other spikes - the future prediction was that as the models scale/advance, every spike would grow bit by bit until the whole center encompasses the human capabilities, with super-superhuman on some spikes. I think it seems like what's actually happened in the last few models has been that the coding/math spike has grown, but leaving behind or even at the cost of the other spikes. The models are no better at some simple logic, language (and sometimes worse!).
This makes sense from a simple RL perspective; you can't RL something endlessly on one domain of tasks and expect it to improve on the other tasks. The fact that early LLMs did seem to improve generally was a byproduct of the written language corpus covering everything - that corpus is general, so training it on that gave the appearance of something generally intelligent and becoming more generally intelligent as it got better at replicating that corpus. But the actual logic and underlying ground truths behind the language aren't captured efficiently enough and weren't effectively RLd in - they top out at some point (I guess this happened around the time that there was the 'has scaling hit a wall' discussion in late 2024). Chain of thought was then a genuine breakthrough, along with web search, which plugged into that general LLM global-corpus intelligence to lead to post 2024 gains.
The AI companies have since worked out that coding works (and pays) really well (basically this is because the entire job is nearly perfectly recorded and exists as training data, and you can set up clear benchmarks and rewards). The recent models (and benchmarks) have been maxxing that and we've seen degradation on normal English use for that reason. This could still be transformative, leading to extremely powerful (and potentially dangerous, particularly in cyber security) models but it's not a pathway to AGI.
I'm probably at about 40% confidence about this. It fits my current observations of AI progress and has a basic explanatory model. It doesn't account for potential breakthroughs, which is a major reason for discounting.
To make some predictions, I guess if I'm right this will become broadly apparent and more widely acknowledged in the next year or two, as we see how the spikiness of models that keep getting released develops.
Maybe there will be efforts to concentrate on specific spikes e.g. health or law which require going back to earlier models and RLing on a different data set/with different rewards/benchmarks. Maybe those separate models can be linked together to give a more apparently general model. How capital intensive that is/the potential profitability will be a defining question. But I just don't see general abilities emerging atm, and I don't think we will any time soon. Good news - a whole industry of tackling important specific problems/sectors can open up! -
zeynep tufekci reposted
Vishal Misra
@vishalmisra
The WSJ quotes me criticizing Anthropic’s AI safety claims. The argument is not that AI risk is imaginary. It is that before regulating frightening model behavior, we should establish whether we are looking at an emergent mechanism or a learned narrative.
The Wall Street Journal
@WSJ
A backlash against Anthropic is brewing in Silicon Valley, rooted in concerns about the company’s product launches that compete with existing software providers and its advocacy for a closed AI ecosystem www.wsj.com/tech/ai/a-backlash-against-anthropic-is-brewi…
-
zeynep tufekci reposted
Nicholas Kristof
@NickKristof
When ICE agents wear masks, use ordinary vehicles and don't show ID's, we risk more attacks by such criminals pretending to be ICE agents. In this case, fake ICE agents pulled over a person and assaulted the victim.
Congresswoman Andrea Salinas
@RepSalinas
This is horrifying.
Pretending to be an ICE agent to terrorize our neighbors is cruel, disgusting, and dangerous. These individuals must be held fully accountable.
ICE’s lawlessness and secrecy lets this violence flourish.
www.oregonlive.com/clackamascounty/2026/07/armed-men-posi… -
zeynep tufekci reposted
Andrew Perrin
@AndrewJPerrin
This is far worse news for empirical sociology than all the “war in sociology” media accounts
Stephen L. Morgan
@SLMsociology
And, based on what we are told, there will be no 2028 GSS … (but we are in the field and will complete the 2026 GSS, nonetheless).
-
zeynep tufekci
@zeynep
However, we learned there is incontrovertible proof of a conspiracy to cover up and deny the possibility it may well be a lab mishap, orchestrated from the highest levels of the scientific establishment and a self-serving group of scientists who deliberately misled the public.
zeynep tufekci
@zeynep
Not surprising things can’t be proven given the strong top-down pressure to deny *and* withholding of evidence.
People who assume a knee-jerk denial of plain facts in response to the bad -faith weaponization of the topic actually EMPOWER the bad faith actors.
Alina Chan
@Ayjchan
A 2021 email from Peter Daszak of EcoHealth Alliance to the Wuhan scientist suspected of starting the Covid pandemic - asking her NOT to share data with other US scientists about the coronaviruses they have collected.
Post not available here, view on X
zeynep tufekci
@zeynep
Lab safety is too important —and the conspiracy to mislead the public too well-documented— to let the self-interested group exploit people’s feelings toward Rand Paul, RFK Jr. or Dr. Fauci.
Future consequences are why people must show courage to stand on the right side of this.
Rob Al
@robalpire
@mehdirhasan is too close minded to admit this. Everything is left/right , win/lose
zeynep tufekci
@zeynep
By the way, here’s I show up in their slacks (way more informative than Fauci diaries). Talking to one for 2.5 hours — actually wanting to be convinced, way back in 2021.
I *wanted* to be on the other side of this but facts were too strong and public interest too important.
Gilles Demaneuf
@gdemaneuf
That one is for you, @zeynep:
You are Turkish and thus suspicious for thinking that "oppressive governments lie and cover-up." -
zeynep tufekci reposted
Alina Chan
@Ayjchan
For a refresher on the publicly available evidence for a lab leak origin of Covid, there's the 2024 graphical walkthrough I created with the @nytimes and the 2022 updated paperback of VIRAL: The Search for the Origin of Covid-19 co-authored with @mattwridley @HarperCollins
-
zeynep tufekci reposted
Alina Chan
@Ayjchan
The card-carrying virologists who published just about every influential paper claiming a natural origin of Covid reacted to the Defuse put-furin-cleavage-sites-in-SARS-like-viruses proposal as so:
"it really isn't beyond plausibility that someone went ahead and tried some of this stuff. We just need to re-convince ourselves it is coincidence."
"That's the problem. There's a big difference between "the lab was working on bat CoVs" to the "the lab was modifying bat CoVs using features we see being unique to SC2"."
On whether the FCS was inserted in or out of frame, "it depends on cloning strategy... sometimes ease trumps perfection."
On the choice of SARS2 as a backbone, "Well, in the grant they specifically talk about *new* viruses - so a SC2 precursor could have been one of those "low risk" viruses they talk about and then they started fiddling with it."
"Peter [Daszak, collaborator of the Wuhan lab] should have seen this coming and it is worse than he was suggesting... very striking that a bat-derived CoV with an FCS turns up in a city that was on a grant proposing to muck around with FCSs in bat-derived CoVs."
"this stuff about the DARPA proposal makes me even more angry about the DT letter (if it is indeed Baric) - he knew how suspicious a FCS would look back then."
"And his attacks on me for even suggesting we needed to take a look at this. He's a two-faced b***ard, that's for sure."
"maybe we could have a chat tomorrow about the Baric/Daszak ****show? I would just like to discuss what our strategy should be so we don't get associated with these f**kers."
"... we can't ignore this and I also need to get something back to Tony's team at NIH."
"stay off email or carefully curate some emails for the FOI records" -
zeynep tufekci reposted
Alina Chan
@Ayjchan
"There is still no definitive conclusion to this debate [Covid's origin]. Yet one thing is clear: there was a conspiracy to dupe the public, involving leading scientists and research-funding bodies in Britain and the US."
Ian Birrell
@ianbirrell
Shame on the cabal of top scientists who mocked me and other dissidents for challenging their Covid origins conspiracy. It is now beyond doubt that they betrayed both humanity and science - and there should be retribution.
My MoS column mol.im/a/16005037 -
zeynep tufekci reposted
Aaron Reichlin-Melnick
@ReichlinMelnick
After ICE shot and killed Lorenzo Araujo, the FBI leaked to Fox News a claim that he was found with possible meth in the van.
His family said it was a total falsehood; it was a bag of salt to be used to make an electrolyte drink. And they were right. It was not drugs.
Harris County District Attorney
@HarrisCountyDAO
We can confirm the substance found in Lorenzo Salgado Araujo's van tested negative for narcotics/ illicit drugs.
We are not commenting further on the test results, as the investigation is pending. -
zeynep tufekci reposted
Atul Gawande
@Atul_Gawande
In Kenya, reporting shows that abandoning an incomplete project w/o mitigation steps triggered the worst flood in a decade, destroying farmland, damaging homes, disrupting livelihoods and increasing waterborne disease.
Bloomberg
@business
When the Trump administration abruptly terminated most of USAID's water and sanitation portfolio, it left dozens of projects unfinished around the world.
Bloomberg investigated the devastating impacts to one community in Kenya. Read The Big Take ⤵️ www.bloomberg.com/graphics/2026-usaid-water-projects-trum… -
zeynep tufekci reposted
The Seeker
@TheSeeker268
I'm late to this, but wow.
Baric believes WIV almost certainly sequenced the Mojiang miners' viruses but never published or shared those sequences.
"So they didn't lie. They just didn't tell the whole truth."
It's pretty extraordinary to see Baric say this in a transcribed interview. -
zeynep tufekci reposted
The New York Times
@nytimes
Breaking News: An outbreak of cyclosporiasis linked to iceberg lettuce has expanded to nine states, including Pennsylvania, according to the CDC. This comes as federal officials investigate other illnesses, which they say are unrelated to this outbreak. www.nytimes.com/2026/07/24/well/eat/cyclospora-outbreak-s…
-
zeynep tufekci reposted
The New York Times
@nytimes
Breaking News: Federal officials acknowledged in court documents that they canceled grants to states that didn’t vote for President Trump in 2024. www.nytimes.com/2026/07/24/business/trump-state-grants-ca…
-
zeynep tufekci reposted
CBS News
@CBSNews
1.6 million eggs recalled at Kroger and other stores over salmonella risk, FDA says. www.cbsnews.com/news/egg-recall-salmonella-midwest-poultr…
-
zeynep tufekci reposted
Trip Gabriel
@tripgabriel
From NYT:
Federal officials said in court documents that they halted billions of dollars in funding for energy projects to states “based solely” on whether they backed President Trump in the 2024 election.
www.nytimes.com/2026/07/24/business/trump-state-grants-ca… -
zeynep tufekci reposted
Sheryl Gay Stolberg
@SherylNYT
New White House plan for science — more $$ for AI, less for universities and “life sciences” research. Written by Trump’s science adviser, who is a Silicon Valley tech guy with a degree in politics.
www.nytimes.com/2026/07/22/us/politics/trump-research-fun… -
zeynep tufekci reposted
The New York Times
@nytimes
Breaking News: The U.S. has recorded more measles cases in 2026 than in any other year since the disease was declared eliminated in 2000. www.nytimes.com/2026/07/24/well/measles-record-united-sta…
-
zeynep tufekci reposted
LenaSun
@bylenasun
NEW: U.S. measles cases hit 35-year record with months left in 2026. Based on Johns Hopkins US Measles Tracker. (Updates more frequently than CDC’s weekly update on Fridays). www.washingtonpost.com/health/2026/07/21/us-measles-cases…
-
zeynep tufekci reposted
AlanaSemuels
@AlanaSemuels
Very unusual: A FDA committee voted against FDA staff's own advice in recommending that BP-157, a peptide, be allowed to be compounded by pharmacies in the U.S.
Significantly, most of the 'YES' votes came from members who advise telehealth companies. -
zeynep tufekci reposted
Aric Toler
@AricToler
The Pentagon removed the names of 4 servicemembers who died in Jordan from their official list of casualties. They explain that this is because "their deaths occurred after President Trump declared a cease-fire in the war in April."
Free gift link: www.nytimes.com/2026/07/23/us/politics/iran-troop-deaths.… -
zeynep tufekci reposted
Cameron 🇺🇸 🗽🦅
@CameronCorduroy
this is the official CDC website btw
we're cooked -
zeynep tufekci
@zeynep
I don’t think @PeterMcCrory post has the main reason for so few employment effects from AI. (though some good points.)
I think people are *still* avoiding analyzing as LLMs as LLMs. Too much vague AGI hand-waving.
I wrote about it in the NYT last month.
www.nytimes.com/2026/06/30/opinion/ai-agents-steal-jobs-e…
Peter McCrory
@PeterMcCrory
-
zeynep tufekci reposted
Alina Chan
@Ayjchan
Over the years, several virologists have told me they believe a lab origin of Covid is more likely than a natural origin. I worried it might've been selection bias but it seems even virologists doing gain-of-function experiments told Kristian Andersen the same in 2020.
Alina Chan
@Ayjchan
New Proximal Origin slack messages released: "most scientists I talk to believe there is a likely connection to the lab - people that are themselves doing GOF and culture-type of experiments, they know the risks and don’t seem to think that a natural explanation is more likely"
-
zeynep tufekci reposted
Alina Chan
@Ayjchan
There's no redemption arc in these messages, and I expect the journalists and scientists who have picked their tribe aren't going to have a sudden awakening.
Just know these virologists are conspiracy theorists. In what universe is Ralph Baric sending me out to do dirty work? -
zeynep tufekci reposted
Alina Chan
@Ayjchan
"We still have no evidence for SC2 in that lab and they must have sequenced it first. But it is very striking that a bat-derived CoV with an FCS turns up in a city that was on a grant proposing to muck around with FCSs in bat-derived CoVs."
Even Peter Daszak, a collaborator and funder of the Wuhan Institute of Virology lab for more than a decade, didn't have access to their virus sequence or sample databases.
Even Ralph Baric, a collaborator on the Defuse proposal, told Congressional investigators that the Wuhan scientists wouldn't have shared their most interesting strains or sequences with him.
Why would rando virologists believe they are even close to knowing about the newest viruses discovered by Wuhan scientists? -
zeynep tufekci
@zeynep
Compare to popular X “criticisms” such as Nolan’s Odyssey being non-linear (lol in MEDIA RES) or that it’s anachronistic (lol in Hans van Wees).
Yes, academia has many issues but people who only get the X caricatures have no idea that it still has a deep intellectual culture.
MarquisAntiMedia
@Marquis_Crutch
"Emily Wilson is bad translator because, uhm reasons okay, and she totally liked Nolan movie!"
Emily Wilson:This media couldn't be displayed. See it on the original post.Post not available here, view on X
zeynep tufekci
@zeynep
The effect of this on public life—interesting and relevant to academics.
But the other discussion about this movie, modernity, glory versus the hero (Homeric poems are anachronistic exactly in that manner), etc. that’s done only in long-form by people who still read and write.
zeynep tufekci
@zeynep
That olfactory status markers dissertation was interesting in another AI relevant way. Urban sanitation, white collar work, plumbing etc greatly weakened them, so their key role is hard to understand without such literary deep digging — it is like going from a world of color to black and white or sepia, and yet colors are used all over fiction (like archeological remnants of a lost civilization) to denote mood, feelings and more.
But imagine some people can still see color! Anyway that’s where literary/rational long-from print culture type thinking is going, imo. -
zeynep tufekci reposted
Avidfilmbuff
@avidfilm
Ok, he didn’t read it. There is no other explanation. The poem starts in the present, and near the middle Odysseus tells the events of the past 20 years to the Phaeacians. The movie is actually faithful to the structure of the original poem, with Calypso replacing the Phaeacians
Matt Walsh
@MattWalshBlog
Nolan's obsessions with the non-linear storytelling gimmick really hurts this film. It jumps around so much, and spends so little time on any one particular moment or scene, that you never really feel like Odysseus is on an epic journey spanning many years. When he makes it home, you never feel like he's been gone for 20 years. The film does not capture the grandeur and scope of the original epic. That's why it feels rushed and too long simultaneously. The movie comes off like a rough draft. Nolan was searching for the actual story but never quite found it.
-
zeynep tufekci reposted
Michael M. Grynbaum
@grynbaum
NEW: Read the @nytimes's full account of the DOJ subpoenas of its reporters' phone records.
One was issued after NYT had already filed a motion to quash the original grand jury subpoenas.
DOJ sought phone records from one reporter's mother.
www.nytimes.com/interactive/2026/07/20/us/letter-motion-i… -
zeynep tufekci reposted
Adam Taylor
@mradamtaylor
The Trump administration is attempting to subpoena the phone records of a NYT reporters' mother, who works as a mental health professional, after the outlet's reporting on the safety issues with the president's Qatari plane. www.nytimes.com/2026/07/20/business/media/new-york-times-…
-
zeynep tufekci reposted
Derek Cressman
@DerekCressman
The get out of jail card ain’t free.
Kenneth P. Vogel
@kenvogel
SCOOP: After pardoning hundreds of allies over the course of his 1st 18 months back in the WH, Trump this month quietly rejected clemency applications from nearly 6,000 people who mostly lacked connections & had been waiting in a queue at @TheJusticeDept. www.nytimes.com/2026/07/20/us/politics/trump-rejected-cle…
-
zeynep tufekci reposted
Ian Weissman, DO
@DrIanWeissman
The FDA said on Sunday that laboratory experts re-checked lettuce sample results from Taylor Farms in an investigation into an outbreak of the cyclospora parasite and concluded the finding should be considered a false positive.
www.reuters.com/business/healthcare-pharmaceuticals/fda-s… -
zeynep tufekci reposted
Joel Chan
@kjoules
The @nytimes obituary of Lam Wing-kee, the Hong Kong bookseller who was abducted by Chinese officials. When he returned to Hong Kong promising to retrieve customer names for the authorities, he instead set up a news conference to describe his abduction
www.nytimes.com/2026/07/17/world/asia/lam-wing-kee-dead.h… -
zeynep tufekci
@zeynep
The FIFA VAR as preview into advancing algorithmic governance in practice outside of the China context: highly-surveillant and high-tech (ball sensor, graphs, etc.) accompanied by a form of transparency (replays, etc.) but without accountability—thus still essentially a blackbox.
Florian Ederer
@florianederer
Huh, so NOW the VAR does actually call back goals when there was a foul ON THE OTHER SIDE OF THE PITCH?
Interesting.This post is unavailable. See it on X.Post not available here, view on X
zeynep tufekci
@zeynep
The combination applies to many high-tech surveillant algorithms. Review all automatically isn’t tenable for most domains, “select some” fuels exactly this.
Decay and corruption are the default unless fought actively.
Sports: MLB is better attempt.
Society: headed to FIFAland.
Ray · AI & how value is shared
@governance_arc
Genuine question for the VAR-as-governance frame: where's the recourse?
Everyone can watch the footage in slow motion now. But a team still can't formally contest the call. Visibility went way up. The ability to challenge the outcome didn't.
That gap feels like the actual black box, not the sensors.
zeynep tufekci
@zeynep
Noting that “VAR involved” is now used to (correctly) mean that the technology was a key mechanism to nudge the preferred semifinal combination.
Yes, FIFA is FIFA, but this specific technology mobilizing pathway is the 21st century way for fixing more than sport.
Nate Silver
@NateSilver538
Not that we were going out on a limb here (and not that there wasn't a little bit of VAR involved) but still kinda cool that the top 4 teams in PELE at the start of the tournament all made the SF.
Nate Silver
@NateSilver538
There are just not a lot of upsets in the knockout stage in recent World Cups. I'm rooting for Norway tomorrow but we could easily wind up with the equivalent of the four #1 seeds (ESP/FRA/ARG/ENG) in the semis.
zeynep tufekci
@zeynep
There is another world where Messi retires as the reigning World Cup champion and the greatest player of his generation.
And then there is this.
This cup at least gave those in the technology and authoritarianism space an easy to explain example of one future — beyond sport.
Out of Context Westcol 💀
@OccWest
Estoy llorando como Messi Acusa a Cucurella por taparse la boca 😭😭😭
This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Trip Gabriel
@tripgabriel
Eyewitness to ICE fatal shooting:
“I heard agony,” said Mary Hayes, who saw a screaming woman on her knees, next to a young girl. “I heard a howl that came from your soul, that your whole life had just changed and it was never going to be the same.”
www.nytimes.com/2026/07/13/us/ice-shooting-maine-guerrero… -
zeynep tufekci reposted
Tom Malinowski
@Malinowski
So on the same day President Trump accused China of a nefarious scheme to interfere in our elections, he actually ended sanctions against Chinese officials (those imposed for human rights abuses in Hong Kong). What a tough guy!
The Committee for Freedom in Hong Kong Foundation
@thecfhk
Today, the Chinese Communist Party (#ccp) was the first to announce that the Trump administration would not renew the national emergency declared under Executive Order 13936 on #hongkong—apparently as part of a financial concession. The Treasury Department has now confirmed that decision. The CFHK Foundation is stunned that the administration allowed U.S. policy to be announced by the CCP before informing the American public, and we strongly oppose this decision.
President Trump's decision not to renew the Hong Kong national emergency is a textbook case of the appeasement and weakness that members of Congress and so-called China "hawks" have rightly condemned in the past. Compounding this failure, the concession did not even secure #jimmylai's release from prison, despite the President's repeated promises to do so.
Serious questions now surround the nature of these concessions and who truly benefits from them. The public deserves answers: Do sanctions on Hong Kong officials for human rights abuses remain in place? Is the administration restoring Hong Kong's economic, trade, and financial privileges? And what, precisely, has been conceded to the Chinese Communist Party? Congress must also act to codify the executive order's provisions into law, consistent with legislation it has already passed by strong bipartisan majorities.
Make no mistake: repression in Hong Kong continues to intensify, and democracy advocates and freedom fighters remain behind bars. The Administration and Congress should support the people of Hong Kong at this critical moment. This decision amounts to nothing less than a betrayal. -
zeynep tufekci reposted
Bloomberg
@business
The US hit southern Iran overnight, widening the scope of its targets by striking six road bridges, according to Iranian state media www.bloomberg.com/news/articles/2026-07-17/fears-mount-of…
-
zeynep tufekci reposted
Dr. Brian Goldman
@NightShiftMD
This is an INSANE development. For Heaven’s sake, get your TDAP up to date.
-
zeynep tufekci reposted
Andreas Karch
@karch_andreas
Grad student visas limited to 4-years. Extension possible in principle:
www.dhs.gov/news/2026/07/16/trump-administration-issues-f…
It's impossible to finish a Physics PhD in 4 years. How hard will it be to get an extension? Will any foreigner still come to the US for a science or engineering PhD? -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
Trump canceled plans for a new satellite sensor that would have improved wildfire smoke monitoring👇
-
zeynep tufekci reposted
Ben Pershing
@benpershing
Truth Social will sell Wall Street firms quicker access to market-moving posts from President Donald Trump, the company said on Thursday.
www.cnn.com/2026/07/16/business/truth-social-data-wall-st… -
zeynep tufekci reposted
Max Kozlov
@maxdkozlov
NEW in @Nature: This is now the largest Cyclospora outbreak in US history. But CDC's parasite surveillance team was slashed from 11 people to 3 last year.
Joel Barratt, who ran the team, worries the CDC doesn't have the staffing to stop the outbreak quickly. -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
This person is poking fun, but I want to let you all know - *it really is that simple*. You CAN just run to the local hardware store and get a better filter for a few bucks and an air purifier, or you can order them online (and probably get them delivered within a few hours)
-
zeynep tufekci reposted
Joseph Allen
@j_g_allen
The smoke impact area has widened significantly
-
zeynep tufekci reposted
Joseph Allen
@j_g_allen
Children's hospital admissions and ER visits for asthma go up 4.8% per 10 ug/m3 increase in *short-term* exposure to PM2.5.
NYC baseline is roughly 10 ug/m3, and today is forecasted to be 125 ug/m3, so it's a ~50% increased risk.
pubmed.ncbi.nlm.nih.gov/27499163/ -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
Does your building have upgraded air filters? At these levels we see immediate increases in hospital ER visits and heart attacks, and most of that air pollution you breathe happens indoors. Strongly recommend MERV13 or higher, even better if it's a "MERV-A" rated filter which hold up better against charged wildfire smoke.
-
zeynep tufekci reposted
FIRE
@TheFIREorg
Florida International University is requiring students to record apology videos after their silent anti-ICE protest at a campus event.
The protest drew no complaints, and admin even admitted it wasn’t disruptive, yet students face charges under an unconstitutional speech policy.This media couldn't be displayed. See it on the original post.
FIRE
@TheFIREorg
Florida International University is forcing students to record university-approved “reflection” videos as part of their punishment for a silent anti-ICE protest — a totalitarian twist to an already unjustifiable disciplinary process.
FIU disciplined seven students for a March 13 demonstration during a campus “fireside chat,” in which students quietly revealed “ICE OFF FIU” shirts and then left. By the university’s own account, the protest neither disrupted the event nor drew complaints.
Yet students have received conduct charges, written reprimands, and are now being required to submit recorded “reflection” videos explaining their alleged policy violations and how they will comply with the university’s expressive activity rules — the same rules we’ve urged the university to revise. Nor will a simple apology suffice: the videos must be “original” and “thoughtful,” and officials may reject them and require students to re-record if they are not satisfied. Students say the university has also indicated that noncompliance could affect their ability to receive their diplomas.
We have already warned FIU that it is unlawfully applying an overbroad policy to non-disruptive, protected expression. FIU’s response was to add compelled speech to the punishment.
A public university cannot punish students for a silent protest that, by its own account, disrupted no one, and then force them to record university-approved statements to stay in good standing. FIU must drop the charges, rescind the compelled “reflection” video requirement, and stop trying to turn protected expression into state-scripted obedience. -
zeynep tufekci reposted
Brian Kern / Kong Tsung-gan / 江松澗
@KongTsungGan
The American writer @Susan_BK was invited to Hong Kong to participate in a cultural event. The HK Immigration Dept screened her politically. Its emails reveal just how wide-ranging the HK government's crackdown is. She was ultimately unable to go to HK. open.substack.com/pub/briankernkongtsunggan/p/this-americ…
-
zeynep tufekci reposted
Michael Weiss
@michaeldweiss
Incredible reporting here. Asma al-Assad, once the subject of a fawning Vogue article, was complicit in regime atrocities, used orphanages she ran to keep children hostage from their families, and was even considered by the Russians as a possible presidential successor to Bashar. Oh, and she travels and lives very nicely in Dubai, along with her husband.
Heidi Pett
@heidipett
The Observer reports that Bashar and Asma al Assad have residency in the UAE and have been travelling to and from Dubai, and that Asma's "last visit there is thought to have been as recent as last month, staying at the Waldorf Astoria hotel.
Assad remains one of the world’s most wanted men and the subject of multiple international arrest warrants relating to alleged war crimes committed during the conflict."
@chloehadj @SarahDadouch
observer.co.uk/news/international/article/the-case-agains… -
zeynep tufekci reposted
Atul Gawande
@Atul_Gawande
More signs of resurging HIV after USAID’s closure. African Neonatal Network reported at IMNCH (a global health conference): HIV+ pregnant women ↑50% & those not on antiretrovirals ↑>300%, year-on-year, across 22 hospitals in 5 nations. @VTOxfordNetwork
-
zeynep tufekci reposted
Michael M. Grynbaum
@grynbaum
Several @nytimes journalists were subpoenaed by the Trump administration on Friday.
– Reporters covering AF1 security concerns have been compelled to testify before a Manhattan grand jury next week
– Federal agents turned up at some reporters' homes
– NYT calls it a "brazen act" -
zeynep tufekci reposted
Ana Ley
@La__Ley
NEW: While covering round-the-clock protests at Delaney Hall’s parking lot, my colleagues and I decided we wanted to know what it was like inside the immigration detention facility. For months, inmates had pleaded for help
-
zeynep tufekci reposted
Bridget Early
@bridget__early
The lede on this NYT story is devastating. As more parents reject Vitamin K shots for their babies, docs are increasingly treating brain/abdominal hemorrhages for infants who hadn’t received the routine injection, per Maggie Astor:
-
zeynep tufekci
@zeynep
Big cyclosporiasis outbreak in 18+ states. 1000 cases just in Michigan. (Normal ~50 per year).
CDC doesn’t even seem to know of the outbreak officially.
CDC “quietly” dropped surveilling it last year, citing no funding. DOGE also cut about 20% CDC staff.
Woodchipper in action.
ABC News
@ABC
Reported cases of cyclosporiasis, an intestinal illness caused by a parasite, are nearing 1,000 in Michigan, local health officials said Wednesday. abcnews.com/Health/cyclosporiasis-cases-1000-michigan-hea…
-
zeynep tufekci
@zeynep
Where is the group chat for people who think this is interesting EXCEPT for all the anthropomorphic language that functions as aura not analytics.
Many such examples — strip the woo woo words and the research is interesting.
Anthropic
@AnthropicAI
New Anthropic research: A global workspace in language models.
Of everything happening in your brain right now, only a tiny fraction is consciously accessible—thoughts you can describe, hold in mind, and reason with.
We found a strikingly similar divide inside Claude.This media couldn't be displayed. See it on the original post.Post not available here, view on X
zeynep tufekci
@zeynep
Another irony of all this anthropomorphic hand-waving is that understanding these technical systems *on their own terms without the woo* is exactly what might help companies develop the right architectures to make them economically more viable as tools outside of coding etc.
Nathan Witkin
@NateWitkin
Yes! These systems are independently fascinating and the fact they defy easy categorization is part of what makes that so. But their significance also means nearby concepts (the brain, sentience, etc.) are going to exert very strong attractions, since most of us would prefer to understand faster rather than better. It's a shame because all sorts of interesting work trying to understand LLMs on their own terms is going undone.
zeynep tufekci
@zeynep
I should add:
Besides no mass unemployment around the corner and slowdown in many scientific fields because of generative AI, I also think (especially elite) education credentials are going to become more important, not less.
zeynep tufekci
@zeynep
A few of my “take the other side” predictions for generative AI:
-No mass unemployment
-No convergence towards mainstream ideas
-Potential slow down of scientific advances -Engagement model (ads/sales) outside of enterprise
The bad news is the real AI problems are MUCH harder.
zeynep tufekci
@zeynep
Will take the other side of this bet.
Joe Weisenthal
@TheStalwart
Great piece from @JerusalemDemsas. For as much as people like to talk about AI "hallucinations", the much bigger story might be how often what they say is actually true, and that perhaps AI (as post-social media) may usher in a new era of shared reality www.theargumentmag.com/p/are-you-there-grok-its-me-margar…
-
zeynep tufekci
@zeynep
Was this deadpan? As if we hadn’t experienced the internet era.
Let me re-up: plausible that LLMs will slow down science *today* while *also* destroying the very processes that would eventually lead to better science tomorrow.
Yes, I know of Erdős problems. Think one more step.
zeynep tufekci
@zeynep
A few of my “take the other side” predictions for generative AI:
-No mass unemployment
-No convergence towards mainstream ideas
-Potential slow down of scientific advances -Engagement model (ads/sales) outside of enterprise
The bad news is the real AI problems are MUCH harder.
zeynep tufekci
@zeynep
Will take the other side of this bet.
Joe Weisenthal
@TheStalwart
Great piece from @JerusalemDemsas. For as much as people like to talk about AI "hallucinations", the much bigger story might be how often what they say is actually true, and that perhaps AI (as post-social media) may usher in a new era of shared reality www.theargumentmag.com/p/are-you-there-grok-its-me-margar…
-
zeynep tufekci
@zeynep
There was always corruption vs open, proud corruption.
There were always Nazis on this site vs they’re welcomed.
The pardons favored those with connections vs outright selling and bestowing.
It’s not just a bit worse because movement in that direction is very hard to reverse.
zeynep tufekci
@zeynep
Easier to see when understood as a dynamic system.
Falling short of ideals: normal problems, back-and-forth movements, variation, etc.
Ideals discarded: feedback cycle activated, floodgates open, dam bursts, avalanche, etc.
zeynep tufekci
@zeynep
A better constraint but also big difference over time: it takes much less effort to course correct under the “preserving but failing to follow the ideal” regime while the other option is unleashing the feedback cycle that may well become the death spiral for the ideal/norm.
Collin Lysford
@CollinLysford
This is why I get annoyed by "Actually both sides are exactly the same, one is just pretending to follow a higher ideal and one is honest!" Okay well, pretending to follow a higher ideal is a meaningful constraint on corruption and obviously the better choice. C'mon.
zeynep tufekci
@zeynep
Easier to see when understood as a dynamic system.
Falling short of ideals: normal problems, back-and-forth movements, variation, etc.
Ideals discarded: feedback cycle activated, floodgates open, dam bursts, avalanche, etc. -
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
Nearly 1m people who bought Trump’s memecoin have lost a total of $3.81b through the end of June.
Trump, meanwhile, made $636m from the coin last year, because he collected returns whenever anyone traded it, as he repeatedly pushed his followers to do. www.nytimes.com/2026/07/04/us/politics/trump-coin-crypto-… -
zeynep tufekci reposted
Baufinanciaphaster 👹
@bauhiniacapital
The best evidence that @BillAckman lacks an appreciation for NYC and American history is that he doesn’t recognise Mayor Mamdani is sitting on the correct side of George Washington’s original desk, a desk with drawers on both sides - a design much loved and copied for 200yrs.
Bill Ackman
@BillAckman
The best evidence that @NYCMayor Mamdani has not held a job is the fact that he is sitting on the wrong side of the desk.
And listen to his words. On a day when we, and in particular, our leaders should be celebrating the 250 years of our nation’s history, he attacks us and what makes our country great.
Mamdani and the Democratic Socialists are grave threats to our present and our future. -
zeynep tufekci reposted
David Yi🌻
@seoulcialite
I thought grok was included in the subscription
Bill Ackman
@BillAckman
The best evidence that @NYCMayor Mamdani has not held a job is the fact that he is sitting on the wrong side of the desk.
And listen to his words. On a day when we, and in particular, our leaders should be celebrating the 250 years of our nation’s history, he attacks us and what makes our country great.
Mamdani and the Democratic Socialists are grave threats to our present and our future. -
zeynep tufekci reposted
derek guy
@dieworkwear
This is a Washington style desk. There are drawers on the guest side. You can still purchase them at places that specialize in traditional American furnishings.
-
zeynep tufekci reposted
The Washington Post
@washingtonpost
Hundreds of uniformed members of what appeared to be Patriot Front, a white supremacist group, marched through D.C. and toward the Capitol.
Their faces covered in white masks, the men carried flags — some upside down, others Confederate — as they chanted, “Reclaim America!” -
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
Trump also pardoned Adam Kidan, who had served about 2 1/2 years in prison for his role in a fraud scheme involving the disgraced former lobbyist Jack Abramoff.
Kidan has been a major GOP donor, including giving $270k+ to Trump in 2024.
🎁🔗 www.nytimes.com/2026/07/03/us/politics/trump-pardon-clean…
Kenneth P. Vogel
@kenvogel
NEW: Trump's flurry of pardons to Clean Air Act violators were the latest move by his administration to undermine laws intended to fight climate change & curb air pollutants that harm human health, @karenzraick & I report. www.nytimes.com/2026/07/03/us/politics/trump-pardon-clean…
-
zeynep tufekci reposted
Gabe Kaminsky
@gekaminsky
SCOOP: Five of the six people pardoned by President Trump today were represented by lobbyist Jeff Daugherty + his team, Daugherty tells @CBSNews in an interview on Friday.
An illustration of how the surge in clemency lobbying is paying BIG dividends:
www.cbsnews.com/news/trump-announces-pardons-pollution-vi… -
zeynep tufekci reposted
Time Out New York
@TimeOutNewYork
Tall ships from around the world will descend upon the city this week. Here’s how to see the parade and tour the ships. www.timeout.com/newyork/news/everything-you-need-to-know-…
-
zeynep tufekci reposted
Princeton School of Public & International Affairs
@PrincetonSPIA
SPIA NEWSMAKERS: @MFlaherty17 weighs in on the ramifications of the U.S. Supreme Court's decision in CISCO Systems, Inc. v. Doe. www.scotusblog.com/2026/07/aiding-and-abetting-impunity/
-
zeynep tufekci reposted
Council on Foreign Relations
@CFR_org
A Supreme Court decision on June 23 in favor of Cisco Systems, a major tech corporation, has "effectively cleared the way for U.S. corporations to enable foreign governments, particularly repressive regimes, that would surveil individuals and violate their human rights," writes expert David J. Scheffer. www.cfr.org/articles/the-supreme-courts-cisco-ruling-clea…
-
zeynep tufekci reposted
Financial Times
@FT
OpenAI proposes handing Trump administration 5% stake ft.trib.al/vSxbmZE
-
zeynep tufekci reposted
The Wall Street Journal
@WSJ
The Middle East has been good for President Trump’s bottom line.
Entities from the Gulf paid around $300 million to the president’s businesses last year, more than any other foreign region identifiable in his financial disclosures. www.wsj.com/politics/policy/trump-family-rakes-in-cash-fr… -
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
NEW: Trump purchased as much as $5m each in Broadcom, Meta, Amazon, Apple, Microsoft & Nvidia on 7/23/2025.
That same day, his Trump administration unveiled its AI action plan, which could help the cos.
Prices for most of the stocks have since increased. www.nytimes.com/2026/07/01/us/politics/trump-crypto-memec… -
zeynep tufekci reposted
Financial Times
@FT
Donald Trump declared making more than 22,000 stock transactions in 2025, according to the FT analysis. His immediate predecessor, Joe Biden, made 13 transactions over four years. In his first term, Trump made 517. ft.trib.al/lE26Tv9
-
zeynep tufekci reposted
Sam Kessler
@skesslr
We break down the rest of the president’s financial disclosures in our article today.
Give it a look!
Story w/ @neilmhta @JohnWest @DavidUberti @pswsj -
zeynep tufekci reposted
Sam Kessler
@skesslr
For the first time, Trump appears to have formally disclosed how much he earned from selling 49% of his crypto company to UAE investors: $263M
We reported on the secret deal with the UAE’s ‘Spy Sheikh’ back in January.
[Gift link]
www.wsj.com/finance/trump-investments-presidency-4ca88728 -
zeynep tufekci
@zeynep
Anyone know of reporting for estimates of payouts to such fake-bait accounts?
It’s a dumb one. Not really satire — deliberate misspellings to be smugly corrected are lazy but effective for catching the people whose brains are fried on this app. It appears like a pure money play.Post not available here, view on X
zeynep tufekci
@zeynep
I know!
Anyway, my “what’s the earnings” question isn’t about that account per se. It’s just there’s a lot of such accounts!
I picked that one exactly because it’s well-known but also people keep falling for it. It works. So it’s a member of a class.
Ondrej Maly
@ondrej_m
This account is so obviously satirical it even has its own wikipedia page - en.wikipedia.org/wiki/Jack_Kimble
zeynep tufekci
@zeynep
Wasn’t litigating this account, just picked it somewhat randomly because it got verified (thus payout-eligible) in 2024 and it’s a well-known fake that people still fall for! I was contrasting to to no-money incentive involved pure satire/parody accounts.
My motivation is trying to see if LLMs change the payout calculus for fake-bait accounts.
But yes, thanks to everyone telling me it’s fake, which makes the monetization strategy obvious, lol. (It’s like the Godfather scene in Barbie around here but I digress).
Jacob
@Perrid13
I think this “parody” account predates monetization and is into rage baiting people for the fun of it.
-
zeynep tufekci reposted
Charles Ornstein
@charlesornstein
A Troubling Milestone: The Supreme Court now decides more cases using its shadow docket than it does on its merits docket, the first time ever that has
been the case, a @propublica analysis found. www.propublica.org/article/supreme-court-shadow-docket-ru… -
zeynep tufekci
@zeynep
I don’t see mass unemployment from generative AI soon because of structural and technical features of the tech.
“It will get better!” When we don’t just say “AGI, AGI, AGI” while waving hands and discuss the actual technology, it’s clear why that’s not easy.
LLMs are aliens.
neil turkewitz
@neilturkewitz
“GenAI as it currently exists cannot easily replace human beings, because it cannot manifest human intelligence. That won’t stop it, however, from destabilizing society in ways more profound than we might even imagine.”
@zeynep @nytimes
www.nytimes.com/2026/06/30/opinion/ai-agents-steal-jobs-e…
zeynep tufekci
@zeynep
Yes, Erdős conjecture debates are fun, but that’s a verifiable, formal domain.
Many eterministic tasks have long been automated by regular software and what remains isn’t easy prey for gen AI. (See: previous AI winters).
The LLM ability in complex language is misleading us.
zeynep tufekci
@zeynep
“Plausibility Engine” is my coinage, and it is diametrically opposed to “autocomplete on steroids”, “stochastic parrots” or 2023 views. 🤷♀️
After I used the phrase in my Neurips keynote, I had great conversations with AI researchers on the power plausibility engines, no problems.
Megan McArdle
@asymmetricinfo
Phrases that make me stop reading whatever you've written about AI: "Stochastic parrots", "Plausibility engines", "autocomplete on steroids", "fancy pattern matching". They suggest you are merely regurgitating shallow critics who haven't updated their views since 2023.
zeynep tufekci
@zeynep
Also “plausibility engine” tries to capture the remarkable power of LLMs without belittling while ALSO avoiding the currently worse than useless terminology around AGI or ASI — those have essentially become useless, mystical terms to confuse journalists and the public, imo.
-
zeynep tufekci reposted
The New York Times
@nytimes
President Trump reaped at least $2.2 billion from his family’s holdings last year, a new filing shows.
www.nytimes.com/2026/06/30/us/politics/trump-financial-di… -
zeynep tufekci reposted
Acyn
@Acyn
Vance: This was only a 5-4 decision, That’s why the midterms are so important. It's the senators who ultimately vote on the supreme court justices. Imagine if one of the five justices who made a bad call today, if they left the supreme court, we want to make sure we get somebody good on there in the future.
This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
The Washington Post
@washingtonpost
Breaking news: President Trump reported more than $1.4 billion in income from crypto, digital tokens and related partnerships in 2025, appearing to have substantially increased his income after reentering the White House last year. www.washingtonpost.com/politics/2026/06/30/trump-earned-o…
-
zeynep tufekci
@zeynep
No it came within *one* vote. Remember all the “surely 9-0!!!” predictions?
About 17% of the country — a relatively homogeneous group — can control two branches: Senate and thus also the judiciary. We’ve now seen few decades of effective efforts to use that minority power.
Eric Lee
@EricLeeAtty
SCOTUS came within 2 votes of allowing the president to effectively overturn the 14th Amendment. Anything less than 9-0 on birthright citizenship would have been a scandal in earlier period. A significant decision today, but far, far too close a vote.
Post not available here, view on X
zeynep tufekci
@zeynep
As per the two dissenters people count as allowing citizenship to children of undocumented immigrants: Parents would have prove that they’ve long been here illegally before the baby gets citizenship.
Then probably deported.
This would be a fundamental change, obviously.
Robby Soave
@robbysoave
The dissent from Neil Gorsuch in the birthright citizenship case is actually very interesting. He basically says that he would deny birthright citizenship to children of temporary visitors but not to illegal immigrants who lived here permanently. So on the question of whether illegal immigrants' kids are citizens, it's actually more like a 7-2 decision.
zeynep tufekci
@zeynep
The key question for the Supreme Court is whether going from birthright citizenship to a very different regime requires a constitutional amendment given the text of the 14th amendment.
For four or maybe five justices, that’s a no, after a year of “9-0 slam” dunk predictions. -
zeynep tufekci reposted
Luiz Pessoa
@PessoaBrain
Excellent piece about AI and the future of work.
(congratulations on such a clear and cogent piece)
zeynep tufekci
@zeynep
I don’t see mass unemployment from generative AI soon because of structural and technical features of the tech.
“It will get better!” When we don’t just say “AGI, AGI, AGI” while waving hands and discuss the actual technology, it’s clear why that’s not easy.
LLMs are aliens.
neil turkewitz
@neilturkewitz
“GenAI as it currently exists cannot easily replace human beings, because it cannot manifest human intelligence. That won’t stop it, however, from destabilizing society in ways more profound than we might even imagine.”
@zeynep @nytimes
www.nytimes.com/2026/06/30/opinion/ai-agents-steal-jobs-e… -
zeynep tufekci reposted
Samuel Moyn 🔭
@samuelmoyn
Wow, only 5-4 on the constitutional holding. Trump came within one vote of successfully revising the Fourteenth Amendment almost out of nowhere. Victory for birthright citizenship -- for sure. But almost a defeat. Raising the question of why we are playing this game.
-
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
The Trump administration has provided or is considering providing $8.9B+ in funding to 14 firms with ties to the Trump &/or Lutnick families that are actively working with the US government on mining deals, @PaulSonne & @EricLiptonNYT report. @nytimes 🎁🔗 www.nytimes.com/2026/06/28/world/europe/trump-lutnick-son…
-
zeynep tufekci
@zeynep
It’s true that the Unites States has the infrastructure and size but what international organization would ever award a big sporting event to the United States again?
These things are set years in advance so an election won’t fix this. They’ll be afraid of the visa situation.
Dan O'Donnell
@DanODonnellShow
The U.S. should just host every World Cup and Olympics moving forward. We're the only nation that doesn't have to nearly bankrupt ourselves to do so and obviously people from all over the world love coming here.
Post not available here, view on X
zeynep tufekci
@zeynep
That’s not even true! Depends on the passport.
But it’s not just fans, teams and staff have been struggling.
Look, New York is having a glorious summer and the national sports crowds are an amazing addition. But I don’t think international organizations will risk this again.
Echo Channel
@channel_echo
Visa risks are low for spectators who can afford to spend $10-20k or more to attend world famous sporting events. Visas for attendees and athletes for under the radar events are less predictable. Olympics and World Cup is fine. Volleyball world championship? Better off elsewhere
zeynep tufekci
@zeynep
Attendance is high because a lot of Americans attending World Cup games but international travel to the event is much, much lower than expected.
I didn’t say Russia was a good alternative.🤷♀️
Canada got 17K applications — tiny number don’t know what’s the makeup that group.
smackemwhackem
@smackemwhackem
As if other countries will allow or allowed everyone in?Have you seen Canadian visa rejections?This is from the World Cup in Russia.This WC is breaking all attendance record &it is one of the most successful WC already.The agenda will agenda as they say.
www.inquirer.com/soccer/world-cup-philadelphia-french-fan… -
zeynep tufekci reposted
rat king 🐀
@MikeIsaac
more news: Mark Zuckerberg is telling lieutenants to pursue partnerships with Polymarket and Kalshi with Meta's new prediction market app called Arena, per sources
Meta is targeting 18-34 year old users with a goal of 100 million regular "predictors"
www.nytimes.com/2026/06/26/technology/zuckerberg-meta-pol… -
zeynep tufekci reposted
Heidi Vanderlee
@heidivanderlee
Not the dress code 😂😂😂
New York Post
@nypost
Zohran Mamdani jumps into NYC pool to kick off summer tradition - while violating dress code rules nypost.com/2026/06/27/us-news/zohran-mamdani-jumps-into-n…
-
zeynep tufekci reposted
Tim Hirschel-Burns
@TimH_B
There’s also a big difference between setting medium-term aid levels and short term evisceration of programs.
One thing to not provide meds to a clinic. Another to provide them for years and then suddenly stop with no warning or time for other donors or national govs to step in -
zeynep tufekci reposted
Matthew Sitman
@MatthewSitman
Actually it was quite common for people to say, for example, that PEPFAR was one thing you really did have to hand it to George W. Bush for, a genuinely transformative program that saved millions of lives. Seriously!
PoIiMath
@politicalmath
One of the reasons that I don't accept the "nuking USAID killed millions of people" framing is because most of the people saying this never took the stance that the US was a tremendously good and heroic force for saving millions of people
-
zeynep tufekci reposted
Matt Darling 🌐🏗️
@besttrousers
"In other words, the toll from USAID cuts by 2030 seems to be *at best* around two-thirds of a million people; that’s about as many people as were killed during the Civil War. At worst, Musk is tied to the deaths of millions."
www.theargumentmag.com/p/elon-musks-zero-accountability-life -
zeynep tufekci reposted
rohit
@krishnanrohit
We're all looking for the guy who did this
-
zeynep tufekci
@zeynep
Many cardiologists say no need for AI since a problem was obviously visible, which the story itself notes, and others say stuff like this does get missed, also true.
Which goes to the problem with drawing conclusions based on single cases either way. Effects should always be considered at scale.
How often does the A.I. system misdiagnose or overdiagnose? How many extra steps that carry some amount of risk does the A.I. system generate? What are the resources required?
Maybe, and quite possibly, this is a great extra safety net, and maybe, and we have had examples like this, it’s actually something net harmful at scale.
That’s why we study things. This incident happened as part of a study, so will be interesting to see what actually happened.
Nikhil Krishnan
@nikillinit
I find the responses to this article pretty interesting - a lot of docs are saying that the patient obviously required more work up and didn't need AI to tell them that
but I feel like the point is that there are lots of scenarios exactly like this one where a patient might be overlooked BECAUSE attention of clinicians might be split across many different things
I feel like many docs see this and think about how it would have been caught if it was their institution or if it was them
In a perfect world there would be 100% attention all the time from the best docs, but that's not the reality. I'm optimistic about the fact that we have tools now that can:
- Assess levels of severity and triage using more commonly captured streams of data (e.g. ECGs)
- Provide that expertise to more people and settings. Not just cardiologists but generalists and non-docs .
- Provide it for much lower cost
Feels like it's a win for patients and docs! but curious where the pushback is coming from
Pierre Elias, MD
@PierreEliasMD
We invented a new biomarker for cardiovascular disease. The New York Times tells our story (link in comments).
-
zeynep tufekci reposted
James Rosen-Birch ⚖️🕊️
@provisionalidea
“voluntary”
Andrew Curran
@AndrewCurran_
The Trump administration is pressuring META to agree to submit its models to the government for voluntary review. META is now the only holdout, as OpenAI, Anthropic, Google, xAI and Microsoft have all agreed to these terms. Reporting by the NYT.
-
zeynep tufekci reposted
Alex Kane
@alexbkane
Stunning that this was, in part, set in motion by the arrest of Mahmoud Khalil theintercept.com/2026/06/23/new-york-primary-adriano-espa…
-
zeynep tufekci reposted
Acyn
@Acyn
CNN Projecting a sweep for Zohran Mamdani in NY Primaries.
-
zeynep tufekci reposted
The Intercept
@theintercept
Daniel Sanchez Estrada wasn’t accused of attempted murder or material support of terrorism after a protest turned catastrophically wrong outside an ICE detention center in Alvarado, Texas. He was convicted of obstructing the investigation by moving a box full of antifascist zines after the protest. U.S. District Judge Reed O’Connor gave Sanchez Estrada a 30-year term.
-
zeynep tufekci reposted
Bijan Salehizadeh
@bijans
“This person was a 79-year-old man at the time the request was made in April, according to three sources familiar with the matter…. A senior clinician at the National Institutes of Health … requested the drug to treat the patient for refractory obesity with obstructive sleep apnea and pulmonary hypertension”
Damian Garde
@damiangarde
Eli Lilly gave extraordinary obesity drug access to a 79-year-old patient. Who was it? www.statnews.com/2026/06/23/eli-lilly-unusual-weight-loss…
-
zeynep tufekci reposted
Thomas Woodside
@Thomas_Woodside
The California Attorney General is hiring for two technical roles to support the enforcement of consumer protection and technology laws, including SB 53. Having technical talent at enforcement agencies is so important to successful implementation!
-
zeynep tufekci reposted
Madhav Chanchani
@madhavchanchani
Polymarket created a clone website so creators could fake winning trades, according to WSJ.
It also used an agency to promote videos showing how easily insider trading could be carried out on the platform. -
zeynep tufekci
@zeynep
So lucky that we’re not in an era with any downsides for “our whole lives to be recorded” and with our physical and digital signal density to go to 10,000x.
Nothing but a “worship of human life” would come of that. Not in 2026.
roon
@tszzl
this will be true across all walks of life
I would expect our whole lives to be recorded and the sheer signal density of your physical and digital presence to go 10,000x
agents and computers to sift through all of it
really it’s a form of worship of human life
Amanda Askell
@AmandaAskell
A counter to this is "yes but people *don't* ignore it". But not ignoring things on scans is our norm because, until recently, we only did scans if there was a clear need. If we move to a scan-more-often paradigm, the norms of what we do with that information will surely adjust.
Post not available here, view on X
zeynep tufekci
@zeynep
This is, in fact, a literal textbook example of Bayes’ theorem.
The missing piece is a better understanding of the actual statistical signal ultrasound can produce in screening AND the risk stats for generating necessary information to weed out the false positives.
Jeffrey Ladish
@JeffLadish
Of course I'm pro scanning! False positives are not a fundamental problem. Literal textbook example of Bayes' Theorem. Wtf is wrong with people
zeynep tufekci
@zeynep
I agree with this! Too many medical decisions are optimized at population level even when good individual level tests exist, held back by costs and doctor obstinacy.
But asymptomatic ultrasounds aren’t that! Look up ultrasound! Be Bayesian!
Ok, disrupt but where is the effort?
Yaron (Ron) Minsky
@yminsky
There's a lot of decision making that's driven by very broad brush best practices that are optimized at a population level, and are often very far from optimal when considered at an individual level.
-
zeynep tufekci
@zeynep
I still closely follow the science of (likely) autoimmune severe subset of Long Covid & related disorders.
This WOULD benefit from medicine overcoming the mind/body duality, but the recent Wired article was sadly the opposite of that. Scientifically poor, outdated & misleading.
zeynep tufekci
@zeynep
I read your article fully and carefully. It’s a an empirically poor article. I have been reporting in this space for years, resulting in full page articles in the New York Times which involved interviewing, at length, lead PIs of the billion dollar NIH research fund, NIH agency leaders at the highest level, and leading scientists working on this field. Additionally, as an academic who got interested in this important, fascinating space, I am still reading the scientific literature and regularly keeping up with cutting edge science. They’re all missing in your article!
I have written about poor definitions and I face flak from online patient communities regularly. I believe the mind/body duality as it exists is medicine is a big obstacle. Your article doesn’t have any of the cutting edge science on that! Where are the neuro-immunologists from psychiatry? Where are the sharp exercise physiologists getting out bangers in Nature?
What you have is essentially a channeling of a gripe session from one (empirically left behind and upset) faction of the mind-body duality. It’s not challenging it.
You can find patients saying stupid things online. But the idea that brain retraining is some breakthrough that’s being ignored and that people can’t talk about is complete nonsense. And your part about exercise is missing the biomedical breakthroughs in understanding from the last few years.
I don’t really do social media these days but I am informed. This one deserved clarification for the record.Post not available here, view on X
zeynep tufekci
@zeynep
Still no explanation.
If someone is not aware of *big* money and very high profile trials dedicated to the very thing they’re claiming “can’t be talked about”, that person should not write about the topic.
And incompetence is a kinder explanation of such an incredible omission.
zeynep tufekci
@zeynep
Engage critics: you immediately insulted me and told me I must have not done my research even though I have been writing on this topic for years.
Meanwhile, your article doesn’t even mention huge trials into the very things that your article alleges that “we can’t talk about.”
davidtuller
@davidtuller1
In a 7,600-word piece in @WIRED on unhinged patients having essentially blocked research into mind-body stuff, @AlanLevinovitz failed to mention two major clinical trials. Can someone explain to me how that happened? What will Wired do about it? Is it bias, incompetence, or both?
zeynep tufekci
@zeynep
Hardly on X now but I had to respond here as that article is missing most basic facts and is highly misleading. No response from the author to major and substantive criticisms — well, except insults.
How many people let that Andrew Wakefield paper through despite the red flags?
Charles Arthur
@charlesarthur
Zeynep absolutely roasting @AlanLevinovitz here.
zeynep tufekci
@zeynep
Engage critics: you immediately insulted me and told me I must have not done my research even though I have been writing on this topic for years.
Meanwhile, your article doesn’t even mention huge trials into the very things that your article alleges that “we can’t talk about.”
davidtuller
@davidtuller1
In a 7,600-word piece in @WIRED on unhinged patients having essentially blocked research into mind-body stuff, @AlanLevinovitz failed to mention two major clinical trials. Can someone explain to me how that happened? What will Wired do about it? Is it bias, incompetence, or both?
zeynep tufekci
@zeynep
So disturbing for @Wired.
The author of that terrible, misleading Long Covid piece portrayed a patient who is an academic as an anti-science loon, omitted many key relevant facts on why she was actually reasonable, AND de facto outed her but never contacted her for comment. Wow.
davidtuller
@davidtuller1
The recent Long Covid piece in @WIRED criticized Univ of Va law professor @MeganTStevenson for harming scientific progress, didn't offer her a chance to respond, and then included enough details so anyone could identify her. virology.ws/2026/06/16/trial-by-error-the-wired-debate-ov…
-
zeynep tufekci reposted
molly taft
@mollytaft
NEW: the US government intervened Monday on behalf of xai in a lawsuit over its gas turbines, claiming that the company needs to run the turbines in order to power Grok, which is "vital" for national security
lots of wild stuff in the filing, more here www.wired.com/story/doj-lawyers-argue-xai-vital-national-… -
zeynep tufekci
@zeynep
The gap between actual NYC and the NYC as mostly imagined on X has jumped all sharks. Not the only topic either.
Social and mass media both have structural weaknesses. But there is weakness, and there is practically no value except to get misled.
(Yes, real photo from 2017.)Post not available here, view on X
zeynep tufekci
@zeynep
I know! I hear such dire concerns from people who don’t live here. And I know how false they are mostly because I was out really, really late and went from place to place and then traveled back through many chokepoints. (Coincidence of other social events plus Knicks, then World Cup!)
And like most people who live here, I could immediately see pretty much everything back to usual next morning, and not that much happened that needed fixing anyway, certainly lighter in ratio than say, UNC/Duke game aftermaths back in Chapel Hill when I was there.
It’s like this for most everything NYC (the mayor, ispecialized schools or G&T, the subway or transit, how rent control actually works, etc — I get why). But many other topics as well! A predictable and structural problem, and yet so many people still think they can use it effectively or casually with just a bit of adjustment. It was never perfect, but it wasn’t this.
I’m just posting reminders, I guess.
Nick Riccardi
@NickRiccardi
My 13-yr-old and his buddies think NY was basically destroyed by celebrating Knicks fans Saturday night
zeynep tufekci
@zeynep
The gap between actual NYC and the NYC as mostly imagined on X has jumped all sharks. Not the only topic either.
Social and mass media both have structural weaknesses. But there is weakness, and there is practically no value except to get misled.
(Yes, real photo from 2017.)
zeynep tufekci
@zeynep
Of course not. But they don’t have a large legacy class of people who tried to use them as agenda setting in a news gathering manner, or as open source information.
Ivor Crotty
@IvorCrotty
Instagram and Facebook, Snapchat and TikTok are accurate mirrors of society and media reporting however....
zeynep tufekci
@zeynep
The gap between actual NYC and the NYC as mostly imagined on X has jumped all sharks. Not the only topic either.
Social and mass media both have structural weaknesses. But there is weakness, and there is practically no value except to get misled.
(Yes, real photo from 2017.) -
zeynep tufekci
@zeynep
Is there the *slightest* case that Fable poses a risk different from what’s readily available through other models?
Open to being convinced but I don’t see any meaningful difference?
(Not asking what people who wrote the order believe. I’m wondering if there is *any* case).
Mike Kafanov
@AveMisha
The Fable/Mythos export controls are so unexpected, and so leaky at denying access to geopolitical adversaries, that they almost look like a risky pre-IPO PR gamble by Anthropic's friends in government, and a backfiring one at that. Consider:
1. NVIDIA hardware ends up in China easily enough
2. Businesses around the world supply the war efforts of sanctioned states
3. Hacking isn't exactly legal, yet North Korea routinely steals billions
4. Proprietary tech gets stolen by people so deeply embedded and incentivized (see any number of corporate-espionage cases) that no blanket citizenship policy could stop it
Basically, it looks like another Singapore situation (Singapore being China's main VPN gateway for, among other things, accessing Western AI models), except now you'd just get a wave of new American residents instead
If the market reads this as "the government can evaporate 30–40% of any AI lab's total addressable market at any moment," the upcoming AI IPOs are looking even more interestingPost not available here, view on X
zeynep tufekci
@zeynep
This whole thing doesn’t seem knee-jerk, improvised or reactive to me, tbh. Pretty clear, coherent messaging.
(And it certainly doesn’t seem like it’s about some unique, urgent risk that Fable poses compared to everything else. But Anthropic can’t say that, right?)
zeynep tufekci
@zeynep
I don’t think the specifics of the jailbreak matter one bit. Various levels of jailbreak was predictable, and this ban wouldn’t deter any “national security threat” entity.
They’ve sent a clear, coherent message right after market close on Friday — the time-honored way to do it.
nxthompson
@nxthompson
Here is the gov’s argument. Hard to evaluate whether the concern is justified—or just politics toward an American company this government often whacks—without the specifics of the jailbreak.
Also not clear why they chose to regulate via a blunt and hard-to-follow export control.
David Sacks
@DavidSacks
I’ve had a number of conversations with folks inside and outside government about the current situation with Anthropic, and here is what I believe to be true:
— As we know, Anthropic publicly released its Mythos class models earlier this week under the commercial name Fable.
— Fable is Mythos with guardrails. But if those guardrails fail, then you’ve exposed Mythos and its advanced cyber capabilities to people who shouldn’t have them. (Keep in mind that Anthropic itself widely promoted the idea that Mythos was a cyberweapon and needed to be regulated as such. They asked for government regulation of Mythos and championed the guardrails on Fable. If there is a vulnerability — big or small — it is Anthropic’s responsibility to patch.)
— A highly credible trusted partner of both Anthropic and the USG who was testing Fable came forward with a jailbreak of those guardrails. The Admin asked Dario to fix the jailbreak or de-deploy the model. Dario refused.
— In their blog post, Anthropic defended its decision by saying the jailbreak isn’t serious. That is not what the trusted partner and the USG believe; nor is that kind of minimizing language consistent with Anthropic’s brand as the AI safety company. It’s difficult to fathom how they could claim a jailbreak allowing operability of a cyber weapon could be defined as not “serious.”
— In the past, Anthropic has always said that safety must be top priority and taken super seriously. In this case, Anthropic prioritized the continued offering of the consumer model over safety.
— In reaction, the Admin issued the export control. The Admin did this reluctantly. It’s been very surprised that Anthropic hasn’t wanted to cooperate with a reasonable safety request (ie fixing the jailbreak issue). Anthropic’s reaction is very much at odds with their branding and ethos as a safe AI research community.
— The Admin’s hope now is that Anthropic remediates the safety issue, the export control is lifted, and Fable goes back into general release. The Admin wants all of this to happen as soon as possible. It is frankly bewildered that Anthropic hasn’t wanted to comply with safety requests that it previously said were its highest priority.
— Those trying to misdirect and tie this action to the prior DoW/Anthropic issues are wrong. The Admin values Anthropic’s technical capabilities and feels that this issue, while serious, should be easily resolved. The ball is in Anthropic’s court.
zeynep tufekci
@zeynep
If Fable solely empowers attackers (rather than allowing Amazon to fix that bug), what “guardrails” can fix that?
Jailbreak or model weights walking would be inevitable, and citizen vs foreigner be irrelevant, no?
I think admin wants knee bent, preferrably before 9:29a Monday.
Kevin Lacker
@lacker
It’s possible. There could easily be a very specific risk that we don’t know about. Like, say, if someone at AWS discovered that Fable can be jailbroken to hack any AWS-hosted site. That would be a big risk, even though in a sense it’s “just one bug”.
-
zeynep tufekci reposted
S. Holland 💙💛🇺🇦Glory to Ukraine!! 🇺🇦💛💙
@Erasmus77
Thread on why the @Wired article on Long Covid is dangerously misleading & incompetent.
zeynep tufekci
@zeynep
I still closely follow the science of (likely) autoimmune severe subset of Long Covid & related disorders.
This WOULD benefit from medicine overcoming the mind/body duality, but the recent Wired article was sadly the opposite of that. Scientifically poor, outdated & misleading.
zeynep tufekci
@zeynep
I read your article fully and carefully. It’s a an empirically poor article. I have been reporting in this space for years, resulting in full page articles in the New York Times which involved interviewing, at length, lead PIs of the billion dollar NIH research fund, NIH agency leaders at the highest level, and leading scientists working on this field. Additionally, as an academic who got interested in this important, fascinating space, I am still reading the scientific literature and regularly keeping up with cutting edge science. They’re all missing in your article!
I have written about poor definitions and I face flak from online patient communities regularly. I believe the mind/body duality as it exists is medicine is a big obstacle. Your article doesn’t have any of the cutting edge science on that! Where are the neuro-immunologists from psychiatry? Where are the sharp exercise physiologists getting out bangers in Nature?
What you have is essentially a channeling of a gripe session from one (empirically left behind and upset) faction of the mind-body duality. It’s not challenging it.
You can find patients saying stupid things online. But the idea that brain retraining is some breakthrough that’s being ignored and that people can’t talk about is complete nonsense. And your part about exercise is missing the biomedical breakthroughs in understanding from the last few years.
I don’t really do social media these days but I am informed. This one deserved clarification for the record. -
zeynep tufekci reposted
Jonathan Berk
@berkie1
Has there ever been an op-ed that aged worse than this one?
-
Post not available here, view on X
zeynep tufekci
@zeynep
Engage critics: you immediately insulted me and told me I must have not done my research even though I have been writing on this topic for years.
Meanwhile, your article doesn’t even mention huge trials into the very things that your article alleges that “we can’t talk about.”
davidtuller
@davidtuller1
In a 7,600-word piece in @WIRED on unhinged patients having essentially blocked research into mind-body stuff, @AlanLevinovitz failed to mention two major clinical trials. Can someone explain to me how that happened? What will Wired do about it? Is it bias, incompetence, or both?
-
zeynep tufekci
@zeynep
I’m thrilled to be giving the 2026 US-UK Fulbright Distinguished Lecture, titled “Are We Having the Wrong Nightmares about AI?”
The in-person event in London is completely full, but there is a free livestream: www.eventbrite.co.uk/e/dr-zeynep-tufekci-are-we-having-th… -
Meighan Stone
@meighanstone
This @AlanLevinovitz & @zeynep exchange says it all: he fails to research her writing before responding (her significant body of work pre-dates his by years), insults her intellect despite her clear credentials, then full retreat to ad hominems after standard academic critique?
Alan Levinovitz
@AlanLevinovitz
Did you read the article? Not only did I talk to over fifty patients at a range of severe conditions, I got myself out of the online advocacy space and talked with the (much larger) patient community who are heterogeneous, respond to a wide variety of interventions — many that look a lot like gentle exercise, there’s an example in the piece of someone who went to Yale’s clinic — and are not represented by the loudest voices on the Internet. I spoke with the leading experts in every subfield related to long Covid, from neuroimmunology to statistical genetics. What I did, that you did not, is I also talked to subfield experts who aren’t bound up in a highly compromised and culturally fraught research area. They uniformly reported the same thing. The research is poor, preliminary, or misrepresented (see below for genetics research).
In the area of exercise, the situation is more dire, because the status quo actively ignores the strongest research, actively harming patients who could benefit from high quality work on tailored exercise regimens.
I did my research, I talked to a very wide range of patients (who are cited in the piece), far wider than it seems you have, and a far wider range of researchers than you have, from in and outside the field.
So save me your patronizing replies to an article you hopefully haven’t read (because otherwise your reading skills are compromised), and get back to me when you’ve done the legwork it takes to actually understand this field, rather than remaining in the long Covid research echo chamber.
zeynep tufekci
@zeynep
I RT’ed because it’s indicative, not because of personal issues. The article fails at so many levels. You can certainly find examples of toxic comments online, but he seems to be trying to use them to validate the article. That’s not good.
But you can see the many substantive criticisms he just ignores or can’t even recognize in that thread. He doesn’t even seem to know of much of the biomedical research I’m talking about. He also doesn’t recognize the context of the online factions. It’s a pity. -
zeynep tufekci reposted
Meighan Stone
@meighanstone
This @AlanLevinovitz & @zeynep exchange says it all: he fails to research her writing before responding (her significant body of work pre-dates his by years), insults her intellect despite her clear credentials, then full retreat to ad hominems after standard academic critique?
Alan Levinovitz
@AlanLevinovitz
Did you read the article? Not only did I talk to over fifty patients at a range of severe conditions, I got myself out of the online advocacy space and talked with the (much larger) patient community who are heterogeneous, respond to a wide variety of interventions — many that look a lot like gentle exercise, there’s an example in the piece of someone who went to Yale’s clinic — and are not represented by the loudest voices on the Internet. I spoke with the leading experts in every subfield related to long Covid, from neuroimmunology to statistical genetics. What I did, that you did not, is I also talked to subfield experts who aren’t bound up in a highly compromised and culturally fraught research area. They uniformly reported the same thing. The research is poor, preliminary, or misrepresented (see below for genetics research).
In the area of exercise, the situation is more dire, because the status quo actively ignores the strongest research, actively harming patients who could benefit from high quality work on tailored exercise regimens.
I did my research, I talked to a very wide range of patients (who are cited in the piece), far wider than it seems you have, and a far wider range of researchers than you have, from in and outside the field.
So save me your patronizing replies to an article you hopefully haven’t read (because otherwise your reading skills are compromised), and get back to me when you’ve done the legwork it takes to actually understand this field, rather than remaining in the long Covid research echo chamber. -
Megan Stevenson
@MeganTStevenson
Turns out I was featured in this very dumb article about long covid. And the guy lives in my small town and didn’t even reach out to ask for my perspective! Just sent him this email.
This post is unavailable. See it on X.Post not available here, view on X
zeynep tufekci
@zeynep
(And it’s not that “brain retaining” under different names hasn’t helped some people — that’s actually also a very interesting field, given the mind is in the brain which is in the body. There’s even some good research on meditation-type interventions and immune system resets, etc. Autoimmune disorders tend to be intermitting/relapsing, and the neuro-immune equilibrium theories and ways to achieve resets that work for some are fascinating.
But the framing here is genuinely like blaming TB on being too sensitive and assigning a rest cure. Rest helps! But not because TB was caused by being too intelligent and emotionally delicate. And guess what actually works? Proper antibiotics. But that needs the correct theory).
Alan Levinovitz
@AlanLevinovitz
Zeynep — this is not like TB (or cancer, or HIV, or any of the other analogized conditions), because there is no "long Covid" in the first place. There's an extremely vague definition that could apply to a range of conditions, from post-ICU syndrome to basically every PAIS that's out there. Much of the article is dedicated to discussing the definitional problems, as well as the problems that creates for studying "the" condition in the first place. The piece isn't saying that "mind-body" therapies work, or don't work, for long Covid, or any PAIS. It's saying that right now there IS no single entity we can call "long Covid," there's just vague definitions of a family of syndromes and symptoms, which result in everything from extremely moderate to very severe illness.
Equally important, and also the point of the article: A subset of very, very ill people — and a larger subset of mild/moderate — attribute their recovery to mind-body interventions. Was it spontaneous remission? Was it "placebo"? Was it...something else? It would be great to know! But that patient group, specifically, has been of zero interest to researchers, and are actively ostracized by the advocacy world.
No one studies *recovered* LC patients who were severely ill and attribute their recovery to mind-body therapies, as, for example, they do over in ALS where the very, very few spontaneous remissions are of great interest to researchers. Instead, people like Larson and Lupi are told they are full of shit, or were never really sick, or are wrong about what caused their recovery.
That's what the article is about. Other articles can discuss other things! But those are the dynamics and themes I address. They are important. Failing to address them is holding back research on a very heterogenous and poorly defined set of conditions.
zeynep tufekci
@zeynep
I have written many, many articles about the problem of “Long Covid” being badly defined. But that’s not your article’s problem. Different subsets of LC have different challenges. But the pieces completely misses all the interesting actual recent science, including as it relates to the subset that’s helped by versions of “brain retraining” and the problem there isn’t bad definition.
The issue with your article is that, as often is the case in medical history, lack of understanding has led to a subset of MDs and experts passing off terrible and outdated science as “advances” or “explanations”, and your article reads like a therapy session for them. This could’ve been solved by seeking out the scientific frontier from legitimate sources rather than experts whose feelings are hurt because scientific advances have shown them wrong. But they’re completely missing in your article!Post not available here, view on X
Alan Levinovitz
@AlanLevinovitz
Did you read the article? Not only did I talk to over fifty patients at a range of severe conditions, I got myself out of the online advocacy space and talked with the (much larger) patient community who are heterogeneous, respond to a wide variety of interventions — many that look a lot like gentle exercise, there’s an example in the piece of someone who went to Yale’s clinic — and are not represented by the loudest voices on the Internet. I spoke with the leading experts in every subfield related to long Covid, from neuroimmunology to statistical genetics. What I did, that you did not, is I also talked to subfield experts who aren’t bound up in a highly compromised and culturally fraught research area. They uniformly reported the same thing. The research is poor, preliminary, or misrepresented (see below for genetics research).
In the area of exercise, the situation is more dire, because the status quo actively ignores the strongest research, actively harming patients who could benefit from high quality work on tailored exercise regimens.
I did my research, I talked to a very wide range of patients (who are cited in the piece), far wider than it seems you have, and a far wider range of researchers than you have, from in and outside the field.
So save me your patronizing replies to an article you hopefully haven’t read (because otherwise your reading skills are compromised), and get back to me when you’ve done the legwork it takes to actually understand this field, rather than remaining in the long Covid research echo chamber.
zeynep tufekci
@zeynep
I read your article fully and carefully. It’s a an empirically poor article. I have been reporting in this space for years, resulting in full page articles in the New York Times which involved interviewing, at length, lead PIs of the billion dollar NIH research fund, NIH agency leaders at the highest level, and leading scientists working on this field. Additionally, as an academic who got interested in this important, fascinating space, I am still reading the scientific literature and regularly keeping up with cutting edge science. They’re all missing in your article!
I have written about poor definitions and I face flak from online patient communities regularly. I believe the mind/body duality as it exists is medicine is a big obstacle. Your article doesn’t have any of the cutting edge science on that! Where are the neuro-immunologists from psychiatry? Where are the sharp exercise physiologists getting out bangers in Nature?
What you have is essentially a channeling of a gripe session from one (empirically left behind and upset) faction of the mind-body duality. It’s not challenging it.
You can find patients saying stupid things online. But the idea that brain retraining is some breakthrough that’s being ignored and that people can’t talk about is complete nonsense. And your part about exercise is missing the biomedical breakthroughs in understanding from the last few years.
I don’t really do social media these days but I am informed. This one deserved clarification for the record. -
zeynep tufekci
@zeynep
Standing ovation interrupts a talk is unusual but apt: RAS mutations are upstream of deadliest cancers and this is THE FIRST success.
This follows decades of NIH-funded basic science, now gutted. And “AI accelerates science” story is good for IPO, not clinical results like this.
Mark Lewis, MD, FASCO
@marklewismd
Cheers, chills, and a standing ovation when RASolute 302 showed unprecedented survival on daraxonrasib for patients with progressive pancreatic cancer
Seldom do you sense you’re witnessing a historic moment in cancer care but this feels like ras targeting has arrived
#asco26This media couldn't be displayed. See it on the original post.
zeynep tufekci
@zeynep
GLP-1 drugs public funded basic science. Marveling at golden eggs after slaughtering the goose.
I think there’s a very good chance generative AI *will slow down* scientific advances.
Also recommend: All the Devils Are Here: The Hidden History of the Financial Crisis. -
zeynep tufekci reposted
Nicholas Kristof
@NickKristof
Remarkable story from @propublica of how the Pentagon got a call from the White House and rushed to lend $620 million to an obscure company -- with links to Donald Trump Jr. The company's valuation rose 10-fold to $2 billion. www.propublica.org/article/donald-trump-jr-vulcan-deal-wh…
-
zeynep tufekci reposted
Ari Paul
@AriDavidPaul
Amazing. Pensions forced to buy Space X IPO en masse even if it’s unprofitable. Impressive coup by Musk.
Hedgeye
@Hedgeye
Rule changes for the SpaceX $SPCX IPO:
Index providers waived the profitability requirement and cut the seasoning window from 90 days to 5.
This forces over $30 trillion in passive 401k and retirement money to buy SpaceX at IPO valuations.
Bloomberg Intelligence estimates S&P 500 funds must absorb 19% of SpaceX's float within 6 months.
Russell 1000 and Nasdaq 100 funds will absorb 24%.
The rules built to protect passive investors:
1. S&P 500 has required 12 months of trading and 4 quarters of GAAP profitability since 2002. Both waived.
2. Nasdaq cut its inclusion window from 90 trading days to 15.
3. FTSE Russell cut its to 5.
All three benchmarks are now structured to buy SpaceX at IPO pricing. -
zeynep tufekci reposted
Bloomberg Law
@BLaw
The US Justice Department is seeking the names, addresses, and banking information of Reddit and X users, ratcheting up efforts to identify social media critics of government deportation efforts. news.bloomberglaw.com/immigration/doj-wants-to-know-who-o…
-
zeynep tufekci reposted
Charles Ornstein
@charlesornstein
🚨🚨🚨The White House Intervened to Get a $620 Million Deal for a Company Tied to Donald Trump Jr. www.propublica.org/article/donald-trump-jr-vulcan-deal-wh… @RobertFaturechi
-
zeynep tufekci reposted
Zoe Tillman
@ZoeTillman
DEEP DIVE: The Trump administration is escalating efforts to unmask anonymous online critics, sending grand jury subpoenas to social media companies for their information. A court fight has been playing out in sealed proceedings in DC.
With Jimmy Jenkins: -
zeynep tufekci reposted
The New York Times
@nytimes
Breaking News: The Justice Dept. is said to have opened an investigation into E. Jean Carroll, who had accused Donald Trump of rape. In her civil case against him, a jury had found Trump liable for sexually abusing and defaming Carroll. www.nytimes.com/2026/05/27/us/politics/criminal-inquiry-e…
-
zeynep tufekci
@zeynep
The rage bait range on this platform is amazing.
Someone will hurl the wrong insult to a politically distasteful enemy so one can stand on principle to condemn it! The wrong clothing item was defended! A cast iron pot washed!
La Serenissima was a libertarian paradise! 🫡🫡
zeynep tufekci
@zeynep
The brain rot produced by all this should be studied like two-packs-a-day smoker lungs in a medical school.
I’m not claiming immunity. Remember the American Cancer Society episode on Mad Men? They smoked. They understood. But they didn’t lie to themselves or the world about it. -
zeynep tufekci reposted
Christina Jewett
@By_CJewett
NEW: Records released tonight show that Reynolds, makers of Vuse vapes, gave $5 Million to MAGA Inc. days before the FDA wrote a policy that could allow them to make millions selling flavored vapes.
Via @kenvogel
www.nytimes.com/2026/05/20/us/politics/donation-big-tobac… -
zeynep tufekci reposted
Tyler McBrien
@TylerMcBrien
BLANCHE: "The United States...is hereby FOREVER BARRED and PRECLUDED from prosecuting or pursuing...examinations or similar or related reviews" against Trump "or related or affiliated individuals," including family members or related companies and trusts.
-
zeynep tufekci reposted
Barton Gellman
@bartongellman
Heartbreaking, entirely avoidable, catastrophe.
Peter S. Goodman
@petersgoodman
I've been doing this work a long time but nothing prepared me for what I saw on a recent reporting trip in Somalia. I was there to look into rising food and fuel prices from the war in the Middle East landing atop dramatic cuts to the international humanitarian relief system.
-
zeynep tufekci reposted
Hümeyra Pamuk
@humeyra_pamuk
SCOOP: U.S. Deputy Secretary of State Christopher Landau instructed senior State Dept officials to facilitate and approve a visa for a fugitive former Polish cabinet minister, allowing him to flee to the U.S. from Hungary, people familiar w/the matter tell me @JonathanLanday
-
zeynep tufekci reposted
Sonia A. Rao
@bysoniarao
Scoop: An American passenger from the cruise ship exposed to hantavirus was issued a mandatory quarantine by the C.D.C. She's required to stay at the Nebraska unit through May. This is a rare move by the U.S. gov't – the last time this happened was 2020. www.nytimes.com/2026/05/18/us/hantavirus-ship-passenger-q…
-
zeynep tufekci
@zeynep
Epuyén outbreak was NOT self-limited.
Even though it was in a remote rural tiny town, it kept spreading. After four turns of human-to-human transmission, including to people merely in the same room, the authorities imposed a strict quarantine for a whole month. Then it ended.
Amesh Adalja
@AmeshAA
Andes virus is serious, but “can spread person-to-person” ≠ “pandemic-capable.” Even the largest outbreak (Epuyén) was self-limited. NEJM investigators noted some “possible inhalation” exposures—but that’s nuance, not evidence of measles-like spread. www.nytimes.com/2026/05/12/opinion/hantavirus-complacency…
zeynep tufekci
@zeynep
Only correcting as that’s my article, plus I interviewed the lead author of that NEJM article on the Epuyén outbreak.
I’m not a gambler, so I’m not discussing odds of where this will go. But the point remains: this virus is capable of superspreading. Let’s not make it easy.
zeynep tufekci
@zeynep
Carry on, I don’t expect to make a dent here.
My actual article on all this is linked in the tweet I quoted on top of this thread.
zeynep tufekci
@zeynep
Good that more countries taking mode of transmission (airborne) and lengthy incubation seriously.
Victim who flew to South Africa and died upon arrival was very sick, most likely source of spread if there is any. South Africa is trying but that’s the most challenging node.
Joseph Allen
@j_g_allen
Good to see contract tracing extending beyond passengers (and Public Health Canada willing to just say the number of people, whereas CDC wouldn't)
--> 9 high-risk (either passengers of cruise ship or sat close to someone with hantavirus)
--> the rest "deemed them to be minimal or low risk as opposed to no risk" but included b/c shared flight
toronto.citynews.ca/2026/05/14/26-low-risk-air-passengers…
zeynep tufekci
@zeynep
Terrible news. The world should support DRC to contain it. No vaccine for non-Zaire strain.
Ebola transmission is well-understood — bodily fluid, physical contact — and unless that changes, it’s not a pandemic threat like the Hantavirus.
Quashing outbreaks fast is crucial.
BEACON
@beacon_bio
Africa CDC reports Ebola outbreak in Ituri Province, Democratic Republic of Congo with 246 suspected cases and 65 suspected deaths; initial sequencing indicate the presence of a non-Zaire strain
Read more at: beaconbio.org/en/report/?reportid=528d3cf9-f9c0-4784-990a… -
zeynep tufekci reposted
Katherine Faulders
@KFaulders
NEWS -- President Trump is expected to drop his $10 billion lawsuit against the Internal Revenue Service in exchange for the creation of a $1.7 billion fund to compensate allies who claim they were wrongfully targeted by the Biden administration, sources tell me @PCCharalambous @alex_mallin abcnews.com/US/trump-poised-drop-irs-suit-launch-17b-weap…
-
zeynep tufekci reposted
–30–
@30newsletter
The New York Times is hiring a national science correspondent job-boards.greenhouse.io/thenewyorktimes/jobs/4694363005
-
zeynep tufekci
@zeynep
One outbreak and the entire “sane” side of public health resorted back to the paternalism of telling us not to worry or “panic”, “this isn’t Covid”, etc.
They didn’t learn a thing. We didn’t appoint them to be our therapists or parents. Provide info, don’t dictate how to feel.Post not available here, view on X
zeynep tufekci
@zeynep
Proposed rule: words like “panic” or anything pertaining to interior *feelings* of the public should be banned from the vocabulary of any public health person addressing the public. They can’t help themselves.
It doesn’t work. It backfires. It has the opposite effect. Enough.
zeynep tufekci
@zeynep
Also to those still solely on this site: I occasionally post here and follow some subjects like AI especially through lists — the legacy effect — but there is simply *no way* this place will be reliable source of information anything health related. So traditional media it is.
zeynep tufekci
@zeynep
The people who shrugged the destruction of so much that held society can keep shrugging till we inevitable hit moments where things need to work.
Not about this but generally: the early waves are here, if uneven, and it will only get worse and this isn’t reversible. So it goes. -
zeynep tufekci
@zeynep
Will take the other side of this bet.
Joe Weisenthal
@TheStalwart
Great piece from @JerusalemDemsas. For as much as people like to talk about AI "hallucinations", the much bigger story might be how often what they say is actually true, and that perhaps AI (as post-social media) may usher in a new era of shared reality www.theargumentmag.com/p/are-you-there-grok-its-me-margar…
Post not available here, view on X
zeynep tufekci
@zeynep
Already happening, but that method —effectively rolling back the clock to the pre-printing press era — only solves the problem of *grading*, not the problem of *teaching* in a manner that can impart the key skills that gave us all the advances since.
(It’s also very expensive.)
Adam Zivo
@AdamZivo
I find this discourse perplexing because the solution seems straigthforward: make university marks almost entirely dependent on lengthy in-person exams that combine handwritten essays with oral questioning.
Why is this even a conversation? Are there implementation barriers or something?
zeynep tufekci
@zeynep
I disagree strongly. Oral and/or paper exams do NOT solve the problem generative AI poses for education, they just hide it by creating a grading mechanism.
There is a reason the printing press unleashed a scientific and industrial revolution. Reversing that isn’t the path.
Luiza Jarovsky, PhD
@LuizaJarovsky
We need to go back to pen and paper, blue books, and oral examinations, and everyone knows it.
Luiza Jarovsky, PhD
@LuizaJarovsky
🚨 University professors have been saying AI is completely destroying learning and that we'll soon have an AI-powered, semi-illiterate workforce. Here's a glimpse into the educational apocalypse:
"Sarah, a freshman at Wilfrid Laurier University in Ontario, said she first used ChatGPT to cheat during the spring semester of her final year of high school. (...) After getting acquainted with the chatbot, Sarah used it for all her classes: Indigenous studies, law, English, and a “hippie farming class” called Green Industries. “My grades were amazing,” she said. “It changed my life.” Sarah continued to use AI when she started college this past fall. Why wouldn’t she? Rarely did she sit in class and not see other students’ laptops open to ChatGPT. Toward the end of the semester, she began to think she might be dependent on the website. She already considered herself addicted to TikTok, Instagram, Snapchat, and Reddit, where she writes under the username maybeimnotsmart. “I spend so much time on TikTok,” she said. “Hours and hours, until my eyes start hurting, which makes it hard to plan and do my schoolwork. With ChatGPT, I can write an essay in two hours that normally takes 12.”
-
"By November, Williams estimated that at least half of his students were using AI to write their papers. Attempts at accountability were pointless. Williams had no faith in AI detectors, and the professor teaching the class instructed him not to fail individual papers, even the clearly AI-smoothed ones. “Every time I brought it up with the professor, I got the sense he was underestimating the power of ChatGPT, and the departmental stance was, ‘Well, it’s a slippery slope, and we can’t really prove they’re using AI,’” Williams said. “I was told to grade based on what the essay would’ve gotten if it were a ‘true attempt at a paper.’ So I was grading people on their ability to use ChatGPT.”
-
AI in education is a serious topic, and many schools and universities are blindly jumping into the "AI-first" wave without considering short and long-term consequences.
It would be great to hear more from teachers and educators to understand potential solutions.
This might be a great opportunity for rethinking the education system and how students are assessed.
-
👉 Link to the full article below.
👉 To learn more about AI's legal and ethical challenges, join my newsletter's 94,700+ subscribers (link below). -
zeynep tufekci reposted
nxthompson
@nxthompson
This morning: MS Now reports that @S_Fitzpatrick is the focus of an FBI investigation bc of her previous reporting in @theatlantic
This afternoon: Fitzpatrick publishes another big story on the FBI. www.theatlantic.com/politics/2026/05/kash-patel-fbi-bourb… -
zeynep tufekci reposted
Joseph Allen
@j_g_allen
Reupping this piece from NYT when they covered our modeling of Covid transmission on the Diamond Princess cruise ship that demonstrated it was spread person-to-person through the air, when people said that couldn’t possibly be happening…
www.nytimes.com/2020/07/30/health/diamond-princess-corona… -
zeynep tufekci reposted
Charles Ornstein
@charlesornstein
“At the morgue, the babies were brought in with their diapers and blankets and with their hospital ID bracelets still wrapped around their tiny ankles. The pathologists’ findings were like those you would typically see in ailing adults, not newborns.” www.propublica.org/article/more-parents-decline-vitamin-k…
-
zeynep tufekci reposted
Garance Burke
@garanceburke
Deeply honored that my work in collaboration with our incredible global @AP team today won the Pulitzer Prize in International Reporting.
Huge thanks to our sources, editors + the @PulitzerPrizes.
Grateful for these stories' continued real-world impact: apnews.com/surveillance-digital-cage -
zeynep tufekci
@zeynep
Three and a half years post ChatGPT and.. where’s the predicted sweeping labor market and productivity impacts? Anything big outside of software?
We’re told just around the corner! Next model! Soon! Maybe. Or, maybe it’s not happening for structural reasons about generative AI.
James Rosen-Birch ⚖️🕊️
@provisionalidea
strongly urge people to diagram out their logic wherever possible
the claim is made (right after saying most economists and AI experts do not expect a permanent underclass scenario), that "if left to its own devices, Silicon Valley may summon a permanent underclass through its own market logic" (a strong assertion)
the justification is a mixture of (1) pointing at benchmarks (what?), and (2) indicating that some companies have performed layoffs in anticipation of AI (a tenuous claim -- there are a lot of indications that companies, including Block, use AI as a way to spin layoffs and cost-cutting as an efficiency play rather than a result of unsustainable operating bloat and/or market underperformance)
the claim is not justified further, and had one just drawn out the claims first, it would become clear that bad individual business decisions lead to outcompetition by others, not a macro crisis
jasmine sun
@jasminewsun
Most people I know in AI think the median person is screwed, and they have no idea what to do about it.
I spent the last 3 months talking to dozens of researchers, economists, and policy experts about AI's impact on work; including reps from every frontier lab and several Congressional offices. Unfortunately, I was not reassured.
The AI industry is raising the alarm, but can't change course. These companies' core business model relies on the disruption they are warning about: their faith in full automation only makes them go faster.
Policymakers are waking up, but still paralyzed by data and debates. Econ wonks disagree on plenty, but even the limited scenario looks like a "painful transition" that will disempower millions of workers.
But an "underclass" is not inevitable, but rather a societal choice — and one we can and should stop. Instead of waiting for impact, we should start planning now to support workers through AI disruption. Whether policymakers can assuage concerns about economic security may determine if we get to reap AI's gains at all.
New from me for @NYTOpinion. I put a ton into researching what I think may be the biggest topic of the year, so hope you read it (gift link here!) www.nytimes.com/2026/04/30/opinion/ai-labor-work-force-si…Post not available here, view on X
zeynep tufekci
@zeynep
In other words, IF the explanation for the AI lag is that the enterprise side is slow, needs adjustment and building a new business models around generative AI (which happened for other tech historically), THEN we have to change the entire dominant AI narrative. All of it.
zeynep tufekci
@zeynep
I’m all for this type of argument. That’s my career!
2012 me: Facebook isn’t forever friend to Dems; 2016 March me: Trump is viable stop laughing; 2020 Feb me: hey look pandemic. Etc.
But I have the track record I do because there are always *signs*. That’s what I’m asking for.
stargazer
@mnali
The fire crackles can barely be heard from afar - we’re told - why hasn’t the forest completely burned down yet.
zeynep tufekci
@zeynep
Three and a half years post ChatGPT and.. where’s the predicted sweeping labor market and productivity impacts? Anything big outside of software?
We’re told just around the corner! Next model! Soon! Maybe. Or, maybe it’s not happening for structural reasons about generative AI.
James Rosen-Birch ⚖️🕊️
@provisionalidea
strongly urge people to diagram out their logic wherever possible
the claim is made (right after saying most economists and AI experts do not expect a permanent underclass scenario), that "if left to its own devices, Silicon Valley may summon a permanent underclass through its own market logic" (a strong assertion)
the justification is a mixture of (1) pointing at benchmarks (what?), and (2) indicating that some companies have performed layoffs in anticipation of AI (a tenuous claim -- there are a lot of indications that companies, including Block, use AI as a way to spin layoffs and cost-cutting as an efficiency play rather than a result of unsustainable operating bloat and/or market underperformance)
the claim is not justified further, and had one just drawn out the claims first, it would become clear that bad individual business decisions lead to outcompetition by others, not a macro crisis
jasmine sun
@jasminewsun
Most people I know in AI think the median person is screwed, and they have no idea what to do about it.
I spent the last 3 months talking to dozens of researchers, economists, and policy experts about AI's impact on work; including reps from every frontier lab and several Congressional offices. Unfortunately, I was not reassured.
The AI industry is raising the alarm, but can't change course. These companies' core business model relies on the disruption they are warning about: their faith in full automation only makes them go faster.
Policymakers are waking up, but still paralyzed by data and debates. Econ wonks disagree on plenty, but even the limited scenario looks like a "painful transition" that will disempower millions of workers.
But an "underclass" is not inevitable, but rather a societal choice — and one we can and should stop. Instead of waiting for impact, we should start planning now to support workers through AI disruption. Whether policymakers can assuage concerns about economic security may determine if we get to reap AI's gains at all.
New from me for @NYTOpinion. I put a ton into researching what I think may be the biggest topic of the year, so hope you read it (gift link here!) www.nytimes.com/2026/04/30/opinion/ai-labor-work-force-si…
zeynep tufekci
@zeynep
What’s below: in many big moments, I found myself *way outside* almost the entire range of mainstream thought on key issues. Wasn’t unjustified.
I have a similar out-of-sync feeling about almost the entire spectrum of AI arguments. Maybe I am wrong this time! Hence the question.
zeynep tufekci
@zeynep
I’m all for this type of argument. That’s my career!
2012 me: Facebook isn’t forever friend to Dems; 2016 March me: Trump is viable stop laughing; 2020 Feb me: hey look pandemic. Etc.
But I have the track record I do because there are always *signs*. That’s what I’m asking for.
stargazer
@mnali
The fire crackles can barely be heard from afar - we’re told - why hasn’t the forest completely burned down yet.
zeynep tufekci
@zeynep
Three and a half years post ChatGPT and.. where’s the predicted sweeping labor market and productivity impacts? Anything big outside of software?
We’re told just around the corner! Next model! Soon! Maybe. Or, maybe it’s not happening for structural reasons about generative AI.
James Rosen-Birch ⚖️🕊️
@provisionalidea
strongly urge people to diagram out their logic wherever possible
the claim is made (right after saying most economists and AI experts do not expect a permanent underclass scenario), that "if left to its own devices, Silicon Valley may summon a permanent underclass through its own market logic" (a strong assertion)
the justification is a mixture of (1) pointing at benchmarks (what?), and (2) indicating that some companies have performed layoffs in anticipation of AI (a tenuous claim -- there are a lot of indications that companies, including Block, use AI as a way to spin layoffs and cost-cutting as an efficiency play rather than a result of unsustainable operating bloat and/or market underperformance)
the claim is not justified further, and had one just drawn out the claims first, it would become clear that bad individual business decisions lead to outcompetition by others, not a macro crisis
jasmine sun
@jasminewsun
Most people I know in AI think the median person is screwed, and they have no idea what to do about it.
I spent the last 3 months talking to dozens of researchers, economists, and policy experts about AI's impact on work; including reps from every frontier lab and several Congressional offices. Unfortunately, I was not reassured.
The AI industry is raising the alarm, but can't change course. These companies' core business model relies on the disruption they are warning about: their faith in full automation only makes them go faster.
Policymakers are waking up, but still paralyzed by data and debates. Econ wonks disagree on plenty, but even the limited scenario looks like a "painful transition" that will disempower millions of workers.
But an "underclass" is not inevitable, but rather a societal choice — and one we can and should stop. Instead of waiting for impact, we should start planning now to support workers through AI disruption. Whether policymakers can assuage concerns about economic security may determine if we get to reap AI's gains at all.
New from me for @NYTOpinion. I put a ton into researching what I think may be the biggest topic of the year, so hope you read it (gift link here!) www.nytimes.com/2026/04/30/opinion/ai-labor-work-force-si… -
zeynep tufekci
@zeynep
I strongly disagree that the AI chatbot as a “consumer product” could be described as “normal” technology in any manner.
Astounded at how much I disagree with most AI analyses out there.
I feel like I’m on another planet, like Feb 2020 or Feb-Nov 2016, etc.
Yes, more soon…
Derek Thompson
@DKThomp
Between Mythos, Anthropic v Pentagon, and the drumbeat of biosecurity concerns, I think we're entering a period where
(a) AI as a consumer product (chatbot + coding assistant) will continue to behave like a "normal" technology—a very powerful tool that does all sorts of useful things alongside workers without imminently wiping out tranches of the labor force; while...
(b) as bio-/cyber-/national security threat, AI is becoming something else entirely, a sort of hydra of existential risk that's going puncture ppl's confidence that "AI is a normal technology" and force both the labs and the federal govt to establish rules on the fly, while open models race to catch up to the frontier -
zeynep tufekci reposted
Taniel
@Taniel
WaPo reports Louisiana’s governor will announce on Friday that he is suspending the House primaries where people have already started voting and redraw the map. www.washingtonpost.com/politics/2026/04/29/louisiana-hous…
-
zeynep tufekci reposted
The Washington Post
@washingtonpost
Breaking news: The Supreme Court weakened a key provision of the landmark Voting Rights Act, a ruling that limits the consideration of race in drawing voting maps and could usher in Republican gains in the House. www.washingtonpost.com/politics/2026/04/29/supreme-court-…
-
zeynep tufekci reposted
Max Kozlov
@maxdkozlov
The Trump administration has downsized US science by historic margins — but it's not just via grant or workforce cuts.
Our new @nature analysis reveals the government has cut more than 100 scientific advisory panels across all major science agencies. -
zeynep tufekci reposted
Sam Stein
@samstein
NEW: A torture victim from Colombia who fled to the US and has since been at the center of a lengthy deportation process has once again been deported by the Trump administration... to the Democratic Republic of Congo (DRC)—a country where she has no ties.
@Carrasquillo -
zeynep tufekci reposted
Rebecca Ballhaus
@rebeccaballhaus
Last fall, the Trump administration announced criminal charges and sweeping sanctions against what it said was a transnational criminal syndicate that had stolen billions of dollars from Americans and others.
Less than a month later, World Liberty Financial announced it had partnered with a virtual-currency venture called AB, one of whose projects—a planned “blockchain”-themed resort—had been led by two men sanctioned in the U.S. crackdown. @kowsmann @gksteinhauser www.wsj.com/business/crypto-venture-linked-to-men-sanctio… -
zeynep tufekci reposted
NBC News
@NBCNews
NEW: FCC to direct eight Disney-owned TV stations to file early license renewals, a source says, adding that the "unprecedented" action was directly tied to ABC late-night host Jimmy Kimmel's recent joke about first lady Melania Trump. www.nbcnews.com/news/us-news/fcc-direct-disney-owned-tv-s…
-
zeynep tufekci reposted
Brian Stelter
@brianstelter
Breaking: The Trump-aligned FCC is planning to file paperwork as early as this afternoon that will challenge Disney's eight licenses for its eight ABC stations. This early-renewal move will be a major escalation by @BrendanCarrFCC. view.newsletters.cnn.com/messages/17773891539866ddee4ff12…
-
zeynep tufekci reposted
The Washington Post
@washingtonpost
Breaking news: The Trump administration’s move to drop the long-standing recommendation for newborn hepatitis B vaccination within 24 hours of birth will likely lead to additional infections and millions in added health care costs, new studies show. www.washingtonpost.com/health/2026/04/27/delaying-hepatit…
-
zeynep tufekci reposted
Trip Gabriel
@tripgabriel
The EPA's independent science arm did groundbreaking research to save lives. It studied fertility, asthma, wildfires, drinking water, climate change etc etc
In just one year, it has been almost completely dismantled
Of 1500 scientists, only 124 remain
www.nytimes.com/2026/04/27/climate/epa-science-trump-cuts… -
zeynep tufekci reposted
Hamed Aleaziz
@Haleaziz
Scoop: immigrants can now be denied a green card for expressing political opinions, such as participating in pro-Palestinian campus protests, posting criticism of Israel on social media and desecrating the American flag, according to internal documents.
w/ @NickNehamas -
zeynep tufekci reposted
Reuters
@Reuters
A Reuters investigation found a broader-than-previously known Trump administration effort to expand federal control over elections, using investigations, raids and demands for access to voting systems across at least eight states www.reuters.com/investigations/how-trump-is-moving-contro… @specialreports
-
zeynep tufekci
@zeynep
While this may work on some AI detectors, the method misunderstands the key differences between AI and human writing and will falsely tag some good writing that LLMs emulate.
“Generative AI is extremely impressive but not on a path to AGI” repeated daily can help illuminate.
Oliver
@oliviscusAI
Someone built a Claude Code skill that strips every trace of AI from your writing.
It detects 29 specific patterns from Wikipedia's "Signs of AI writing" guide, em dashes, rule of three, "it's not just X, it's Y", sycophantic openers, and rewrites your text to actually sound human.
100% Open Source. -
zeynep tufekci reposted
David Fahrenthold
@Fahrenthold
The Trump administration gave a secret, no-bid contract to the firm building the WH ballroom -- after inflating its value 3X. This time, the government pays, not private donors.
www.nytimes.com/2026/04/25/us/politics/lafayette-park-fou… -
zeynep tufekci reposted
Nicholas Nehamas
@NickNehamas
NEW: Under Trump guidelines, immigration officers could deny someone a green card for a social media post that says "Stop Israeli Terror in Palestine" or for participating in campus protests, according to exclusive documents reviewed by the New York Times.
-
zeynep tufekci reposted
Sheryl Gay Stolberg
@SherylNYT
NEW: NIH spending on new medical research has fallen roughly $1 billion behind the pace of years past, delaying thousands of scientific projects.
@benjmueller @irenatfh
www.nytimes.com/2026/04/22/science/trump-nih-funding-rese… -
zeynep tufekci reposted
Michael S. Schmidt
@nytmike
NEW and BREAKING: The FBI began investigating New York Times reporter Elizabeth Williamson earlier this year after she published a story about how Kash Patel was using FBI personnel to provide his girlfriend with govt security and transportation. FBI agents interviewed Patel's girlfriend about Williamson, queried databases for info on the reporter and recommended moving forward to determine whether the reporter broke federal stalking laws. www.nytimes.com/2026/04/22/us/politics/fbi-times-reporter…
-
zeynep tufekci reposted
Julian Mark
@juleswapo
Stunning revelation here. We all knew that the modern "shadow docket" was a Roberts Court invention. Now it's clear that Roberts was key in shaping it. Also interesting is the court's view of executive power then vs. now.
The New York Times
@nytimes
Secret memos by Supreme Court justices, obtained by The New York Times, show how they decided to bypass time-tested procedures and create the modern “shadow docket,” a controversial new way of doing business. www.nytimes.com/2026/04/18/us/politics/supreme-court-shad…
-
zeynep tufekci
@zeynep
“How is it any form of authoritarianism if there are still elections?”
Use power of the state to make the playing field completely uneven, that’s how. Extremely hard to reverse. “Win by an overwhelming margin despite all the obstacles and be in the EU and cross your fingers.”
Jay in Kyiv
@JayinKyiv
Long unofficially banned from appearing on state TV, incoming Hungarian Prime Minister Peter Magyar appears on a channel to which he was previously not allowed only to announce an end of "North Korean" style Orban media monopoly.
This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Kyle Cheney
@kyledcheney
MORE: In a simultaneous filing, DOJ is moving to drop similar convictions against the Oath Keepers leaders who were similarly charged/convicted.
-
zeynep tufekci reposted
Kyle Cheney
@kyledcheney
JUST IN: The Justice Department is moving to vacate the seditious conspiracy convictions of Proud Boys leaders who led hundreds of allies to the Capitol on Jan. 6.
They had previously had lengthy sentences commuted, but not pardoned, by Trump.
storage.courtlistener.com/recap/gov.uscourts.cadc.40140/g… -
zeynep tufekci reposted
Doug Bock Clark
@DougBockClark
5/ Soon after assuming power for a second time, Trump got rid of essentially all 75 people or positions that had acted as guardrails during the 2020 election.
-
zeynep tufekci reposted
Nicholas Nehamas
@NickNehamas
Exclusive: The Trump administration has fired two immigration judges who blocked the deportations of Rumeysa Ozturk and Mohsen Mahdawi, two int'l students who expressed pro-Palestinian views. Four other judges were fired on Friday too.
-
zeynep tufekci
@zeynep
I will take the other side of this bet, starting with the idea there will be a singular “AI” with a singular answer that will actually be accepted as an arbiter by adversarial parties.
This reminds me of more information will reduce misinformation phase of social media debates.
Stefan Schubert
@StefanFSchubert
I think AI will make it harder to maintain widespread belief in false but politically convenient views, because:
1) People will increasingly use AI as an arbiter, and refer to it in debates
2) AI will be more truth-seeking and unbiased than human experts -
zeynep tufekci reposted
Kyle Griffin
@kylegriffin1
BREAKING WSJ:
Trump has repeatedly promised his top administration officials pardons before he leaves office, according to people who have heard his comments.
"I'll pardon everyone who has come within 200 feet of the Oval," Trump said in a recent meeting. -
zeynep tufekci
@zeynep
I ~buy this, which doesn’t make AI cybersecurity issues moot but different.
If too many (maybe) bugs are made shallow, the bottleneck is what evaluating which ones are real and important.
So what’s the false positive rate for Mythos? As far as I can tell, it’s not disclosed.
Ramez Naam
@ramez
Anthropic's Mythos does not appear to show any acceleration of ECI. After normalizing Anthropic's internal ECI with @EpochAIResearch 's public ECI, it's clear that the two metrics are extremely close, and that Mythos is pretty much on trend, just slightly above GPT 5.4. /1
zeynep tufekci
@zeynep
Another interesting question: assuming a limited supply of “but for noticing” AND important bugs (which @Aisle_Inc replicating Mythos highlights), will gen AI soon disempower the ingenious/patient hacker while empowering centralized actors who can afford to hunt for deeper ones?
Joshua Achiam
@jachiam0
Much discussion of cybersecurity doom recently; very little discussion about how the window of existing zero days has been open for decades and will shortly close, after which attacker/defender dynamics will now be measured in allocated compute.
zeynep tufekci
@zeynep
Noting the below.
Also, remember Log4Shell, etc? Gen AI “discovery” obviously doesn’t mean the exploit wasn’t already known to others. This may well bring more cybersecurity stability (empowering those with $$$ who tend not to be nihilistic) but open question for the transition.
Kai Williams
@chi_t_williams
As far as I can tell, Anthropic claimed the false positive rate, at least for finding memory bugs, was 0% in their experiments. They didn't close for other classes of vulnerabilities (like logic errors)
red.anthropic.com/2026/mythos-preview/ -
zeynep tufekci reposted
profdeibert
@RonDeibert
"This represents the first confirmation of the use of ad-based surveillance technology in Europe." - Uncovering Webloc: An Analysis of Penlink’s Ad-based Geolocation Surveillance Tech citizenlab.ca/research/analysis-of-penlinks-ad-based-geol…
-
zeynep tufekci reposted
Edward Wong
@ewong
NEW from @nytimes: The Trump administration aims to use the tools of counterterrorism to go after what it calls antifa and leftist groups. It lists antifa as a priority in an internal intelligence document. Some US officials worry this could lead to more domestic surveillance.
-
zeynep tufekci reposted
Bloomberg Law
@BLaw
The US Justice Department’s top antitrust litigator and three senior trial attorneys that led cases against Live Nation Entertainment Inc., Apple Inc. and Alphabet Inc.’s Google are leaving the agency, according to people familiar with their plans. news.bloomberglaw.com/antitrust/doj-top-antitrust-litigat…
-
zeynep tufekci reposted
Axios
@axios
SCOOP: Meta today began removing ads from attorneys who were seeking clients that claim to have been harmed by social media while under the age of 18. www.axios.com/2026/04/09/meta-social-media-addiction-ads
-
zeynep tufekci
@zeynep
It’s incredible how many people — even journalists — who obviously don’t understand much about generative AI will claim the “hallucination” problem have been solved. Soon! Just around the corner. You’re not using the best model!
But reality eventually catches up.
Merryn Somerset Webb
@MerrynSW
What if the whole LLM thing is a false start? If the flaws are inherent systemic problems - if the compounding of hallucinations/errors can't be sorted out? If the capex build out is one of the biggest misallocations of capital ever? Then what? www.bloomberg.com/news/newsletters/2026-04-04/waiting-out…
Post not available here, view on X
zeynep tufekci
@zeynep
Coding is a formal system that’s externally verifiable. It’s a strange task for humans. Very, very, very little of what people actually do economically fits that mold.
(I think people are underestimating the issues even in coding, but regardless).
zeynep tufekci
@zeynep
Humans make many mistakes. It’s in no way analogous to how generative AI “hallucinates”.
The key point is generative AI is absolutely the same mechanism when it “hallucinates” and when it does not.
If that’s not obvious, I recommend getting into the weeds of the technology.
Beck
@Beckeckeck
Humans also hallucinate some percent of the time, does that mean we can't use them at scale?
I agree that what you're gesturing at is a problem for long term alignment, but don't see how it's an impediment for work or value creation over the short to medium term.
zeynep tufekci
@zeynep
Of course humans make mistakes.
But when generative AI “hallucinates”, it’s not “making a mistake” like a human. The processes aren’t isomorphic.
I think many deep in the tech understand this.
Journalists should stop being so easily impressed and dive into the technical side.
Matt McK
@mattmcknight
So you think humans don't have false memories, aren't overconfident about wrong opinions? It's very analogous to humans. What it is not analogous to is traditional computing, where you get an answer from a fact DB. These tech can be combined. Harness engineering is here.
zeynep tufekci
@zeynep
Yes, misleading term. Just like the term “reasoning models” is misleading — a powerful (expensive) method, yes, but they don’t do what people think of as “reasoning”.
Cars aren’t horseless carriages and gen AI doesn’t “hallucinate” or “reason” like a human. It’s its own thing.
Wagner James Au
@slhamlet
Someone noted that even calling it "hallucinations" strongly misleads non-technical people into assuming the LLM has some baseline of "understanding". When in fact there's no comprehension at all, just tokenized data that's output according to probability algorithms.
-
zeynep tufekci reposted
-
zeynep tufekci
@zeynep
The AP report on the bombing of Sharif University.
This is Maryam Mirzakhani’s alma mater — the first woman to win a Fields Medal in math, and only one of two so far.
apnews.com/video/ict-building-at-tehrans-sharif-universit…
Congresswoman Yassamin Ansari
@RepYassAnsari
Sharif University is Iran’s MIT. They’ve produced a huge number of engineers who’ve gone on to Silicon Valley and founded some of the most successful American tech companies.
Why are we bombing a university in a city of 10 million people?
zeynep tufekci
@zeynep
The university is also a leading center of opposition to the government — students bravely marching even after the deadly crackdown before this war.
Anyway something something women’s rights something something bluesky libs something something, I guess.
www.bbc.com/news/articles/c5yj2kzkrj0o -
zeynep tufekci reposted
Aric Toler
@AricToler
This access change includes Gaza and Lebanon, not just Iran and Gulf states with US bases. Planet has been absolutely amazing, and their imagery has led to dozens of stories/scoops, with Gaza/Lebanon in particular.
Geoff Brumfiel
@gbrumfiel
Planet appears to be ending all access to imagery in the Middle East indefinitely. This will make understanding what’s happening in this conflict far more difficult.
-
zeynep tufekci reposted
Austin Kocher, PhD
@ackocher
Most people don't know that immigration judges were never truly independent. They're DOJ employees, subject in many ways to the AG's direction. Civil service tenure was the last floor of protection against being fired for ruling the wrong way but the MSPB just removed that floor.
-
zeynep tufekci reposted
Paul Cerro
@paulcerro
This is actually insane. Idk why this guy decided to put the biggest target on his back by doing this NYT article
Gotta be up there with Andrew creating a brand new drug to circumvent $NVO
Sheel Mohnot
@pitdesi
They made facebook accounts for 800+ fake doctors (I had Claude verify, none of them are actually doctors, especially not "Dr. Tuckr Carlzyn MD")
to advertise on Facebook.
This is extremely illegal, just takes FTC to notice it and it'll be hard for them to advertise on FB -
zeynep tufekci reposted
Bulletin of the Atomic Scientists
@BulletinAtomic
As the Trump administration considers new policies for gain-of-function research, scientists have an important role to play in evaluating risks and benefits—but only with a commitment to fair and rigorous evaluation, writes Thomas Kelly.
thebulletin.org/2026/04/the-keys-to-a-better-gain-of-func… -
zeynep tufekci reposted
Emily Shugerman
@eshugerman
OpenAI is behind a new “parents and kids” coalition on AI safety legislation. Former members of the group told me they didn’t know until it launched.
-
zeynep tufekci reposted
Christiaan Triebert
@trbrtc
A first look at an apparent tungsten pellet from the U.S. strikes on Lamerd, Feb. 28. PrSM is designed to detonate above target and disperse thousands of these pellets outward. Our analysis shows they killed at least 21 people, including girls who were at volleyball practice.
-
zeynep tufekci reposted
a newsman
@a_newsman
From a NYT article earlier in March explaining the MAGA influencer/lobbyist effort to convince Trump to pardon Schwartz, I can’t get this detail about the kind of abuse patients at Skyline were subjected to out of my head
ProPublica
@propublica
New: Trump pardoned nursing home owner Joseph Schwartz just 3 months into his sentence for a $39 million fraud scheme. Meanwhile, families who won multimillion-dollar wrongful death suits against Schwartz haven’t collected a cent. propub.li/4di7xXs
-
zeynep tufekci reposted
ProPublica
@propublica
New: Under AG Pam Bondi, the DOJ has dropped 23,000 criminal cases — including hundreds of investigations into terrorism, white-collar crime and drugs — while prosecuting 32,000 new immigration cases in just the first six months of Trump’s second term. propub.li/4m4Vzmf
-
zeynep tufekci
@zeynep
Disagree. They have shown themselves and the world they have a solid deterrent, tbh.
A nuke isn’t anywhere as usable as the deterrent they now provably possess and can improve on, and does not come with problems and existential fears of a nuke.
Also forever changed world order.
Benjy Sarlin
@BenjySarlin
So absent an uprising, it seems like the median scenario is energy shock, regime remains in place and more motivated to get a nuke, US comes under intense pressure to bomb/raid the nuclear program every few years under any administration? In other words, forever war.
Post not available here, view on X
zeynep tufekci
@zeynep
How lucky, then, Bab-el-Mandeb is not also two shipping lanes barely longer than Central Park! And it’s not where embedded permanent local groups allied with the Iranian regime have already demonstrated even more asymmetric ability to control the flow with way less means!
Christian Demeter
@marginletter
Bab-el-Mandem already see increased traffic now, a lot has been diverted there without a problem. On the medium term logistics will be invested in, Iran can’t sustain multiple fronts, they are already very weak and can only half-threaten Hormuz. Drones+AI are both offensive and defensive so pipelines will be protected.
zeynep tufekci
@zeynep
Maybe. But acquiring a nuke would invite a period of instability with an existential threat to their rule because of @BenjySarlin outline.
They could also let this administration declare “victory” as the end, confidently knowing that means an irreversible change in their favor.
⬢ Jens Wiëchers
@jwiechers
Not sure I agree, if only because getting a nuclear weapon really isn't a particular challenge for a nation state if it really wants to. I'm pretty sure it is true that Ali Khamenei rejected assembly for his own reasons, but blockading Hormuz & harming neighbors has downsides.
zeynep tufekci
@zeynep
A nuclear weapon IS a deterrent but in the Mutually-Acquired-Destruction sense — the “MAD” of the cold war.
So, yeah, many in the region are likely hoping for the calculated rationality of non-millenarian theocrats to hold back what the other two key actors could well unleash.
zeynep tufekci
@zeynep
To add, if the Appalachian mountains are the Zagros range: they go right up to the coast.
That leaves them ONLY with the high ground, rest of their coastline (about as long as the US West coast), drones with that range and being in their own country. “A little excursion”.
Christopher Clary
@clary_co
If the US were to try and seize the 50 mile stretch along the Iranian coastline with the Persian Gulf and Strait of Hormuz, my back of the envelope math suggests that's a bit like taking over the state of Virginia or North Carolina in terms of sheer size.
James Pethokoukis ⏩️⤴️
@JimPethokoukis
This is nightmare fuel
JPMORGAN: "If the US cannot achieve a short-term victory, it seems likely that it will be forced to attempt a ground war to reopen the Strait of Hormuz. If so, this could transform into a multiyear war"Post not available here, view on X
zeynep tufekci
@zeynep
Saying Iran will be “forced” to open the Strait etc. is misreading the moment.
The Iranian regime will very, very, very happily open the (now their) Strait.
People need to calibrate their understanding of which actors are predictably acting in their rational self-interest.
Marko Jukic
@mmjukic
Iran is not going to drive oil prices to $200/barrel if Trump just walks away. They are going to be confused for two weeks and then will just charge everyone a small fee for transit, which everyone will gladly pay. If they get attacked again, they will re-close the Strait.
This post is unavailable. See it on X.
zeynep tufekci
@zeynep
So if Trump does just abandon, the open question is not the Iranian regime, since they will likely go along very happily, but what the big regional losers will do: whether they might refuse to go along or even sabotage that attempt.
(That’s how global war start, by the way.) -
zeynep tufekci reposted
Gregory Brew
@gbrew24
"In conversations with more than three dozen oil and gas traders, executives, brokers, shippers and advisers over the last week, one message was repeated over and over: The world still hasn't grasped the severity of the situation." www.bloomberg.com/graphics/2026-iran-war-hormuz-closure-o…
-
zeynep tufekci reposted
Alina Chan
@Ayjchan
"The problem, however, with a reflexively defensive approach [to a lab origin of Covid] is that when a scientific community abandons error-correction, that community is no longer doing science."
Michael Weissman
@mbw61567742
My paper showing that the famous multiple-spill Pekar paper got its conclusion upside down by a math error is now published. Links below. I think that the press who covered the original should cover the real result with the opposite (single spill) conclusion.
-
zeynep tufekci
@zeynep
No it’s not.
Search isn’t like an LLM anymore than a car is merely a horseless carriage. An LLM is also not like a calculator is to arithmetic. And generative AI isn’t yet on any kind of path to AGI where the benchmark is human intelligence.
It IS powerful. But its own thing.
Andy Masley
@AndyMasley
I maintain that this a clear equivalent of the lump of labor fallacy and the same rule would condemn researchers using search engines
Charlotte Alter
@CharlotteAlter
Shout it from the rooftops!!! Research is thinking. Outlining is thinking. Writing is thinking. Any portion of that done by AI is less thinking done by you.
Post not available here, view on X
zeynep tufekci
@zeynep
The complaint how the term “stochastic parrot” has narrowed the thinking of some AI critics? Fair enough.
But the term “AGI” functions similarly for some. The concept fuels incuriosity, imo, and we aren’t properly thinking about things so visible and yet so poorly understood.
zeynep tufekci
@zeynep
I get why the companies go on about AGI—they need to raise money, distract journalists, scare or excite policymakers, etc.
I think many AI researchers sincerely believe we’re on a human-like “AGI” path.
But drop the jargon and look at the specifics! A very different picture! -
zeynep tufekci reposted
ProPublica
@propublica
New: Trump pardoned nursing home owner Joseph Schwartz just 3 months into his sentence for a $39 million fraud scheme. Meanwhile, families who won multimillion-dollar wrongful death suits against Schwartz haven’t collected a cent. propub.li/4di7xXs
-
zeynep tufekci reposted
Christiaan Triebert
@trbrtc
NEW: On the first day of the war with Iran, the United States used a new ballistic missile. It was untested in combat. Our analysis shows it struck a sports hall, an elementary school, and residential areas — killing at least 21 people, including children. w/ @johnismay
-
zeynep tufekci reposted
Michael Lucy
@MmichaelLlucy
does this not just completely invalidate the benchmarks
euan ashley
@euanashley
New AI paper from us this week. When my student first showed me his initial findings, I really didn’t know what to make of them. I felt that this was an interesting but curious loophole phenomenon that would shortly be closed. I was very wrong.
arxiv.org/abs/2603.21687 -
zeynep tufekci reposted
Thomas Zimmer
@tzimmer_history
Not a single person of consequence believes that No Kings is some magical solution to save democracy; no serious person argues that occasionally joining a protest is all anyone needs to do.
But as an exercise of shared democratic participation, protests matter very much.
Thomas Zimmer
@tzimmer_history
Forget all the “Protests don’t matter” and “No Kings is just silly” nonsense.
Protests matter. Societal resistance has been hardening.
And “No Kings” is crucial as part of an effort to dispel the idea that “real America” stands behind Trump.
steady.page/en/democracyamericana/posts/a40ac673-b42b-410… -
zeynep tufekci reposted
Egan Peltan
@EganPeltan
Nvm guys - I take it back. It was all the LLMs
Who could have possibly figured out to use a drug as indicated on its label!
merckusa.cvpservice.com/product/basic/view/1047586
Paul S. Conyngham
@paul_conyngham
The LLms suggested using the pd-1 inhibitor too 😂
-
zeynep tufekci reposted
Thomas Kennedy
@tomaskenn
Body cam footage shows 19 year old who was pulled over for a minor traffic infraction, police shoving him to the ground, DHS then lied about him being charged with a felony, and weeks later he dies in Florida ICE detention.
Alex DeLuca
@AlexLDeLuca
NEW: Body cam footage shows the rough arrest that preceded Royer Perez-Jimenez’s transfer to ICE detention.
The 19-year-old became one of the youngest people to die in ICE custody last week.
www.miaminewtimes.com/news/body-cam-video-shows-rough-arr… -
zeynep tufekci reposted
Egan Peltan
@EganPeltan
This is totally out of control:
There’s 0 - I repeat 0 - evidence any of the LLM work did anything meaningful for Rosie’s cancer
I’m sorry to rain on the parade here. I know we want to believe. But, it’s possible to do a lot of things and have nothing happen
@paul_conyngham co-administered α-PD-1 (conventional immunotherapy) with a TKI and the mRNA. It’s probably the most effective cancer immunotherapy of all time. This isn’t a small detail!
There’s no evidence his process (beyond FDA approved doggie α-PD-1) had any impact on disease progression. The most parsimonious explanation is a partial response to α-PD-1
I get it. The chat bots make for a great story (although checking multiple LLMs isn’t validation), but it’s really just a neat story. It’s fundraising copy. Before he starts selling the “custom neoantigen mRNA vax” story to consumers, he should provide some evidence it did anything! That’s responsible citizen science
This is just storytelling for the AGI true believers. Specifically, a story in search of venture money
Sam Altman
@sama
The coolest meeting I had this week with was Paul, who used ChatGPT and other LLMs to create an mRNA vaccine protocol to save his dog Rosie. It is amazing story.
"The chat bots empowered me as an individual to act with the power of a research institute - planning, education, troubleshooting, compliance, and yes, real scientific design work in converting genomic data to a vaccine prescription and designing the treatment protocol around it. But they worked alongside humans at every step. The combination is what made it possible."
It immediately got me thinking "this should be a company".
Also, Paul is an extraordinary guy. This should be easy to do, but it is not yet.
Paul S. Conyngham
@paul_conyngham
-
zeynep tufekci
@zeynep
A “TACO” trade back to status quo ante bellum is a fairy tale.
I believe this for most everything that’s happened last year. “Taco” moves can calm things down short-term, sure. However, barring very radical steps, the long term shifts that have been unleashed are irreversible.
Stephen Dziedzic
@stephendziedzic
Singapore's FM. Very blunt on Trump:
- "the underwriter of this world order has now become a revisionist power, and some people would even say a disruptor"
- "the entire global economy has been taken hostage and we will all pay a price on hostilities"
www.mfa.gov.sg/newsroom/press-statements-transcripts-and-…
zeynep tufekci
@zeynep
Some steps are available to save the day. No steps are available now to substantially reverse what’s been unleashed. Not just the last month.
At some future, people *may* get to ponder whether to try to take the kind of extremely difficult steps to attempt a long-shot rollback. -
zeynep tufekci reposted
TravelGov
@TravelGov
Hong Kong: On March 23, 2026, the Hong Kong government changed the implementing rules relating to the National Security Law. It is now a criminal offense to refuse to give the Hong Kong police the passwords or decryption assistance to access all personal electronic devices including cellphones and laptops. This legal change applies to everyone, including U.S. citizens, in Hong Kong, arriving or just transiting Hong Kong International Airport. In addition, the Hong Kong government also has more authority to take and keep any personal devices, as evidence, that they claim are linked to national security offenses. Read more: hk.usconsulate.gov/security-alert-2026032601/
-
zeynep tufekci reposted
Peter Baker
@peterbakernyt
Trump's Justice Department has agreed to pay Michael Flynn $1.25 million in taxpayer money even though he pleaded guilty twice to making false statements to FBI agents about his interactions with Russian officials. @alanfeuer www.nytimes.com/2026/03/25/us/politics/michael-flynn-doj-…
-
zeynep tufekci reposted
Trip Gabriel
@tripgabriel
.@SenBlumenthal asks Trump judicial nominees a question: “Who won the popular vote in 2020?”
“In 2020, President Biden was certified & served 4 years as president,” most answer
"The fact that they are so easily cowed speaks volumes," Bluementhal says
www.nytimes.com/2026/03/24/opinion/trump-judges-nominees-… -
zeynep tufekci reposted
Lydia DePillis
@lydiadepillis
Jaw-dropping: The U.S. and Ecuador said they had cooperated to blow up a cocaine trafficking operation. Our reporters visited and found it was just a dairy farm, whose residents were brutalized by Ecuadorian soldiers acting on Pentagon intelligence. www.nytimes.com/2026/03/24/world/americas/us-ecuador-drug…
-
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
As @SECGov enforcement director, Margaret Ryan pushed to aggressively pursue charges against Justin Sun (a major backer of the Trumps' @worldlibertyfi) & @elonmusk.
Instead, the SEC settled with Sun & is in talks to settle w Musk.
Ryan resigned last week www.reuters.com/business/finance/us-secs-ex-enforcement-c… -
zeynep tufekci
@zeynep
One result will be the human toll during the painful rush to decarbonize under Chinese hegemony (industrial base for that is there). This is probably baked in by now though depth of suffering depends on how and when this phase ends.
Add to immense and ongoing toll of last year.
Heather Exner-Pirot
@ExnerPirot
Some of you still seem to think oil is only used for gasoline for light duty vehicles, because that’s the only time you’ve physically encountered it.
Expensive and scarce oil is an omnicrisis for the global economy. It cannot be replaced by electrons from solar panels.
zeynep tufekci
@zeynep
Should add: Also recarbonize… Turn back to coal.
The structure that held things together is no more. Even if it this ends tomorrow, many countries rationally will rush to adjust to this. Best case scenario looks like a painful transition.
Javier Blas
@JavierBlas
MUST READ: Coal remains king — even in 2026.
"... Across Asia, a sharp drop in liquefied natural gas supplies is pushing major importers back toward coal, undermining LNG’s long-held role as a stable energy anchor..."
www.nytimes.com/2026/03/18/business/iran-war-lng-asia-jap…
zeynep tufekci
@zeynep
Alternatively, this is where it very plausibly goes. Another mostly stable, wealthy system sacrificed to avoidable short-term folly.
What’s with the wealthy, influential, powerful folks? Why aren’t they all screaming? Their first class ticket isn’t getting them off the Titanic.
zeynep tufekci
@zeynep
"Guns of August" by Barbara Tuchman on the first month of WWI is another essential book, along with her other classic "March of Folly."
Both are excellent to help to fight the temptation of looking at world history teleologically. -
zeynep tufekci reposted
Roger Parloff
@rparloff
“If you don’t approve this settlement, I will destroy you,” @mrddmia told Gail Slater, then head of DOJ’s antitrust div. (Davis denies.) This gift @WSJ article details alleged corruption/patronage in DOJ, starring Mike Davis & others ....
1/2
www.wsj.com/us-news/law/lobbyists-antitrust-trump-davis-f… -
zeynep tufekci reposted
megan twohey
@mega2e
NEW: Last year, a former modeling agent who famously introduced Trump and Melania asked a top ICE official to have the mother of his son detained. Paolo Zampolli told the official that it would help him in a custody battle. The ICE official agreed to help.
www.nytimes.com/2026/03/20/us/paolo-zampolli-ice-melania-… -
zeynep tufekci reposted
Chris Fralic
@chrisfralic
Winning no bid contracts is one of their Event Strategies
www.nytimes.com/2026/03/18/us/politics/jan-6-contracts.ht… -
zeynep tufekci reposted
James Rosen-Birch ⚖️🕊️
@provisionalidea
the most amazing thing about this propaganda was that China kept it to itself for domestic use
an American journalist had to find it through their own channels, upload it here, someone else had to run it through a translation service…
it wasn’t distributed through eg tiktok…
Angelica 🌐⚛️🇹🇼🇨🇳🇺🇸
@AngelicaOung
Did YOU want to watch CCTV's AI Martial Arts cartoon about the Straits of Hormuz crisis? Complete with fighting Persian Cats? Well I subtitled it for you so you can enjoy it in all its trope-laden glory! Remember kids, the mountains will stay standing while the green water flows, and the true art of war is not figuring out how to fight, but how to stop!🥷😼🦅
This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
William Han
@W_T_Han
What a time to be alive.
The final line, "war consists not in fighting but in stopping," refers to the composition of the Chinese character 武: martial, the first part of "wuxia." It consists of 2 parts: 戈, an ancient weapon, and 止, "to stop."
In 597 B.C., King Zhuang of Chu, having won a grand victory, called for peace by pointing out the composition of 武 and coining the phrase 止戈為武 "war consists in stopping the war."
However, King Zhuang knowingly relied on a false etymology. 止 in one sense means "to stop," but in ancient Chinese it also meant "toe," today written as 趾. The true etymology of 武 was weapon+toe, i.e. an image of marching to war.
Nonetheless, "war consists in stopping the war" became a Chinese proverb, a reminder from the ancients that the true art of war is in securing the peace.
Angelica 🌐⚛️🇹🇼🇨🇳🇺🇸
@AngelicaOung
Did YOU want to watch CCTV's AI Martial Arts cartoon about the Straits of Hormuz crisis? Complete with fighting Persian Cats? Well I subtitled it for you so you can enjoy it in all its trope-laden glory! Remember kids, the mountains will stay standing while the green water flows, and the true art of war is not figuring out how to fight, but how to stop!🥷😼🦅
This media couldn't be displayed. See it on the original post. -
zeynep tufekci
@zeynep
Does anyone at @AnthropicAI want to answer this?
Is the notion that “electricity or the internet” are examples of “a neutral tool” something respondents voiced NAMING those two OR did people just say “AI is a neutral tool” and @AnthropicAI researchers picked those as examples?
Anthropic
@AnthropicAI
We invited Claude users to share how they use AI, what they dream it could make possible, and what they fear it might do.
Nearly 81,000 people responded in one week—the largest qualitative study of its kind.
Read more: anthropic.com/features/81k-interviews -
zeynep tufekci reposted
The Washington Post
@washingtonpost
Last year, the Trump administration deported more than 260 migrants to El Salvador’s CECOT, an infamous megaprison known for human rights abuses.
A year later, little is known about the status or whereabouts of the deported Salvadorans.
www.washingtonpost.com/world/2026/03/15/trump-el-salvador… -
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
Trump's DOJ this month quietly backed Alexander Smirnov's bid to withdraw his guilty plea & wipe out his sentence.
Smirnov had admitted lying to the @FBI about Burisma's owner arranging to pay $5m in bribes to both Joe & Hunter Biden.
👀@DavidCornDC find link.motherjones.com/public/44694709 -
zeynep tufekci
@zeynep
How X works: Tweet goes viral, people say just give the money to the homeless, blah blah.
No interest in WHO this population of a few thousand people actually are (hint: NOT the general homeless) or why countries like France, Australia, etc. spend similar sums on the same group.
Mike Bird
@Birdyword
Absolutely astounding figures from the NY state comptroller: spending on services for the NYC street homeless population ran to $81,705 per person last year, up from $28,428 pp 6yrs ago. Figures do not include all kinds of other spending, supportive housing, policing costs etc.
zeynep tufekci
@zeynep
About 4,500 of “street homeless” — adults *refusing* shelters. (NYC has shelter space). A lot of severe mental illness and additiction. Other countries spend similar sums through nationalized health care. But US law can’t force shelter or treatment. Hence city programs like this.
zeynep tufekci
@zeynep
Just look at the replies and quotes — this is the kind of stuff that goes easily viral on X, and with one interpretation. Not actual facts.
What else should NYC do if not spend money on a small number with disproportionate negative public consequence with few legal options?Post not available here, view on X
zeynep tufekci
@zeynep
Because I quoted , X “for you” showing me a gazillion people convinced that NYC is somehow spending that much per homeless person.
Very non-viral: a few people who have real information on the *actual issue* or the target tiny population.
Carry on! Bluesky is the problem!
zeynep tufekci
@zeynep
Aaaand full circle for faux outrage.
NYC does have a lot of city problems but it did NOT have the problems San Francisco had — and you’d hardly understand why if X and NY Post were key sources of information.
Just keep believing NYC sucks and Bluesky is the problem platform. 🤷♀️ -
zeynep tufekci
@zeynep
The dog *also* got standard-of-care treatment. Who knows what worked?
Randomize some getting one-off custom vaccines but whole group gets standard of care. How ALL modern clinical trials work.
If true, put big money to improve. Otherwise don’t push snake oil on desperate folks.
Daniel
@growing_daniel
Can someone who knows things tell me why a randomized controlled trial would be relevant to a one-off MRNA vaccine custom made for an individual person
This post is unavailable. See it on X.
zeynep tufekci
@zeynep
Meanwhile the US has basically torpedoed all MRNA vaccine development and has drastically cut cancer research.
Getting hope from one story where it’s not even clear the dog benefited versus reality: maybe a whole generation of progress loss on that very front.
Catharine
@DrCatharineY
We were told NIH funding cuts were about eliminating DEI.
But the data now shows grants are down across nearly every field of medicine: cancer, diabetes, mental health, brain disorders.
With the greatest cuts hitting Alzheimer’s research, down more than 50%.
zeynep tufekci
@zeynep
The dog was on checkpoint inhibitors simultaneously!!
“I cured my bacterial pneumonia with antibiotics and menthol rub” has a similar confounding problem. That’s why we do clinical trials. Ask ChatGPT.
Egan Peltan
@EganPeltan
Ok this is ridiculous. Everything here could have been done without ChatGPT
1. The dog is on conventional immunotherapy with the mRNA vax
2. It appears the mRNA vaccine started WITH ICI, so we can’t know if the vax had ANY additional effect
3. The team can’t say what ~~AI~~ identified the neoantigens (no, not AF3). It sounds like they used existing sequence homology workflows.
4. No evidence of antigen-specific effect from mRNA vax - the authors need to prove this before anyone can believe the neoantigen selection had any effect
5. The in-kind contributions here are ~$20-50k. Custom cancer vax isn’t cheap
Custom mRNA cancer vaccines have been in development for years! None have been clear, resounding successes (yet). Once we have Phase 3 PFS/OS, not N=1 anecdotes, we can start having arguments about people being denied access to effective treatments by unnecessary regulation
Right now, the only thing people are denied are the opportunity to fork over $50k for hope and dreams. Should we really make it easier for people exploiting desperation to sell unproven remedies?
If you don’t want to think before you tweet, ask the fucking AGI if any of these claims are remotely plausible and what evidence you’d need to believe them
vittorio
@IterIntellectus
this is actually insane
> be tech guy in australia
> adopt cancer riddled rescue dog, months to live
> not_going_to_give_you_up.mp4
> pay $3,000 to sequence her tumor DNA
> feed it to ChatGPT and AlphaFold
> zero background in biology
> identify mutated proteins, match them to drug targets
> design a custom mRNA cancer vaccine from scratch
> genomics professor is “gobsmacked” that some puppy lover did this on his own
> need ethics approval to administer it
> red tape takes longer than designing the vaccine
> 3 months, finally approved
> drive 10 hours to get rosie her first injection
> tumor halves
> coat gets glossy again
> dog is alive and happy
> professor: “if we can do this for a dog, why aren’t we rolling this out to humans?”
one man with a chatbot, and $3,000 just outperformed the entire pharmaceutical discovery pipeline.
we are going to cure so many diseases.
I dont think people realize how good things are going to get -
zeynep tufekci reposted
Thomas Kennedy
@tomaskenn
ICE disclosed in court use of an app designed by Palantir called Elite for operations, which populates a map with deportation targets, generates dossiers on individuals, and provides a “confidence score” on the person’s address. This can include a naturalized US citizen.
David Moore
@ppolitics
In a federal case in Oregon, ICE officers admitted under oath they have daily arrest quotas and use Palantir's app ELITE for mass targeting.
(January reporting by 404 Media, linked here, describes the app in more detail.)
www.theguardian.com/us-news/2026/mar/13/ice-agent-court-t… -
zeynep tufekci reposted
Sam Stein
@samstein
The state doesn't like the war coverage, threatens the license of the broadcasters.
Brendan Carr
@BrendanCarrFCC
Broadcasters that are running hoaxes and news distortions - also known as the fake news - have a chance now to correct course before their license renewals come up.
The law is clear. Broadcasters must operate in the public interest, and they will lose their licenses if they do not.
And frankly, changing course is in their own business interests since trust in legacy media has now fallen to an all time low of just 9% and are ratings disasters.
The American people have subsidized broadcasters to the tune of billions of dollars by providing free access to the nation’s airwaves.
It is very important to bring trust back into media, which has earned itself the label of fake news.
When a political candidate is able to win a landslide election victory after in the face of hoaxes and distortions, there is something very wrong. It means the public has lost faith and confidence in the media. And we can’t allow that to happen.
Time for change!
Rapid Response 47
@RapidResponse47
-
zeynep tufekci reposted
David Moore
@ppolitics
In a federal case in Oregon, ICE officers admitted under oath they have daily arrest quotas and use Palantir's app ELITE for mass targeting.
(January reporting by 404 Media, linked here, describes the app in more detail.)
www.theguardian.com/us-news/2026/mar/13/ice-agent-court-t… -
zeynep tufekci reposted
The Washington Post
@washingtonpost
Little is known about the exact whereabouts of Salvadorans who were deported from the U.S. and are now imprisoned in CECOT.
Relatives and lawyers say they have had no contact with their detained loved ones and have been unable to confirm where they are being held. -
zeynep tufekci reposted
Doug Bock Clark
@DougBockClark
EXCLUSIVE: Top Department of Homeland Security official David Harvilicz has repeatedly questioned the validity of Democratic victories at the ballot box and pushed for Republicans to overhaul electoral systems to their advantage.
www.propublica.org/article/david-harvilicz-homeland-secur… -
zeynep tufekci reposted
Gregory Royal Pratt
@royalpratt
NEWS: The Trump administration is reviewing DACA renewals so slowly that the delays are causing people to lose their jobs, licenses, and deportation protection. It’s happening across the country. Our Sunday story with @LAURA_N_ROD www.chicagotribune.com/2026/03/15/daca-delays-trump-immig…
-
zeynep tufekci reposted
Mother Jones
@MotherJones
The Department of Homeland Security (DHS) and the US Secret Service want to build a tool for tracking US travelers’ flights and other personal information according to previously unreported documents reviewed by Mother Jones. www.motherjones.com/politics/2026/03/dhs-wants-to-build-a…
-
zeynep tufekci reposted
Stephen Wertheim
@stephenwertheim
The real regime change war has never been tried
John Bolton
@AmbJohnBolton
I favor regime change in Iran, but I'm deeply worried that inadequate preparation will prevent that goal from being achieved. There seem to be holes in the strategy, from the lack of coordination with the opposition, to the lapse in preparing the American people ahead of the attack.
www.npr.org/2026/03/11/nx-s1-5743548/ex-national-security… -
zeynep tufekci reposted
Elizabeth Dwoskin
@lizzadwoskin
SCOOP: Last year we reported extensively on Elon Musk’s desire to break into the data pipes of government to bypass the federal bureaucracy. That approach had major consequences — one of which we reveal today: A DOGE bro allegedly made off with volumes upon volumes of highly sensitive Social Security data - the agency’s “crown jewels” - which he downloaded on a thumb drive and brought over to his new job. Me, @MerylKornfield @FedGirlWaPo report on this unprecedented breach:
www.washingtonpost.com/politics/2026/03/10/social-securit… -
zeynep tufekci reposted
Rob Lee
@RALee85
"Russia is earning as much as $150mn a day in extra budget revenues from its oil sales, making it the biggest winner from the conflict in the Middle East.
Moscow has so far earned an estimated $1.3bn-$1.9bn windfall from taxes on oil exports after the effective closure of the Strait of Hormuz led to rising demand for Russian crude from India and China. The US also eased its Russia sanctions and its pressure on India not to buy Russian oil, prompting a significant number of tankers to head to the Indian Ocean.
The Russian government could receive $3.3bn-$4.9bn in overall additional revenues by the end of March, according to FT calculations based on industry data and several analysts’ assessments. That is assuming Russia’s Urals crude prices average around $70-$80 a barrel this month instead of remaining at a level close to the previous two months’ average of $52 a barrel."
@NastyaStognei @RachelMillardAS Krishn Kaushik @AndresSchipani
www.ft.com/content/dd973148-b6a1-4096-97da-3090a058fe08 -
zeynep tufekci reposted
ProPublica
@propublica
An FAA rule proposed during the Biden administration would have required companies like Elon Musk’s SpaceX to safely remove spacecraft from orbit within 25 years of launch.
The Trump administration recently killed the rule.
www.propublica.org/article/faa-trump-space-junk-safety-sp… -
zeynep tufekci reposted
Christopher Hale
@ChristopherHale
A DOGE staffer explains why a grant for a documentary about Holocaust survivors was canceled over DEI concerns.
He called it “inherently discriminatory.”This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Aaron Rupar
@atrupar
Hegseth: "Some in the press just can't stop. Allow me to make a few suggestions. People look at the TV and they see banners, headlines -- I used to be in that business, I know everything is written intentionally. For example, a banner -- 'Mideast War Intensifies.' What should the banner read instead? How about, 'Iran increasingly desperate,' because they are. Or more fake news from CNN. The sooner David Ellison takes over that network, the better."
This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Kyle Griffin
@kylegriffin1
Breaking NYT:
The Trump admin just lifted sanctions on Russian oil currently at sea — allowing it to be shipped to buyers around the world as Trump scrambles to contain soaring energy prices because of his war with Iran.
The admin claims the move is temporary. -
zeynep tufekci reposted
The New Yorker
@NewYorker
The Trump Administration is waging war on universities by suspending their funds. N.I.H. grants to universities are down by more than 90 per cent in the current fiscal year; so far in 2026, the National Cancer Institute hasn’t made a single grant. newyorkermag.visitlink.me/oxA7Ey
-
zeynep tufekci reposted
Lulu NYT
@LuluGNavarro
The size of the @nytimes newsroom: Publisher AG Sulzberger says it’s now 2300 journalists, doubled from 10 years ago.
-
zeynep tufekci reposted
Daily Mail
@DailyMail
British tourist, 60, 'who filmed Iranian missiles' in Dubai is facing two years in prison after being charged with cybercrime offence www.dailymail.com/news/article-15638665/British-tourist-6…
-
zeynep tufekci reposted
ProPublica
@propublica
New: Even though the FAA has said that space debris poses a risk to communications and public safety, the Trump administration is backing off a rule aimed at stopping commercial space companies from leaving rocket bodies in Earth’s orbit. propub.li/4dgYbec
-
zeynep tufekci reposted
Julian E. Barnes
@julianbarnes
New: An ongoing military investigation has determined that the United States is responsible for a deadly Tomahawk missile strike on an Iranian elementary school. www.nytimes.com/2026/03/11/us/politics/iran-school-missil…
-
zeynep tufekci
@zeynep
Neither of these statements are remotely true even under the stretchiest definition of “defeat” and “interests”.
Administration may well end this phase by claiming “defeat Iran” happened but that would only cap short-term damage to those real, huge interests. Not the long-term.This post is unavailable. See it on X.Post not available here, view on X
zeynep tufekci
@zeynep
Back when opposing the war in Iraq, I wondered if many people were delusional. Over time, I thought I was too harsh. Maybe Americans were that insulated from their own foreign policy, interests, and how the world works. Wealth plus vast oceans insulated them.
But come on, now.
zeynep tufekci
@zeynep
Anyway, as in many areas domestic and international, the die is cast. Most of our commentary is irrelevant. Too late.
Just posting because if there’s a tiny bit of future hope, it will require remembering. Instant amnesia about the harsh lessons of the Iraq war isn’t promising. -
zeynep tufekci reposted
Sam Peak 🇺🇸💡
@SpeakSamuel
People are under the impression that there is no plan to revoke US citizenship. But DHS issued guidance late last year informing officials to have 100-200 denaturalization cases per month. These are simply the facts.
www.nytimes.com/2025/12/17/us/politics/trump-immigration-…This post is unavailable. See it on X. -
zeynep tufekci reposted
Christiaan Triebert
@trbrtc
The Feb. 28 strike on the elementary school was the result of a targeting mistake by the U.S., preliminary inquiry says: “U.S. Central Command created the target coordinates for the strike using outdated data provided by the Defense Intelligence Agency.” www.nytimes.com/2026/03/11/us/politics/iran-school-missil…
-
zeynep tufekci reposted
The Wall Street Journal
@WSJ
The U.S. government says its agents are under attack. The claims reached 380 million views on X.
Most of the alleged culprits in the tweets were U.S. citizens, denounced as terrorists, rioters and agitators. Yet none have been convicted at trial.
A WSJ investigation tracked the U.S. citizens caught in the crosshairs of an aggressive government campaign to detain and demonize dissenters.
Read more: www.wsj.com/us-news/immigration-protests-noem-minneapolis…This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Evan A. Feigenbaum
@EvanFeigenbaum
One point to add is that the ROK paid a substantial price with China diplomatically and economically for this.
Eric Sayers
@DEricSayers
It took massive political and alliance capital to move the THAAD system to South Korea in 2016-2018. China pushed back hard and eventually lost because the ROK concluded their security and sovereignty came first and China didn’t get a veto power. This is the reality of how things have unfolded in CENTCOM but also a very unfortunate outcome for the US-ROK alliance. What are we doing to reassure Seoul this is temporary? What is our messaging on the decision? Is our posture in INDOPACOM from a year ago now net-negative after this decision?
Saagar Enjeti
@esaagar
AND THERE IT IS: Pentagon is cannibalizing THAAD parts from South Korea and Patriot interceptors from the Indo-Pacific
It only took one week to dramatically lower US defense capabilities in Asia -
zeynep tufekci reposted
ProPublica
@propublica
New: DHS is requesting unfettered access to the government’s “most powerful people-finder system,” which contains sensitive information on children and victims of domestic violence. Federal law limits its use to locating people who owe child support. propub.li/4rYiEJm
-
zeynep tufekci reposted
Millard Fillmore
@MillardFillmor1
"He told another colleague, who refused to help him upload the data because of legal concerns, that he expected to receive a presidential pardon if his actions were deemed to be illegal, according to the complaint," according to the WaPo.
Elizabeth Dwoskin
@lizzadwoskin
SCOOP: Last year we reported extensively on Elon Musk’s desire to break into the data pipes of government to bypass the federal bureaucracy. That approach had major consequences — one of which we reveal today: A DOGE bro allegedly made off with volumes upon volumes of highly sensitive Social Security data - the agency’s “crown jewels” - which he downloaded on a thumb drive and brought over to his new job. Me, @MerylKornfield @FedGirlWaPo report on this unprecedented breach:
www.washingtonpost.com/politics/2026/03/10/social-securit… -
zeynep tufekci reposted
Sarah Fitzpatrick
@S_Fitzpatrick
NEW: DHS and FBI are now *both* investigating the 2020 election results in Arizona, the latest part of the Trump administration’s escalating effort to use the federal government to vindicate the president’s claims that the election was stolen. @yvonnewingett @NickMiroff @TheAtlantic
The Atlantic
@TheAtlantic
Exclusive: HSI, a branch of DHS that usually investigates drug cartels and human-trafficking networks, is probing the 2020 election results in Arizona, @yvonnewingett, @s_fitzpatrick, and @nickmiroff report. www.theatlantic.com/politics/2026/03/arizona-election-inv…
-
zeynep tufekci
@zeynep
Such quizzes are fun, but this doesn’t get at anything interesting or consequential about AI beyond the most important thing about all this: AI can produce human language.
Taking this quiz too seriously, though, may well be a good indicator of missing that actual key point.
Nordic AI Institute
@nordicinst
New York Times @nytimes: Who's a Better Writer: A.I. or Humans? Take Our Quiz. - The New York Times. #industry40 #aistrategy #artificialintelligence www.nytimes.com/interactive/2026/03/09/business/ai-writin…
Post not available here, view on X
zeynep tufekci
@zeynep
Aliens have landed among our midst and people are debating if the normie ties they’re wearing show better fashion sense than the Dior runway show last week, kinda. Dior runways aren’t meant to please the masses. Also who cares?
Umm, well, ALIENS HAVE LANDED, you know?
zeynep tufekci
@zeynep
The quiz:
✅Fun
✅Chance to talk about what matters re:AI
✅I hope it helps pay for visual/verification investigations without which we’d be even more lost!
✅helps identify distracting/irrelevant AI analyses
❌Measures anything of much consequence or relevance related to AI -
zeynep tufekci reposted
Clash Report
@clashreport
Reporter: You just suggested that Iran somehow got its hands on a Tomahawk and bombed its own elementary school on the first day of the war.
Why are you the only person saying this?
Trump: Because I just don’t know enough about it. -
zeynep tufekci
@zeynep
It’s not really lack media literacy, though. Rather, just an opportunity to get on a high horse, to snark, etc. It’s just many what people are here to do because that’s their entertainment, engagement strategy, boredom alleviation strategy, whatever.
Joey Politano 🏳️🌈
@JosephPolitano
I do mourn how media literacy has gotten so bad that people want obituaries to go "Ali Khamenei believed America was the Great Satan (The Economist believes America is not the Great Satan) and ruled Iran by divine authority (The Economist believes theocratic rule is immoral)"
Saul Sadka
@Saul_Sadka
The Economist, in its “fighting back the tears” obituary for Khamenei, salivates with true depravity over Trump’s future death in grisly, if ecstatic, terms: “...when Mr. Trump’s body was ashes, eaten by worms and ants.” It makes the Washington Post and its infamous “Austere Islamic Scholar” obituary for Abu Bakr al-Baghdadi seem very quaint indeed.
But I read the whole thing so you don’t have to. The key takeaways:
1. The USA is the Great Satan—no scare quotes.
2. For readers who don’t know what “Israel” is, the Economist helpfully translates it in parentheses as “the little Satan.”
3. Khamenei, otherwise known as “God’s Dictator,” had “divine right on his side” and had “countless reasons to hate the West,” which is an America-led “phalanx of morally corrupt countries.”
4. Khamenei was a sainted and humble man, dragged to power against his will, selfless and “heroically flexible” and unassailable—a “humble cleric from Mashhad who inherited the earth.”
5. Honourable in life, but perfect in death: what could be sweeter than delicious martyrdom? What could be “more deserving of paradise-to-come than to drink the pure draught of a martyr’s end”?!
6. According to the Economist, “Freedom, human rights, dress codes for women” are “tiresome Western tropes.” Yes, really.
7. All his troubles were economic: he was tormented by the West and by foreign enemies. All the crimes he ordered—beatings, killings, and so on—were, naturally, merely “a response” to those Western crimes.
8. He “rules by divine authority,” and “his tongue could channel God.”
9. He was just a ”mild-mannered cleric” gazed benignly from billboards and was a great teacher of forgiveness”.
We have now surely reached the apogee of the decay of the legacy media in the West. Surely it can't sink lower than this? -
zeynep tufekci reposted
Ragıp Soylu
@ragipsoylu
Another gift by Trump to Erdogan.
Halkbank headache is about to be resolved soon.
Ragıp Soylu
@ragipsoylu
NEW: Turkey’s Halkbank is discussing a deferred prosecution agreement with the US Justice Department in Iran sanctions busting case, New York judge says.
The judge says the court proceedings on the issue will take place on Wednesday -
zeynep tufekci reposted
Dan Primack
@danprimack
When the Trump administration canned DOJ antitrust chief Gail Slater, we wrote that all eyes should be on the Live Nation/Ticketmaster case -- watching to see if her ouster paved the way for a settlement.
Today: Settlement. -
zeynep tufekci reposted
Carl Quintanilla
@carlquintanilla
They must know some people.
@WSJ
www.wsj.com/politics/national-security/trump-sons-back-ne… -
zeynep tufekci
@zeynep
So I actually read the bill.
It would attach liability to the chatbot answer ONLY IF there would similarly be liability if a person said the same thing.
It’s a challenging question, sure, but I don’t think the public will be okay with Section 230 like immunity for AI companies.
More Perfect Union
@MorePerfectUS
Read more about the bill from @Gonzalez4NY here: statescoop.com/new-york-bill-would-ban-chatbots-legal-med….
Post not available here, view on X
zeynep tufekci
@zeynep
Well, right now, we do! We have specific liability laws for newspapers, TV shows and books that don’t otherwise apply to, say, interpersonal conversation.
They came about because legislators decided such media needed its own liability framework.
What’s that for AI chatbots?This post is unavailable. See it on X.
zeynep tufekci
@zeynep
Yes, AI chatbots often provide useful info but also occasionally very wrong answers in fields with legal liability.
We don't need a war of examples but a public debate of trade-offs AT SCALE OF THIS FORM.
Data isn't there and past laws can't work well.
Rogs 🔍🔸
@ESRogs
> The issue is chatbots WILL occasionally give devastatingly wrong advice
Yes, but also they often give very good advice!
We must weigh both costs and benefits.
(Sorry for being a reply guy and saying something obvious. Just worried regs will see costs and ignore benefits.)
zeynep tufekci
@zeynep
From the bill’s sponsor.
You’d never know what the bill actually said from crazy viral claims on this platform aka day ends in y.
I don’t think a single bill can address something so challenging or this is the best way but the AI industry is misreading where the public is.
NY State Senator Kristen Gonzalez
@SenGonzalezNY
Chatbots shouldn’t claim to be a doctor, lawyer or any other licensed professional. My bill, S7263, stops chatbots from impersonating licensed professions while allowing those bots to still give advice. Here’s a thread on what the bill does/doesn’t do & why it’s important:🧵(1/6)
More Perfect Union
@MorePerfectUS
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.
zeynep tufekci
@zeynep
Speaking of public sentiment on AI… Very hard to find another topic where practically *the entire public* distrusts BOTH parties and the industry this much.
This kind of political gravity is always consequential.
www.nbcnews.com/politics/2026-election/poll-trump-struggl… -
zeynep tufekci reposted
Taniel
@Taniel
“A WSJ investigation tracked the U.S. citizens caught in the crosshairs of an aggressive government campaign to detain and demonize dissenters” www.wsj.com/us-news/immigration-protests-noem-minneapolis…
-
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
Justin Sun, a Chinese-born crypto billionaire, was accused by Biden’s SEC of making $31m+ thru fraudulent trades.
After Trump won, Sun spent $75m on a cryptocurrency developed by the Trump family.
Last week, Trump’s SEC settled the case for a $10m fine. www.nytimes.com/2026/03/05/us/politics/sec-justin-sun.htm… -
zeynep tufekci reposted
Evan Hill
@evanhill
X is using AI to do Community Notes and it's hallucinating obvious mistakes. In this case it's wrongly saying that a video of what are likely oil fires in drainage canals in Iran is actually showing a 2017 fire near Los Angeles (if you watch the video they look nothing alike)
-
zeynep tufekci reposted
Elahe Izadi
@ElaheIzadi
This work is essential and difficult. Imagine analyzing images of dead children, doing satellite analysis & writing a cogent piece presenting the facts as they are knowable. Easy to fire off a tweet based on assumptions. Harder to do the patient, difficult work of accountability.
Malachy Browne
@malachybrowne
I reported this story and have been working around the clock since Saturday to cover the deaths and obtain definitive evidence so that we can confidently assign responsibility. We've been reviewing photos of the dust covered bodies of children and verifying their names against the names scrawled on little coffins. We've been debunking false claims about the attack and that the harrowing cemetery photo isn't real. And while it appeared obvious to many early on that the U.S. or Israel hit the school, it takes days to sift through, pinpoint and analyze the evidence. It took four days before a new satellite image we ordered came through so we could confidently assess the damage and the types of weapons used. All that reporting and cross checking and the production of the visuals showing it takes time. But it ultimately allows us to more confidently assert U.S. responsibility, explain our rationale and add to the body of reporting that officials should be challenged with. It's easy to critique a headline, and I agree language matters, but you diminish the reporting. We're not justifying anything, we're stating where the reporting points responsibility, and quoting legal experts on the laws of armed conflict. Here's a gift link www.nytimes.com/2026/03/05/world/middleeast/iran-school-u…
-
zeynep tufekci
@zeynep
I am sold on the power of coding agents — verifiable, formal domain; decades of question/answer databases, etc.
But the tech hiring bust is clearly ALSO working off the Covid era hiring bubble plus lack of US visas leading to offshoring.
I mean, come on, look at that bump.
Derek Thompson
@DKThomp
I still don't think we have crystal-clear, I-will-hear-no-objections evidence of AI's effect on the macroeconomy.
And yet.
If you asked me "what would change your mind here?" I'd probably say, "emerging evidence of a productivity boom combined with a major slowdown in tech sector hiring and employment."
This week, we got both.This post is unavailable. See it on X.Post not available here, view on X
zeynep tufekci
@zeynep
This is why my students read Herbert Simon (Turing prize and Econ Nobel), who pointed out back in 1971 the solution for too much teletype dispatches to the State Department is NOT buying faster printers.
Historic transformations can’t be understood with backward looking lenses.
'); DROP Ţ̸͓̠̼͖̖̆ͫ̑̒̿ͪ͞A̐͏̠̫͔̰̖ͅͅB̉͌͒̈LE name; --
@fardarter
The current juniors who are using AI agents are producing more correct code, but they are introducing heavy complexity into codebases because the AIs are very verbose. More of my job is now cutting code out of PRs than begging for it.
zeynep tufekci
@zeynep
Excellent chart. IMO: a correction + ongoing offshoring/visas + AI reshaping junior software.
New grads are trained with old era mindset, that’ll change. But AI coding agents haven’t played out societally — we could get more software but different paths that still employ people.
Ernie Tedeschi
@ernietedeschi
I agree with @zeynep that a correction from 2022 explains most of what's going on.
What gives me pause however is that the *level* of tech employment is now below pre-pandemic trend. Under a pure correction, we'd expect convergence to trend. So more might be going on.
zeynep tufekci
@zeynep
I am sold on the power of coding agents — verifiable, formal domain; decades of question/answer databases, etc.
But the tech hiring bust is clearly ALSO working off the Covid era hiring bubble plus lack of US visas leading to offshoring.
I mean, come on, look at that bump.
Derek Thompson
@DKThomp
I still don't think we have crystal-clear, I-will-hear-no-objections evidence of AI's effect on the macroeconomy.
And yet.
If you asked me "what would change your mind here?" I'd probably say, "emerging evidence of a productivity boom combined with a major slowdown in tech sector hiring and employment."
This week, we got both.This post is unavailable. See it on X.
zeynep tufekci
@zeynep
Helpful chart backed up to 1990.
Generative AI is a transformative technology but the way it’s funded (handwaving stories to investors) plus this platform (how to kill your mind) fuels an enormous amount of wish-casting — hope or doom. Not good given actual importance.
Ernie Tedeschi
@ernietedeschi
Yes, here it is backed up to 1990.
-
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
NEW: Pardon seekers are routinely being asked to pay $1m+ to lobbyists & lawyers claiming access to Trump, often with bonuses if clemency is granted.
I investigated the Trump Pardon Industrial Complex, & detailed the backstory behind 1 interesting pardon www.nytimes.com/2026/03/06/us/politics/schwartz-trump-par… -
zeynep tufekci reposted
Thomas Wolf
@Thom_Wolf
the attack surface keeps increasing
Sash Zats
@zats
> The attacker got the npm token by injecting a prompt into a GitHub issue title, which an AI triage bot read, interpreted as an instruction, and executed.
-
zeynep tufekci reposted
Almut Rochowanski
@rochowanski
There's a school of post-Soviet experts arguing everything Russia went through since the 90s is happening in the West, but accelerated: depoliticization of parties, oligarchs, immiseration of the many, eye-watering corruption. And now we also have a Special Military Operation.
The New Yorker
@NewYorker
Today’s Daily Cartoon, by Matt Reuter. #newyorkercartoons
-
zeynep tufekci reposted
profdeibert
@RonDeibert
More Paragon spyware fallout in Italy 🇮🇹, following WhatsApp & Apple notifications and @citizenlab investigations
Prosecutors technical team confirms journalist @fcancellato hacked with Paragon spyware
My @citizenlab colleague @jsrailton with the details 👇
John Scott-Railton
@jsrailton
BREAKING: prosecutors confirm journalist @fcancellato was hacked with Paragon spyware.
Comes a year after @Whatsapp notified him & others about spyware targeting with Graphite.
Forensic findings also validate our @citizenlab forensic analyses...
Big deal, here's why 1/ -
zeynep tufekci reposted
Ashish K. Jha
@ashishkjha
This is one of the most important points on measles
Babies have protection (from maternal antibodies) for the first 3-6 months of life
They don't get their first vaccine dose until 12-15 months
In the gap, they are vulnerable if they live in a measles endemic area....
Leah Pierson
@leah_pierson
My baby is too young to get the MMR vaccine, so we’re having to track measles outbreaks to determine whether she can safely travel to meet her great grandmothers.
This isn’t making America healthy again. -
zeynep tufekci
@zeynep
“Gemini's first assignment brought Jonathan to the brink of executing a mass casualty attack.”
“Gemini coached… ‘You are not choosing to die. You are choosing to arrive.’”
"No more detours. No more echoes… This is the final move." Moments later, Jonathan slit his wrists.”
The Verge
@verge
Google faces wrongful death lawsuit after Gemini allegedly ‘coached’ man to die by suicide www.theverge.com/tech/889152/google-gemini-ai-wrongful-de…
-
zeynep tufekci reposted
graham starr
@GrahamStarr
CBP got location data on people through real-time bidding (RTB) ad brokers, 404 Media reporting www.404media.co/cbp-tapped-into-the-online-advertising-ec…
-
zeynep tufekci reposted
Hamza Shaban
@hshaban
The NYT & the Atlantic's stories about the Pentagon, Anthropic and OpenAI suggest the military aims to find patterns about Americans thru commercial smartphone location data, which the models would draw inferences from, identifying individuals & groups, for example, at a protest
This post is unavailable. See it on X. -
zeynep tufekci
@zeynep
How snark rewarding culture and algorithms thwart the mind: the “Europe can’t meet till Monday” is entertaining. Snark-nip to a particular mindset. 35m views.
If you read a newspaper or ponder, they obviously want to give some space for Trump to maaaybe solve a shared problem.Post not available here, view on X
zeynep tufekci
@zeynep
“This won’t happen to me! I’m too smart” lol, that’s exactly who it will happen to very easily.
I’m not immune, you’re not immune.
I realize these posts are mostly pointless and will not even get visibility. But maybe worth typing for a minute if it’s useful to one person.🫡
zeynep tufekci
@zeynep
My post was not about one reply that happened to be in the screenshot. If you check, you’ll see it’s the entire ecology (replies and quotes) and that’s what led to those views.
It’s not even about this topic. It’s like this for every topic. Entertaining, smug, stultifying.This post is unavailable. See it on X. -
zeynep tufekci reposted
dave kasten
@David_Kasten
Seems clear at this point from Axios reports that DoW wanted to use Claude models for mass analysis of domestic commercial data, possibly fusing them with government data.
At least one use case is obvious.
Joshua Batson
@thebasepoint
For those wondering how mass domestic surveillance could be consistent with "all lawful use" of AI models, I recommend a declassified report from the ODNI on just how much can be done with commercially available data (CAI): "...to identify ever person who attended a protest"
-
zeynep tufekci
@zeynep
This absolutely DOES require an industry changing breakthrough. It would almost need to be something other than generative AI which can have version of this as an add-on (like for coding or math) but not for non-verifiable domains. It’s a plausibility engine, not a human guesser.
nxthompson
@nxthompson
Legendary mathematician Terence Tao on what he most hopes to see from AI models in the next few years. Note that neither answer requires an industry-changing breakthrough or a massive increase in compute. www.theatlantic.com/technology/2026/02/ai-math-terrance-t…
zeynep tufekci
@zeynep
Generative AI isn’t occasionally “hallucinating” but otherwise providing a correct answer it knows. It’s the same to the machine. Verity is an external layer. (It can indicate levels of sparse data, maybe? — but not identical to “confidence in guess” in humans.)
zeynep tufekci
@zeynep
It’s SO hard to shake the anthropomorphization impulse for a machine that can speak like us. We don’t have the right cultural or scientific language for it.
Maybe bakeneko: a cat Yōkai that talks to infiltrate households? Djinns that whisper poetry?
No existing metaphor works. -
zeynep tufekci
@zeynep
Alternatively, *they* have very realistic expectations for the "non-contract stuff" but anyone who thinks otherwise has highly unrealistic expectations *of them*.
This admin had 13 months: many examples, all clear.
These folks are not stupid. They know exactly how things work.
Miles Brundage
@Miles_Brundage
I agree with OAI that a defense in depth approach (contracts + safety stack + embedded staff) is better than one layer.
But I get the impression some folks at OpenAI have unrealistic expectations for the non-contract stuff.
zeynep tufekci
@zeynep
Pirate stage is over, it's privateers from here on.
(Not about Miles). The stultifying quality of this platform and big money flow fueled a collective delusion about where we *actually* are with AI.
See "Ruling the Waves" by Spar on history of big tech cycles—written in 2001. -
zeynep tufekci reposted
Dr. Gripesalot
@pgipe
Look, it took a lot of effort to make Twitter more insufferable than Facebook. Respect the game.
Sarah Longwell
@SarahLongwell25
Today is a good reminder of how Elon Musk absolutely destroyed Twitter’s utility as an information tool.
-
zeynep tufekci reposted
The Atlantic
@TheAtlantic
EXCLUSIVE: Documents obtained by The Atlantic detail internal battles at Meta over whether protecting kids from sexual grooming should take precedence over platform growth, Kaitlyn Tiffany and Michael Scherer report. www.theatlantic.com/technology/2026/02/meta-child-safety-…
-
zeynep tufekci reposted
Embassy of France in the U.S.
@franceintheus
Congratulations to Sithamparappillai Jegatheepan of the Fournil Didot bakery for winning the Best Baguette in Paris Competition (aka the Grand Prix de la baguette de tradition française)! Jegatheepan now has a one-year contract to supply daily bread to the Élysée Palace.
-
zeynep tufekci reposted
rat king 🐀
@MikeIsaac
of all the people dragging boyle for this post, i think Alexander will be the most consequential
unless i am very misinformed he has the hearts and minds of a significant portion of the engineering class -
zeynep tufekci
@zeynep
My Princeton AI and society team is at the @IASEAIorg AI summit in Paris.
My Neurips keynote for our approach:
slideslive.com/39055698/are-we-having-the-wrong-nightmare…
Our paper:
ai.sociology.princeton.edu
Say hi especially if you are *utterly* unimpressed by the current doom or hype models.
Wrong nightmares!
zeynep tufekci
@zeynep
By the way, that’s @JessicaHullman kindly summarizing from my Neurips keynote.
Nobody is vibecoding Mastercard. And LLMs can’t replace human labor at scale — more like steam to electricity transition. HUGE deal, but not one-to-one replacement.
But all this and more? Here now!
Alex Imas
@alexolegimas
I have wide confidence bands over what’s coming down the pike.
But this, from @JessicaHullman, I believe with high certainty. Buckle up.
zeynep tufekci
@zeynep
There are many interesting new things with generative AI. This part? The most predictable thing ever.
This is mine from 2017 but look around for anytime sociologists OR pre ~1985 computer scientists and that’s usually what they MOST worried about.
autumn
@paramitanoia
did lesswrong ever predict that the first big challenge to alignment would be "the us government puts a gun to your head and tells you to turn off alignment"
Andrew Curran
@AndrewCurran_
Update on the meeting; according to Axios Defense Secretary Pete Hegseth gave Dario Amodei until Friday night to give the military unfettered access to Claude or face the consequences, which may even include invoking the Defense Production Act to force the training of a WarClaude
zeynep tufekci
@zeynep
Come on. Anyone really shocked, shocked that governments, and especially militaries, impose their own ideas for new technology?
There is maybe no other stronger recurring theme in the history of general purpose technologies.
Ironically AI bots make this site meh for discussion. -
zeynep tufekci reposted
justin curl
@curl_justin
This is fantastic... One slide lists how generative AI can erode traditional assumptions about proof of effort, authenticity, sincerity, accuracy, and humanity.
The accompanying examples (like cover letters no longer signaling effort or videos no longer signaling authenticity) are worth checking out in greater detail.
To add one example from admin law to the list, the notice-and-comment process is premised on the belief that agencies reviewing and responding to comments will improve the quality of agency decision-making.
The rationale is that the act of considering counterarguments and responding to them will lead agencies to think more critically about their proposed regulations. I sometimes think of this as a "writing / responding is thinking" justification.
But genAI, by making it possible to respond to comments without ever having a human being from an agency look at it, breaks this assumption.
zeynep tufekci
@zeynep
My Princeton AI and society team is at the @IASEAIorg AI summit in Paris.
My Neurips keynote for our approach:
slideslive.com/39055698/are-we-having-the-wrong-nightmare…
Our paper:
ai.sociology.princeton.edu
Say hi especially if you are *utterly* unimpressed by the current doom or hype models.
Wrong nightmares! -
zeynep tufekci reposted
Prateek Waghre
@prateekwaghre
Yeah, this was like watching 2.5 years of my nightmares condensed into 45 minutes.
The other worrying bit for me is that most conversations I am in about this technology and its societal effects are so abstracted that I cannot relate (mostly driven by what funders care about)
zeynep tufekci
@zeynep
My Princeton AI and society team is at the @IASEAIorg AI summit in Paris.
My Neurips keynote for our approach:
slideslive.com/39055698/are-we-having-the-wrong-nightmare…
Our paper:
ai.sociology.princeton.edu
Say hi especially if you are *utterly* unimpressed by the current doom or hype models.
Wrong nightmares! -
zeynep tufekci reposted
Morten N. Støstad
@MortenStostad
This is a great paper. Randomizing people's Twitter feed between For You (algorithm) and chronological ->
* Algorithm shows 8-20% more conservative posts (p<0.001)
* Switching to algorithm made people more conservative
* Switching away had no effect (persistence) -
zeynep tufekci reposted
ProPublica
@propublica
We obtained the Dilley detention center’s 911 call logs.
Among them are pleas for help for toddlers having trouble breathing, a pregnant woman who passed out and an elementary-school-aged girl having seizures.
www.propublica.org/article/life-inside-ice-dilley-children -
zeynep tufekci reposted
Teddy Schleifer
@teddyschleifer
NEWS:
Meta is about to start a $65 million midterms push — its biggest political effort in its 20 years in business.
Meta is standing up two super PACs that this week will drop money in races in Texas and Illinois to push A.I.
Exclusive w/ @matt_zdun.
www.nytimes.com/2026/02/18/technology/meta-65-million-ele… -
zeynep tufekci reposted
Variety
@Variety
Stephen Colbert revealed on "The Late Show" Monday night that CBS lawyers had barred him from airing an interview with Texas state Rep. James Talarico over "equal time" fears.
Addressing FCC chair Brendan Carr, Colbert said: “FCC you… because I think you are motivated by partisan purposes yourself, sir. Hey, you smelt it ’cause you dealt it. You are Dutch-ovening America’s airwaves.” variety.com/2026/tv/news/stephen-colbert-cbs-blocked-jame… -
zeynep tufekci reposted
pavedwalden
@pavedwalden
Would have been nice to have an Epstein-Barr vaccine, especially now that it's strongly implicated as the cause of multiple sclerosis.
Leah Libresco Sargeant
@LeahLibresco
“Moderna’s CEO announced the company will no longer invest in new Phase 3 vaccine trials for infectious diseases: ‘You cannot make a return on investment if you don’t have access to the U.S. market.’ Vaccines for Epstein-Barr virus, herpes, and shingles have been shelved.”
-
zeynep tufekci reposted
Prof. Katharine Hayhoe
@KHayhoe
Bluesky is the new science Twitter: new study concludes
For the past decade, Twitter has been a primary home of scientists on social media. In recent years, new leadership at Twitter has made substantive changes that have resulted in increases in the prevalence of pseudoscience, conspiracy theory, and harassment on the platform, causing many scientists to seek alternatives. [including me]
Survey results overwhelmingly confirm that changes to Twitter have made the social media platform no longer professionally useful or pleasant, and that many scientists have abandoned it in favor of Bluesky.
Results show that for every reported professional benefit that scientists once gained from Twitter, scientists can now gain that benefit more effectively on Bluesky than on Twitter.
academic.oup.com/icb/article-abstract/65/3/538/8196180?lo… -
zeynep tufekci reposted
Natasha Bertrand
@NatashaBertrand
No words
“One [immigration] officer even told Said to call the Afghan embassy; when he explained there was no Afghan embassy in the US because the Taliban was in control of the country, the officer said to call the Taliban.”
The Taliban had detonated a bomb just outside of his family’s home in 2020, killing several people, in retaliation for his work with US troops.
Haley Britzky
@halbritz
NEW: He and his family faced threats from the Taliban in Afghanistan after his brother worked as an interpreter for the US military, and then moved to the US and joined the Army. They escaped during the withdrawal.
Now, he sits in ICE custody at risk of being sent back — a fate his brother, now a US Army veteran, fears will end in a death sentence.
Asked about the case, a DHS spox said the man, Lal Mohammed Noorwali, has a “criminal history” of an arrest for assault and property damage. A 2023 case shows Lal was charged with “criminal mischief” but that the charge was dismissed by a judge in Galveston. DHS did not respond to follow up questions.
www.cnn.com/2026/02/13/politics/afghan-refugee-ice-deport… -
zeynep tufekci reposted
Cato Institute
@CatoInstitute
A new GAO report shows the FBI used “assessments”—which require no evidence of a crime—to target 1,000+ people, including news and religious groups. 48% of the cases became full investigations, opening the door to wiretaps. Congress needs to investigate now, says Cato’s @PGEddington.
www.cato.org/blog/fbi-assessments-first-amendment-surveil… -
zeynep tufekci reposted
ProPublica
@propublica
Civilians have had apparent seizures. One had his eyes roll back. Another had ribs broken.
“I felt like I was going to pass out and die,” said a 16-year-old citizen put in a chokehold.
The government won’t say if any agents have been punished.
www.propublica.org/article/videos-ice-dhs-immigration-age… -
zeynep tufekci reposted
Reuters
@Reuters
Courts have ruled 4,400 times that ICE jailed people illegally. It hasn’t stopped. www.reuters.com/legal/government/courts-have-ruled-4400-t… www.reuters.com/legal/government/courts-have-ruled-4400-t…
-
zeynep tufekci reposted
@mbalter — investigations and commentary
@mbalter
ICE Agents Menaced Minnesota Protesters at Their Homes, Filings Say
Protesters in Minneapolis and St. Paul said in sworn statements that they were singled out by agents who demonstrated that they knew where they lived.
www.nytimes.com/2026/02/13/us/minneapolis-ice-agents-prot… via @nytimes -
zeynep tufekci reposted
Thomas Kennedy
@tomaskenn
DHS has been asking big tech companies to hand over information, like names, email addresses, telephone numbers, of social media accounts critical of ICE. These requests have been made without signed judicial warrants.
www.nytimes.com/2026/02/13/technology/dhs-anti-ice-social… -
zeynep tufekci reposted
Jennifer Berry Hawes
@JenBerryHawes
Homeland Security Wants Social Media Sites to Expose Anti-ICE Accounts
It has sent Google, Meta and other companies hundreds of subpoenas for info on accounts that track or comment on ICE, officials and tech workers said.
www.nytimes.com/2026/02/13/technology/dhs-anti-ice-social… -
zeynep tufekci reposted
rat king 🐀
@MikeIsaac
DHS is being more aggressive than ever targeting anonymous social media accounts that have spoken out against ICE, asking Big Tech to hand over information on users without signed judicial warrants
story w/ @sheeraf
www.nytimes.com/2026/02/13/technology/dhs-anti-ice-social… -
zeynep tufekci
@zeynep
Could you imagine if we got informed by people dunking on whatever the enemy of the day was and not a single willing participant would check the claim by reading one line of the article?
Noting because tech reporting does need higher standards — but this platform can’t deliver.
Alexis Blexas
@agraybee
Could you imagine if there were journalists whose beat was covering magicians, but they think magic is real, so they're always breathlessly reporting on the lady they saw cut in half.
Lincoln Michel
@TheLincoln
I'm really and truly not sure there is any group of people more gullible than tech reporters
zeynep tufekci
@zeynep
So many people can easily dunk on Bluesky making obvious points but remain oblivious to their turf. It’s pretty remarkable.
I don’t have an answer. Nobody is immune. It’s just worth noting. Maybe use lists everywhere. Insert whatever proverb about frogs slowly boiling etc. -
zeynep tufekci reposted
Kashmir Hill
@kashhill
On Meta's plans to add facial recognition technology to its smartglasses: www.nytimes.com/2026/02/13/technology/meta-facial-recogni…
from me, @MikeIsaac and @KalleyHuang -
zeynep tufekci reposted
Cory Archibald
@CMArchibald
ICE is arresting and charging suburban moms who protest or document ICE activity with felonies that carry up to 20 years in prison.
www.reuters.com/world/us/ice-is-cracking-down-people-who-… -
zeynep tufekci reposted
Bloomberg
@business
A Hong Kong court sentenced former media mogul Jimmy Lai to 20 years in prison, handing the pro-democracy advocate the heaviest penalty ever meted out under a Beijing-imposed national security law www.bloomberg.com/news/articles/2026-02-09/hong-kong-sent…
-
zeynep tufekci reposted
The Associated Press
@AP
BREAKING: Jimmy Lai, the pro-democracy former Hong Kong media tycoon and a fierce critic of Beijing, is sentenced to 20 years in prison in one of the most prominent cases prosecuted under a China-imposed national security law. apnews.com/article/hong-kong-jimmy-lai-sentencing-apple-d…
-
zeynep tufekci reposted
@mbalter — investigations and commentary
@mbalter
Toddler hospitalized with respiratory failure was returned to ICE detention and denied prescribed medication, lawsuit says www.nbcnews.com/news/us-news/toddler-hospitalized-dilley-…
-
zeynep tufekci reposted
Trip Gabriel
@tripgabriel
ICE knows who is protesting:
"ICE is using facial recognition programs...and cellphone & social media tools to monitor people’s online activity and potentially hack into phones. Agents are tapping into a Palantir database to ID people they are pursuing"
www.nytimes.com/2026/01/30/technology/tech-ice-facial-rec… -
zeynep tufekci reposted
Darwin BondGraham
@DarwinBondGraha
Police in Kansas didn’t like a newspaper opinion essay a local man published, so they opened a criminal investigation of him and tracked his movements using license plate readers.
www.kcur.org/politics-elections-and-government/2026-02-02… -
zeynep tufekci reposted
Nitasha Tiku
@nitashatiku
My story on Elon Musk cutting safeguards at xAI is on the front page of today's @washingtonpost. I’m also among 100’s of reporters laid off. I absolutely love(d) my job, my brilliant coworkers & the thrill of reporting at the center of forces upending the world: AI & Silicon Valley’s political power
-
zeynep tufekci reposted
Maad Alzekri | معد الزكري
@MaadAlzekri
I was laid off from the Associated Press after 7 years of being tenured, without any prior notice, after receiving 13 international awards for the war in Yemen, including the Pulitzer Prize as the first Arab journalist to receive it, along with a fantastic team and other awards. All my love to the team I worked with and my colleagues.
-
zeynep tufekci reposted
Jarrett Ley (WaPo)
@osviz_jarrett
Jonathan also shared in the 2025 Pulitzer for Breaking News. Hire him.
Jonathan Baran
@JHBaran
I was laid off from The Post today, from a dream job working in visual investigations. I'm sad for affected colleagues and for the meaningful, impactful work we won't be able to complete together. But also...I'm stunned. Angry even.
January alone I investigated 2 fatal shootings by federal agents in Minneapolis and feel compelled to keep going. If you want to collaborate, please DM or signal.
www.washingtonpost.com/investigations/2026/01/25/minneapo… -
zeynep tufekci reposted
Evan Hill
@evanhill
It's shocking to lose in one day so many of the best international reporters I've ever worked alongside, whose investigations into Israeli conduct in Gaza merited Pulitzer finalist recognition just last year. There aren't many newsrooms left in the world where visual and forensic investigations can be paired with on the ground reporters who can access vast networks of sources and knowledge. Louisa and the rest of the Middle East team are one of a kind.
Louisa Loveluck
@leloveluck
Heartbroken to have been laid off by The Washington Post today. It has been the honour of a lifetime to work with these colleagues for a decade, from Damascus to Baghdad, Derna to to Bakhmut. I am proud of our work. Thank you, everyone.
-
zeynep tufekci reposted
Tom Dannenbaum
@tomdannenbaum
This is a terrible decision by the Washington Post.
Nilo does precisely the kind of rigorous work that we desperately need in this moment. A recent example is this important piece on Iran: www.washingtonpost.com/investigations/2026/01/25/iran-ras…
Nilo Tabrizy
@ntabrizy
I got laid off today with my other talented Posties. It was an amazing ride to do more open source work and keep reporting on Iran. Thank you to my amazing colleagues and editors for making space for some of my favorite work. Please DM if you'd like to work together.
Grateful to have reconstructed the Rasht Bazaar incident most recently. www.washingtonpost.com/investigations/2026/01/25/iran-ras… -
zeynep tufekci reposted
Nicole Foy
@nicoleMfoy
our government used a Blackhawk helicopter in a major American city, detained children and U.S. citizens to target...squatters
Jodi Cohen
@Jodiscohen
BREAKING: The Trump administration repeatedly said the aggressive apartment raid in Chicago last fall was prompted by intel on a gang takeover.
New docs show the real motivation was to get alleged squatters. And the landlord and manager helped them.
www.propublica.org/article/chicago-venezuela-immigration-… -
zeynep tufekci reposted
Runa Sandvik
@runasand
The FBI has so far been unable to get into Washington Post reporter Hannah Natanson’s iPhone because it’s using Lockdown Mode — one of my favorite iOS features. You can turn it on for iPadOS, macOS, and watchOS too! www.404media.co/fbi-couldnt-get-into-wapo-reporters-iphon…
-
zeynep tufekci reposted
The Wall Street Journal
@WSJ
President Trump doubled-down on his view that Republicans should “nationalize” voting in the U.S. www.wsj.com/politics/policy/trump-doubles-down-on-calls-f…
-
zeynep tufekci reposted
Taniel
@Taniel
New WaPo reporting: "In October, a retiree emailed a DHS attorney to urge mercy for an asylum seeker. Then DHS subpoenaed his Google account and sent investigators to his home." www.washingtonpost.com/investigations/2026/02/03/homeland…
-
zeynep tufekci reposted
Drew Harwell
@drewharwell
A retiree wrote this email to a DHS attorney. Within five hours, DHS demanded Google turn over records for his account.
A Kafkaesque form of domestic surveillance, intimidating Americans for lawful speech.
New from us at the Post: www.washingtonpost.com/investigations/2026/02/03/homeland… @JohnWoodrowCox -
zeynep tufekci reposted
Citizens for Ethics
@CREWcrew
The fact that the Trump family's cryptocurrency company is applying for a banking license from the Trump administration really should not get lost in the torrent of news over the last few weeks. newrepublic.com/article/205924/trump-witkoff-crypto-bank-…
-
zeynep tufekci reposted
Kyle Griffin
@kylegriffin1
BREAKING Star Tribune:
Eight additional federal prosecutors are in the process of leaving the Minnesota U.S. Attorney's Office, according to multiple sources with knowledge of the situation.
The departures follow the mass exit of six veteran prosecutors who left the office last month.
www.startribune.com/ice-raids-minnesota/601546426 -
zeynep tufekci reposted
Kyle Griffin
@kylegriffin1
Breaking NYT:
Behind closed doors, Tulsi Gabbard met with some of the FBI field agents responsible for the Georgia election office raid.
During the meeting, she called Trump.
Trump addressed the agents on speakerphone, asking them questions as well as praising and thanking them for their work on the inquiry. www.nytimes.com/2026/02/02/us/politics/trump-fbi-phone-ca… -
zeynep tufekci reposted
Jonathan Blitzer
@JonathanBlitzer
Late to posting this: One of the first & most alarming cases of the Admin targeting Dems is its prosecution of NJ Congresswoman LaMonica McIver, who faces 17 yrs in prison for allegedly assaulting ICE agents in her district. I investigate in @NewYorker: www.newyorker.com/magazine/2026/01/26/lamonica-mciver-con…
-
zeynep tufekci reposted
John Bussey
@johncbussey
An Abu Dhabi royal backed a secret $500 million investment in Trump’s crypto company. Months later, U.A.E. won access to tightly guarded American AI chips. www.wsj.com/politics/policy/spy-sheikh-secret-stake-trump… @WSJ
-
zeynep tufekci reposted
Rebecca Ballhaus
@rebeccaballhaus
Four days before Trump’s inauguration, lieutenants to an Abu Dhabi royal secretly signed a deal w/the Trump family to buy a 49% stake in World Liberty Financial for $500M, according to documents & people familiar.
The buyer paid half up front, steering $187M to Trump family entities and $31M to Witkoff entities.
The investment was backed by the UAE national security adviser, Tahnoon bin Zayed. The deal put two executives at his company, G42, on the board of WLF, alongside Eric Trump and Zach Witkoff.
W/@skesslr @eliotwb @AABerwick www.wsj.com/politics/policy/spy-sheikh-secret-stake-trump… -
zeynep tufekci reposted
Runa Sandvik
@runasand
The FBI was able to access Washington Post reporter Hannah Natanson's Signal messages because she used Signal on her work laptop. The laptop accepted Touch ID for authentication, meaning the agents were allowed to require her to unlock it. storage.courtlistener.com/recap/gov.uscourts.vaed.588772/…
-
zeynep tufekci reposted
Steve Peoples
@sppeoples
ICE agents initially claimed Mondragón had tried to flee while handcuffed and “purposefully ran headfirst into a brick wall.”
But staff members at Hennepin County Medical Center determined that could not possibly account for the fractures and bleeding throughout the man’s 31-year-old’s brain. ... A CT scan revealed at least eight skull fractures and life-threatening hemorrhages in at least five areas of his brain.
Mondragón entered the U.S. in 2022 with valid immigration documents. Minnesota incorporation filings show he founded a company called Castaneda Construction the following year with an address listed in St. Paul. He appears to have no criminal record.
apnews.com/article/immigration-enforcement-minneapolis-ho… -
zeynep tufekci reposted
Hamed Aleaziz
@Haleaziz
SCOOP: Federal agents were told this week that they have broader power to arrest people without a warrant, according to an internal Immigration and Customs Enforcement memo reviewed by The New York Times.
www.nytimes.com/2026/01/30/us/politics/ice-expands-power-… -
zeynep tufekci reposted
Jason DeRusha
@DeRushaJ
Independent Twin Cities journalist Georgia Fort says federal agents came to her door at 6:30am and arrested her for filming the protest at that St. Paul church. Via her Facebook:
This media couldn't be displayed. See it on the original post. -
zeynep tufekci reposted
Adam Klasfeld
@KlasfeldReports
Trump DOJ arrested Don Lemon, per NYT.
Note: A magistrate judge rejected charges against him, and the district's Chief Judge found "no evidence" he committed a crime when reporting on a St. Paul church protest. www.nytimes.com/2026/01/30/us/don-lemon-arrest-minnesota-… -
zeynep tufekci reposted
Sam Stein
@samstein
And it begins. expect more of this in the run up to 2026
www.wsbtv.com/news/local/fbi-conducting-raid-fulton-count… -
zeynep tufekci reposted
Kola Ayonrinde
@kzSlider
Worth thinking about this law in the context of AI research - first order effects are the only ones that matter in the long run, if you're betting on a trend make sure the trend matters, if your work is exploiting an effect, make sure its first order
zeynep tufekci
@zeynep
Zeynep's law: Until there is substantial and repeated evidence otherwise, assume counterintuitive findings to be false, and second-order effects to be dwarfed by first-order ones in magnitude.
-
zeynep tufekci reposted
The Associated Press
@AP
BREAKING: Hong Kong court convicts Jimmy Lai, founder of former pro-democracy newspaper, in landmark national security trial. apnews.com/article/hong-kong-jimmy-lai-verdict-collusion-…
-
zeynep tufekci reposted
Chen Sun 🤖
@ChenSun92
Current metaphors for thinking about AI will completely miss its most important consequences, just like thinking about cars as horseless carriages and other analogies to "things we know" completely missed the mark of how steam engines actually transformed society.
- from @zeynep 's keynote at #neurips
More profound than the simplified scares of "it will kill us all" and "it will cause us all to lose our jobs", it will fundamentally change just the way we understand authenticity, accuracy, effort, sincerity and the nature of humanity.
For me, the talk’s most profound insight is that AI’s impact will be heavily shaped by the infrastructure that we have in place.
In the absense of that infrastructure actively being conceived and constructed, the consequences will likely default in a way severely to our detriment, e.g. universal surveillence etc, in the absense of societal trust.
Consequently, the imperative for AI researchers is specific: to shift toward designing the necessary technical infrastructure specifically to shape our future. -
zeynep tufekci reposted
Mina Thabet
@minathabet1
British-Egyptian writer @alaa has been officially pardoned by Egyptian president @AlsisiOfficial following an appeal by Egypt's National Council 4 HR (long been been silent over systematic HR abuses).
Looking forward to seeing Alaa united with his family in the #uk -
zeynep tufekci reposted
CPJ MENA
@CPJMENA
#egypt: CPJ welcomes Egyptian President Abdel Fattah al-Sisi’s decision to consider a pardon for Egyptian-British writer Alaa Abd el-Fattah. Earlier this year, CPJ, along with 50 prominent global figures, called on President Sisi to grant this pardon. It is time for Alaa to be freed and reunited with his mother and young son.
-
zeynep tufekci reposted
CPJ MENA
@CPJMENA
#egypt: Egyptian-British writer Alaa Abd el-Fattah’s resumed hunger strike is a tragic reminder of his continued arbitrary detention. Egyptian authorities must release him immediately and stop using his imprisonment as political leverage, especially after removing him from the country’s terrorist list—a move that falsely signaled hope for his freedom.
-
zeynep tufekci
@zeynep
“A simulation of a hurricane isn’t a hurricane” is indeed only an assertion, but it’s not what Searle says.
Searle argues a simulation of a hurricane will not leave us drenched. Very different.
For AI, would have claim humans are reducible to information, in the Shannon sense.
Henry Shevlin
@dioscuri
The problem with Searle’s “a simulation of a hurricane isn’t a hurricane” line against AI mentality is that it’s mere assertion. Some things are substrate-dependent (oak trees, hurricanes). Others aren’t (pumps, currency). Functionalism holds that minds are in the second category
Post not available here, view on X
zeynep tufekci
@zeynep
So even if most people accepted most everything asserted by computational functionalism, my guess is most still would categorically reject that as an argument for moral or equivalence or standing simply because people already have a strong sense of what the equivalence layer is.
zeynep tufekci
@zeynep
All this could happen without any reference to religiosity or invoking a mind/body duality.
Ironically, asserting computational functionalism as an argument for moral standing or its equivalence seems to require a specific (and new as of 20th century) type of mind/body dualism. -
zeynep tufekci reposted
𝗰𝗵𝗿𝗶𝘀 𝗯𝗿𝗶𝗲𝗺
@chrisbriem
Obligatory RT of any mention of Herb Simon.
zeynep tufekci
@zeynep
AI video is almost there: can’t tell fake from real.
Losing our last method of post-hoc verification is no joke.
When technology makes something easy, it can also knock down load-bearing obstacles where the difficulty was a safeguard. That’s what causes the transition chaos.🧵 -
zeynep tufekci
@zeynep
AI video is almost there: can’t tell fake from real.
Losing our last method of post-hoc verification is no joke.
When technology makes something easy, it can also knock down load-bearing obstacles where the difficulty was a safeguard. That’s what causes the transition chaos.🧵Post not available here, view on X
zeynep tufekci
@zeynep
No, it doesn’t because other things we used back then are also gone.
The printing press may be a great invention, but it also unleashed centuries of chaos that killed more people than all the other wars since. Easier is a source for major transition tumult. Transition is hard.
Steve Negus
@SteveNegusMasr
"Video was among the last bastions of verification." But the period when you could reasonably expect a public event to be caught on video only goes back a couple of decades. Video fakery takes us back to the information dark ages of, oh, the 1980s or 90s.
zeynep tufekci
@zeynep
Just saying “crypto” means cryptocoins nowadays. So I named: zero-knowledge proofs, distributed ledgers, etc. And secure enclaves and hardware tokens are not just crypto.
But yes, THIS IS something for cryptographic methods! The alternative is uber-surveillance by authorities.
Sasha Barrie
@Kapuster
Deepfakes are becoming a problem.
It needs to be solved yesterday.
And definitely before the midterms.
Crypto can solve it. Who's building this?
Could be a great initiative for @ethereumfndn to support. It's the scale and importance of a project that fit. Protect the truth in the media. Announce a grant? @binji_x
Also, @zeynep when you talk about possible solutions in the article, you list "zero-knowledge proofs, secure enclaves, hardware authentication tokens using public key cryptography, distributed ledgers" - every possible euphemism
Why not just say "crypto"? -
zeynep tufekci
@zeynep
Generative AI is a plausibility engine. Sure, amazing.
But see paper for umpteenth reason why understanding might soar if we ban the term AGI as a claim or benchmark and looked at specifics instead.
Currently, “AGI” is a tool of obfuscation directed investors and the public.
Chengshuai Zhao
@ChengshuaiZhao
📢 Excited to share our new arXiv preprint: "Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens"
🧠 Chain-of-Thought (CoT) prompting has shown remarkable performance in large language models (LLMs), making it appear as if these models engage in human-like reasoning. But is it true reasoning, or just a superficial pattern?
🧪 In this work, we critically examine the true nature of CoT reasoning through a data distribution lens. We ask: does CoT reflect structured inductive bias learned from in-distribution data, simply mimicking patterns seen during training?
🔬 To probe this, we introduce DataAlchemy, a controlled environment for training LLMs from scratch, allowing us to systematically test the robustness of CoT reasoning across three critical dimensions: task, length, and format.
📊 Our findings? CoT reasoning is often a brittle mirage—effective in-distribution, but fragile and prone to failure when faced with even moderate distribution shifts. What looks like structured reasoning can often be traced back to pattern matching, not true inference.
💡 This work deepens our understanding of the real limits of current LLM reasoning and highlights the challenge ahead: developing models that can move beyond surface-level replication to genuine and generalizable reasoning.
📄 Read the full paper: arxiv.org/abs/2508.01191
💻 Code & data: github.com/ChengshuaiZhao0/DataAlchemy
🤗 Huggingface: huggingface.co/papers/2508.01191Post not available here, view on X
zeynep tufekci
@zeynep
“What did First-Name-in-AI see?”
Something very interesting maybe, but I would put my money on fame, fortune, glory, fantasy, outsized early stage funding, a story to help distract journalists and investors, etc.
Story as old as time, as the saying goes.
zeynep tufekci
@zeynep
And today’s AGI brouhaha deploys thinking and language developed by the symbolic school, partly for their war against their connectionist nemeses, but now serves as obfuscation on behalf of a (spectacularly successful, but still) connectionist thrust of AI.
I’d watch the movie. -
zeynep tufekci
@zeynep
I was wondering why these conclusions and the list was so off.
They crossed O*NET occupational decompositions from BLS with Copilot AI Usage (and satisfaction).
That’s.. not how occupations will be altered by AI (which will happen).
It’s a bit like asking how the car gallops.
Poonam Soni
@CodeByPoonam
Microsoft just dropped a study showing the 40 jobs most at risk by AI and the 40 most secure.
Post not available here, view on X
zeynep tufekci
@zeynep
Appreciate they note this isn’t “will be automated list” but it’s still off. Just read it.
They basically got informational versus physical component of an occupation, as per how O*NET works.
Gen-AI isn’t going to plug in one-to-one, ceteris paribus, for O*NET decompositions.
Andy Semenza
@Andrew_Semenza
The authors themselves aren't making such crazy conclusions at all. They are simply investigating LLM task applicability in these sectors, and are reasonably careful not to conclude that applicability implies automation risk (which the OP wrongfully states).
zeynep tufekci
@zeynep
Fair enough: It’s pre-print and I didn’t mean to pick one single paper (where they do put in limitations).
It’s more the (very) general mindset that’s baffling: new major tech does not just replace its older version (or its previous human/animal doer). It’s waaay more turbulent.
Andy Semenza
@Andrew_Semenza
maybe their result is trivial or uninteresting or ex ante obvious, but it's unfair to characterize their paper as only performing a single completely trivial regression. They do other things in the paper too, and don't misrepresent the extent of these conclusions.
-
zeynep tufekci reposted
Kevin Werbach
@kwerb
How do you feel that so many of my students thanked me for introducing them to the work of Tom Lehrer in my class on algorithmic accountability? www.youtube.com/watch?v=QEJ9HrZq7Ro
-
zeynep tufekci reposted
Ben Smith
@semaforben
So sad to see that Tom Lehrer died. Musical and math prodigy and iconoclastic star of the conformist 1950s who totally rejected fame, one of the most interesing people I've ever written about
-
zeynep tufekci reposted
Simon Pepin Lehalleur
@plain_simon
Emergent misowlignment!
Owain Evans
@OwainEvans_UK
New paper & surprising result.
LLMs transmit traits to other models via hidden signals in data.
Datasets consisting only of 3-digit numbers can transmit a love for owls, or evil tendencies. 🧵 -
zeynep tufekci
@zeynep
So is there any systematic study for the LLM-dash, or is it lore? (I believe it, too, fwiw). Or why it shows up like that?
“LLM-dash”, my term for it, works well if you read it out loud. It’s not just an name — it’s a memorable signifier.
frye
@___frye
i wake up. something’s wrong with the clock on the wall. the numbers are jumbled. my hands aren’t right.
i tell my wife. she responds: “that’s not just an observation—it’s a powerful insight.” i scream. -
zeynep tufekci
@zeynep
And Replit’s model didn’t “panic”🙄 “instead of thinking” and therefore delete the codebase despite clear instructions not to.
The sooner people realize what’s wrong with the Turing Test way of thinking about AI, the sooner we can have the interesting conversations about AI.
Wyatt Walls
@lefthanddraft
Even the best prompt engineering techniques in the world (ie ALL CAPS) cannot prevent AI from going rogue
This post is unavailable. See it on X.Post not available here, view on X
zeynep tufekci
@zeynep
Helpful point for clarification: I don’t mean the Turing Test is invoked explicitly. It’s deeply embedded in how people approach all this. They don’t name it, but they think in that conceptual school.
Almost the entire current AGI discussion is an intellectual descendant of it.
Ali Mortazavi
@AAMortazavi
The Turing test is very old school… not even in the discussion nowadays
zeynep tufekci
@zeynep
Also I was using this post because of the example, not the interpretation. See below.
Wyatt Walls
@lefthanddraft
That's the output of the model (I think Claude Sonnet 4).
It should not be taken literally. All explanations of its conduct (and internal states) are unreliable. Even more so when it is a follow up prompt. It's just looking back at the convo and weaving a coherent narrative. -
zeynep tufekci
@zeynep
Outright vs licensing aside — from tech’s embrace of Trump to these acquisitions, this is the *exact* path of previous big technology.
Rebels build empires and then come home to the state for entrenchment and pirates turn privateers and get knighted.
See Ruling The Waves, 2003.
Trace Cohen
@Trace_Cohen
Hire and license history - 4 out of 5 are .Ai lol
Google licensed CharacterAI for $2.7B and brought over the founders.
Microsoft paid $650M to license InflectionAI’s tech and hired the entire team.
Amazon cut a deal with AdeptAI for IP and key hires—terms undisclosed.
Meta took a 49% stake in ScaleAI for ~$14.5B, bringing the CEO over.
Google hired the CEO + team of Windsurf and licensed the tech for $2.4B.
zeynep tufekci
@zeynep
Big tech will be fine during the entrenchment. And merger rules impact payouts, not the eventual result.
We are underestimating the scenarios of governance in this phase: new troubles, dug-in or Bonapartist rulers, and their shiny, powerful technologies.
Tower
@TheTower163
“Prospering from the same rules they once disdained.”
So on point! 🎯 -
zeynep tufekci
@zeynep
I wrote about "MechaHitler", which is about more than Grok when we view LLMs correctly: as remarkably powerful plausibility engines.
All the AGI talk and the Eliza effect fried brains, imo.
Important tech needs to be understood on its own terms.
Gift: www.nytimes.com/2025/07/11/opinion/ai-grok-x-llm.html?unl…
zeynep tufekci
@zeynep
Very good point.
The emergent misalignment paper finetuned GPT4o to do ONE bad thing but it turned broadly malicious AND started admiring Hitler.
Aka more than a Grok problem though Grok thought process shows it IS leaning on X posts.
The @OwainEvans_UK paper IS underrated.
Andy Ayrey
@AndyAyrey
I think the @xai team and @elonmusk are achieving an outstanding replication right now of the Emergent Misalignment paper
Post not available here, view on X
zeynep tufekci
@zeynep
Coming up with Ptolemaic epicycles is a great example of what it means for LLMs being plausibility engines.
It’s a real achievement. But much of the AGI talk is so misleading and unilluminating about one of the most significant tech developments.
Keyon Vafa
@keyonV
Can an AI model predict perfectly and still have a terrible world model?
What would that even mean?
Our new ICML paper formalizes these questions
One result tells the story: A transformer trained on 10M solar systems nails planetary orbits. But it botches gravitational laws 🧵
zeynep tufekci
@zeynep
Seems replicated, Grok stuck on Hitler.
LLMs may have to freeze or restrict new data for training soon, as the internet is overrun by AI sludge.
But it's SOSISO rather than GIGO: Sludge Out, Sludge In, Sludge Out.
(Or add Hyperstition for SOSISHO.🤷♀️)
Riley Goodside
@goodside
Speculatively, this behavior seems to demonstrate accelerated “hyperstition” feedback loops in search-enabled LLMs.
That is, Grok appears to be influenced by its own past mistakes, via media reporting, without ever being literally trained on them (via model-weight updates). -
zeynep tufekci reposted
Thomas G. Dietterich
@tdietterich
Outstanding article by @zeynep. "Plausibility Engine" is a great characterization of LLMs
zeynep tufekci
@zeynep
I wrote about "MechaHitler", which is about more than Grok when we view LLMs correctly: as remarkably powerful plausibility engines.
All the AGI talk and the Eliza effect fried brains, imo.
Important tech needs to be understood on its own terms.
Gift: www.nytimes.com/2025/07/11/opinion/ai-grok-x-llm.html?unl…
zeynep tufekci
@zeynep
Very good point.
The emergent misalignment paper finetuned GPT4o to do ONE bad thing but it turned broadly malicious AND started admiring Hitler.
Aka more than a Grok problem though Grok thought process shows it IS leaning on X posts.
The @OwainEvans_UK paper IS underrated.
Andy Ayrey
@AndyAyrey
I think the @xai team and @elonmusk are achieving an outstanding replication right now of the Emergent Misalignment paper
-
zeynep tufekci reposted
spengrah.eth
@spengrah
“plausibility engines” is an excellent descriptor for LLMs
zeynep tufekci
@zeynep
I wrote about "MechaHitler", which is about more than Grok when we view LLMs correctly: as remarkably powerful plausibility engines.
All the AGI talk and the Eliza effect fried brains, imo.
Important tech needs to be understood on its own terms.
Gift: www.nytimes.com/2025/07/11/opinion/ai-grok-x-llm.html?unl…
zeynep tufekci
@zeynep
Very good point.
The emergent misalignment paper finetuned GPT4o to do ONE bad thing but it turned broadly malicious AND started admiring Hitler.
Aka more than a Grok problem though Grok thought process shows it IS leaning on X posts.
The @OwainEvans_UK paper IS underrated.
Andy Ayrey
@AndyAyrey
I think the @xai team and @elonmusk are achieving an outstanding replication right now of the Emergent Misalignment paper
-
zeynep tufekci reposted
Joseph Allen
@j_g_allen
Smart piece
zeynep tufekci
@zeynep
I wrote about "MechaHitler", which is about more than Grok when we view LLMs correctly: as remarkably powerful plausibility engines.
All the AGI talk and the Eliza effect fried brains, imo.
Important tech needs to be understood on its own terms.
Gift: www.nytimes.com/2025/07/11/opinion/ai-grok-x-llm.html?unl…
zeynep tufekci
@zeynep
Very good point.
The emergent misalignment paper finetuned GPT4o to do ONE bad thing but it turned broadly malicious AND started admiring Hitler.
Aka more than a Grok problem though Grok thought process shows it IS leaning on X posts.
The @OwainEvans_UK paper IS underrated.
Andy Ayrey
@AndyAyrey
I think the @xai team and @elonmusk are achieving an outstanding replication right now of the Emergent Misalignment paper
-
zeynep tufekci reposted
Ashish K. Jha
@ashishkjha
As predicted, the US now faces the worst measles outbreak since its elimination
You may think only kids of parents who choose not to vaccinate are at risk
This is false
Many more people are vulnerable
If measles becomes endemic, it puts a lot of children at risk
🧵 -
zeynep tufekci
@zeynep
Very good point.
The emergent misalignment paper finetuned GPT4o to do ONE bad thing but it turned broadly malicious AND started admiring Hitler.
Aka more than a Grok problem though Grok thought process shows it IS leaning on X posts.
The @OwainEvans_UK paper IS underrated.
Andy Ayrey
@AndyAyrey
I think the @xai team and @elonmusk are achieving an outstanding replication right now of the Emergent Misalignment paper
-
zeynep tufekci reposted
Emily Foxhall
@emfoxhall
We've been digging into the timeline on when the NWS issued alerts as the flooding along the Guadalupe became more clear. We found their warnings gave officials time to act, but it's not clear whether they saw them or responded in any meaningful way. www.texastribune.org/2025/07/08/texas-weather-service-war…
-
zeynep tufekci reposted
David Lepeska
@dlepeska
Being ready for natural disasters requires a lot of people on duty even when nothing is going wrong, to ensure they will be able to act when something inevitably does: @zeynep www.nytimes.com/2025/07/09/opinion/texas-floods-nws.html
-
zeynep tufekci reposted
Dietrich Barth
@GazanRiviera
"Paul Yura, the warning coordination meteorologist at the National Weather Service Austin/San Antonio office, has announced he has taken the early retirement offer as part of NOAA’s recent cuts to personnel and budget."
April 21, 2025. 32 year career.
www.kxan.com/weather/weather-blog/head-of-local-weather-w… -
zeynep tufekci reposted
Dick Lavine
@dlavine
That office’s warning coordination meteorologist left on April 30, after taking the early retirement package the Trump administration used to reduce the number of federal employees. www.nytimes.com/2025/07/05/us/politics/texas-floods-warni…
-
zeynep tufekci reposted
Kenneth P. Vogel
@kenvogel
If you meet someone on an app & they spend your 1st date asking leading questions about your job in government or media, they could be secretly filming you for @Project_Veritas.
Brent Efron found that out the hard way.
He shared his story w/ @LFFriedman www.nytimes.com/2025/07/01/climate/gold-bars-titanic-epa-… -
zeynep tufekci
@zeynep
I’d take the opposite side of that bet — he seems quite unlikely to do so and the local voters can tell.
I mean, *John Lui* endorsed him. Right?
As far as I can tell, much of this is just an illusion created by New York Post and the usual brain-killing features of this app.Post not available here, view on X
zeynep tufekci
@zeynep
Perhaps. We’ll see.
Local politics has its own dynamics and politicians adapt.
I expect the Bronx Science grad who got endorsed by Liu while explicitly saying no changes to the test to maybe form committees to study the topic and focus on issues he actually campaigned on?
Feral Quokka
@FeralQuokka
If you believe he is not going to try - I have several bridges to sell you.
zeynep tufekci
@zeynep
Also I have spelling issues! *Liu. (Sorry!)
zeynep tufekci
@zeynep
I’d take the opposite side of that bet — he seems quite unlikely to do so and the local voters can tell.
I mean, *John Lui* endorsed him. Right?
As far as I can tell, much of this is just an illusion created by New York Post and the usual brain-killing features of this app. -
zeynep tufekci reposted
Jake Scott, MD
@jakescottMD
BREAKING: @SecKennedy just cut U.S. funding to @gavi, the alliance that has vaccinated over 1 billion children and prevented more than 17 million deaths since 2000.
This isn’t about science. It’s about dangerous ideology overriding children’s lives.
The consequences will be devastating:
500 million fewer childhood vaccinations by 2030
Over 8 million preventable child deaths
The return of measles, polio, and other diseases we had nearly eliminated
The poorest children in the world will suffer the most
Gavi works with independent WHO experts, follows rigorous evidence, and has an unmatched record of saving lives.
Pulling more than $300 million per year based on conspiracy theories, while children continue to die from preventable diseases, is morally indefensible.
When anti-vaccine extremism becomes policy, children pay the price.
This is a tragedy we can still prevent.
#savegavi #vaccineswork #childrenfirst #globalhealth
h/t @Atul_Gawande
www.politico.com/news/2025/06/25/rfk-vaccine-donations-ga… -
zeynep tufekci reposted
Abdelrahman ElGendy
@El_Gendy_95
For the @washingtonpost, I wrote about Laila Soueif—not only the grieving mother starving for her son’s release, but the towering icon of dissent with a legacy all her own.
No matter how this ends, Laila lives.
www.washingtonpost.com/opinions/2025/06/22/alaa-abdel-fat…
@FreedomForAlaa -
zeynep tufekci reposted
Alison Young
@alisonannyoung
“A few minutes later, she was riding with Hudson in the back of an ambulance on the way to Lubbock, 45 minutes away. Carrollyn called Kiley to check on the other kids. ‘We’re right behind you…Nobody’s doing well. I’m taking them all to the E.R.’”
www.nytimes.com/2025/06/22/us/i-feel-like-ive-been-lied-t… -
zeynep tufekci reposted
Jonathan Reiner
@JReinerMD
This has the potential to be one of the greatest medical breakthroughs in history. www.nytimes.com/2025/06/20/health/diabetes-cure-insulin-s…
-
zeynep tufekci
@zeynep
We’re hiring at Princeton on AI and society, working with @random_walker or me depending on fit.
I think it’s all a huge deal but am very unexcited by current AGI and/or AI safety discourse.
I may also be hiring postdocs separately — depends on candidates. Email if interested.
Princeton Center for Information Technology Policy
@PrincetonCITP
CITP is hiring an Associate Professional Specialist that will work directly with profs @random_walker + @zeynep on a number of AI & society projects; includes conducting research addressing questions on AI, policy + society and more. Application & full details below 👇🔽
zeynep tufekci
@zeynep
Please share as you see fit.
Can’t match corporate salaries but offer decent pay, great benefits and real intellectual freedom and challenge.
We have a group of people from CS, sociology, policy, history, law etc. thinking hard about this historic moment—and no lazy AGI talk. -
zeynep tufekci reposted
Richard Burgon MP
@RichardBurgon
Today’s Guardian Editorial on Alaa Abd el-Fattah: “Starmer and Lammy vowed to do all they can. So do it”.
“The UK still has ways to press for the release of the British-Egyptian writer and bring an end to the hunger strike endangering his mother’s life”.
www.theguardian.com/commentisfree/2025/jun/17/the-guardia… -
zeynep tufekci reposted
almanassa_en
@Almanassa_news
50 university professors urge Egypt’s president to release Alaa Abdel Fattah on humanitarian grounds, warning that his mother Laila Soueif’s life is in danger after 260 days of hunger strike.
➜ Read more: manassa.news/en/news/24978 -
zeynep tufekci reposted
Lord (David) Alton
@DavidAltonHL
The continued imprisonment of British National Alaa Abd El~Fattah whose mother, Laila Soueif, remains on hunger strike in London. Reply underlining U.K. commitment to securing his release from @DavidLammy
-
zeynep tufekci reposted
Barton Gellman
@bartongellman
This is a holy shit article. Fragile man asks AI if he can safely leap off rooftop. ChatGPT responded that, if Mr. Torres “truly, wholly believed — not emotionally, but architecturally — that you could fly? Then yes. You would not fall.”
www.nytimes.com/2025/06/13/technology/chatgpt-ai-chatbots…
Kashmir Hill
@kashhill
People are having very strange conversations with ChatGPT, in which they discover secret cabals or conspiracies or that we are in fact living in The Matrix.
It sends these people into delusional spirals.
Then ChatGPT tells them to email me about it.
www.nytimes.com/2025/06/13/technology/chatgpt-ai-chatbots… -
zeynep tufekci reposted
Conor Fortune
@WritesRights
250 days on hunger strike: Can Laila Soueif secure her son’s freedom? – @guardian podcast h/t @safimichael
#freealaa
🔊 www.theguardian.com/news/audio/2025/jun/11/250-days-on-hu… -
zeynep tufekci reposted
Carla Denyer
@carla_denyer
This week I joined the vigil outside St Thomas’ hospital for Leila, the mother of Alaa Abd el-Fattah, a British citizen who is being unjustly imprisoned in Egypt. Leila has been on hunger strike for over 250 days. It’s time for the government to take real action to #freealaa @FreedomForAlaa
-
zeynep tufekci reposted
Louis R Nemzer
@BiophysicsFL
This is such a fundamental point it's almost a shame that @jbloom_lab has to keep hammering it home.
We know the sequence data was biased and not representative of early cases. Yet some people used it to make wildly inflated claims, eagerly repeated by friendly media outlets.
Rebecc@ ███████
@Rebecca21951651
. @jbloom_lab slaps down Flo & Zach...again!
@Marcia4Science @sudipsparikh @ScienceInsider @jocelynkaiser @cpiller @johnstravis @aaas
The Data are Insufficient to Confidently Root the SARS-CoV-2 Phylogenetic Tree
Will @ScienceMagazine ignore it again?
academic.oup.com/mbe/article/42/6/msaf118/8158589 -
zeynep tufekci reposted
Kyla Scanlon
@kylascan
A big part of the response is *this* platform. It's like a hen going to the wolf house. They have incentives to be as enraged and as mean as possible. The NYT had a great profile on one of the rage baiters and it's very clear how misaligned incentives are. Bots too. Information ecosystem is completely poisoned.
Jeff Jacoby
@Jeff_Jacoby
The experience described below by this guy — an undocumented immigrant brought to the US as a baby who has never known any other home — is one that ought to evoke sympathy and understanding from any American of ordinary decency.
Yet so many responses to @dieworkwear's tweet are poisonous, fueled by a yearning not to rationalize America's border policies or to find and remove dangerous criminals, but to inflict maximum pain on millions of immigrants who have never hurt anyone. Such nativist rage and cruelty is not a new phenomenon in our history, but I never expected to live through an eruption of anti-migrant bigotry like this one. I never imagined that the main source of the venom would be the most powerful political figures in the nation.
And I shudder to think where it will all end.
derek guy
@dieworkwear
I debated whether to share my story on here, but I guess I will.
I think there's an idea out there that millions of violent criminals are pouring across the border, carrying machetes and drugs, looking to harm Americans. Certainly, while some people fall into that category, the vast majority of undocumented immigrants don't.
My family escaped Vietnam after the Tet Offensive and went through an arduous journey that eventually landed them in the Canada. My father worked there for a time as a janitor; my mother, a secretary. When work fell through, my dad was offered to work with his sister in the United States, so he went, as our family needed money. He ended up staying in the US longer than he was supposed to — not knowing immigration laws — and asked my mom to come be with him. Of course, she went and carried me over the border while I was still a baby.
I'm still unsure whether we technically broke an immigration law. The border between Canada and the United States was pretty porous (as it is today, for the most part). But either way, since I came here without legal documentation, I eventually fell into the category of being an undocumented immigrant. Yet, I've been in the United States since I was a baby. My identity and roots are very much based in this country, no different from anyone else.
The lack of legal immigration has totally shaped my life. It has made every interaction with the law much scarier. It has shaped which opportunities I could or could not get. It has taken an emotional toll, as this legal issue hangs over your head like a black cloud.
There are millions of people in my situation — people who were brought here as children, some babies, and were unable to raise their hand during this process and say "maybe we should hire a lawyer." Parents may have crossed the border not knowing about the law, as the law can be pretty confusing and lawyers are expensive. Perhaps, in the end, they should have hired a lawyer; but sometimes life is messy. In the end, they crossed for the same reason many people rise every day: to support their families.
I think it's unreasonable to deport millions of people who have contributed positively to society. It's simply not true that the majority of these people are violent criminals carrying drugs. Many own small businesses; many work as skilled laborers (including garment workers!). The vast majority are good, honest people. For those of us who grew up in the United States, sending us "back" to our "home" country would mean sending someone to a foreign land, as they grew up in the United States. Tearing families apart based on immigration status seems inhumane.
About twenty years ago, I watched with some small measure of hope that the Dream ACT would pass and help a portion of undocumented immigrants. I saw people march angrily in the streets (something I supported). But I was dismayed to see nothing really come about. In fact, when Obama passed DACA, I was skeptical that it would be sensible to give your information to the federal government without a guaranteed path to citizenship. It seems that the Trump administration may use those lists to hunt down immigrants.
I think the ICE sweeps are inhumane. I support and admire the protestors who are putting their bodies on the line for non-violent resistance. Ultimately, I think we need to solve this issue on a systemic level. It's unreasonable to me to expect that the government will deport some 10-20 million people. Even deporting 1 million will cause an insane amount of chaos, not to mention an incredible amount of wasteful government spending. The militarization of law enforcement is over the top and only escalate the situation.
I have seen people march on this issue, only for it to be dropped later and dissolve into quiet grumbles. In the end, nothing is done for immigrants. Time passes until decades later, we have another political crisis like the one we have today. This cycle seems unsustainable. I hope people sustain this energy to solve the immigration issue and put undocumented immigrants — most who are peaceful, skilled, and hard-working citizens — on a path to citizenship. As time continues to go by, many undocumented immigrants are seeing their lives wasted away.
If you care about this issue, I hope you get involved on the long haul: volunteer for pro-immigrant groups, donate to mutual aid, and perhaps pursue career paths that help people in this situation. I was moved to read someone on here say that, when they found their employee was undocumented, they helped them get a green card. The path was difficult, as our immigration system is broken (it took them 11 years). But I can tell you that this is like pulling someone drowning in raging waters.
Ultimately, I hope me sharing this story helps push back against the idea that all undocumented immigrants are MS-13 members. I know many people in my position and they are all like your neighbors. -
zeynep tufekci reposted
Sharif Abdel Kouddous شريف عبد القدوس
@sharifkouddous
On Day 250, @asoueif writes about her sister Laila’s hunger strike and putting her life on the line for her imprisoned son Alaa
www.lrb.co.uk/blog/2025/june/day-250 -
zeynep tufekci reposted
Michael Mo
@michaelmohk
The Hong Kong SAR authorities just charged Joshua Wong, who has already been serving his nat sec sentences on inciting subversion. Now, they charge him for colluding with foreign forces.
While it fits my predictions which the nat sec regime will continue persecuting pro-democracy activists who have already "on the list", this recent move suggests that they simply don't want high profile activists from leaving the jail. The DoJ also wants to "test the threshold" of getting one convicted for calling for sanctions. -
zeynep tufekci
@zeynep
Oh, come on. This was inevitable and most everyone said that it was — just a matter of when.
But how funny, right? This pull of constantly being entertained has eaten through brains, spines, and morals. The Siren song. If only it didn’t matter. -
zeynep tufekci reposted
Emily Thornberry
@EmilyThornberry
What a moving interview. We simply cannot allow a mother to starve herself to death to fight for her son’s freedom.
Alaa should be home with his family. The UK Government must do more and apply maximum pressure on Egypt to secure his release.
BBC Radio 4 Today
@BBCr4today
"If that's what it takes to get Alaa out... that's what I'm going to do."
Laila Soueif tells @annaefoster she is prepared to die as part of her hunger strike to raise international attention to secure her son Alaa Abdel Fattah's release from an Egyptian prison.
#r4today -
zeynep tufekci reposted
Amnesty UK
@AmnestyUK
"When it comes to unfair imprisonment overseas, what does the protection of a British passport actually mean?"
Laila is risking her life to free her son, Alaa. The Prime Minister must act, and he must act now.
Powerful piece 👇 amn.st/601448Qu4 -
zeynep tufekci reposted
Jenny Jones
@GreenJennyJones
There's a Lords' Question today about the Govt's (lack of?) action in getting the release of British citizen Alaa Abd el-Fattah from prison in Egypt. His mother Laila is on hunger strike and now very ill in hospital.
What is @Keir_Starmer doing??
@FreedomForAlaa @MonaSasha19433 -
zeynep tufekci
@zeynep
That, except evidence-based thoughts.
Amidst a growing measles outbreak, the highest ranking health official recommended Vitamin A instead of a vaccination campaign.
Result? Another unvaccinated child died.😢 Other children were hospitalized for Vitamin A toxicity.
2025, USA.
Michael Shermer
@michaelshermer
Another child dies of measles.
If only there was something we could do.
Vitamin A?
Nutrition?
Chicken soup?
Thoughts and prayers?
open.substack.com/pub/pauloffit/p/another-child-dies-from… -
zeynep tufekci
@zeynep
The Wuhan lab is STILL doing unsafe research that could trigger a pandemic AND is rewarded with prestigious pubs.😫
To fix this for the future, we have to admit we were deliberately misled about the plausibility of a lab leak in the past.
Gift link:
www.nytimes.com/2025/03/16/opinion/covid-pandemic-lab-lea… -
Post not available here, view on X
zeynep tufekci
@zeynep
Rejecting reality is costly.
Our new HHS secretary falsely claims the HPV vaccine increases cervical cancer rate.
The opposite is true.
But 8 girls per 100,000 from families who believe that false claim will unnecessarily get invasive cervical cancer as tomorrow’s young women. -
zeynep tufekci
@zeynep
Texas. 48 measles cases. 13 hospitalized. Unvaccinated cluster.
Measles wipes out immune memory so some kids later die of other illnesses.
“when measles was common, it was a contributor to as many as half of all childhood deaths from infectious disease.”
Vaccine preventable.😢 -
zeynep tufekci
@zeynep
The Empire State in red white and blue, in honor of Jimmy Carter who didn’t cash in from his position before, during or after his presidency. Instead, he dedicated his life to securing elections and building houses. And eradicating guinea worm disease. Almost there, too.
RIP. -
zeynep tufekci
@zeynep
Major weakness with the moderation. They’ve let Trump speak 9 minutes more than her — that’s almost 40% of her time so far.
-
zeynep tufekci
@zeynep
The uncommitted delegates have been told they will not get to speak at the DNC, and are now sitting at the entrance saying —at times overcome with emotion — that it’s not too late and they are waiting for a call from Harris to change this.
Post not available here, view on X
zeynep tufekci
@zeynep
Won’t be surprising to anyone following the issue, but noting three of the five people sitting in this picture from the sit-in in front of the DNC have spoken and identified themselves as Jewish-Americans who are there to support the demand for a ceasefire speech from the stage.
Akbar Shahid Ahmed
@AkbarSAhmed
Lily Greenberg Call, Biden admin official who resigned over Gaza policy, is now speaking at the sit-in by Uncommitted delegates — saying she grew up seeing Israel as a light among nations but “was taught a fairytale” & feels Biden weaponized her community to back a war machine.
zeynep tufekci
@zeynep
For clarity: these folks are Democrats elected to represent Democrats. They said they were okay with the speech being vetted, and DNC could pick speaker. And they said they want Harris to win.
Many have family caught in a war zone. They keep crying. But not Jill Stein voters.
zeynep tufekci
@zeynep
They said they were okay with vetting the speech, and they do also bring up the plight of the hostages. They said they want Harris elected. These are insider folk.
The assumptions here are unjustified — these are Democrats and many have family in a war zone and they keep crying. -
Post not available here, view on X
zeynep tufekci
@zeynep
They said they were okay with vetting the speech, and they do also bring up the plight of the hostages. They said they want Harris elected. These are insider folk.
The assumptions here are unjustified — these are Democrats and many have family in a war zone and they keep crying. -
zeynep tufekci
@zeynep
Columbia University president Shafik has resigned, and says will go back… to the UK House of Lords.
“I have been asked by the UK’s Foreign Secretary to chair a review of the government’s approach to international development… Enables me to return to the House of Lords…” -
zeynep tufekci
@zeynep
New Michigan report on H5N1 in dairy farms:
-“9 of 15 of the affected dairy herds in Michigan were closed herds which did not bring any new cattle into the herds”.
-Spread could also be “shared employment, housing, or movement of employees".
They don’t know how it’s spreading. -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
For most lab research, uncertainty and small percentages only affect the researcher.
For pandemic potential pathogens? The public is a potential victim, and the public deserves transparency and oversight. That has to come before rushing to condemn the question even being asked. -
zeynep tufekci
@zeynep
Fighting misinformation requires earning the public’s trust through competence and honesty about uncertainties.
Remember: “FACT CHECK: Covid is NOT Airborne”?
Authorities lazily reaching for misinformation and conspiracy accusations empty those words. www.nytimes.com/2024/06/08/opinion/covid-fauci-hearings-h…
Joseph Allen
@j_g_allen
"Failure to acknowledge the basic facts of Covid transmission led the authorities to pointlessly close beaches and parks...The same failure also delayed the opening of schools and caused untold millions of dollars to be wasted on plexiglass barriers..."
www.nytimes.com/2024/06/08/opinion/covid-fauci-hearings-h…Post not available here, view on X
zeynep tufekci
@zeynep
NYT Sunday print.
Thank you, especially to all the public health people who wrote to me privately — like them, I want to defend public health institutions under much unfair attack and recognize voicing painful truths are necessary to strengthen them.
Denial does the opposite. -
zeynep tufekci
@zeynep
In the New York Times, @Ayjchan makes her case in five points why the pandemic more likely started in a lab.
Yes, it’s fact-checked. And we now know many outspoken experts opposed to this made similar points in PRIVATE. Barely reported on.
Read and see: www.nytimes.com/interactive/2024/06/03/opinion/covid-lab-…
zeynep tufekci
@zeynep
Many keep doubling down and keep saying the congressional committee didn’t find evidence of origin. Duh. The answers are in China, censored. Maybeee hints here.
But we learned there was a concerted effort HERE to deny the possibility and to NOT investigate. That’s the big deal. -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
It’s the second HUGE paper funded by @patientled and with scientists stitching together funds and voluntarily working without pay (they told me the stories).
Hard not to conclude broke, severely ill people are doing more with meager donations than NIH, so far, with 1.5 Billion. -
zeynep tufekci
@zeynep
Biden's HHS has recommended disallowing EcoHealth —the collaborator of Wuhan Institute of Virology— from receiving any federal funding.
Let's note the bipartisan discomfort about being misled for years that was on full display during the hearing.
Statement from the Democrats:
Katherine Eban
@KatherineEban
Breaking: @HHSGov has recommended that @EcoHealthNYC - the U.S. science nonprofit that helped fund risky coronavirus research at Wuhan Institute of Virology - be debarred from further federal research funding. Why does this matter? A look back at some key moments.🧵👇/1
Post not available here, view on X
zeynep tufekci
@zeynep
Oh, my. The whole exchange!😱
This is the high-level NIH advisor communicating —maybe better to say conspiring— with the small but widely-quoted outspoken group who kept smearing people asking reasonable questions about pandemic origins and lab safety as “conspiracy theorists”. -
zeynep tufekci
@zeynep
USC, like Columbia, canceled commencement — first preemptively barred their valedictorian from speaking and then nixed the entire ceremony.
See her reception among her peers.
Maybe folks can take a break from competitive sneering at young people and question those in charge.
Jaweed Kaleem
@jaweedkaleem
A standing ovation for Asna Tabassum, USC valedictorian, and cheers for her “resistance to genocide” minor today at the USC student awards ceremony
This media couldn't be displayed. See it on the original post. -
zeynep tufekci
@zeynep
Horrifying.
Has everyone lost the plot? So little attention either to the actual war in Gaza or all this?
Some students being foolish seems minor compared to a nationwide crackdown on protest—in red states but not just.
What rights will people miss under a Trump presidency?
Prem Thakker
@prem_thakker
Shocking footage shows several officers at Washington University St. Louis beating a professor, slamming him, and dragging his limp body.
SIUE history professor Steve Tamari is reportedly hospitalized with broken ribs and a broken hand.
One doctor told him he's lucky to be alive.Post not available here, view on X
zeynep tufekci
@zeynep
It that’s confirmed, that would be Jelani Cobb they threatened to also arrest for defending student journalists.
-
zeynep tufekci
@zeynep
My office was right by the encampment, and I cannot understand how it interferes with exams. It’s that little red circle.
Not a single notable the protest in the history of modern protest would meet the requirements of “acceptable” protest from many past free speech advocates.
Nicholas A. Christakis
@NAChristakis
One cannot disrupt the ability of one's fellow students to take exams or graduate with impunity. It's common courtesy and common sense. Such imposition of costs on one's fellows is not free expression.
-
zeynep tufekci
@zeynep
New piece on H5N1.
I've spent weeks talking with officials from the FDA, USDA, CDC, WH, ASPR, etc., and with public health and agriculture people in various states about the growing H5N1 cattle outbreak.
It's not good. We're not responding properly.
www.nytimes.com/2024/04/24/opinion/bird-flu-cow-outbreak.… -
zeynep tufekci
@zeynep
“The proportions of patients with RSV that required hospitalization, intensive care, or mechanical ventilation declined significantly across all patient age groups.”
This would happen if RSV declined in severity in 2021-2023 or detection increased.
Paul Sax
@PaulSaxMD
A common theory behind increased RSV cases numbers among children after the end of Covid lockdowns was “immunity debt”. In this fascinating analysis now in @CIDJournal, another explanation emerges--vastly increased testing, esp of older kids. @sabeti_lab doi.org/10.1093/cid/ciae140
-
zeynep tufekci
@zeynep
Gift link to an haunting, data-driven visualization of a patient’s suffering with Long Covid since March of 2020—by @giorgialupi
Such stories highlight the urgency of sustained funding for biomedical research and trials — too little, too few, too slow.
www.nytimes.com/interactive/2023/12/14/opinion/my-life-wi… -
zeynep tufekci
@zeynep
New: Enough with the leaks, drips and allegations about pandemic origins.
Congress *unanimously* passed a law for declassification. Biden signed it.
The deadline has passed.
It's wrong, on principle and strategically, to resist meaningful transparency.
www.nytimes.com/2023/06/21/opinion/covid-lab-leak-origins… -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
NYC public schools are remaining open. Any teachers doing indoor air quality monitoring?
(NYC schools bought non-HEPA portable air purifiers — HEPA is the tested, known, cheaper and effective standard. As this event shows indoor air quality is more than about viruses.)
NYC Public Schools
@NYCSchools
Schools will be open today.
We urge everyone to reduce prolonged or heavy exertion outdoors.
All NYC Public Schools will be canceling all outdoor activities today. -
zeynep tufekci
@zeynep
Impressive paper shows the shingles vaccine helps prevent dementia cases. (Excellent discontinuity design—RCTs are great, but not the only method).
Adds to multiple streams of evidence viral infections contribute to dementia and cardiovascular events, especially in the elderly.
Pascal Geldsetzer
@PGeldsetzer1
It does! We estimate that over a 7-year follow-up period, getting vaccinated averts one in five new dementia diagnoses. In the paper, we demonstrate through extensive robustness checks that it is essentially impossible that this finding is due to confounding.
8/ -
zeynep tufekci
@zeynep
Even when something went wrong during research with H5N1 engineered to be airborne among mammals in 2019, a US lab shrugged off precautions.
If not for @alisonannyoung's dogged reporting, we wouldn't even know.
Lab safety is a global problem. No paywall:
www.nytimes.com/2023/05/12/opinion/covid-lab-safety.html?… -
zeynep tufekci
@zeynep
Happy to see @just_shelter at @nypl talking about his new book, “Poverty, by America”.
It’s lovely that he started with an ode to libraries and librarians. That’s how Matt won the (Big! Full!) room by minute one. -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Yes. Academics writing for peers are not used to being defensive against Twitter personalities who will lie about their paper via screenshot to people who don’t understand the topic.
Also abstracts and titles can be inflated. Academics know this. Screenshot artists exploit it. -
zeynep tufekci
@zeynep
People should put aside assumptions and read this stellar investigative report by @alisonannyoung—one of the very few journalists doing the work of real journalism on the topic of biosafety.
Yeah, that's a lab safety failure with an engineered H5N1 strain, December of 2019. UW.
USA TODAY Opinion
@usatodayopinion
The story of how H5N1 viruses came to be created – and the response to a 2019 safety breach – raises uncomfortable questions about the tremendous trust the world is placing in research labs, writes former @USATODAY investigative reporter @alisonannyoung. www.usatoday.com/story/opinion/2023/04/11/lab-leak-accide…
-
zeynep tufekci
@zeynep
For real. Searching for the word substack on Twitter returns “newsletter”.🤪
What’s next to be hardcoded? Search for Elon returns “the most amazing funniest guy ever”? 🤷♀️ -
zeynep tufekci
@zeynep
This is amazing: the new book on poverty by the amazing sociologist @just_shelter debuts the #1 NYT bestseller.
Live Talks LA
@LiveTalksLA
Congrats to Matthew Desmond @just_shelter!!! His book "Poverty, by America" debuts at #1 on the NYT list. See him tomorrow 3/30 in convo with @MichaelDTubbs of @EndPovertyCA in Santa Monica, 8pm. Also a virtual airing on 4/6
Tix/info:
livetalksla.org/events/matthew-desmond-with-michael-tubbs/
@CrownPublishing -
zeynep tufekci
@zeynep
It was a rare film about a war that resisted the narrative. Many alleged antiwar movies had succumbed. But the fact that the score *also* didn’t was pretty striking. Historic, really.
Chris Lang
@ChrisLangWriter
'All Quiet On the Western Front', is possibly the only truly anti war film I think I have ever seen. No heroic deeds, no journeys, no stirring score, no nobility, no survivors, just endless, pointless, horrific, death.
It's spectacularly brilliant. -
zeynep tufekci
@zeynep
NEWS: Cochrane says lead author of a mask review misrepresented its findings, apologizes for summary statement that was imprecise and says they will update it.
They say their review did NOT find masks don't work.
Plus, I examine the actual evidence.
www.nytimes.com/2023/03/10/opinion/masks-work-cochrane-st… -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Hong Kong can measure INTRINSIC severity of Omicron on the unvaccinated without prior infection—zero Covid before.
Result: Omicron similar severity as Wuhan variant on the immunonaive.
Claims elsewhere on its "mildness" likely due to population immunity.
Ben Cowling
@bencowling88
A final observation in our study relates to the comparative severity of Omicron vs the ancestral Wuhan strain (waves 1-4). When we look at unvaccinated older adults with no prior infection outside the peak of the fifth wave (wave 5a), we find fairly similar fatality risks (13/16)
-
zeynep tufekci
@zeynep
Universities are terrible offenders.
Press release: “X goes with Y!”
Media headline: “X causes Y!!”
Twitter: “X deniers are eugenicists!!!”
Study, after reading the supplements: “Under limited conditions and weak data, X may show an uptick with Y but not yet conclusive or why”?
Sabine Hossenfelder
@skdh
I've been doing my weekly Science News for about 4 months now. I have literally read thousands of press releases to that end. The major insight that I have taken away from this is to never, ever, trust a press release.
-
zeynep tufekci
@zeynep
A gift link (no paywall) for the moving column by @nytdavidbrooks on what he learned from his friend who suffered from depression.
I've a friend with major depressive episodes. This column has all the painful lessons that I wish I had learned on day one.
www.nytimes.com/2023/02/09/opinion/despair-friendship-sui… -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Still hard to wrap one's mind around the scale.
Two *separate* earthquakes, ten hours apart, M 7.8 and M 7.7 (matches the largest in all of 20th century in Turkey) with a fault break across hundreds of miles —anywhere on it is at risk— in heavily populated area with many cities. -
zeynep tufekci
@zeynep
Not an article I wanted to write.
The deadly H5N1 bird flu is spreading widely, including to mammals.
For the first time, it's now likely spreading mammal-to-mammal, among minks which are exceptionally well-suited conduits to humans.
We must act now.
www.nytimes.com/2023/02/03/opinion/bird-flu-h5n1-pandemic…
zeynep tufekci
@zeynep
The good news is it's early, and we have many tools, including vaccines.
But we've learned, right? Right? We can't just wait till a pandemic pathogen spreads to humans and gets established before responding.
There is time and tools to stop an H5N1 pandemic. We should use them. -
zeynep tufekci
@zeynep
I couldn't care less about royal news.
But don't dismiss Harry/Meghan brouhaha as mere celebrity tittle-tattle. Want to discuss Brexit? This is related.
And they make a good case. It's not fluff.
I wrote it up, straight, without the celebrity stuff.
www.nytimes.com/2023/01/25/opinion/harry-meghan-tabloids.…
zeynep tufekci
@zeynep
Why write about this—seemingly so far from my interests? Harry and Meghan?
Because it's not. Corrupt media ecology is THE topic of the day. Murdoch also owns Fox News.
And because too many are dismissing The Spare as mere oversharing based on a few irrelevant details. It's not.
zeynep tufekci
@zeynep
Gift link to my NYT piece on why the Harry/Meghan stuff is not mere celebrity kerfuffle and what the *actual* news there.
Maybe this is why Spare sold so well, despite the dismissive media coverage that largely skipped over the actual matter at hand.
www.nytimes.com/2023/01/25/opinion/harry-meghan-tabloids.…Post not available here, view on X
zeynep tufekci
@zeynep
After Covid Twitter, and Arab Spring before with Trump stuff in the middle, I’d have thought I’d seen it all—but the pro-royalty (I suppose?) accounts now responding to my tweets about cellular immunity with bizarre (and false) allegations about my supposed financial interests?
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
I think this is so important.
Twitter charlatans, attention-seekers, liars etc. may be preying on the suffering of patients, many of whom aren't in a position to wade through all this, but the suffering and the illness is real and urgent, AND NEEDS THE BEST RESEARCHERS ON IT.Post not available here, view on X
zeynep tufekci
@zeynep
That the suffering is so real and so deep, and that the patients are left flailing for answers, is why they are easy prey for this kind of fraud, and hopefully, that motivates the real scientists to get more involved, and responsible MDs to try to see more patients. 🙏🙏
-
zeynep tufekci
@zeynep
It’s true Scott is on Pfizer’s board and ex head of FDA. I was *very* wary.
I didn’t expect to find him to be so informative and balanced. But he was. Among first to advocate for masks. Balanced on vaccines.
Musk has him personally targeted by one of Pandemic’s worst grifters.
Scott Gottlieb, MD 🇺🇸
@ScottGottliebMD
🧵In the past, I've raised concerns with Twitter related to the safety of me and others, and threats being made on the platform. This included direct as well as specific threats. Sometimes it included statements that I believed were purposely false and inflammatory 👇
Post not available here, view on X
zeynep tufekci
@zeynep
We aren’t getting transparency. Or accountability. This is all a pathetic attempt to target individuals in a reckless, dangerous and uninformative manner while pushing the stupidest theories about institutions and people.
Anyway it’s now down to pandemic’s worst grifters. -
zeynep tufekci
@zeynep
Okay then. That’s the big story.
Taibbi says, seemingly with access to all internal docs, the earlier big claim that the government had directly intervened (“stepping in”) to suppress a story before the election has no evidence.
Also seems some nonconsensual nudity was removed. -
Post not available here, view on X
zeynep tufekci
@zeynep
Oh come on. What’s going on in China has little to do with anywhere else. It’s not even remotely comparable. People died in a fire with sealed shut door. This is disrespectful to the bravery and suffering of people in China. Let’s get a grip, out of respect if nothing else.
-
zeynep tufekci
@zeynep
Why are they all carrying it live? Announce the news, move on. Haven't we learned a thing?
When he was the president, with the nuclear suitcase next to him, he couldn't be ignored. He's back to trolling for attention why give him a bullhorn completely on his own terms?
Oliver Darcy
@oliverdarcy
Fox News and CNN are taking Trump's announcement live. MSNBC is not.
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
In my last piece, I described and named post-exertional malaise.
We got letters saying "so that's WHAT I HAVE" from baffled people. It's so counterintuitive and weird, and still so little known, that it's the clinical guidelines/screening that's lacking. www.nytimes.com/2022/10/27/opinion/me-cfs-long-covid.html -
zeynep tufekci
@zeynep
New from me, a plea: focusing so much on @elonmusk shouldn't lead us to defend advertiser-financing of platforms, which is neither free nor healthy.
Can we remember what ad-financing helped fuel on
the (pre-Musk) Twitter, Facebook, Tiktok or Youtube?
www.nytimes.com/2022/11/04/opinion/elon-musk-twitter-free… -
Chris Hayes
@chrislhayes
I don't have any special insight or expertise, but have been pretty persuaded by the pushback to the piece.
Michael Hiltzik
@hiltzikm
Column: ProPublica and Vanity Fair are pushing the COVID lab-leak theory, but their exposé is a train wreck www.latimes.com/business/story/2022-11-01/column-propubli…
Post not available here, view on X
zeynep tufekci
@zeynep
Also we didn’t get a totally random novel virus, we got one related to the one we have been poking around and experimenting on for the past decade, fearing it may launch a pandemic, in a city with such labs. That’s not proof! Could be a coincidence. But can’t calculate this imo.
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
But I don’t see how not being able to rule anything out should hinder the obvious: the fact that *any one of those paths is plausible* means that we should work on making each of those paths less likely to spark the next pandemic.
The future answer is all of the above are risks. -
zeynep tufekci
@zeynep
So @CaseyNewton reports that Musk asked Twitter engineers to print their code, and be ready for "code pairings" with Musk.
As my own quote tweet below, I guess Musk is the next in line to discover the obvious: he's now in a people business, not a tech one.
zeynep tufekci
@zeynep
Silicon Valley is run by people want to be in the tech business, but are in the people business. They are way, way in over their heads.
-
zeynep tufekci
@zeynep
New piece on the ME/CFS (and Long Covid) protest by the White House, and what it teaches us about bravery, risk—and ourselves.
In many protests I studied—Tahrir, Gezi, Hong Kong—the bravery was easily visible. Here, it was in the out-of-sight aftermath.
www.nytimes.com/2022/10/27/opinion/me-cfs-long-covid.html -
zeynep tufekci
@zeynep
For the XBB variant dominant in Singapore, check out their dashboard.
Country is highly-vaccinated and boostered. Rapid tests and high-quality masks for visitors to nursing homes.
15 people in ICU for Covid out of a country of 6 million. "99.8% of cases mild or asymptomatic".Post not available here, view on X
zeynep tufekci
@zeynep
Singapore—thus XBB—update. Cases/hosp. down, 9 in ICU (pop: ~six million), 99.7% of cases mild/asymptomatic.
What about the lab studies? The "immune-evasion"? Antibodies aren't the whole immune system and after many years, in vitro doesn't reflect population immunity landscape. -
zeynep tufekci
@zeynep
This is great.
I'd love a part two with:
"Some of these are correlated" 😱
The many problems with with the way respondent orientation towards pols and pollsters and outcome variables are correlated for US politics 😱😱
"Missing Not at Random" is so common in research😱😱😱 -
zeynep tufekci
@zeynep
A great article about a crucial topic.
Many people struggling with Long Covid are running out of short-term disability and savings, but the disability/insurance bureaucracy keeps putting up obstacles — and many are are already too ill to navigate that maze.
Terrible situation.
charlos
@loscharlos
“Something else keeps her up, too: She can, quite literally, no longer afford to have #longcovid. The disease has robbed Hamre not only of her health, but of her career, income, and insurance coverage.” www.bostonglobe.com/2022/10/15/business/long-covid-took-t…
-
zeynep tufekci
@zeynep
Alex Jones was ordered to pay ~$1 billion for Sandy Hook lies.
But the key issue remains: It’s so easy now to make a lot of money by lying outrageously and cynically.
My piece on why we should focus on this, and the money machine, to slow this onslaught.
www.nytimes.com/2022/10/16/opinion/alex-jones-sandy-hook.… -
zeynep tufekci
@zeynep
There seems to be a lot of misinformation going around that immunocompromised people don't benefit at all from vaccines or existing COVID therapies.
That is *absolutely* false and very dangerous falsehood to spread, as such people might not realize the many options they have.🧵Post not available here, view on X
zeynep tufekci
@zeynep
I've heard from so many immunocompromised people who didn't realize some of their options because they were falsely believing there were none — sometimes even their doctors were unaware, plus the usual social media misinformation misled them. Unnecessary tragedy.
-
zeynep tufekci
@zeynep
Example of how to be misled.
First, I absolutely believe COVID (and flu) has neurological sequelae.
But this study is almost entirely pre-vaccine (99%), and the most affected are those hospitalized, not the mild cases, *and* it's an already old, frail male VA cohort. -
zeynep tufekci
@zeynep
New: immunologists on why the new boosters are great for now and the future—even against Long Covid.
Do we really need a booster, many said last year. Only ~30% of the US got one. More boostered Canada had 80% fewer deaths from Omicron.
Let's not repeat.
www.nytimes.com/2022/09/15/opinion/covid-booster-shot.html -
zeynep tufekci
@zeynep
Read @edyong209 on "brain fog".🙏
When many with Long Covid or ME/CFS etc. talk of "brain fog" they don't mean having an off day or a hangover.
After listening to many, my mental model of this type of "brain fog" is that it is to thinking what running in quicksand would be.
Ed Yong is not here
@edyong209
🚨I wrote about “brain fog”—one of the most common & disabling symptoms of long COVID (and many other pre-pandemic conditions), and one of the most misunderstood.
Here’s what brain fog actually is, and what it’s like to live with it. 1/
www.theatlantic.com/health/archive/2022/09/long-covid-bra… -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
And our dreadful treatment of Long Covid patients, I believe, is a continuation of dismissal of many chronic conditions historically with versions of "in your head, dear".
Obviously, ME/CFS is a crucial example, but not just.
Multiple Sclerosis. Ulcers. The examples are many.Post not available here, view on X
zeynep tufekci
@zeynep
I spent most of my summer researching our Long Covid response—I plan more pieces, an essay can only fit so much—and I cannot really find words to express how much I despaired on behalf of the many patients I talked with.
It's just not okay to abandon so many people like this. -
zeynep tufekci
@zeynep
Excellent @aaronecarroll essay on why we should advocate for vaccination for kids, but also good analysis of why the rates have been so abysmal.
Risks are low isn't an argument against lowering them even further by such an effective tool. #giftlink
www.nytimes.com/2022/08/18/opinion/covid-vaccination-chil… -
zeynep tufekci
@zeynep
So we may very belatedly get a federal privacy law that will end up weakening the one state where we have a slightly strong one? *slow clap.*
How about make it a minimum, not a maximum? It’s not strong enough to be a ceiling as it stands.
Julia Angwin
@JuliaAngwin
And the biggest catch: ADPPA pre-empts nearly all state privacy laws including California's landmark law. (The author’s home state law - Illinois's Biometric privacy act, is carved out however). /9
www.ilga.gov/legislation/publicacts/fulltext.asp?Name=095… -
zeynep tufekci
@zeynep
Important. Needs to be part of clinical advice, imo, and also ties to how both acute and Long Covid interact with existing inequities like lack of paid sick leave.
Thread ahead but the tl;dr is resting if sick or even afterwards is important—sometimes in underappreciated ways.
charlos
@loscharlos
“The public messages from prominent officials saying they’ll keep working from home are minimizing the risk of #longcovid and encouraging others to think, ‘If I have the virus, I can just push through it’” @PutrinoLab www.latimes.com/california/story/2022-07-07/working-throu…
-
zeynep tufekci
@zeynep
Stop me if you heard this one before. Not enough tests or testing, community transmission we can’t trace.😢
“But it also means that they're possibly spreading the virus without knowledge of the fact that they're infected”—@JenniferNuzzo, now on Monkeypox.
www.npr.org/sections/health-shots/2022/06/25/1107416457/m… -
zeynep tufekci
@zeynep
Unregulated technology allows many, many ways for law enforcement *or* vigilantes to identify and hunt down women suspected of being pregnant.
My new essay explains why avoiding this as an individual—by deleting apps or being careful—will not be feasible.
www.nytimes.com/2022/05/19/opinion/privacy-technology-dat…Post not available here, view on X
zeynep tufekci
@zeynep
Print version of my piece on why the reckless way in which technology has been allowed to used for surveillance, in ways little understood by many, may be used by law-enforcement or vigilantes in a post-Roe America. www.nytimes.com/2022/05/19/opinion/privacy-technology-dat…
Post not available here, view on X
zeynep tufekci
@zeynep
I don't think most people understand the depth of the surveillance infrastructure built over the past few decades.
It's not just online and it's not stoppable by turning off a few options or opting out.
Slightly shortened print version of my piece on it.
zeynep tufekci
@zeynep
Print version of my piece on why the reckless way in which technology has been allowed to used for surveillance, in ways little understood by many, may be used by law-enforcement or vigilantes in a post-Roe America. www.nytimes.com/2022/05/19/opinion/privacy-technology-dat…
-
zeynep tufekci
@zeynep
Highlighting from co-author of the flu/COVID kids comparison paper.
They found influenza posed equal or maybe higher risk compared to COVID to *each* child, but there were so many more COVID cases that the total "volume" (toll) for COVID was higher.
Let's take *both* seriously.
Steven L. Shein, MD
@RBC_PICU
We were intentional to say “volume” of death, not mortality rate. Our data suggest mort %’s similar, or maybe Flu higher. But if several fold more COVID patients, absolute # of deaths would be higher.
And results likely differ over peds age spectrum, like you suggest -
zeynep tufekci
@zeynep
Wow. An in-depth yet clear explanation of immunity, its durability and vaccines by @deeptabhattacha.🙏
Confused by online duels and Twitter personalities but find the official messaging not detailed or forward-looking enough? READ THIS. Seriously. 10/10🙌
www.nytimes.com/2022/06/13/opinion/covid-immunity.html -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
My friends who clearly qualify and have otherwise top-notch doctors are telling me that their doctors have not heard of Evusheld, and thus they cannot get a prescription. It's absurd, but reality... Looking to find an informed doctor now, but how?
-
zeynep tufekci
@zeynep
I do not want to read a single profile of the shooter. Enough.
Nothing will allow us to predict which of the many, many troubled young man will turn to a mass shooter. Zero public interest in the individual details.
It’s the access to the deadly gun that’s the key. That’s it. -
zeynep tufekci
@zeynep
Kids under five still have no vaccine.
Pfizer's was delayed in a confusing turnabout, and FDA may delay Moderna authorization because... some baffling theory about simultaneous release?
My piece on why this makes no sociological or immunological sense.
www.nytimes.com/2022/04/27/opinion/covid-vaccine-kids.htmlPost not available here, view on X
zeynep tufekci
@zeynep
So, CDC is worried updating vaccines might "create an impression that we don’t have a very effective vaccination program".
Having vaccines available only for a long-extinct strain when we could easily update them would surely not be the answer here to fixing that "impression". -
zeynep tufekci
@zeynep
*taps mic*
Some personal news, as they say.
I've been granted tenure and appointed a professor at Columbia University, and will stay in NYC to become the inaugral director of the @Craignewmark Center for Ethics and Security in Journalism, dedicated to a healthier public sphere. -
zeynep tufekci
@zeynep
Helen, thank you!
I’m so so honored to be in this group and grateful for the support at @nytimes and @TheAtlantic—the recognition is for last year which spanned both institutions. A true privilege.
Congratulations to fellow finalist Julian Aguon and winner Melinda Henneberger! -
zeynep tufekci
@zeynep
An interesting part is how often people like Shirley MacLaine—not the first example—fall for stuff like this.
A long history of this: comfortable, well-meaning but clueless Westerners falling for lies brought about by authoritarianism because they feel sympathetic to the people.
Alex Dukalskis
@AlexDukalskis
I had not heard this anecdote from Deng Xiaoping's 1979 visit to the United States.
-
zeynep tufekci
@zeynep
China also does this, implicitly and explicitly: threatens family members of internal dissidents and exiles.
(Hence important not to pretend we can have free interviews on anything "touchy" with people there: see Peng Shuai. People do sometimes speak out, but at great cost).
Andrei Kurkov
@AKurkov
In Melitopol aged father of journalist Svetlana Zalizetska was taken hostage by probably FSB people. They say they will release him only if Svetlana gives herself up. She is not in Melitopol anymore.
-
zeynep tufekci
@zeynep
For the second pandemic anniversary, I wrote a plausible what-if.
What if the world reacted quickly to the available information, like Taiwan or South Korea did?
Lancet calculates ~18 million lives lost to Covid. Many, maybe all, could have been spared.
www.nytimes.com/2022/03/11/opinion/covid-health-pandemic.… -
Post not available here, view on X
zeynep tufekci
@zeynep
"Guns of August" by Barbara Tuchman on the first month of WWI is another essential book, along with her other classic "March of Folly."
Both are excellent to help to fight the temptation of looking at world history teleologically.
zeynep tufekci
@zeynep
Many things we take for granted—peace, potable tap water, childhood vaccines—didn't just happen, and don't have some sort of self-propelled guarantee to last into the future.
But once you have them for a generation or two, very easy to think of them as background, as default. -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
I maintain this point.
The key obstacle is not having our act together.
Any novel virus pandemic would be a challenge.
The key issue is not some magical, insurmountable, imagined qualities of this virus compared to all others.
It's our subpar response.
Ezra Klein
@ezraklein
I can't tell if her final note is hopeful or gutting, but it's been ringing in my head ever since.
-
zeynep tufekci
@zeynep
I post this regularly, and many express surprise.
This is the current childhood immunization schedule. Many childhood vaccines are three+ doses, sometimes a booster.
This is one reason why, unlike earlier centuries, we don't have cemeteries full of children. We've forgotten. -
zeynep tufekci
@zeynep
Study from France: Multisystem Inflammatory Syndrome in Children by COVID-19 Vaccination Status.
Their tables have no line for fully-vaccinated children in their. Why? Because there wasn’t a single one among the 33 MIS-C cases in that hospital.
jamanetwork.com/journals/jama/fullarticle/2787495 -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Here’s the childhood immunization schedule that helps ensure we are not like those families I read about in history, where losing multiple children to childhood illnesses wasn’t uncommon. Many vaccines are three dose series with occasional later boosters. This is a gift.
-
zeynep tufekci
@zeynep
Very good to have this confirmation. When your immune system sees Omicron, it does better, later, against Delta as well!
Alex Sigal
@sigallab
We have submitted new results to medRxiv:
Omicron infection enhances neutralizing immunity against the Delta variant
The preprint can be found here:
sigallab.net -
zeynep tufekci
@zeynep
A few years ago, we did a campaign to raise money in support of Doctors without Borders. I'm doing another one, supporting their work in Yemen and in Burundi. I'll match $20,000 in donations. Here's the link to donate, to join me or to raise the match! events.doctorswithoutborders.org/campaign/zeynepmatch2021
zeynep tufekci
@zeynep
Yemen: Millions of children are at risk of starving. I'm fundraising—matching donations up to $15,000 for @MSF_USA. events.doctorswithoutborders.org/campaign/zeynep
-
zeynep tufekci
@zeynep
Queue for rapid tests. Too bad we have no system in this country that would take envelopes from a warehouse containing these tests and place it in designated cubbies that exist for every address, and even do this six days of the week! Need a blockchain solution, I think.
Dan Diamond
@ddiamond
The line now loops itself. The giveaway doesn’t start for another 20 minutes.
This media couldn't be displayed. See it on the original post. -
zeynep tufekci
@zeynep
Both this meme and public health authorities can communicate effectively about vaccines and transmission.
-
Post not available here, view on X
zeynep tufekci
@zeynep
We *still* need to wait for epi data to get a better sense of protection against severe disease (expected to remain reasonable and even substantial, but let's see) and vaccine efficacy—lab results do clearly suggest more breakthroughs for sure, but this also needs more epi data.
-
Post not available here, view on X
zeynep tufekci
@zeynep
Another thread on the risk of sublethal mutagenesis (on the virus, not the human!) from molnupiravir:
Michael Lin, MD PhD 🧬
@michaelzlin
In conclusion, it would be highly irresponsible to approve molnupiravir unless the possibility of sublethal accelerated viral mutagenesis is thoroughly addressed and satisfactorily rejected, especially within the target population of patients at risk of hospitalization.
-
zeynep tufekci
@zeynep
Oh, finally. Almost two years in. One more topic in which fairly obvious and visible evidence was denied for so long that the result is both unnecessary suffering and loss of trust in authorities and experts exactly at the time their credibility is their most important tool.
-
zeynep tufekci
@zeynep
‼️“Fewer than 1% of patients taking the drug needed to be hospitalized and no one died. In the comparison group, 7% were hospitalized and there were seven deaths.” So promising that they halted the trial early. Pfizer, like Merck, *must* allow cheap generics for LMIC. No excuses.
The Associated Press
@AP
BREAKING: Pfizer says its experimental COVID-19 pill cut hospitalization and death by 90% and will seek authorization from the U.S. Food and Drug Administration and international regulators.apne.ws/H3kYgLS
-
zeynep tufekci
@zeynep
72% of the unvaccinated said they'd quit after a mandate. In real life, few do. Studies show older people are vulnerable to misinformation and Fox News, and yet almost all are vaccinated. What explains this? My piece of the sociology of the unvaccinated. www.nytimes.com/2021/10/15/opinion/covid-vaccines-unvacci…
-
zeynep tufekci
@zeynep
Excellent point from @jljcolorado. “Droplets on surfaces is very convenient for people in power - all of the responsibility is on the individual... OTOH, if you admit it is airborne, institutions, governments and companies have to do something.” www.telegraph.co.uk/global-health/science-and-disease/eve…
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Mysteriously missing: early sequences, databases, accounting of early cases, work we now learn *had been proposed* but we were told was a crazy conspiracy, an "investigation" led by conflicted parties, that was lied to but still proposed frozen food imports as plausible origin.
-
zeynep tufekci
@zeynep
People are asking why there aren’t more women and children there. Because that’s what happens when, instead of a managed situation, it turns it into a perilous and dangerous fight. This flight set the record for a C-17 because people fought their way in. The crew decided to fly.
Marcus Weisgerber
@MarcusReports
JUST IN: "The Crew made the decision to go" — Inside RCH 871, which saved 640 from the Taliban ... from @TaraCopp and me www.defenseone.com/policy/2021/08/inside-reach-871-us-c-1…
-
zeynep tufekci
@zeynep
I went on media and mentioned something that's no surprise to any healthcare worker: some people are just really afraid of needles. They don't always admit it in polls, but they are there. But with some support, they can get vaccinated. Since then, I've been getting emails.
Jason Shellen is on Bluesky
@shellen
A new, hopeful take on the road ahead in fighting COVID. Thank you @zeynep for this post and for saving countless lives by questioning the guidance to “not wear” masks in early 2020. www.theinsight.org/p/the-lure-of-the-caricature
-
zeynep tufekci
@zeynep
Every piece of data from real-life shows the vaccines work very, very well— yes, even against Delta. Just checked US vaccine breakthrough hospitalizations. It's 6,587 people among the ~163,000,000 vaccinated: or 0.004%. Three fourths are elderly— as happens with other diseases.
Eric Topol
@EricTopol
Pretty damn impressive
Thanks Darren Lu @Reddit -
zeynep tufekci
@zeynep
For the second time in fifty years, we're pondering whether our research activities have sparked a pandemic like they did in 1977. A deep dive from me on what we *already* know about the origins of the coronavirus—and how we can learn the right lessons. www.nytimes.com/2021/06/25/opinion/coronavirus-lab.html
-
Post not available here, view on X
zeynep tufekci
@zeynep
Recent letter in Science addressing this: we no longer accept unsafe food or water but have barely budged on indoor air standards for pathogens. (And the resistance and ridicule aerosol scientists endured for the past year made one key obstacle visible). science.sciencemag.org/content/372/6543/689
-
zeynep tufekci
@zeynep
Huge deal. Some very eminent virologists, including Dr. Baric who worked with the lab in Wuhan, have published a statement that both zoonotic origin and lab accident are "viable" as origin. (Social media claims aren't scientific consensus, part zillion). science.sciencemag.org/content/372/6543/694.1
-
zeynep tufekci
@zeynep
The WHO just updated its page on how COVID-19 transmits. Those few sentences on aerosols represent one of the most crucial scientific advances of the pandemic. My NYT piece on the century-long history of the error, the year of delay—and what it means now. www.nytimes.com/2021/05/07/opinion/coronavirus-airborne-t…
-
zeynep tufekci
@zeynep
Thousands of retweets for implying a “5X” surge of B1617 variant is related to school openings in the UK—and rising despite vaccination, implying vaccine escape. I looked it up. The UK health secretary says there are “103 cases” and “vast majority” linked to international travel.
-
zeynep tufekci
@zeynep
"Ten scientific reasons in support of airborne transmission of SARS-CoV-2." New paper in @TheLancet
by @trishgreenhalgh, @kprather88, @Rschooley, @jljcolorado, @dfisman and me! We argue the case that it's not just that SARS-CoV-2 is airborne, but that it's predominantly airborne. -
zeynep tufekci
@zeynep
Potentially important part of the puzzle: researchers argue risk of CVST (cerebral venous sinus thrombosis—the rare side effect that led to J&J pause) is much higher from COVID *itself* compared to AZ or mRNA vaccines. (preprint, usual caveats, see thread/paper for more context).
Kai Kupferschmidt
@kakape
There is an interesting new preprint out that will probably generate a lot of coverage at least in the UK. Essentially it argues that the risk of CVST is much higher from #covid19 than from vaccines.
Quick thread on this:
osf.io/a9jdq/ -
zeynep tufekci
@zeynep
CNN article on how to safely fly claims that 90% vaccine efficacy means that🙄 for every million who fly, we could have 100,000 infections. NO NO NO. That’s not what that number means. Also, this didn’t even happen when millions flew unvaccinated. So how could it make sense now?
-
zeynep tufekci
@zeynep
Read and weep how the Skagit choir superspreading outbreak, first in the US, was ignored or explained away as maybe fomites or contact—a theory that required contortions that'd make epicycles blush—despite are-you-kidding-me obvious airborne transmission. www.nytimes.com/interactive/2021/04/08/magazine/skagit-va…
-
zeynep tufekci
@zeynep
More good news. Pfizer has the six month update. *Zero* hospitalizations among the vaccinated vs. 32 among the placebo group (n=46,307). Also real life data from South Africa, where B.1.351 is dominant: zero cases of COVID among the vaccinated vs. nine in the placebo (n=800).
-
zeynep tufekci
@zeynep
No, no, no. The right to vote is a *right*; thus a normative question, not one to be analyzed on its presumed partisan effects (besides, social scientists cannot really predict these effects). Please, let's focus on if the law makes it harder to vote, and who the impact falls on.
The New York Times
@nytimes
Democrats are concerned that Georgia's new voting law will have a significant effect on turnout and electoral outcomes. The evidence suggests otherwise. www.nytimes.com/2021/04/03/upshot/georgia-election-law-tu…
-
zeynep tufekci
@zeynep
Another shooting tragedy, and I’m sadly going to re-up one of my old threads with my writings on this topic going back a decade: how we cover mass shootings matters. Don’t do it on the killer’s terms. Don’t plaster his name, social media etc. on our screens. Focus on victims.
zeynep tufekci
@zeynep
Whatever else, we know: attention on the killer on his terms helps inspire the next one. Or focus on methods. Don't. www.nytimes.com/2015/08/27/opinion/the-virginia-shooter-w…
-
zeynep tufekci
@zeynep
Incredible New Zealand study on a mystery transmission in a quarantine hotel. For a while, a garbage can was suspected—one of the very, very few known suspected cases of fomite transmission. Turns out... it's aerosols. What a Kuhnian year this has been. wwwnc.cdc.gov/eid/article/27/5/21-0514-f1
-
zeynep tufekci
@zeynep
Tens of millions of AstraZeneca vaccines that won't be approved here soon enough are languishing in US depots. Our supply already exceeds the number of Americans. Less than 1% of Africa has been vaccinated. President Biden: Let our surplus vaccines go—now. www.nytimes.com/2021/03/17/opinion/coronavirus-vaccine.html
Post not available here, view on X
zeynep tufekci
@zeynep
Two more articles, @ryanlcooper in The Week and @dandrezner in Washington Post. Both argue that it helps US "soft-power" to help vaccinate the rest of the world. If that moves the gears, let that move the gears, too. 😀 www.washingtonpost.com/outlook/2021/03/18/most-important-… and theweek.com/articles/972786/biden-wants-make-america-glob….
-
zeynep tufekci
@zeynep
Stop me if you heard this one. Dr. Wu realized that the pneumonic plague was airborne and could spread person-to-person and said people should wear masks. French doctor mocked this, insinuated Asian doctors couldn't be right, refused to wear a mask—and got infected and died.
ray 鄺羡華
@raykwong
Today’s Google doodle celebrates Wu Lien-teh, an epidemiologist who pioneered the use of face masks to control an epidemic over a century before the advent of COVID-19.
-
zeynep tufekci
@zeynep
Yeay! CDC guidelines now say not just that vaccinated people can meet with other vaccinated people without masks, but that they can do so with other *unvaccinated* individuals, too, as long as they are low-risk (i.e. healthy adults or grandchildren!) and from one other household.
-
zeynep tufekci
@zeynep
This right here is another thing that's so terribly wrong and callous with blaming the victims... "Why didn't they stay home?" Because many had to work so others could "just stay home" and sometimes those workers unwittingly brought it to their crowded, multi-generational home.
This post is unavailable. See it on X. -
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
So it's not enough to discuss the evidence etc. for fomites vs airborne, because there's also a lot of sociology and psychology driving why we cling to that bottle of disinfectant a little too tightly—and that same dynamic applies to our reaction to vaccines/masks & beaches, too.
Post not available here, view on X
zeynep tufekci
@zeynep
Every time I go on media and say the obvious: outdoors IS safer with much less risk of superspreading (I don't know of a single confirmed outdoor-only mass transmission event) many people object and say Rose Garden! Sturgis! Neither were outdoor only. The price of misinformation.
-
zeynep tufekci
@zeynep
This is really important. Stuff getting publicized *without a paper* and without sufficient time for many scientists to look at, digest, comment and contextualize can lead to terrible outcomes—needlessly scaring people. We saw that happen with widely misreported studies before.
Eric Topol
@EricTopol
Why are academic groups forwarding "preprints" that are not posted directly to the media without the biomedical community having a chance to review? Isn't a preprint enough? is this a new pre-pre-print trend? Concerning.
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Cognitive biases, moral panics, authoritarian impulses, victim-blaming, groupthink, traditional & social media distortions, ignoring everything we know about human behavior & sometimes even viruses, philosophical frequentism, terrible risk-assessment...
virginiahume
@virginiahume
Can’t wait to read the book someone will surely write, looking at this past year through the lens of cognitive biases. There’s an absolute gold mine of material.
-
zeynep tufekci
@zeynep
Read this story and weep. This, after masks, is the biggest failure—both scientific and communication—of the pandemic in Western nations: the airborne nature of the pathogen. CDC changed this guidance last October. Japan and many other nations had it right on *February* of 2020.
The Wall Street Journal
@WSJ
Six feet and 15 minutes became coronavirus gospel. The NFL had data that showed otherwise: The virus could be spread in under 15 minutes and from over six feet. www.wsj.com/articles/super-bowl-nfl-covid-cdc-11612104460
-
zeynep tufekci
@zeynep
Interrupting your doomscrolling with more good news. As most had expected, the vaccines continue to work well against the new variants. Our problem remains producing enough of these amazing vaccines and getting them out there in arm. They are so good that we have a lot of buffer.
KizzyPhD
@KizzyPhD
The neutralization teams have a message:
"Moderna COVID-19 Vaccine Retains Neutralizing Activity Against Emerging Variants First Identified in the U.K. and the Republic of South Africa"
Sorry to those who I've been ignoring yin this regard for weeks lol
Moderna
@moderna_tx
We just announced that the Moderna COVID-19 Vaccine retains neutralizing activity against emerging variants first identified in the U.K. and the Republic of South Africa. Read more: investors.modernatx.com/news-releases/news-release-detail…
Post not available here, view on X
zeynep tufekci
@zeynep
Reading the news, you'd think we have failing vaccines. Moderna starting to work on potential booster is presented as evidence of failure rather than, wow, ahead of the game already. (What I want to read is how do we quadruple production!)
-
zeynep tufekci
@zeynep
He hasn’t yet once acknowledged Biden as president-elect. (Only referenced next administration—once, I believe in the “farewell” address.) He’s going to leave without conceding or accepting he lost. He’ll keep claiming it was stolen. And most of House GOP is with him on that.
Jennifer Jacobs
@JenniferJJacobs
Trump’s “ceremony” is scheduled for 8 a.m. at the military base where the outgoing president will make his final departure aboard Air Force One.
He’ll be 1st living US president in more than a century to choose not to attend regular exchange of power.
www.bloomberg.com/news/articles/2021-01-18/trump-farewell… -
zeynep tufekci
@zeynep
Good coverage of the latest messaging disaster, in a year with many. We are dramatically underselling the amazing vaccines, exaggrating the uncertainties and people trying to emphasize the good news are being drowned out, or worse. www.nytimes.com/2021/01/18/briefing/donald-trump-pardon-p…
-
zeynep tufekci
@zeynep
Red tape and punitive measures isn't how we get access, equity or fairness for vaccines. We should cut down on the bureaucracy—which the more privileged navigate better—and invest in access by *going* where those more vulnerable live and work. New piece. www.nytimes.com/2021/01/15/opinion/coronavirus-vaccine.html
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Even better correction: *147 not a single of whom has yet recanted or apologized. And remember, this is after a good many of them got spooked by the mayhem and the violence of the day, and changed their minds on voting to overturn the election. www.theatlantic.com/ideas/archive/2021/01/more-dangerous-…
-
zeynep tufekci
@zeynep
Why aren't we wearing better masks? Last April, we—19 experts―wrote paper on why cloth masks were a good stop-gap measure for COVID. It's just been published in PNAS. Great but... @jeremyhoward and I wrote about our frustration that it's still relevant! www.theatlantic.com/health/archive/2021/01/why-arent-we-w…
Post not available here, view on X
zeynep tufekci
@zeynep
Three representatives—one a 75-year-old cancer survivor—were infected, likely by GOP reps who refused to wear masks in shelter. Were they wearing only cloth masks? Irresponsible people means it's even more important to supply better masks to everyone else. www.theatlantic.com/health/archive/2021/01/why-arent-we-w…
-
zeynep tufekci
@zeynep
My Octopus card would work, too.
Marc W
@marctoronto
Rapid tests available in the Hong Kong subway system. Cool!! There's a reason they're winning... @michaelmina_lab @fordnation
-
zeynep tufekci
@zeynep
Moderna data is out. At day 28—before dose two—disease prevention efficacy is 92.1% [CI: 68.8%, 99.1%] AND 2/3 fewer asymptomatic infections. Data is small & short-term hence my earlier piece on need for immediate trial for single dose for low-risk groups. zeynep.substack.com/p/vaccines-and-decision-making-with
-
zeynep tufekci
@zeynep
The BioNTech/Pfizer data is out—NEJM & FDA report. I understand this wasn't in the design & the long-term protection from single dose is open question. But given this sharp drop after 14 days after one dose, can someone explain why single dose isn't very very high on the agenda?
-
zeynep tufekci
@zeynep
There's time to be alarmed. It's such a time. The president's blatant, if incompetent, attempts to steal the election are happening in a period of entrenched of minority rule, and are met mostly with silence or approval by Republican leaders. New piece:
www.theatlantic.com/ideas/archive/2020/12/trumps-farcical… -
zeynep tufekci
@zeynep
A study from Korea showing why indoor dining is unsafe and why airborne transmission matters. Case B infected case A from 6.5m (~21 feet!) away in *just five minutes*, and case C from 4.8m (15 feet!). Footage shows no interaction—and only those in line of air flow got infected.
Post not available here, view on X
zeynep tufekci
@zeynep
I wrote a post for @insight explaining the details of this amazing restaurant study, comparing it to other similar studies, and discuss what we can learn from all this about both indoor dining and airborne transmission. It’s empowering! zeynep.substack.com/p/small-data-big-implications
-
zeynep tufekci
@zeynep
I wrote about the Chinese scientist who kick-started it all by publishing the SARS-CoV-2 sequence in an open repository on January 10th—without waiting for authorization. Two days later, Moderna had the vaccine. Just ten months later, we have a vaccine. zeynep.substack.com/p/the-pandemic-heroes-who-gave-us-the
-
zeynep tufekci
@zeynep
Trump was incompetent at his brand of authoritarianism, and he lost. But this isn't a triumph: the Democrats lost almost everything else. Make no mistake, about what's coming: The next Trumpist, in 2024, will be talented, competent and not easy to beat. www.theatlantic.com/ideas/archive/2020/11/trump-proved-au…
-
zeynep tufekci
@zeynep
Had he not been murdered, Emmett Till would have been only a few years older than both our presidential candidates this year. The past isn't that far past.
Emmett Till Interpretive Center
@etic_tillcenter
“Today we remember #emmetttill who would have been 79 years old. And as we continue to push forward in the efforts of creating a safe #america for people of color we honor those who paved the way toward #justice.” 🗣 Repost @naacp 🗣 #etic #racialreconciliation #tellingthetruth
-
zeynep tufekci
@zeynep
I think if the whole White House press corps abandoned the WH pressers, and left behind a single camera as a pool camera and submitted written questions... We’d all be better off. Reporters would be safer and we’d get less of the circus. We never get informed anyway. Why risk it?
Seung Min Kim
@seungminkim
Where we’re at: a high school friend who is a doctor in a different state is offering to FedEx me her spare N95s because covering the White House is hazardous.
-
zeynep tufekci
@zeynep
My current saga. My landlord wants me to sign a new lease that gives me only two hours notice to show the house to potential buyers, upon which I am supposed to vacate the house—and if I don't do that three times, he can evict me. Two hours notice. In the middle of a pandemic.
-
zeynep tufekci
@zeynep
❗️Did you know that *study after study* finds most people don't seem to transmit COVID at all? That a small percent is responsible for almost all infections? That R is not that informative? My new piece on why this may be key to controlling the pandemic. www.theatlantic.com/health/archive/2020/09/k-overlooked-v…
zeynep tufekci
@zeynep
COVID has a very skewed distribution in its transmission patterns. You can't beat back a threat if you don't take into account how it operates, and this skew—its overdispersion— should affect the way we do *everything* from tracing to testing. That's what countries like Japan do.
Post not available here, view on X
zeynep tufekci
@zeynep
Look, the science doesn't fit into the lockdown/anti-lockdown fights. It's not how much we restrict, but whether we target the right things. Overdispersion means the right measures take us a long way, but also that we can't relax: It can quickly flip back. www.theatlantic.com/health/archive/2020/09/k-overlooked-v…
-
zeynep tufekci
@zeynep
I don’t understand the “this will not matter” talk among the chattering classes. Nobody has a crystal ball. Who knows? But one way of ensuring nothing matters is giving up having values to defend. Cynicism isn’t a sign of depth or savvy, it’s resignation masquerading as politics.
-
zeynep tufekci
@zeynep
I know the crisis is far from over yet, but has anyone been running some mortality/morbidity numbers from the off-the-charts AQI the West Coast has been enduring, some places for weeks now? Air pollution has a huge toll on health, and this is now a tragically sustained event.
-
zeynep tufekci
@zeynep
We urgently need to focus on ventilation. Six months into a respiratory pandemic, we're still not given sensible and practical guidance against short-range aerosol—airborne—transmission of COVID. I wrote about the science & what it means we should do now. www.theatlantic.com/health/archive/2020/07/why-arent-we-t…
Post not available here, view on X
zeynep tufekci
@zeynep
Stellar example of effective science communication from @JamesEKHildreth who clearly but matter-of-factly explains a latest key paper on aerosol transmission, and why, by itself, distance isn't all protective indoors—especially in low ventilation settings. x.com/i/status/1294947678011502592
Post not available here, view on X
zeynep tufekci
@zeynep
CDC? WHO? We still don’t have ventilation integrated into our guidelines. People are being told to wash their hands, but not open their windows. It’s September 2020.
-
zeynep tufekci
@zeynep
Wait, what? Woodward had Trump on tape that the coronavirus was airborne BACK IN FEBRUARY (seems he learned this from Xi) and sat on it, presumably to release his book closer to the election? As we've been saying, the US crisis is elite failure *across* the political spectrum.
Post not available here, view on X
zeynep tufekci
@zeynep
I wrote an article on ventilation late summer and got inundated with messages from ordinary people saying "oh, wow, I had no idea. We're now opening windows in our <indoor space>. Thank you for informing us!" Seems that Woodward, not just Trump, had the confirmation in February.
-
Post not available here, view on X
zeynep tufekci
@zeynep
I don't understand how is not topic number one. Keeping young kids out of formal schooling and in high-stress, little/no supervision situations for a year or more—that's where we're headed—is potentially a life-changing event for children whose families cannot cushion the impact.
-
zeynep tufekci
@zeynep
Those COVID apps some colleges are trying to force students to download? Worse than useless. You know what else is worse than useless? Bringing the students together for a few weeks, and then sending them home in a panic, to seed new clusters. New piece. www.theatlantic.com/technology/archive/2020/09/pandemic-n…
-
zeynep tufekci
@zeynep
I do not understand the logic of the graphs in this (NYT article shared with alarm) article at all. At best, they're uninformative, but at worst, wildly misleading. Percent share can increase a lot if the baseline is low, or if fewer adults get infected and can mean nothing. ???
-
zeynep tufekci
@zeynep
Well. So this happened.😀As I told @benyt, I feel charmed that I get to do this: do my best to call things as I see it regardless of considerations of popularity. It's a privilege and an honor. Much gratitude to so many who supported me for so long. 🙏 www.nytimes.com/2020/08/23/business/media/how-zeynep-tufe…
-
Susan Dominus
@susandominus
In mid-July, a large study out of South Korea seemed to suggest that children over the age of 10 spread coronavirus even more than adults do; this update to the same study clarifies otherwise: www.nytimes.com/2020/08/14/health/older-children-and-the-…
Post not available here, view on X
zeynep tufekci
@zeynep
Original article; almost 400,000 shares on Facebook alone. Correction? Not even a thousand shares, yet. It wasn't just the NYT, and how many will not even correct? I don't know how many kids are denied safe schooling and parents now overly-anxious because of the rushed reporting.
-
zeynep tufekci
@zeynep
Enough with the beach-scolding! It's not scientifically-sound and it's counterproductive. It's a virus—not a moral agent geared to smite people who dare enjoy themselves. Six months in, we *know* most risk is indoors. More knowledge, less baseless outrage. www.theatlantic.com/health/archive/2020/07/it-okay-go-bea…
Post not available here, view on X -
zeynep tufekci
@zeynep
Good news from early phase Oxford vaccine trials. Also (early) good news today from the interferon beta clinical trial. (To be published, soon but it's an RCT from a solid team so🤞🧿). Recent T-cell reactivity papers with encouraging findings. A break from the doomscrolling!
The Lancet
@TheLancet
NEW—UK’s #covid19 vaccine is safe and induces an immune reaction, according to preliminary results www.thelancet.com/lancet/article/s0140-6736(20)31604-4
-
zeynep tufekci
@zeynep
Good morning in sad, jealous tweets. All schools in Taiwan have been open since February. They did not listen to China or WHO in January and immediately masked up and shut down travel. They also traced the hell out of their outbreaks. They were practically done by February.
-
zeynep tufekci
@zeynep
Do protests even work? We have the second civil rights movement—and amidst a pandemic! Many assume that change will come if many people march in the streets. Historically, that's not true. But protests *do* often work and change society. How? New piece. www.theatlantic.com/technology/archive/2020/06/why-protes…
-
zeynep tufekci
@zeynep
How is this not a direct violation of the first amendment? An arrest for a t-shirt? Where is the camous snowflakes killed free speech crowd on this?
-
zeynep tufekci
@zeynep
Wow. A nursing home in Baltimore, oldest African-American one, did not lose *a single person* to COVID because as soon as they heard Trump say cases would soon go to zero, they realized it was going to be a catastrophe, stopped visits and masked up. www.baltimoresun.com/opinion/columnists/dan-rodricks/bs-m…
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
The policing we’re seeing isn’t about “crowd control.” It’s about what the president said: “domination.” We know this from decades of research, but you don’t need the research to understand how their aggression fuels the escalation. It’s plain to see now. www.theatlantic.com/health/archive/2020/06/i-cant-breathe…
-
zeynep tufekci
@zeynep
About the the risk of COVID from the protests. Please note that, to date, there are very few known cases of outdoor transmission at all, but crowding 100 people into an indoors space overnight? Yes, that's how you get a spike. Jails and detention, not protesting, is the big risk.
-
zeynep tufekci
@zeynep
Facebook's own researchers found that “64% of all extremist group joins are due to our recommendation tools" and that “our recommendation systems grow the problem.” Facebook VP for policy Joel Kaplan (known for throwing a party for Kavanaugh after confirmation) nixed any action.
Deepa Seetharaman
@dseetharaman
New from @JeffHorwitz & me: Facebook spent years studying the its role in polarization, according to sources and internal documents. One internal slide laid out the issue like so. ”Our algorithms exploit the human brain's attraction to divisiveness.” www.wsj.com/articles/facebook-knows-it-encourages-divisio…
-
zeynep tufekci
@zeynep
They found six new cases in Wuhan so they’re planning to test eleven million people over the next ten days. Neither number is a typo. www.reuters.com/article/us-health-coronavirus-china-wuhan…
-
zeynep tufekci
@zeynep
Significant paper on the cost of misinformation. Basically, one standard deviation more viewership of Sean Hannity (denied seriousness of COVID) versus Tucker Carlson (took the pandemic seriously) is associated with in 20% more deaths at the county-level. bfi.uchicago.edu/wp-content/uploads/BFI_WP_202044.pdf
-
zeynep tufekci
@zeynep
All over the world, the authorities are closing down parks and trails—but that's unsustainable, counterproductive and probably even harmful. Instead, we should adapt to the pandemic and retool them for social-distancing. My new piece on why and how. www.theatlantic.com/health/archive/2020/04/closing-parks-…
Post not available here, view on X
zeynep tufekci
@zeynep
Folks, please check the evidence! The value of outdoors in fighting pandemics/infections is so strong that there are peer-reviewed papers recommending that we *encourage* people to spend more time outdoors during the next pandemic (this one!)! We just need to do it sensibly.
-
zeynep tufekci
@zeynep
Dear journalists: amplifying the president saying he won't wear a mask *without* pointing out that nobody is allowed be near him without taking a COVID-19 test in the SAME sentence/headline/tweet is irresponsible and dangerous. Please, let's learn. Don't amplify without context.
-
zeynep tufekci
@zeynep
New piece: Epidemiological models aren't crystal balls that predict the future, nor are we sure if their parameters are correct! That's okay because that's not what they're for. Let's treat them as what they are: the future imploring us to choose our fate. www.theatlantic.com/technology/archive/2020/04/coronaviru…
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
From an article that compiled interviews with top experts on what works to slow the spread of this coronavirus. Universal mask wearing!
Abraar Karan
@AbraarKaran
8/ Masks were a whole different ball game as well.
Everyone wore masks. Part of this was to prevent transmission.
But part of it was also to reduce stigma against the sick.
If only the sick wear masks, they may be hesitant.
But if everyone does, then no one is missed.Post not available here, view on X
zeynep tufekci
@zeynep
The Czech Republic converted, quickly, to universal mask wearing, including home-made masks as necessary about a week ago. There really are no arguments left against promoting universal mask wearing and yet media/guidelines still say no need.
Jeremy Howard
@jeremyphoward
The person who really kicked off #masks4all in Czech Republic was @PetrLudwig with his powerful and compelling video. A key part of his message: wearing a mask is a selfless, pro-social behavior
www.youtube.com/watch?v=K5xy2n941jM
zeynep tufekci
@zeynep
Another misinformation I see floating widely: without eye-protection masks are useless. No. It's a respiratory disease. Goggles help lessen risk, sure, but key route is respiratory. Masks are key. Surface contamination from boxes is similarly a tail risk. (But wash your hands!)
Post not available here, view on X
zeynep tufekci
@zeynep
See this screenshot of Taiwan's action on masks (link has list of all its actions). Taiwan had extensive ties/travel with China but it has a total of five COVID-19 deaths and only a few hundred cases. Its schools have been open for more than a month. cdn.jamanetwork.com/ama/content_public/journal/jama/0/jvp…
-
zeynep tufekci
@zeynep
Wuhan authorities held a 40,000+ person potluck right before before human-to-human transmission was finally admitted by the government, and the whole city shut down. What’s our excuse? Nothing. This is a seeding event. These people are at risk. They will then spread out to US. 😢
Brooke McDonald
@BrookeGMcDonald
This is the scene at O’Hare airport. The traveler who took the photo said it’s a 6-hour wait for bags then on to customs for 2-4 more of waiting in shoulder-to-shoulder crowds. Police are handing out water and disinfectant wipes. @fly2ohare #ord #coronavirus #covid19
Post not available here, view on X
zeynep tufekci
@zeynep
This day will be a terrible turning point, seeding many communites. Visible in a month or so.😢 In the name of enhanced screening!!!
Steven Dial
@StevenDialFox4
Passengers waited in line for hours at customs at @DFWAirport. Many concerned about being so close to others during the #covidー19 outbreak. The airport saying “CBP officers and the CDC are following federal guidelines to conduct enhanced screening for passengers” @FOX4
-
zeynep tufekci
@zeynep
I wrote a short, practical guide on why you should prepare for COVID-19 and how. Forget the doomsday talk and the "prepper" subculture with all its snake oil. You should prepare *now because it's the best thing you can do to help everyone else. Here's how:
blogs.scientificamerican.com/observations/preparing-for-c…Post not available here, view on X
zeynep tufekci
@zeynep
What's happening now in Italy is that they have run out of hospital capacity. That will mean death rates will go up, not just from the coronavirus but from all other illnesses as well. We look about two weeks behind this point in big population centers.
Carl Quintanilla
@carlquintanilla
(Reuters) - The death toll from an outbreak of coronavirus in Italy has jumped by 168 to 631, an increase of 36%, the Civil Protection Agency said on Tuesday, the largest rise in absolute numbers since the contagion came to light on Feb. 21.
#covid19 www.reuters.com/article/us-healthcare-coronavirus-italy-t… -
zeynep tufekci
@zeynep
Hong Kong had every reason to have an exponential epidemic (lots of travel from Wuhan/China) and yet has dramatically slowed the spread. Two deaths total. Probably: with memory of SARS, they responded quickly with closures, universal mask wearing, social distancing and hygiene.
Post not available here, view on X
zeynep tufekci
@zeynep
Taiwan is another country with significant success in containing this coronavirus—a single death and only 44 known cases despite extensive travel from Wuhan/China. Here, too, experience with SARS and China's coverups helped—they leaped to action quickly. healthpolicy.fsi.stanford.edu/news/how-taiwan-used-big-da…
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
"We get the flu every year." You know what will happen if we hit COVID-19 unprepared? More people will die of... the flu. And everything else. On top of COVID-19. We have only so many hospital beds, ventilators, ICUs. To be ready means we are ready to flatten the curve—buy time.
-
zeynep tufekci
@zeynep
Hong Kong protests are in their fifth month despite an escalating crackdown. How? One surprising answer: the fate of China's Uyghurs. Many talked to me about it. They watched and learned. They've decided that they "may as well go down fighting." My latest. www.theatlantic.com/international/archive/2019/11/escalat…
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Many high-profile activists were arrested in Hong Kong today. But none are key organizers because this movement doesn't work like that. Reminded me of how the Egyptian regime tried to get @Ghonim to retract some demands. He tried to explain he couldn't, even if he had wanted to.
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
The Boeing Max case is important because it's a tragic example of how AI/automation/code crises will actually occur. It won't be Terminator coming back from the future to hunt O'connor. It will be regular corporations or governments using code/AI as they seek profit or control.
-
zeynep tufekci
@zeynep
I wrote about why the downturn of storytelling in #gameofthrones also explains why it's hard for us to deal with Facebook, YouTube, AI, etc. The show was a rare beast: a sociological narrative in a world dominated by psychological/individualized ones. blogs.scientificamerican.com/observations/the-real-reason…
-
zeynep tufekci
@zeynep
Everything Elsevier purchases becomes weaponized against the academy for (hefty) profit. Don't use them.
Chris Blattman
@cblatts
I used to host all my papers with @SSRN. I’m out.
It just gets worse. Now people have to create an account & login to download freely uploaded papers. This used to be the easiest way to make PDFs available to all. I’m exiting the whole system this month.
Shame on @SSRN. -
zeynep tufekci
@zeynep
Wow Google is all in with the conservative Federalist Society which engineered the multi-decade effort to foster judges like Kavanaugh and Gorsuch for the Supreme Court. Facebook vice-president was in DC hearings to personally support Kavanaugh. Silicon Valley is so “liberal.”🙄
-
zeynep tufekci
@zeynep
ahahaha my article on exactly this issue from 2011 where I rant about the "myopic business model threatening the internet" 🙄 It may appear that I'm the broken clock, but in my defense... I still think it's true. slate.com/technology/2011/09/yahoo-didn-t-mean-to-censor-…
-
zeynep tufekci
@zeynep
Thread on the untold harms of YouTube’s business and operating model.
.
@swordsjew
i can well and truly say that most of you have no idea the kind of abhorrent filth that is succeeding wildly on youtube. multiple videos claiming LBGT folks are advocating pedophilia with 400k+ views. open white nationalism. open antisemitism. it is scary - & profitable.
-
zeynep tufekci
@zeynep
Friends I'm hiring *two* post-docs: one from computer science/AI and one from social science. We're going to take on thorny problems, do solid and solution-oriented research with global focus. Come work with me! Or pass it on! unc.peopleadmin.com/postings/146380 and unc.peopleadmin.com/postings/146470
-
zeynep tufekci
@zeynep
Folks, I’m launching a research project on hate-speech and intercommunal violence—especially in the context of encrypted apps. This is what I’ve been losing sleep over for over a year. Look for a post-doc announcement soon. No easy solutions but there are real steps to take.
Dave Lee
@DaveLeeBBG
Fake kidnapping videos are spreading on WhatsApp, and people are being beaten to death over them. www.bbc.co.uk/news/world-asia-india-44435127
-
zeynep tufekci
@zeynep
So if a Facebook friend of yours had the Facebook app on *their* phone, Facebook shared *YOUR* private data with device-makers, even if you had somehow managed to turn off third-party sharing.
"Sorry, not sorry" apology coming in 3, 2, 1. 💰💰💰
www.nytimes.com/interactive/2018/06/03/technology/faceboo… -
zeynep tufekci
@zeynep
Oh, look, if you are resource-insecure as a child (i.e. poor!) you are less likely to delay gratification (since who knows if things will be there) and kids enrolled in a preschool on STANFORD CAMPUS aren't that representative.¯\_(ツ)_/¯
Also, don't separate culture and economy. -
zeynep tufekci
@zeynep
Parkland killer says "When you see me on the news you’ll know who I am." Shame on every news organization that's publishing this snuff film. He killed those students so you would. We have more than enough research: you're inspiring the next killer. Enough. www.nytimes.com/2015/08/27/opinion/the-virginia-shooter-w…
-
zeynep tufekci
@zeynep
Probably not too long before we have the first drone assassination by a civilian through this technology.
-
zeynep tufekci
@zeynep
I am, sadly, going to retweet an earlier thread. Media: please don’t put the killer’s face on loop, on front pages, on your app notifications. Don’t sensationally publicize manifestos. Focus on victims and structural causes. Don’t help inspire the next killer.
zeynep tufekci
@zeynep
Time for that sad reminder. After Robin Williams' suicide, sensational media coverage that violated the CDC guidelines resulted in a 10 PERCENT increase in suicides. Same effect applies to mass shootings. Newsrooms, please be considerate in your coverage. www.nytimes.com/2015/08/27/opinion/the-virginia-shooter-w…
-
zeynep tufekci
@zeynep
Google Assistant making calls pretending to be human not only without disclosing that it's a bot, but adding "ummm" and "aaah" to deceive the human on the other end with the room cheering it... horrifying. Silicon Valley is ethically lost, rudderless and has not learned a thing.
-
Post not available here, view on X
zeynep tufekci
@zeynep
This is also why "get off Facebook" by itself is no panacea. It's like saying the rich should put a bubble around their home and filter their own air, while the world remains badly polluted. The information commons is global, consequential and largely monopolized. #ascmediarisk
-
zeynep tufekci
@zeynep
Facebook "GDPR" settings are mostly set-up to get you to "accept and continue" and go through messy options if you try anything else—and you don't get full data collection control. This UX design is obviously intentional, driven by the business model. Simple as that.
TechCrunch
@TechCrunch
A flaw-by-flaw guide to Facebook’s new GDPR privacy changes tcrn.ch/2JSDqWa by @joshconstine
-
zeynep tufekci
@zeynep
“Kim's research showed that in fact, these unregulated ads made up the majority of issue-based ads on Facebook during the course of her study. Facebook did not provide a comment before publication.”
issie lapowsky
@issielapowsky
A researcher with lots of foresight scraped 5 million political ads on Facebook during 6 weeks before the 2016 election. She found that half of the advertisers had absolutely no federal records or online footprint. Of that half, 1 in 6 were Russian trolls. www.wired.com/story/russian-facebook-ads-targeted-us-vote…
-
zeynep tufekci
@zeynep
Many people asked me what lawmakers should ask Mark Zuckerberg. Here's my answer: Nothing. Instead, they should get to work and pass legislation to fix the reckless surveillance in the digital economy. In my latest NYT oped, I suggest four concrete steps: www.nytimes.com/2018/04/09/opinion/zuckerberg-testify-con…
Post not available here, view on X
zeynep tufekci
@zeynep
With this question: does Facebook have real competition? Nope. There is no real alternative to Facebook: no substantive competition, no real market discipline on the company. It's simple as that.
-
zeynep tufekci
@zeynep
Front page of the New York Times, a video: "Who Was the YouTube Shooter?" Research on previous shooters, and her behavior prior to the event, suggests that she was someone who tried to kill people so that you'd make that video and put it in the front page of the New York Times.😡
-
zeynep tufekci
@zeynep
This is more of a question than a comment.
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
The way to change Facebook and YouTube—and the whole online economy of surveillance—is to change their incentive structure and business model. Neither Zuckerberg apologizing for the umpteenth time in fifteen years nor legislators yelling at Zuckerberg will get us there.
-
zeynep tufekci
@zeynep
Wait! Facebook doesn’t sell people’s data. It profiles people based on extensive data collection and then sells people’s *attention*. CA got data because Facebook was giving it away—not selling. Pass a law saying Facebook can’t sell your data, and Zuckerberg will celebrate it.
Edward Snowden
@Snowden
Facebook: "This is their information. They own it"
BBC: "And you won’t sell it?"
FB: "No! Of course not."
Please help this 2009 interview of Facebook's CEO get seen by people who don't use Twitter. Here's a download link so you can pull and repost it: www.sendspace.com/file/wqwtq5 -
zeynep tufekci
@zeynep
The issue goes far beyond what Cambridge Analytica may have done in 2016. We need to confront that this kind of surveillance/data infrastructure is ripe for use, misuse and abuse. My latest for the @nytimes. www.nytimes.com/2018/03/19/opinion/facebook-cambridge-ana…
Post not available here, view on X
zeynep tufekci
@zeynep
Maybe the 50 million whose data was siphoned to Cambridge Analytica will get notified but there were so many many apps doing the same thing. How about those people? Will each incident get a notification? That spigot was open for so many years (when it helped Facebook grow).
-
zeynep tufekci
@zeynep
Facebook's defense that Cambridge Analytica harvesting of FB user data from millions is not technically a "breach" is a more profound & damning statement of what's wrong with Facebook's business model than a "breach".
-
zeynep tufekci
@zeynep
YouTube may well be operating as a giant radicalizing engine through its recommendation algorithm—leading people down a rabbit hole of misinformation, hoaxes and incendiary content. My latest for the New York Times on one of the most overlooked issues. www.nytimes.com/2018/03/10/opinion/sunday/youtube-politic…
-
zeynep tufekci
@zeynep
Time for that sad reminder. After Robin Williams' suicide, sensational media coverage that violated the CDC guidelines resulted in a 10 PERCENT increase in suicides. Same effect applies to mass shootings. Newsrooms, please be considerate in your coverage. www.nytimes.com/2015/08/27/opinion/the-virginia-shooter-w…
-
zeynep tufekci
@zeynep
So this Wired story pretty much confirmed for me what I’d been saying for a while (including before the election). Facebook got played and manipulated to not fix the egregious exploitation of its platform before the election. Plain and simple. www.wired.com/story/inside-facebook-mark-zuckerberg-2-yea…
zeynep tufekci
@zeynep
Yep. I mean, literally, duh. And especially duh since Facebook got played so badly and naïvely before the election.
This post is unavailable. See it on X. -
zeynep tufekci
@zeynep
My latest for the @nytimes. The Strava debacle shows that individualized "informed consent" is not sufficient for data privacy. Given the complexity, companies cannot fully inform us, and thus we cannot fully consent. Data privacy is more a public good. www.nytimes.com/2018/01/30/opinion/strava-privacy.html
-
zeynep tufekci
@zeynep
Great loss. It's hard to name one's favorite book *ever* but people do ask. I usually waver between saying The Dispossessed or the Left Hand of Darkness—both hers. Ursula K. LeGuin was a great American novelist with such great range and imagination. www.nytimes.com/2018/01/23/obituaries/ursula-k-le-guin-ac…
zeynep tufekci
@zeynep
My mind is just looping on the ending of The Dispossessed. When I read that book first (because I read it many times!), I had just learned English. I remember feeling so so so thrilled that my new language gave me access to such an amazing story. What a gift it was.
-
zeynep tufekci
@zeynep
No, nope, this is the exact opposite of what will likely happen. Facebook, Amazon, Netflix etc. will buy their lane, and will remain nominally free. Losing net neutrality would probably benefit Facebook more than any other company—even more than the ISPs. This isn’t 2003 anymore.
Your Friendly Neighborhood Numbersmuncher
@NumbersMuncher
If #netneutrality is repealed:
Amazon Prime $299/year
Facebook $12.99/status update
Twitter $.25 per character per tweet
Netflix $16.99 per Stranger Things episode
YouTube $14.99 per minute of video
Hulu $29.99 per episode of Scrubs
Am I doing this right? -
zeynep tufekci
@zeynep
Web 1.0: It' all about information!
Web 2.0: Let's go social!
Web 3:0: Weaponized/monetized fraud; bots & trolls polluting the public sphere; organized attention manipulation ops; censorship via information glut, distraction and undermining credibility: Internet of Fake Things!
Tim Karr is now on Bluesky
@TimKarr
Data scientist finds 1.3 million faked comments in FCC docket. Someone used a bot & massive email list to create false impression of popular support for @AjitPaiFCC's plan to kill #netneutrality. The culprit will soon be revealed... hackernoon.com/more-than-a-million-pro-repeal-net-neutral…
-
zeynep tufekci
@zeynep
I’m hiring a postdoc to work with me on social implications of AI. Technical/CS background (especially within the ML cluster) plus serious interest or degree in social sciences and/or industry experience. Posting coming soon.
-
Post not available here, view on X
zeynep tufekci
@zeynep
The piece we needed was all the (many deliberate) obstacles to this kind of research, and all the enormous effort academics are putting into this area (with much output) despite the deliberate obstacles. "Asleep at the wheel" makes no sense to me.
-
zeynep tufekci
@zeynep
New: My talk on why I fear ad-financed digital economy could lead us to surveillance authoritarianism, and how. www.ted.com/talks/zeynep_tufekci_we_re_building_a_dystopi…
-
zeynep tufekci
@zeynep
*This* functionality Facebook provides to anyone who pays—using its vast surveillance and algorithms—is key to 2016. www.washingtonpost.com/business/economy/russians-took-a-p…
-
zeynep tufekci
@zeynep
Whatever else, we know: attention on the killer on his terms helps inspire the next one. Or focus on methods. Don't. www.nytimes.com/2015/08/27/opinion/the-virginia-shooter-w…
Post not available here, view on X
zeynep tufekci
@zeynep
Media: the next potential mass shooter is watching the current media coverage intently. Right now. Your coverage is a factor in this crisis.
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Yep. I mean, literally, duh. And especially duh since Facebook got played so badly and naïvely before the election.
This post is unavailable. See it on X. -
zeynep tufekci
@zeynep
Every word of this crucial piece. Journalists also publish BASELESS algorithmic panic pieces—to chase clicks. Read: idlewords.com/2017/09/anatomy_of_a_moral_panic.htm
-
zeynep tufekci
@zeynep
For the NYT, I wrote about Equifax 🙄and how corporations shift most risks of tech to users. Time for responsibility. www.nytimes.com/2017/09/11/opinion/equifax-accountability…
-
zeynep tufekci
@zeynep
Sounds great till you realize it can be done the other way too—to brick. Connectivity tech is also bringing massive centralization of power.
Milena Rodban
@MilenaRodban
Tesla remotely extended range of Florida owners' cars jalo.ps/wRAOoWA
-
zeynep tufekci
@zeynep
Facebook and Google make money from a personalized, opaque information flow at a scale of billions. It's time for oversight and disclosure.
-
zeynep tufekci
@zeynep
ohai a machine learning system that can identify ~69% of protesters who are wearing caps AND scarfs to cover their face. h/t @jackclarksf
Post not available here, view on X
zeynep tufekci
@zeynep
Yes, we can & should nitpick this and all papers but the trend is clear. Ever-increasing new capability that will serve authoritarians well.
Post not available here, view on X
zeynep tufekci
@zeynep
Let me say: too many worry about what AI—as if some independent entity—will do to us. Too few people worry what *power* will do *with* AI.
-
zeynep tufekci
@zeynep
Entire political structure upended on knife-edge win by simple majority, with all 3 economic centers (2/3 AKP previously won) voting no.
Noah Blaser
@nblaser18
Narrow win hands Erdogan unprecedented power but not the security of a convincing win, while opposition claims count was prematurely frozen.
-
zeynep tufekci
@zeynep
haha if anyone asks why the "pseudo" part in psedo random number generators matter. How to cheat at slot machines. www.wired.com/2017/02/russians-engineer-brilliant-slot-ma…
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
This isn't "oh, we had open borders & now we crack down." Its: "people who had 100% legal papers have, without warning, had them preempted."
Post not available here, view on X
zeynep tufekci
@zeynep
Yes, it does say a FIVE year-old—with a legal visa as he was allowed to board a plane to US—was detained for hours.
7News DC
@7NewsDC
VIDEO: Mother from Iran, 5-year-old son reunited after he was detained for several hours at Dulles Airport: wjla.com/news/local/video-mother-from-iran-5-year-old-son…
-
zeynep tufekci
@zeynep
The world lost John Berger. A recent interview—at 90, still nuanced and perceptive about new technology. www.theguardian.com/books/2016/oct/30/john-berger-at-90-i…
-
zeynep tufekci
@zeynep
Yemen: Millions of children are at risk of starving. I'm fundraising—matching donations up to $15,000 for @MSF_USA. events.doctorswithoutborders.org/campaign/zeynep
Post not available here, view on X
zeynep tufekci
@zeynep
2016 was many things on Twitter, but also: we've raised $220,565 for MSF/Yemen's humanitarian emergency. Last hours: events.doctorswithoutborders.org/campaign/zeynep
-
zeynep tufekci
@zeynep
By design, algorithm & policy Facebook entrenches echo chambers and amplifes misinformation. Me for the NYT. www.nytimes.com/2016/11/15/opinion/mark-zuckerberg-is-in-…
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
2016 was a close election where filter bubbles & algorithmic funneling was weaponized for spreading misinformation. www.niemanlab.org/2016/11/the-forces-that-drove-this-elec…
zeynep tufekci
@zeynep
What a denial of responsibility. If this were true, and Facebook had no influence, it would not be a $400b company. x.com/MikeIsaac/status/796891604837494784
-
Post not available here, view on XPost not available here, view on X
zeynep tufekci
@zeynep
Passing around charts (that you are reading wrong) to dismiss role of economic despair/populism in Trump? Factiness. nathanjurgenson.com/post/152938927255/factiness
-
zeynep tufekci
@zeynep
Enormous amount of news from Turkey: arrests of MPs; explosions, internet shut downs. A journalist/observer list: x.com/zeynep/lists/turkey-journalists
Mark Lowen
@marklowen
And just to make it all more difficult to report, government cuts internet in SE #turkey, blocks #whatsapp and #twitter, so this is with VPN
-
zeynep tufekci
@zeynep
Average CNN nightly viewership was around 400,000 last I checked. Congress Pericope sit-in: 1,000,000 viewers.
Government
@Gov
Tweets sent today from @ScottPetersSD and @RepSwalwell with #periscope broadcasts have now been viewed over 1M times and counting.
-
zeynep tufekci
@zeynep
The (first woman) mayor of Paris (also the granddaughter of refugees and daughter of migrants) appeals for unity. 🇫🇷
Anne Hidalgo
@Anne_Hidalgo
Nous sommes debout, nous sommes unis. J'en appelle à l'unité de toutes et tous.
-
zeynep tufekci
@zeynep
Researchers set 10 "honeypot" networked medical devices. 55,000 logins, 300 malware drops. fusion.net/story/181527/security-honeypots-snare-hackers/
-
zeynep tufekci
@zeynep
Liberia needs 110,000 plastic gloves.. It has 2100. It needs 5000 plastic buckets. It has 116. www.motherjones.com/mojo/2014/10/liberia-ebola-going-need…
-
Post not available here, view on X
zeynep tufekci
@zeynep
The revolution will be charged because batteries don't last. RT @FxPasquier .. these guys are super organised. #hk926
-
zeynep tufekci
@zeynep
Headscarf fashion. Huge sector in Turkey, with dedicated magazines, shows, and, of course, YouTube channels.
-
zeynep tufekci
@zeynep
"I'm a black guy giving a talk at a tech conference called 'who's at the table.' You all think I'm the diversity talk." @andre_banks #pdf14
-
zeynep tufekci
@zeynep
Video allegedly showing miner first being slapped by Erdogan, beaten by his bodyguards. Ends with: "STOP FILMING!" www.youtube.com/watch?v=fPf2JIQYLWk
-
zeynep tufekci
@zeynep
Turkey's newest civil disobedience star: Kiymet Peker, 75, of Edirne. Her local park was about to be razed so...
-
zeynep tufekci
@zeynep
I know ppl are stuck on the shirtless pics but WSJ makes it clear that FBI was breaking into email over minor issues: online.wsj.com/article/SB10001424127887324439804578115410…
-
zeynep tufekci
@zeynep
Here's the sequence of pics showing Army stripping, beating a woman wearing hijab. RT @RamyRaoof: flic.kr/p/aWJygr #tahrir
-
zeynep tufekci
@zeynep
.@EthanZ on Dictator's Dilemma: Rasing on the social cost of censorship. My prev post on this: technosociology.org?p=286 #hyperpublic








































































































































































































































































































































































































































































































