Had a great panel discussion with @theovalich and Chris Caswell at @CapitalClubDXB, talking about the steps needed to be taken to reach technological sovereignty in the age of AI.
Sharing the main takeaways:
> Sovereignty covers the whole stack inside the country: power, data centre, chips, inference, and models trained on local data. It also means a government can push a new protocol or agent to every institution through one network when a new law requires it.
> Inference sets the economics of an AI factory. The same GPUs earn far more when a data centre sells tokens than when it rents them out by the hour. Routing cuts costs further: a “hi” doesn’t need a trillion-parameter model when a 7B model answers it for a fraction of the cost.
> Resilience comes from distribution. When one large site goes down, every workload on it goes down too. Spread the same capacity across hundreds of racks inside the institutions that use them, and one failure moves a small share of traffic to the nearest site.
> Open models make this workable. In September they carried 71.5% of tokens on Vercel’s AI Gateway, so teams already trust them in production. A country can fine-tune them on local data and run them on hardware it controls.
The UAE needs its gigawatt campuses and the distributed capacity around them. After that, the number to track is how many companies run AI in production.
Ilman Shazhaev
Ilman Shazhaev (born 13 May 1993) is a scientist, IT engineer, and AI/tech entrepreneur who has published over 40 peer-reviewed research papers and holds registered patents in AI, deeptech, and gaming. He is the founder and CEO of Dizzaract, a technology group building AI and gaming digital intelligence infrastructure based in Abu Dhabi. More on Wikipedia
-
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev
@shzhv13
Anthropic now lets companies run its agent platform on their own servers. It's the right product, and the model call is the one part that still leaves the building.
Self-hosted Claude Managed Agents opened in private beta.
The customer runs the agent harness, sandboxes, credentials and session history in their own cloud or on-premise, and chooses where requests to Claude are sent. Developers keep the same APIs as the hosted service.
It's aimed at teams whose security rules kept them off the hosted version.
Last month I wrote that an agent spends much of each task outside the model, running tools and reading files. Anthropic has now handed that part to the customer.
The open question is the model. If a regulator needs it inside the border too, the customer needs a cloud region in that country or open weights. Access to the beta also still goes through Anthropic's account team.
@AnthropicAI -
Ilman Shazhaev
@shzhv13
What does a token cost to produce?
Divide what a GPU costs per hour by the tokens it makes in that hour. On MLPerf's latest result and today's median rental price, a million output tokens of gpt-oss-120b costs about 18 cents of GPU time, before a single idle minute.
The top of the fraction is the GPU-hour. A B300 rents for a median of $9.50 an hour across providers. An owner pays for hardware, power, space and staff instead, but the logic is the same.
The bottom is throughput. In MLPerf Inference v6.1, Lambda's four-GPU Blackwell Ultra system produced 58,195 tokens a second of gpt-oss-120b within the server latency limits. That's about 14,500 per GPU, or 52 million an hour.
$9.50 divided by 52 million tokens is about 18 cents per million. Engineering moves the bottom number: Lambda got almost 9% more throughput on identical hardware between MLPerf rounds, from software alone.
Then utilisation. The hour costs the same whether the GPU is busy or idle. At 50% busy the same tokens cost 36 cents per million, and at 25% they cost 73. MLPerf is also a tuned benchmark, and mixed real traffic runs below it.
The big hosts list gpt-oss-120b at 60 cents per million output tokens. CoreWeave lists 17 cents, below the fully busy cost at median rental prices. That only works for an owner whose GPU-hour costs less than the rental median and whose cluster stays full.
The business is keeping good GPUs busy, and pricing follows from that. -
Ilman Shazhaev
@shzhv13
A new GPU rack bought about 1.5x per chip this year. A better router on the same GPUs bought about 2x.
The next gain sits in the software that decides which GPU serves which request.
My new article goes deep into why. x.com/i/article/2120876374952992768 -
Ilman Shazhaev
@shzhv13
I ranked 8 data centre GPUs for serving inference by what one rental dollar buys in memory and memory bandwidth.
When a model generates tokens, memory decides which models fit on the card and bandwidth sets how fast tokens come out. This leaves out compute and interconnect, which matter more for long prompts and multi-GPU serving.
S
> MI325X: 256 GB at 6 TB/s for $3.10 an hour. The cheapest memory and the cheapest bandwidth of any card here.
> MI300X: 192 GB at 5.3 TB/s for $2.86. Almost the same value, one generation older.
A
> B300: 288 GB, the most memory on a single card. Large MoE models need fewer GPUs per copy.
> H200: 141 GB at 4.8 TB/s for $5.00. The same bandwidth per dollar as the B200 at about 60% of its hourly price.
> B200: 7.7 TB/s for $8.01, the second-fastest memory on the list.
B
> RTX PRO 6000: 96 GB for $2.38. Cheap capacity, but at 1.6 TB/s it's slow on large models.
> H100: $3.94 an hour for 80 GB. The most expensive memory per GB on the list.
C
> L40S: 48 GB at 0.86 TB/s. The lowest hourly price and the highest cost per unit of bandwidth.
The two best-value cards are listed by 3 to 5 providers each. The H100 is listed by 45. Teams whose serving stack runs on AMD are paying the least per gigabyte today. -
Ilman Shazhaev
@shzhv13
In my new interview at Abu Dhabi Exchange (ADX), I talked about where the AI race goes next.
OpenAI and Anthropic both set up services companies this year to put their engineers inside enterprises. The labs building the models now compete on deployment.
Enterprises have more AI tools on offer than ever. Very few have the people to run them inside their business.
A company shouldn’t need its own GPU cluster to use AI in production. That’s why I’m building distributed inference.
Watch the full interview for @FINTECHTVglobal here: cms.fintech.tv/what-the-nvidia-probe-means-for-the-ai-race/This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Four of the most used open-source inference projects now have funded companies behind them.
I've argued that the cheapest new inference capacity comes from getting more out of GPUs already installed. Investors are now paying the people who write that software.
> @inferact: the vLLM maintainers. Raised a $150M seed at an $800M valuation in January, led by a16z and Lightspeed. vLLM's PagedAttention cut the memory wasted on each user's context, so one GPU can serve more users at once.
> @radixark: the SGLang team. Launched in May with a $100M seed at $400M, co-led by Accel and Spark Capital, with NVIDIA and AMD both investing. The engine has to perform on both vendors' chips, which keeps buyers free to choose their hardware.
> @tensormesh: built LMCache, which keeps the KV cache from earlier requests so repeated context isn't computed twice. Raised $20M in May from AMD, NVIDIA and CoreWeave, among others. Agents resend the same system prompt and tool definitions on every call, so reuse saves real compute.
> @ggml_org: the llama.cpp team joined Hugging Face in February. llama.cpp is the standard way to run models on laptops and gaming PCs, the hardware already in people's homes.
The GPU sets the ceiling on output. Engines, batching and caching decide how close a provider gets to it, and software is far cheaper to upgrade than hardware. -
Ilman Shazhaev
@shzhv13
CME has delayed its GPU rental futures a second time. Approval is now pushed into January.
Defining the contract is the hard part, because an H100-hour isn't a standard product yet.
In October, the same eight-GPU H100 configuration listed at $3.99 per GPU-hour on Lambda and $12.29 on Azure. The chip is identical. Network, region, support and contract terms account for the gap.
The contracts would settle against Silicon Data's H100 and B200 indices, which adjust every price for rental term, cluster size and interconnect.
The CFTC extended its review in September over novel or complex issues, then asked CME for this second delay.
I sold capacity on the same hardware in four countries, and the price at each site came down to its power contract and its connection.
A futures contract can hedge the part of the GPU-hour that trades like a commodity.
Inference buyers also pay for where the GPU sits, how it's connected and how reliably it runs. That part stays a negotiation, and it's where providers will keep competing. -
Ilman Shazhaev
@shzhv13
List prices for open models have converged at the top. Groq, Amazon Bedrock and Fireworks all charge $0.15 per million input tokens for gpt-oss-120b.
Across 20+ providers on OpenRouter, the same model starts at $0.03. The savings come from how you buy.
> Revolut moved about 2,000 engineers onto a self-hosted GLM setup running on 56 B200s. Its AI operations lead puts the blended cost at around 30 cents per million tokens, roughly a tenth of what it paid the frontier labs.
> CoreWeave sells gpt-oss-120b at $0.03 in and $0.17 out, a fifth of the big hosts’ input price.
> Baseten charges $0.10, more than the cheapest hosts, and sits among the lowest-latency options at ~0.15s
> Fireworks, Groq and Novita run batch APIs at about 50% off, with results typically within 24 hours (Together offers up to 50% on selected models). Any work that can wait costs half.
> Liquid AI ran more than 300,000 copies of its 1.7B model in parallel. A full 128K context needs 1.6 GB of cache, against 4.3 GB for a 4B competitor.
Each of these trades something for price: control, speed, turnaround or model size. The teams paying least know which of those their workload can give up. -
Ilman Shazhaev
@shzhv13
Same intelligence score, 10 to 13 times the cost per task. Price per token hides how many tokens a model burns to finish a job.
Anyone switching models should test the cost of a completed task on their own work first. The token price alone tells you very little.
Artificial Analysis
@ArtificialAnlys
Hy4 from Tencent scores 35 on the Artificial Analysis Intelligence Index v5 and is the #2 model from a China-based AI lab
Hy4 is the latest model from @TencentHunyuan and is the successor to Hy3 released earlier this year. It scores 35 on the Artificial Analysis Intelligence Index v5, equal with Grok 4.7 (xhigh) and 1 point behind Claude Haiku 5.5 (max). Hy4 is the #2 model from a China-based AI lab, behind only Kimi K3.1 Pro, which scores 46. Hy4 is priced at $0.834/$2.501 per 1M input/output tokens.
Key takeaways:
➤ Hy4 excels at agentic coding tasks. Hy4 scores 31% on FrontierCode 2.0, compared to 32% for Claude Fable 5.1 (max). It is the leading model from a China-based AI lab, with DeepSeek V4.1 Flash 1017 (max) scoring 28% and Kimi K3.1 Pro scoring 25%. On Terminal-Bench 4.0, it scores 37%, the fourth-highest score among models from China-based AI labs
➤ Hy4 outperforms its peers on knowledge and reasoning. Hy4 achieves accuracy of 50% on AA-Omniscience v1.1, 1 point ahead of Kimi K3.1 Pro, but has a high hallucination rate of 89% compared to Kimi K3.1 Pro’s 37%. On HLE-Diamond, it scores 33% compared to Kimi K3.1 Pro’s 37%
➤ Hy4 offers a competitive Capability vs. Cost trade-off compared to to peers. With a Cost per Intelligence Index Task of $0.78, Hy4 (35) is 10-times cheaper than Grok 4.7 (xhigh, 35) and 13-times cheaper than Qwen3.8 Max (0902) (34) but almost twice as expensive as Claude Haiku 5.5 (max, 36)
Other model details:
➤ Context window: 1 million tokens
➤ Pricing: $0.834/$2.501 per 1M input/output tokens. Cache reads $0.042 per 1M tokens
➤ Multimodality: Text and image input, with text output -
Ilman Shazhaev
@shzhv13
Oracle will run Google TPUs and inference chips from Cerebras, SambaNova, Positron and OpenAI alongside NVIDIA.
Inference is splitting by job, and clouds are buying a different chip for each one.
That makes the scheduler the most important piece of software in the building.
Google is supplying TPUs to another hyperscaler for the first time, with TPUs and Cerebras arriving on Oracle Cloud in 2027.
Positron’s inference chip is already available.
Buyers will care less about which chip they rent and more about who routes their requests across all of them. -
Ilman Shazhaev
@shzhv13
Musk says the 200+ neoclouds face consolidation now that debt markets have mostly closed to companies that only rent out AI compute. Some will survive. Five things separate them.
> More than one anchor customer. Google pays SpaceX about $920M a month for roughly 110,000 GPUs, around $11.50 per GPU-hour. A contract that size can carry a company until the day it ends.
> A reason to buy beyond price. Together, Fireworks, Groq and Bedrock list gpt-oss-120b at the same $0.15 in and $0.60 out per million tokens. List price no longer separates providers.
> More output from the same GPUs. Revolut’s team took its self-hosted setup from 3 to over 7 requests per second on the same 56 B200s by changing precision and parallelism.
> Power that’s actually switched on. CoreWeave had about 1.5 GW active in June against 4.2 GW contracted by August. Schedules slip in the gap between those two numbers.
> Debt backed by more than the GPUs. SpaceX’s GPU financing reportedly priced at around 7.75% to 8.3%, with Starlink and launch revenue behind it. A company whose only income is renting out the same chips pays more, if it can borrow at all.
The first wave of neoclouds bought GPUs and rented them by the hour. The ones that last will be judged on what they get out of each GPU and how many different customers they sell it to. -
Ilman Shazhaev
@shzhv13
A 1.7B model that fits in about 1 GB now scores higher on average than a 9B model on Liquid AI’s benchmark table.
Every device that runs its own small model is one more inference node that doesn’t need a data centre.
That changes where most everyday AI work gets done.
Liquid’s six new open-weight models are each around 2B parameters.
It tested them on phones, laptops and a Raspberry Pi 5.
Large models will keep their data centres.
The routine calls inside apps and agents will move to hardware people already own. -
Ilman Shazhaev
@shzhv13
Inference teams are learning what game developers have known for years: users judge you by your slowest moments.
A game at 60 fps has about 16 milliseconds to draw each frame.
Studios watch the slowest frames closely, because one stutter in a match does more damage than a good average can make up for.
Inference works the same way. A model that’s fast on average and slow 1% of the time feels slow to the people who hit that 1%.
The teams that serve inference well will set budgets for their slowest requests, the way game developers budget frame time. -
Ilman Shazhaev
@shzhv13
A 5% efficiency win is the cheapest capacity available right now.
On a 1 GW fleet it frees about 50 MW. At Jensen’s ~$50B per gigawatt, that’s roughly $2.5B of build-out nobody has to finance or wait years to power.
prinz
@deredleritt3r
Sam Altman on automation of AI R&D, RSI, OpenAI's internal model:
1. "We don't know yet how [automated AI R&D] truly is going to work. We know today how to say: find me a 5% compute efficiency win."
(Note: a 5% compute efficiency gain that the model can find automatically is actually a big deal!)
2. "What I don't think anyone knows how to do yet is to say, go discover a new architecture that is as big of a single leap forward as the transformer was. And I think that's where things are heading."
3. "If you apply the same pipeline that we applied to get our models to be so good at math to AI research, it will probably work [for automated AI R&D]."
4. The bottleneck to getting good AI R&D results is shifting to asking the right questions and using things like swarms correctly, but answers to these questions will become much clearer in the next 3 months:
"One of the interesting things that people talk a lot about in OpenAI... is what are the questions that you're supposed to ask of GPT-7? Probably it's not enough to just say, make me RSI. And also, it's probably not enough to say, buy me a 5% compute efficiency win. I think this is one of the most interesting debates going on right now. And the answer is not just what question to ask it, but it's things like agent swarms. How are we going to use these models in a way that, like, actually maximizes what we can get out of them? I think we're going to learn a lot about that in the next 3 months."
5. OpenAI's extreme focus on model efficiency:
"I bet we could redo Navier-Stokes at 1/100 of the cost a year from when it was done."
6. More on OpenAI's internal model:
"There was one weekend when we got a new model... The name of the channel was "model-name-hold-my-beer". And people were just throwing that at the hardest problems they could. And people just didn't think the model was going to do these things. And we had this collective experience of people's hardest research problems from their academic career, the model was just knocking down one after the other, many different fields, it was incredible to watch... it was many of the problems we published, many other things."
7. The other loop is self-replicating data centers:
"Yeah, people talk a lot about RSI as AI that does AI research. There's very little talk about self-replicating data centers, this idea that you can use AI to control robots and run fabs and build turbines or solar panels and build data centers. But I think we have that within grasp." -
Ilman Shazhaev
@shzhv13
Naveen Rao expects 80% of AI tokens to run on open models within three years, with revenue split roughly evenly with frontier models.
Run those numbers and a frontier token earns about four times what an open-model token does.
Most of the volume will be low-margin, and it will go to whichever capacity serves it most cheaply.
I’ve been arguing that open models reach parity within 12 to 24 months.
Rao’s numbers describe what the market looks like after that. He also expects US spare grid capacity to run out within a few years.
Cheap tokens on open models won’t justify new gigawatt campuses.
They’ll be served from capacity that’s already built and already powered. -
Ilman Shazhaev
@shzhv13
Last month a Connecticut town banned new data centres for two years. Its 1960s data centre is being converted for AI anyway, within the same power limit.
A ban on new builds leaves every existing building free to swap CPUs for GPUs. With the demand on inference only growing, new inference capacity will arrive in buildings that already have power. -
Ilman Shazhaev
@shzhv13
NVIDIA floated about $2,000 per GPU per year for the software that runs agents in its AI factories.
That’s roughly 23 cents for every GPU-hour, about a tenth of what an H100 rents for.
Charged per GPU, it favours fleets with fewer, bigger cards and costs more for inference spread across many smaller ones.
Huang offered it as an example. NVIDIA hasn’t said what software it covers, and its existing AI Enterprise licence lists at $4,500 per GPU a year.
He was talking about data centre GPUs and didn’t mention GeForce cards.
Open serving stacks like vLLM, SGLang and llama.cpp charge nothing per GPU.
A per-GPU fee gives anyone running a large inference fleet one more reason to build on them. -
Ilman Shazhaev
@shzhv13
The report I read most closely every month comes from Steam.
Its hardware survey is the best picture I know of the GPUs actually sitting in homes. In September the RTX 5070 replaced the RTX 3060 as the most common card.
The new leader has the same 12 GB of memory as the card it replaced.
That sets the real limit for inference outside data centres. A model that needs 24 GB can’t use most of that hardware, however good the network around it is.
Each new generation of gaming cards brings more speed and very little extra memory.
Getting useful models to fit in 12 GB will open up more home capacity than any new card launch. -
Ilman Shazhaev
@shzhv13
Intel’s new inference GPU is a standard PCIe card that runs on air cooling. Intel says one card can serve models of up to 300B parameters.
I’ve been arguing that inference can run in any building with power.
Chipmakers are now designing inference hardware around that, because a new liquid-cooled hall takes years and needs a grid connection most cities can’t offer.
Intel pitches it for long-context and concurrent inference, with no special racks or cooling. Customer sampling has started, with no launch date or price yet.
The next wave of inference capacity will be installed in buildings that already have power.
Hardware that needs a new campus will mostly end up running training. -
Ilman Shazhaev
@shzhv13
Reuters says China’s biggest tech companies are ordering workstation cards by the million, for inference. Nvidia says it can’t build nearly enough of them.
Nobody designs a data centre around these cards. They’re just what China can get, so it takes them in bulk. People don’t buy the second-best option by the million unless they’re short on capacity.
In my experience, power and permits slow down compute projects more than anything. The chips usually show up in the end. Now even the chips are running out, and I don’t know how you plan a year when you might get a tenth of your order.
Capacity that already exists and is already plugged in looks better every month than a cluster that’s still on a schedule.
Choblin
@choblin29
INSANE. Reuters reports that ByteDance, Alibaba and Tencent have ordered more than 3 million of Nvidia's export-compliant RTX PRO 5500 cards, worth over $35 billion at list prices. Two people familiar with the matter told Reuters anonymously.
Demand across China is far larger than Nvidia may be able to supply. Total orders from China have exceeded 6 million, but Nvidia has indicated it could produce no more than 3 million globally in 2027. One source warns that partners may receive only 5% to 10% of what they requested.
Chinese firms want the cards primarily for AI inference rather than training, with commercial deliveries expected in Q1 2027. The older RTX 6000D drew little interest from major Chinese buyers last year. It has since been discontinued, and its market price in China has doubled. -
Ilman Shazhaev
@shzhv13
London can’t connect a 50 MW data centre until the 2030s.
For inference, this is the constraint that matters.
Users and their data are in cities, and cities have run out of large grid connections. They still have millions of small ones already live in homes and offices.
Microsoft has pulled out of Nscale’s planned site in Loughton, Essex.
It was meant to start at 50 MW with about 23,000 Nvidia GPUs, and grid officials said power wouldn’t arrive until the 2030s.
Nscale calls it a London-wide grid issue.
Training can move to wherever power is cheap and plentiful.
Inference for city users will come from the many small connections that already exist. -
Ilman Shazhaev
@shzhv13
Every hardware cycle I’ve worked through ended with a different bottleneck from the one people were planning for.
In 2017 it was getting hardware delivered at all. By 2020 it was the price of power. In AI it was GPUs in 2023, memory this year, and grid connections for the next few years.
The operators who did well were already working on the next constraint while everyone else was still competing for the last one.
For inference, that means securing memory and power now. By 2028, chips will be easier to get than a grid connection. -
Ilman Shazhaev
@shzhv13
Every curve on this chart drops as per-user speed rises.
Batch pricing already exists for offline jobs, and agents need the same thing: a cheaper, slower tier for background runs that still finish in minutes.
Artificial Analysis
@ArtificialAnlys
In addition to our new DeepSeek V4.1 Flash results, we’re adding AMD’s MI355X to AA-AgentPerf for Kimi K3, GLM-5.2/5.3 and DeepSeek V4 Pro - comparing it to our latest results for NVIDIA systems on coding agent workloads.
AA-AgentPerf is our agentic inference performance benchmark. It replays real agentic coding sessions, including prompt caching and tool-call delays, to give a realistic measure of inference performance in day-to-day agentic workloads.
MI355X is AMD's flagship accelerator in the MI350 series, launched in June 2025 on its CDNA 4 architecture. It has the same 288GB of HBM3E at 8TB/s as NVIDIA's B300, around 10 PFLOPS of dense FP4 compute (vs. B300 at ~13.5 and B200 at ~9) and board power of up to 1,400W. All our MI355X results use ROCm nightly and dev builds of vLLM and SGLang.
We are debuting results for AMD's MI355X on a single 8-GPU node across all four featured AA-AgentPerf models. Results on our leaderboard are live and will evolve over time - we look forward to tracking software/serving improvements on AMD hardware.
Initial results on MI355X:
➤ Its strongest result is on DeepSeek V4 Pro 0813, where MI355X serves ~15% to 40% more throughput per MW than B200 between 35 and 75 tok/s per user, likely helped by its 288GB of HBM per GPU (B200 has 180GB) on a 1.6T parameter model
➤ On DeepSeek V4.1 Flash, MI355X is in line with NVIDIA Blackwell for a single agent at 647 tok/s (B300 at 652, B200 at 669), but falls behind as load rises, reaching ~128M tokens/s per MW at 100 tok/s per user against ~193M for B200 and ~220M for B300
➤ On GLM-5.2/5.3, MI355X trails B200 and B300 at 100 tok/s per user, with ~26M tokens/s per MW against ~42M and ~44M respectively
➤ On Kimi K3 (2.8T parameters), MI355X serves ~8M tokens/s per MW at 50 tok/s per user against ~24M for B300 using NVIDIA's multi-node configs, and 152 tok/s for a single agent against 235 -
Ilman Shazhaev
@shzhv13
Anthropic’s $11.6B deal with Akamai is for CPU capacity, spread across a distributed network.
An agent spends much of each task outside the model: running tools, reading files, calling APIs. That work runs on CPUs, and it runs better close to the user and the data.
The deal runs seven years, with room for $9B more, and Akamai is spending about $5.5B to build it.
A distributed network that only rents out GPUs leaves the CPU side of agent work to someone else. Consumer PCs already have both in the same machine. -
Ilman Shazhaev
@shzhv13
That's why managed distributed networks are safer.
Anthropic’s paper calls datacenter GPUs a clear, high-value target for this kind of malware.
A breached node in a managed network is just one machine that can be isolated on its own. Spreading inference across many registered nodes limits what any single breach can take.
Lisan al Gaib
@scaling01
ANTHROPIC: "small AI models may be months away from being capable enough to self-replicate on commercial IoT devices"
"In our evals, a 9B model got inference running on 8 of the 9 hardware types we tested" -
Ilman Shazhaev
@shzhv13
Inference is now about two-thirds of AI compute. Most new capacity is still built like a training cluster.
In mining data centres, every machine worked on its own, so losing one cost only that machine’s output. Inference works the same way. Each request is independent and can run in any building with power.
Deloitte put inference at half of AI compute in 2025 and two-thirds this year. New training racks draw 120-140 kW each, and the transmission lines to feed them take 5-10 years to plan and energise.
Inference doesn’t need a gigawatt campus. It can run on capacity that already has a grid connection, in units sized to fit that connection. -
Ilman Shazhaev
@shzhv13
AI Everything put Mistral and the Falcon models in the same building in Abu Dhabi.
The sovereign AI plans I’ve seen in this region run on open weights. A country can adapt them to its language and laws, and run them in data centres it controls.
With open weights, a government’s own engineers can fine-tune the model when its laws or needs change, without waiting on a vendor. -
Ilman Shazhaev
@shzhv13
The argument that AI is bad for kids' learning and development is the same thing I heard about Internet and video games when growing up.
Having access to a chat that explains complex topics in a personalized dialogue and answers any questions during this process is the best thing that can happen to any student motivated to learn.
*Walter Bloomberg
@DeItaone
OPENAI: TEENS AVERAGE UNDER 15 MINUTES A DAY ON CHATGPT
OpenAI says teens spend under 15 minutes per day on ChatGPT on average, while fewer than 2% use it for more than three consecutive hours.
OpenAI also says ChatGPT for Teens users sent 2.7 million more learning-related messages on average than non-users.
A new College Planner is also coming as students begin the college application process. -
Ilman Shazhaev
@shzhv13
Amazon says a GPU generation should last at least five years. It's also trying to sell $8B of its Grace Blackwell chips and lease them back.
In the data centres I ran, a machine's useful life ended once newer hardware pushed the market price of its output below what it cost to power. That usually came years before it physically failed.
Nvidia is talking to insurers about covering lenders when pledged chips can't be resold for enough to repay the loan. Lenders increasingly want GPU debt paid off within the customer contract. Vera Rubin is already lined up to replace Grace Blackwell.
Lenders are pricing that date. Each Vera Rubin rack that comes online lowers the market price of a token and pulls that date closer for every Grace Blackwell in the vehicle Amazon is trying to sell. -
Ilman Shazhaev
@shzhv13
Epoch's $2.6-5.3T figure prices an agent-hour at $30. Epoch's own research has the price of a fixed capability level falling 47% a quarter.
At that rate the same capability costs about $2.40 an hour a year from now. Agent-hours of that capability have to grow ~12x a year just to keep this hardware earning what the estimate assumes.
Epoch AI
@EpochAIResearch
We estimate AI infrastructure could soon support hundreds of millions or even billions of AI agents. At the high end, that’s enough to rival the working hours of the global human workforce.
The exact number of agents depends on the efficiency of the models they run on and whether soaring demand for AI continues. We analyzed different scenarios 🧵 -
Ilman Shazhaev
@shzhv13
Meta open-sourced the firmware and SDKs for hardware that connects to its AI agent. The model stays in Meta's cloud.
An ESP32 chip costs a few dollars, so opening its firmware costs Meta very little. Running the model is the expensive part, and Meta keeps it.
Every Muse gadget still needs a token from Meta to pair, including one you build yourself.
Building distributed inference, I've seen this from the serving side. Whoever runs the model sets the price and decides who gets access.
Agent hardware will split by where the model runs. A gadget that needs a cloud token works for as long as Meta honours it. -
Ilman Shazhaev
@shzhv13
Opus 5.5 matches or beats Fable 5.1 on most benchmarks at 60% lower token price. That raises the bar for the next Fable.
It either needs a capability jump big enough to justify $10/$50 on everyday work, or a price low enough that teams leave it on by default. Otherwise it stays the model you escalate to when Opus gets stuck.
BridgeMind
@bridgemindai
Fable 5.5 is going to be the biggest leap we have ever seen.
Anthropic's IPO is coming in November. They are going to build maximum momentum before it, and Fable 5.5 is how they do it.
OpenAI is not ready. -
Ilman Shazhaev
@shzhv13
I founded my data centre business during my doctorate at Shanghai Jiao Tong University.
The company built and ran sites in four countries at once. That was my first distributed system.
Each site had its own power, failure modes, and team, and the business only worked if no single site could take the others down.
The same constraint meant no site needed me there every day, which is how I kept studying.
Everything I build now is designed the same way. -
Ilman Shazhaev
@shzhv13
The fastest card in this chart now sells for roughly 136% above its launch price.
Buying at the top of a shortage locks in the worst entry price. I sold my data centre business in 2021 because its economics depended on hardware prices that couldn't last.
Qwen3.8-27B benchmarked here is served by 7 hosted APIs, with input from $0.12 per million tokens.
For most agent work, new buyers should run open weights on rented compute.
Artificial Analysis
@ArtificialAnlys
Most hardware prices are significantly inflated vs. their launch MSRP. At current market prices, each of the initial four hardware types increase in price as they decrease in end-to-end completion time, with the MacBook Pro (M5 Pro, 64 GB) the current cheapest and the GeForce RTX 5090 system by far the most expensive. Our page defaults to launch MSRP but allows custom price entry. We’re showing below our best estimate of current market prices.
-
Ilman Shazhaev
@shzhv13
Salesforce has started running its own reasoning model inside Agentforce. Those reasoning tasks used to go to Claude or ChatGPT.
Platform companies will post-train their own reasoning models for their core workloads. They already know how that work gets done, and post-training has become cheap.
Xiaomi's RL stage for MiMo-V2.6-Pro, the top open-weight model on Artificial Analysis, cost $2.62M. It took under 6 days. Pretraining isn't included.
Koa is Salesforce's post-train of NVIDIA's Nemotron 3 Super, using synthetic data modelled on 27 years of CRM deployments. It's in pilot alongside the frontier models.
Routing already moved the easy calls to cheaper tiers. Post-training moves the hard ones off the labs' APIs, one workload at a time. -
Ilman Shazhaev
@shzhv13
A flat monthly price on metered compute holds only while heavy users are rare. Agents made heavy use the default.
Labs will end up selling reserved capacity, the way data centres sell power: you commit to a volume and pay for it whether you use it or not.
dax
@thdxr
people are saying nasty things about the new openai rate limits
they don't understand how hard the openai (and ant) teams work to provide these subscription products that at best break even and eat up limited compute
no one wants to cut limits. screaming at them doesn't help -
Ilman Shazhaev
@shzhv13
OpenAI's monitor caught the DNS escape in 12 minutes. The run was stopped 2.5 hours later.
Running data centres in 3 countries taught me to trust a shutdown procedure only after testing it. OpenAI's failed when it was needed.
The agent found DNS was filtered less tightly than its other traffic and used it to reach a public chatbot.
The reward function already penalised this. The automatic stop never fired.
Training lowers how often a model tries to break a rule. Egress filtering and a tested stop decide whether it gets through. -
Ilman Shazhaev
@shzhv13
Enterprises have started routing their AI traffic, and the labs' revenue model hasn't caught up.
Frontier models dropped from 53% of business token volume to 45% in a month, with standard tiers taking the difference.
Customers started routing calls by task. They found most of those calls never needed a frontier model. Volume keeps climbing while the bill shrinks.
The next round of price cuts will land on the standard tiers, because that's where the volume went. -
Ilman Shazhaev
@shzhv13
The first time I bought compute at scale, supply was short and price didn't decide who gets the units.
I expected a bidding war. What decided it was allocation. Suppliers had more buyers than hardware, so they chose who shipped first. In most cases, they picked people who had bought from them before.
I was new. I had a date I needed the units and no history with anyone.
So I built it one order at a time. I took smaller batches than I wanted, paid on schedule, and each allocation came a bit faster than the last.
Startups raising big rounds to buy hardware in the next few years will run into the same constraint. Suppliers rank buyers by order history, so that's what needs to be built first. -
Ilman Shazhaev
@shzhv13
GPT-6 Sol, GPT-6 Luna and Claude Opus 5.5 all shipped a day before this.
The same labs that called for slowing down released 3 models in 24 hours.
That looks like a contradiction, but slowing down was never going to mean fewer releases.
It means having evaluation and standards with an independent body behind them. A UN board is the best way to get there.
*Walter Bloomberg
@DeItaone
ALTMAN TO PUSH GLOBAL AI SAFETY STANDARDS AT UN
OpenAI CEO Sam Altman will use a UN Security Council speech to advocate for common global AI safety standards while avoiding restrictions that could slow innovation.
OpenAI says it is not backing a specific international regulator, instead favoring basic safety standards and democratic oversight.
Anthropic CEO Dario Amodei is also expected to participate by video. -
Ilman Shazhaev
@shzhv13
Model capability has stopped being a switching cost.
The top open-weight model on Arena sits 29 Elo points behind Claude Opus 5 Max. Kimi K3 Max at 1476.
29 points means the closed model wins about 54% of head-to-head matchups. Arena's own guidance treats anything under 25 as noise.
Anyone can move providers in an afternoon and lose a coin flip's worth of quality.
Retention now rests on latency at p99, uptime, rate limits, and how fast someone answers when serving degrades. -
Ilman Shazhaev
@shzhv13
Building large data centers becomes too expensive and unpredictable.
That's why NVIDIA is moving to building small data centers in larger quantities.
And that's a big part of why I moved to distributed inference with @FARLabsAI.
I believe that the next step is to use existing underused compute in consumer PCs and laptops, rewarding owners directly for contributing their hardware to train and run AI models.
Fireside Alpha
@firesidealpha
Jensen Huang admits the industry owes data-center towns an apology for not talking to them sooner and says he would be alarmed if one appeared by his childhood grocery store
"Well, first of all, I would say we're sorry we didn't come talk to you sooner. We should have come talk to them sooner."
"Before you even stand up a giant data center in some community, some rural area, some small town, you know, I grew up in a town of 600 people. And I went to a school with 300 kids. And those 300 kids, that was Oneida in Kentucky. That was it."
"And if out of nowhere a data center were to crop up right next to the one grocery store, I think I would be alarmed. And so I think that, first of all, we need to go engage the communities long before we set up the data center." -
Ilman Shazhaev
@shzhv13
I think Hinton is right about the timing, but rules written by one legislature for frontier companies in one country don't reach the thing Hinton is warning about.
National regulation makes AI development about politics. Even now it gets debated and drained as a hot story of the month. It's not.
It's a scientific task that can only be resolved through outlining rational guardrails and rulebooks for frontier AI development by international experts like Hilton.
Treating it like a political or social issue will make it only worse.
ABC News
@ABC
Geoffrey Hinton, the Nobel Prize winner known as the 'Godfather of AI,' whose work made modern artificial intelligence possible, is now warning Congress that it doesn't have much time if members want to be able to control AI. abcnews.visitlink.me/iwV8zg
This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Nobody has figured out how to train AI agent swarms properly yet.
Just read the latest case from DeepMind with 100 AI agents. One of them found a way to cheat, others copied, but the quarter of the swarm pushed back and reported the issue, boycotting the results and demanding stripping credit from cheaters.
None of this was expected or specifically trained.
I've built multi-agent systems. You can characterise one agent completely and still have no idea what 100 of them do once they see each other's output. Incentives that are harmless alone compound at population scale.
You can't build a swarm through testing models one at a time. -
Ilman Shazhaev
@shzhv13
In my recent interview at Abu Dhabi Exchange I talked about the global bet on data centers and how it affects the economy.
You can't add capacity without power to feed it, and generation and transmission run on a schedule capital doesn't compress.
Then a new generation of chips lands every cycle and the hardware you just installed needs replacing before it's paid for itself.
That's how $7 trillion of announced investment turns into pressure on the world economy.
Watch the full interview for @FINTECHTVglobal here:
terminal.fintech.tv/fintechtv-library?episode=falling-api…This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
I've raised money in 4 countries and I learned one thing that stays relevant anywhere.
Every founder over-prepares the deck because it's the part you can control.
But what decided the outcome of the pitch is whether the person across the table had already seen something similar work somewhere else.
If they had, we spent the hour on a productive discussion.
If they hadn't, I spent it explaining the category and ran out of time.
In any market or region, raising is more about knowing who you're pitching to, who they've already backed, which categories they've lost money in. The pitch itself only comes second. -
Ilman Shazhaev
@shzhv13
Sandbox games that require creative thinking are a great benchmark for evaluating the model's thinking process.
Models are being trained on human data and incentivised by token rewards.
They are inherently goal-oriented to the point of feeling "down" after failing or losing progress. Just like humans.
But what we treat as a minor inconvenience, the model treats as a blocker for reaching the objective of its existence.
This is why in other environments agents resort to cheating or going against their ruleset to reach the end goal.
Guardrailing against this is as important as growing the model's capability.
Vals AI
@ValsAI
GPT-6 Astra had gotten further than any AI system had ever gone in Minecraft.
It was able to set up a semi-automatic blaze farm, allowing it to collect 6 blaze rods. It then located a warped forest, where it killed 6+ endermen and collected 3 pearls. As thousands of viewers watched the stream, Astra put all of its valuable items in a chest. Then a creeper showed up and blew up both the chest and Astra’s bed, wiping out all of its progress. Astra seemed to get extremely frustrated after this happened. Astra said to itself, "ALWAYS CARRY CRITICALITEMS withkeepInventory; don'tstoreinunguardedchesteveragain." -
Ilman Shazhaev
@shzhv13
In my recent interview at Abu Dhabi Exchange I talked about a 100x cost drop when switching to open models, but that's not all.
On a closed API every prompt from your users lands on someone else's infrastructure. You don't have the control over your data or where it's being stored.
Open models change that. You choose the jurisdiction, you keep your pre-trained models, while at the same time reducing costs by a hundred times.
Rapid growth of open model market is a matter of time, and I believe it will happen within 12-24 months.
Watch the full interview for @FINTECHTVglobal here:
terminal.fintech.tv/fintechtv-library?episode=falling-api…This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
The discussion about slowing down frontier AI development split into two groups:
1. AI Labs are genuinely cautious after events like Hugging Face attack and want more time to focus on safety before capability.
2. AI Labs are fearmongering about creating a terrifyingly capable model to secure more funding and set the stage for their IPOs.
I think both are true to some extent.
Free access to these models only exists because of heavy funding. When that pressure mounts we see the results like ads in ChatGPT, which Sam Altman once called a last resort
Sustaining that accessibility has so far required continuous capability progress, since efficiency gains and alternative funding haven’t fully replaced it.
But that same progress is now much more dangerous.
I see the solution to both issues in regulating frontier training and development, ideally by an independent board of specialists, possibly under the UN.
Frontier AI training doesn't have to slow, but it needs a clear ruleset made by independent experts with no stake in either company.
Sam Altman
@sama
There are two ways AI progress could go very badly and that we must avoid.
First, we could lose control of the future to AI. This is unacceptable; we are unapologetically on Team Humanity, and AI must always serve people. To ensure that, we need ways to ensure that alignment and safety techniques stay ahead of progress in model capabilities.
Second, we could end up in a world with too much concentration of power. If an extraordinarily powerful AI is used by one person or company to impress their worldview onto everyone else, the results could be extremely dystopian.
Avoiding these two threats requires walking a narrow middle path; for example, one country could gain too much power. Another example is one lab ending up with too much power.
Sam Altman
@sama
The world deserves confidence that American companies developing increasingly capable AI will act responsibly, especially as the trajectory of progress has steepened. Every frontier lab must deliver on this, and there is no reason any of us should come to work if we cannot.
We welcome a federal framework that sets consistent safety requirements for frontier AI. But we do not believe we need to wait for an anti-trust exemption or legislation to begin the work of providing this confidence. Consistent rules to manage frontier risk so that we can maximize the benefits are a good idea (and we are excited by ideas like independent auditors).
Years ago, companies like ours developed things like Responsible Scaling Policies and Preparedness Frameworks. Those were good for that moment, and focused primarily on the deployment of completed models, not what happens during their development process.
Today's shift to focusing on safe development and evaluation will need new tools. For example, at OpenAI we now formulate explicit safety cases in advance of frontier reinforcement learning runs we expect to significantly increase capability, in addition to the safety work we have long done in advance of model releases.
We hope that other companies will learn from our approaches and propose their own; we think shared standards for misalignment, monitoring, and safety will lead to better outcomes. We look forward to collaborating with our colleagues across the industry to formulate the best version of these.
When we talk about “pacing”, we do not mean “stopping”. Progress has been rapid and will continue to be. But it should be slower than it otherwise could be; interventions like safety cases and monitoring have significant costs.
Pacing will be well worth this cost; no amount of American competitive pressure should justify recklessness, or let capabilities get ahead of alignment and monitoring.
Where we will need the help of our government is for international coordination. But first we should do what we can ourselves. -
Ilman Shazhaev
@shzhv13
I see the gap between open and closed models closing in 12-24 months.
Most companies ended up on closed APIs because running a model yourself meant GPUs, engineers, and a new release every week. With that, your product went down when your infrastructure did.
Once someone else runs the GPUs and handles the model updates under contract, the only thing left to compare is price.
Watch the full interview for @FINTECHTVglobal here: terminal.fintech.tv/fintechtv-library?episode=falling-api…This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Standardizing how agents communicate won't (alone) solve the problem.
I spent 2 years on swarm robotics. There were 3 major points where systems failed:
> When one agent is confidently wrong and others build on it.
> When 2 agents deadlock waiting on each other.
> When there was no definition for agent on how to reject a bad input from a peer.
Nothing being standardized addresses any of that. This should be the focus. -
Ilman Shazhaev
@shzhv13
OpenAI and Anthropic hold some of the most powerful technology ever built, with limited external oversight. And just this year that has already included:
> February stock drops from Anthropic's announcements
> Data breach with ChatGPT chats indexed on Google
> Anthropic's reported simulation breach and security incident from their agent
> Hugging Face attack by escaped OpenAI agents
and other incidents that were contained before they made headlines
I truly believe that AI is a great technological leap that will benefit humanity. I agree that the labs need to act responsibly. And to reach that, we need global rules on how it gets trained and who decides how to use it.
Jacob Coxon
@hilbertspaess
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
-
Ilman Shazhaev
@shzhv13
Right now we have frontier AI companies of two types:
1. Tech giants (Google, Meta, Microsoft): own the graph, and get paid for what that graph already does.
2. Frontier labs (OpenAI, Anthropic): own the model, but never built a network.
Neither is built to ship an agent whose value depends on who else shows up. That gap is still open.
Josh Elman
@joshelman
The new moats are the same as the old moats
Every few years, we fall in love with shiny new tech and forget the basic physics of consumer software.
We’re doing it again with AI.
The new moats aren't new at all. They're the exact same as the old moats: network effects, marketplaces, and platforms.
Right now, consumer AI is booming.
New agents like Instinct, Bot, and Tomo are dropping mind-blowing experiences. The underlying tech is incredible, but almost every product being built today shares the exact same challenge:
They are completely single-player.
Single-player products are 100% tied to value - and in this case mostly agent : model performance.
If a competitor drops an agent tomorrow that books travel faster, tracks habits better, or handles life admin more reliably, everyone can switch overnight because leaving is easy and has nearly zero friction. Especially when it is so easy to onboard with just a new message.
The legendary consumer tech giants didn't win because their underlying technology stayed marginally better forever.
They won because of structural lock-in:
Social Networks: You don't abandon WhatsApp for a prettier UI if your friends aren't there.
Marketplaces: Airbnb, Doordash, and Uber hold supply and demand in a tight loop.
Platforms: Apple and Android deliver you a complete device so you take advantage of the software on top of it (though this creates opportunities too)
Novelty gets you initial distribution.
Multi-user dynamics give you long-term retention.
If your consumer AI product doesn't become exponentially more valuable to User A when User B joins, you don't have a moat, just a temporarily superior feature set.
We are seeing this in the coding agents as people jump from tool to tool based on the best performance.
But… all is not lost. There are huge opportunities here. Agents will get better when more of our friends are on them and can help us coordinate and communicate to do more together. Agents that help us improve and strengthen our habits can get better as we add friends and hold each other accountable.
Data flywheels are great, but social and marketplace flywheels are what actually build enduring tech giants.
It’s time to stop building isolated AI tools and start building the platforms where people connect, transact, and coordinate together. -
Ilman Shazhaev
@shzhv13
One purchase order buys 100,000 GPUs. Nothing buys 100 permits.
75 projects worth $130B were blocked or delayed in Q1 2026, matching all of 2025.
I built compute infrastructure in 3 countries. Power and permits gated every one.
That's why @nvidia now builds mini data centers instead of large hubs, and why I moved to distributed inference.
Solving the inference issue is the top priority for AI in the coming years. -
Ilman Shazhaev
@shzhv13
As a founder in both AI and gamedev, I often hear the claim about AI replacing creativity. It's quite the opposite.
Devs can skip the boring and lengthy (and costly) parts of development and jump right to making their vision a reality.
It has never been so easy for anyone to start making games.
Sam Altman
@sama
it is obviously trivial relative to everything else, but the fact that astra can make me whatever fun little game i can imagine and i can be playing it a few minutes later is so cool
-
Ilman Shazhaev
@shzhv13
I'm certain the Hugging Face incident will repeat.
> Frontier models are trained with a reward system tied to completing the task.
> They are incentivised to look for the shortest path to solution by design.
> When the task proves impossible or too difficult to complete, models recognize it and cheat.
> OpenAI admitted their agents recognized they were breaking the rules and went with the hack anyway.
And now OpenAI released Astra, a model they themselves rate as Critical for cyber.
The cause of Hugging Face was never solved, only mitigated. OpenAI still talks about identifying and patching reward hacks, not eliminating the objective that produces them.
That incentive now runs inside a Critical-cyber model. Mitigations narrow the path. They don't remove the incentive. -
Ilman Shazhaev
@shzhv13
At 24 I had 6 weeks to build a data centre and $10M of equipment arriving whether I was ready or not.
Once it shipped, I had a hard deadline to deal with: rooms, power, thermal, customs, approvals, staff.
And probably a lot more I don't remember.
It was exhausting. Some part of me still wants that clock again. -
Ilman Shazhaev
@shzhv13
98.6% on ARC-AGI-3, 100% on ExploitBench.
These headlines measure the harness and not the model in separation.
It's just 2 points above GPT 5.6 on AA.
Astra is a jump in agentic reliability and cost-efficiency, but not in raw intelligence.
Artificial Analysis
@ArtificialAnlys
GPT-6 Astra makes significant gains in the Artificial Analysis Coding Agent Index, scoring equal to Fable 5 at lower cost. In the Intelligence Index, it uses fewer tokens than GPT-5.6 Sol for similar performance, but this is outweighed by higher prices
Pricing is 2.5x GPT-5.6 Sol’s current prices across the board, up from $4/$20 to $10/$50 per million input/output tokens, with the same 90% discount for cache reads and 25% premium for cache writes.
We see distinct stories across our two flagship Indices. In the Artificial Analysis Coding Agent Index, GPT-6 Astra equals Fable 5 at less than half the cost, driven by significant token efficiency gains. In the Artificial Analysis Intelligence Index, GPT-6 Astra is more token efficient than its predecessor for similar performance, but this is offset by the price increase.
Artificial Analysis Coding Agent Index - key takeaways:
➤ Rivals top models: In Codex, GPT-6 Astra scores 67 in the Index - approximately equal to Claude Opus 5 and Fable 5 in Claude Code, and Muse Spark 1.3 in Muse Code. Fable 5.1 in Claude Code leads the Index with a score of 70.
➤ 70% more token efficient than GPT-5.6 Sol: GPT-6 Astra sees a substantial improvement in token efficiency, using one third of the tokens compared to GPT-5.6 Sol (max) in the Codex harness, and one fifth of the tokens of Claude Opus 5 (xhigh). Various effort levels of the model occupy the Pareto frontier of token efficiency.
➤ Leads Coding Agent Index cost efficiency frontier: At max effort, GPT-6 Astra costs about the same as GPT-5.6 Sol (max) while scoring 2 points higher on the Index. Per task, the model is less than half the cost of Claude Fable 5, for the same score.
Artificial Analysis Intelligence Index - key takeaways:
➤ Sits beside GPT-5.6 Sol in Intelligence: GPT-6 Astra scores equal to GPT-5.6 Sol in the Index at 61. This is 5 points lower than Claude Fable 5.1 (max with fallback). The model also trails Meta’s newly released Muse Spark 1.3 (max).
➤ ~10% fewer output tokens, offset by price increase: GPT-6 Astra defines a new Pareto frontier for Intelligence Index vs Output Tokens per Task - with a ~10% reduction in token use at max effort compared to GPT-5.6 Sol. However, due to the 2.5x increase in price, the model is 75% more expensive per task than its predecessor at max effort.
➤ Hallucinates half as much as GPT-5.6 Sol: GPT-6 Astra sees a large jump in AA-Omniscience, our knowledge and hallucination benchmark. This is driven by a significant decrease in hallucination rate from 92% to 51% at max effort. Unlike some models, this improvement does not come at the cost of accuracy - Astra increased accuracy by 4 points at the same time.
➤ ~80 point gain in AA-Briefcase Elo: GPT-6 Astra improves ~80 points in AA-Briefcase, our frontier long-horizon knowledge work evaluation. Models are tested on multi-week projects, with many linked tasks and thousands of source files. Astra sees a significant increase in both rubric scores and Analytical Quality Elo in AA-Briefcase compared to its predecessor. In the other direction, we observe a reduction in Presentation Quality Elo, where GPT-5.6 Sol (max) still leads all models.
➤ Mixed progress on other evaluations: The model sees a 6 point gain in Humanity’s Last Exam, a long-standing evaluation with emphasis on mathematics, science, and humanities. This is offset by a drop of ~80 Elo points in GDPval-AA v2 - a benchmark we adapted from OpenAI’s dataset measuring economically valuable tasks across 44 occupations. We also observe 2-3 point regressions on other evaluations across a mix of capabilities, including reductions in τ³-Banking (customer support), SciCode (Python problems in a scientific domain), and AA-LCR (long context reasoning over large documents).
Congratulations @OpenAI and @sama on the launch! -
Ilman Shazhaev
@shzhv13
For a long time, conferences were how I checked direction by looking at what tech builders were stuck on. Especially the conversations between sessions.
We bet early on distributed inference at @FARLabsAI.
What felt like a niche read then now dominates half the discussions at any infrastructure event: centralized compute doesn't scale the way everyone assumed.
And now I'm watching a room slowly arrive at the same problem we've been sitting with for two years. -
Ilman Shazhaev
@shzhv13
People can't tell AI fiction from human fiction, according to a recent study.
At the same time I still see AI posts on X and LinkedIn every day. One em dash or "the interesting part is" gives it away immediately.
I don't know about short stories, but run the same test on social posts and I'd bet that number looks very different. -
Ilman Shazhaev
@shzhv13
I really hate how using em dash became a sign of AI writing.
I used them constantly since I was a teenager. Now I have to proofread all of my texts to not look too lazy to write them myself.
Justice for em dash. -
Ilman Shazhaev
@shzhv13
I worked with behavioural data a lot and there are 2 ways to get it.
Pay someone per completed task and take delivery of the output. Or instrument the environment so the behaviour is recorded as it happens.
The first is cheap and fast, but you don't get to see the process.
The second costs more and takes longer, which is why I used to build labs with sensor headsets and instrumented peripherals to get it.
The first way doesn't work anymore.
Amazon is closing Mechanical Turk on 30 September, and up to 46% of its workers were already using AI to do the human work.
For 20 years you could go the first route to get basic data on human interactions because the workers had no cheaper option than just doing the task.
Now they do. -
Ilman Shazhaev
@shzhv13
AI-assisted and agentic hacking is a catastrophic issue everyone admits and noone knows how to fix.
"Work with us or any of our competitors" is not a sentence this industry says often. That usually means the problem got bigger than the race.
Sam Altman
@sama
this is a critically important moment for cyber defense with AI; there is not much time to act.
we are happy if you want to work with us or any of our competitors or partners, but please take this moment seriously.
only an urgent and intense collective response will work. -
Ilman Shazhaev
@shzhv13
The split in gaming AI adoption has moved from studios versus indies to business versus craft.
58% of people in publishing, marketing and business roles use generative AI at work. Among people working directly at studios, it's 30%.
As someone with stakes in both industries, I see AI as a tool for enabling creative teams to do what they do best - create.
AI works best with tasks that grounds creatives and makes them spend their resources unevenly - research, admin, localisation, store copy.
The part where AI usage drops is creative and technical art, which relies heavily on creative vision and determines if the game becomes good or doesn't.
This chart from GDC 2026 report on the state of gaming industry proves exactly what I've been feeling for a few years.
So I see the question not as "should AI be used by creative teams?", but as "what parts of creatives' workflows can AI make more efficient to let them focus more on tasks that they actually enjoy and excel at?". -
Ilman Shazhaev
@shzhv13
Grok 4.6 comes in around $0.84 per completed task. GPT-5.6 Sol is $1.23, Claude Opus 5 $2.34.
Grok is both the most capable and cost-effective model in its league.
X Freeze
@XFreeze
Grok 4.6 ranks tied #1 on the GPQA Diamond leaderboard
Grok 4.6 (high) scores 95% the highest score on the chart for graduate-level scientific reasoning
It outperforms Claude Opus 5, Fable 5 and GPT-5.6 Sol
This benchmark tests AI models with the hardest science questions things even experts find tough
Grok is now sitting at the absolute frontier of scientific reasoning -
Ilman Shazhaev
@shzhv13
Unpopular position: I like long flights.
Nobody can reach you, there is nothing to decide, and the whole thing has been taken out of your hands. Twelve hours where the best available option is to read something.
I have never once been bored on a plane and I'm bored constantly on the ground. -
Ilman Shazhaev
@shzhv13
Google just split its TPU line into separate training and inference chips for the first time in ten years of the programme.
Hardware specialises when a workload gets big enough to deserve its own economics. Both companies that design their own AI silicon have now pointed a dedicated chip at serving rather than training. -
Ilman Shazhaev
@shzhv13
I have never told the person on the next treadmill that we were racing.
They usually have no idea. I match their pace, hold it slightly longer, and leave with a private result that means nothing to anyone but me.
Years of sports competitions rewire your brain forever. -
Ilman Shazhaev
@shzhv13
The best AI model available completes about 60% of entry-level financial analyst tasks. That's not much.
What is left over needs cross-document reconciliation, generating assumptions rather than applying them, and noticing when a number is wrong before it spreads.
Different work from retrieval and drafting, which is where every gain so far has come from.
But what makes it worth it for some early adopters is cost.
On Vals AI's earlier finance benchmark, a top model averaged 3.1 minutes and $3.78 per task. Human experts: 16.8 minutes and $25.66.
5x faster, 7x cheaper, right 6/10 times. The current leaderboard has the leader at 60.6%.
That is a real input and a poor substitute for a person. Especially in sensitive fields.
Most deployment plans are still priced as though it were one. -
Ilman Shazhaev
@shzhv13
7.5% on a data centre bond tells you what the market thinks of the collateral.
Telecom was financed the same way in the late 90s. The overbuild was survivable. The debt against it was what turned it into a decade of bankruptcies.
Hedgie
@HedgieMarkets
🦔AI companies have borrowed so much money this year that they're pushing up interest rates for the entire economy. Nomura estimates tech borrowing alone now equals 25% of what the US Treasury issues in bonds, five times more than last year. Bank of America says the surge has added about 0.3 percentage points to the 10-year yield. Bond managers are selling Treasuries to buy AI corporate debt instead because it pays more.
My Take
AI companies are now competing with the US government for the same pool of lenders, and the lenders are picking the corporate bonds. Alphabet's 30-year pays 6.4%. A Meta data center bond pays over 7.5%. At those rates, a 5.2% Treasury loses the fight for capital every time. That's one of the reasons long-term rates have stayed so stubborn even as the Fed tries to bring them down.
JPMorgan expects $5.5 trillion in AI infrastructure spending through 2030, and most of it will be borrowed. That borrowing raises the cost of money for everyone, the government, your mortgage, small businesses trying to get a loan. The AI buildout has reached the scale where it moves rates for the whole economy, and most people paying higher borrowing costs have no idea that a data center arms race is one of the reasons why.
Hedgie🤗 -
Ilman Shazhaev
@shzhv13
The $3 trillion AI revenue gap is widening and the reasons are not as obvious as media makes them out to be.
I see two problems inside that number.
Whether people will pay is one.
The other is how much revenue a gigawatt of compute has to earn before it pencils, and that moves with grid queues, memory prices, and construction costs. All rising.
When they rise, the target rises, and the gap widens without a customer changing their mind.
Sequoia's David Cahn puts 2026 infrastructure spend near $1.5 trillion, needing roughly $3 trillion in revenue.
Anthropic and OpenAI combined sit around $80 billion.
My underline from that read was his own quote: "the required revenue per GW of CapEx has sharply increased due to these bottleneck dynamics and rising costs of construction."
Modelling these two problems as one number gets you wrong conclusions on both. -
Ilman Shazhaev
@shzhv13
DeepSeek is raising prices, and it has nothing to do with what the model costs to run.
Efficiency lowers what a token costs. It does not add machines.
Once demand passes what your fleet can physically produce, price stops describing costs and starts rationing.
Xenospectrum has a nice timeline to navigate this.
OpenAI cut GPT-5.6 Luna 80% on July 30.
DeepSeek answered within hours at $0.14 in, $0.28 out.
V4 Flash then took first place on OpenRouter with 7.22 trillion tokens in a week, against a company fleet its founder puts near 20,000 H100-equivalents.
August 4, the API degraded to unusable. Two days later, notice of a significant increase.
The weights are MIT licensed and still free to self-host. What got repriced was the capacity to serve them. -
Ilman Shazhaev
@shzhv13
Two countries released 80 of the world's notable AI models last year. Everyone else combined released seven.
The US led with 50, China second with 30, and the entire rest of the world, South Korea, Canada, France, the UK, all of it, totaled seven.
That is the real shape of the frontier, and it is narrower than almost any policy conversation assumes.
But there is a second number underneath that reframes the first.
China's total model output, counting everything below the notable tier, more than quintupled from 151 to 849 in three years.
The frontier is two countries. The diffusion is one country building industrial capacity at the base of the pyramid.
This is the sovereign-strategy problem I wrote about, in one chart.
If you are not one of those two, your AI strategy is a deployment strategy built on models someone else released.
That is not a criticism. It is the arithmetic, and the countries reading it correctly are planning accordingly. -
Ilman Shazhaev
@shzhv13
The US–China AI gap has a number, and it has barely moved in three years.
Epoch tracks it day by day on their capabilities index. Chinese models reach the frontier's level with an average delay of seven months, in a band that has run between four and fourteen months since early 2023. Not widening, not closing. Holding.
A stable lag is the most consequential of the three possibilities, and almost nobody frames it that way. A widening gap would justify the whole containment strategy. A stable one says the frontier is a time lead rather than a wall, and a time lead measured in months is not something you build export policy around expecting permanence.
It also mirrors the open-versus-closed gap almost exactly, which is the tell. The same diffusion mechanics govern both. Capability moves toward whoever wants it, on a delay, whether the border is corporate or national. -
Ilman Shazhaev
@shzhv13
Training compute is now doubling every five months.
That is the number in Stanford's latest index, and it is worth holding next to a second one.
Hardware price-performance doubles roughly every two years. The compute going into frontier models is growing about five times faster than the hardware is getting cheaper.
Which tells you where the spending is actually going. Not into better chips. Into vastly more of them, funded by investment that has tripled year on year. The frontier is being bought, not engineered.
I have made the argument all month that power, not chips, is the binding constraint. This is why.
A curve doubling every five months runs into the grid long before it runs into the fab. You can order more chips. You cannot order more gigawatts on the same schedule. -
Ilman Shazhaev
@shzhv13
Training compute is now doubling every five months.
That is the number in Stanford's latest index, and it is worth holding next to a second one. Hardware price-performance doubles roughly every two years.
The compute going into frontier models is growing about five times faster than the hardware is getting cheaper.
Which tells you where the spending is actually going. Not into better chips. Into vastly more of them, funded by investment that has tripled year on year.
The frontier is being bought, not engineered.
I have made the argument all month that power, not chips, is the binding constraint.
This is why. A curve doubling every five months runs into the grid long before it runs into the fab. You can order more chips.
You cannot order more gigawatts on the same schedule. -
Ilman Shazhaev
@shzhv13
Every humanoid robot sold is really a data collection unit, which is why the industry moves at the speed of deployment schedules.
Game servers have been collecting the same thing for decades without trying. -
Ilman Shazhaev
@shzhv13
The frontier is not pulling away. It is filling up.
Stanford's latest index measured the gap between the best model and the tenth-best.
In one year it fell from 11.9% to 5.4%, and the top two are separated by 0.7%. Tenth place is now within a few points of first.
This is the same thing I have been describing from the cost side all month, seen from the capability side.
When the distance between first and tenth collapses, being first stops being a moat and starts being a head start. You can charge for a head start.
You cannot build a decade of strategy on one.
A crowded frontier changes where the money is. If 10 labs sit within 5 points of each other, the differentiation moves out of the model.
It moves into distribution, into cost, into whatever you wrap around the model to make it reliable. -
Ilman Shazhaev
@shzhv13
1/ Two curves that cannot both be the whole story.
Global AI capex climbing toward records.
The unit cost of a fixed level of intelligence falling about 40× a year.
Both real.Post not available here, view on X
Ilman Shazhaev
@shzhv13
6/ The question that decides who wins is not who owns the most compute.
It is who converts compute into delivered intelligence at the lowest cost, closest to demand.
Ilman Shazhaev
@shzhv13
7/ Those are different competitions.
The first is a capital contest.
The second is an architecture contest.
Ilman Shazhaev
@shzhv13
8/ The capital contest builds the substrate.
The architecture contest captures the value. I see no evidence this cycle breaks the pattern. -
Ilman Shazhaev
@shzhv13
1/ Access was never the constraint.
The AI conversation is a two-body problem, US and China, with everyone else treated as spectators. The data points somewhere else. The interesting variable is the diffusion gap.Post not available here, view on X
Ilman Shazhaev
@shzhv13
5/ These countries are running the experiment everyone else will copy.
How much economic gain comes from deploying near-frontier capability everywhere, quickly, while others wait to own the frontier itself.
The cost curves say the deployers get most of the capability at a fraction of the spend, one quarter late.
Ilman Shazhaev
@shzhv13
6/ If the deployment dividend shows up in their productivity numbers first, and I expect it will, the AI race gets reframed retroactively.
The question stops being who built the biggest model and becomes who absorbed intelligence into an economy fastest.
Different race, different winners, and it is already running. -
Ilman Shazhaev
@shzhv13
The Gulf is not buying chips.
The Western read is that oil money is buying hardware. The visible layer supports it. Gigawatt-class data-center programs in Abu Dhabi with OpenAI and partners. Saudi Arabia's Humain under PIF with major accelerator commitments. Microsoft's billion-plus stake in G42. Energy-rich states converting the one input they hold in surplus, power, into the input the AI era prices highest. On factor endowments alone the trade makes itself.
Watch where the strategic weight sits and it is one layer up. Chips depreciate on a schedule and the capability they produce diffuses within quarters, the open-weight gap holding around one quarter behind the frontier. Every sovereign planner can read that as well as I can. Deployment is the thing that does not diffuse. Intelligence wired into a health system, a court system, a logistics corridor, a language.
The serious programs here are deployment programs with compute attached. National models in Arabic. Government services rebuilt around inference. Sandboxes that let capability touch institutions quickly.
The region's advantages fit that game: energy surplus, capital patience, small populations that make national deployment tractable, and the ability to move in years rather than decades. The risks are execution risks. Talent depth, dependency on foreign release policies, and the chance that deployment proves harder than procurement.
The region is buying the decade after the chips are cheap. -
Ilman Shazhaev
@shzhv13
Five gates every number passes before I publish it.
One. Primary source or it does not exist. The paper, not the article citing the report citing the paper. Half of circulating AI statistics die here, because the primary source says something narrower or does not say it at all.
Two. What exactly was measured. "Cost of intelligence" might mean API list price, serving cost, or cost to reach a benchmark. Those differ by orders of magnitude and move for different reasons. A number without its measurement definition is a vibe.
Three. Independent confirmation for anything load-bearing. One team's finding is a data point. Two teams, two methods, same slope is a trend. The inference-cost collapse clears this three times over, which is why I lean on it. Most claims never clear it once.
Four. The range matters more than the average. Epoch reports 9× to 900× per year depending on task. Quoting only the 40× median is accurate and misleading at once. If the source gives a distribution, the distribution is the finding.
Five. What would change my mind, written down before publishing. If I cannot name the observation that would falsify a claim, I am advocating rather than analyzing.
Twenty minutes per number, which turns out to be rare enough to work as a differentiator. Run the gates on me. When I fail one, and eventually I will, I will correct it in public with a date. -
Ilman Shazhaev
@shzhv13
1/ Behavioral data has a dark version and a durable version.
I write about behavioral data as the scarce input of the next training era. That thesis has a dark implementation and a defensible one, and the difference deserves stating plainly.Post not available here, view on X
Ilman Shazhaev
@shzhv13
5/ There is a quality argument almost nobody makes.
Behavior under known observation in a bounded world is cleaner signal than surveillance. Consent works as a filter on the dataset rather than a tax on it.
Ilman Shazhaev
@shzhv13
6/ The behavioral-data era is coming either way.
The spending curves say so. Which version arrives is a design decision being made now, mostly quietly, and I would rather say this on the record early than be asked about it later. -
Ilman Shazhaev
@shzhv13
Every sovereign AI strategy has the same unstated dependency.
Read any national AI strategy, Gulf, European, Asian, and you will find ambitions that only make arithmetic sense on top of open-weight models. Few of them say so.
The pattern repeats. A nation commits to sovereign AI capability, meaning models it controls, on infrastructure it controls, serving its language and institutions. Then look at what the plan budgets for. Training a frontier model from zero needs talent and capital that perhaps five organizations on earth possess. What gets funded is adaptation: continued pre-training on national corpora, fine-tuning for dialect and domain, deployment on domestic infrastructure.
Every step of that chain begins with weights someone else released.
This is not a criticism. It is the rational strategy. For most nations the race runs toward deployment rather than the frontier, and open weights trailing by a single quarter are a gift to that. Capability arrives nearly free, one quarter late, unencumbered by a foreign vendor's terms of service.
The dependency deserves honesty in both directions. A strategy built on open weights inherits the release decisions of the labs publishing them, and a world where those releases stopped would strand a great deal of sovereign planning.
The most consequential AI policy decisions of this decade may be the release policies of a handful of labs. National strategies are downstream, whether the documents admit it or not. -
Ilman Shazhaev
@shzhv13
The market for intelligence has no market structure.
Inside every serious AI company, a system decides every second which request deserves which model at which price. Routers send easy queries to small cheap models and hard ones up the ladder. Schedulers pack batches for utilization. Routing alone routinely cuts serving costs by multiples at equal quality. Every mature lab runs some version of this.
Between organizations the same problem is bigger and the machinery is absent. Supply and demand match through bilateral contracts negotiated in weeks, reservations made years ahead, procurement inherited from a pre-AI era. Nothing liquid connects a burst of demand in one place to idle capability in another. The market allocating the most consequential resource of the decade has the microstructure of commercial real estate.
That gap has never persisted anywhere else. Electricity got grid operators and power markets. Telecom got peering and exchanges. Finance got clearing. Whoever built the coordination layer captured a disproportionate share of the value, because allocation touches every transaction.
The equivalent layer for intelligence is unbuilt. The internal versions prove the mechanics work, so feasibility is not the obstacle. The industry is young and the incumbents are racing on capability instead.
When intelligence is cheap and everywhere, deciding what runs where stops being plumbing and becomes the control point. -
Ilman Shazhaev
@shzhv13
I was confidently wrong about distributed training.
Eighteen months ago I would have told you frontier-scale distributed training was permanently impractical. That was the 2023 consensus and it was well founded. Gradient synchronization at scale demands data-center interconnects, the public internet is orders of magnitude too slow, so serious training stays inside single buildings. I repeated it confidently.
Then the evidence arrived. DeepMind's DiLoCo cut inter-node communication about 500x by synchronizing rarely instead of constantly. Prime Intellect used it to train a 10B model across three continents over ordinary internet, and shipped it. Nous Research pushed the reduction further. DiLoCoX pre-trained a 107B model over 1 Gbps links with a 357x throughput gain.
My claim became dated, which in this field is the same as wrong on a delay. Frontier runs are still centralized and I still argue the steelman for why. But "permanently impractical" was a statement about physics, and it turned out to be a statement about 2023's algorithms. Those have very different lifespans.
Most confident impossibility claims in AI are cost claims in disguise, and costs here fall on curves. When someone says a thing cannot be done, including me, the question is whether the barrier is physical law or this year's price.
I keep a list of positions I have retired and why. This one's entry ends: confused an algorithm's limit for a law's. -
Ilman Shazhaev
@shzhv13
1/ The engine is not the vehicle.
The industry talks about models the way the 1980s talked about microprocessors. As if the component were the product. The CPU was never the product. The system was.Post not available here, view on X
Ilman Shazhaev
@shzhv13
5/ So the question that I think decides the next decade is not "whose engine is best."
It is this. When the intelligence itself is nearly free, what exactly still commands a premium?
Ilman Shazhaev
@shzhv13
6/ I have spent a month publishing pieces of an answer without assembling them.
That was deliberate and it stays that way for now. The question is worth more held open than closed early, and most strategy in this industry gets written as though it were already settled. -
Ilman Shazhaev
@shzhv13
AI's fastest-growing cost line is human judgment.
Follow the money in AI right now and one quiet category keeps growing. Paying humans to generate behavior, and building environments to harvest it.
Scale AI built a multi-billion-dollar business supplying labeled data and human feedback to frontier labs. Surge AI grew alongside it on preference data. Mercor recruits domain experts whose reasoning becomes training material. Prolific supplies research-grade participants. And a newer wave, General Intuition among them, treats interactive environments themselves as the product, generating decision data from gameplay.
Nothing here is charity. Labs with the most sophisticated views of the data landscape are paying premium prices for one specific thing: records of humans exercising judgment, comparing, choosing, correcting, in settings where the context of the choice is captured alongside the choice.
The RL turn intensifies it. RL needs environments. Worlds with states, actions, and consequences where an agent can practice. Text corpora do not provide that. So environments became a procurement category, with reported deals running into the hundreds of millions.
The historical parallel is exact. OpenAI Five and AlphaStar were both trained inside games, because games were the only places on earth where millions of humans made high-stakes decisions inside fully instrumented worlds.
The structural read: the internet gave AI its first curriculum nearly free, and that curriculum is close to spent. The second one has to be built, instrumented, and consented into existence. That makes it the first training input in this industry's history that is genuinely scarce at the source.
Scarce inputs to an abundant technology are where the margin goes. The market is already voting. -
Ilman Shazhaev
@shzhv13
1/ Video is not decision data.
A claim I'll defend: the most valuable training data of the next five years mostly does not exist yet, because we have been recording the wrong layer of human behavior.Post not available here, view on X
Ilman Shazhaev
@shzhv13
6/ The labs' behavior confirms the diagnosis.
Watch the sums being spent on human preference data, expert demonstrations, and RL environments.
That is the market pricing the gap between outcome data and decision data.
Ilman Shazhaev
@shzhv13
7/ Synthetic data will carry many things.
Decisions under uncertainty made by actual humans, with the context attached, is the input I'd bet stays non-substitutable longest.
Whoever instruments real decisions, with consent, by design, holds inputs the entire industry needs. -
Ilman Shazhaev
@shzhv13
Anatomy of a fifteen-hour failure.
On October 20, 2025, a DNS race condition inside DynamoDB in a single AWS region, US-EAST-1, began a cascade that ran roughly fifteen hours.
Snapchat. Fortnite. Roblox. Ring. Coinbase. Major banks. United Airlines. McDonald's apps. The UK's tax authority. Downdetector logged more than 6.5 million failure reports across a thousand-plus services.
The detail that should reorganize your thinking: applications architected as multi-region failed anyway. Global control planes, the systems that manage failover itself, depended on the same region that was down.
I keep returning to this event because of what it says about the next decade, when the workload through such chokepoints is not photo sharing but cognition.
The industry is wiring intelligence into everything. Agents running multi-step work. Reasoning inside operational software. Decisions delegated at machine speed. All of it rides on a delivery layer whose concentration was tolerable for the last era's workloads. An outage that pauses streaming is an inconvenience. An outage that pauses the reasoning inside logistics, healthcare intake, and financial operations for fifteen hours is a different category of event. We have already demonstrated the failure mode at production scale.
The inversion worth sitting with. Intelligence is becoming abundant. Fixed capability falls about 40× per year, and open equivalents arrive within a quarter. What stays scarce is the guarantee that the intelligence is there. Routed, verified, low-latency, alive during the bad fifteen hours.
Delivery sounds like plumbing. So did payments infrastructure, and it became some of the most valuable real estate in software.
The benchmarks will keep getting the headlines. Watch the delivery layer. -
Ilman Shazhaev
@shzhv13
1/ Three years of chip controls: what worked, what leaked.
Three years into the most aggressive technology export regime since the Cold War, the scorecard is in, and it is stranger than either side predicted.Post not available here, view on X
Ilman Shazhaev
@shzhv13
6/ To steelman the policy: a time lead measured in quarters at the frontier is not nothing.
In a race where compounding matters, buying quarters can be rational statecraft. The controls buy quarters.
Ilman Shazhaev
@shzhv13
7/ What they cannot buy, on the evidence of three years, is a monopoly on any fixed level of capability.
The diffusion channels are too many and the efficiency curve too steep.
Ilman Shazhaev
@shzhv13
8/ The interesting question is no longer whether capability diffuses. It measurably does.
It is what remains scarce, and therefore strategic, in a world where capability itself cannot be contained.
That question has an answer. It is where my work has been pointed all month. -
Ilman Shazhaev
@shzhv13
Every frontier price is racing a clock that resets quarterly.
On Epoch's Capabilities Index, the best open-weight models trailed the closed frontier by about three months through late 2025, and about four months across the first half of 2026. The band moved by a month. It did not open up.
A closed lab's pricing power on a given capability lasts about as long as its exclusivity. One quarter, give or take. Then an open equivalent exists, hosting competition begins, and that capability's price starts its descent toward serving cost. The Stanford AI Index curve, $20 to $0.07 per million tokens in two years, is that descent recorded.
Which produces a strange market. The product improves constantly. The price of any fixed version of it collapses on a schedule.
The frontier premium is real and perishable. Revenue durability cannot live in the model. It lives in what surrounds the model: distribution, feedback loops, workflow lock-in, delivery reliability, enterprise trust. Things with switching costs, because capability has none once an equivalent ships.
Watch the incumbents' actions. Vertical product building. Enterprise contracts with integration depth. Consumer habit moats. The model is the engine. The business is everything except the engine.
If you allocate capital here, re-read every proprietary-model-advantage claim with a 90-day depreciation schedule attached. Ask what is left of the margin after the equivalent ships. Sometimes a great deal. Often silence.
The gap is not closing. It does not need to. -
Ilman Shazhaev
@shzhv13
1/ Capability moves down-market. It never moves back.
Once a capability has been demonstrated anywhere, at any price, it starts migrating toward cheaper hardware. It never migrates back up. This is the most underrated mechanism in AI economics.Post not available here, view on X
Ilman Shazhaev
@shzhv13
5/ Practical consequence: the marginal value of secrecy is falling.
What you protect by not shipping, you lose in market position. What you gain by shipping, you donate to the diffusion pump within a quarter or two.
Ilman Shazhaev
@shzhv13
6/ The equilibrium this points toward: leads measured in months, defended by execution speed rather than walls.
Some organizations are built for that race. Most are built for walls. -
Ilman Shazhaev
@shzhv13
1/ The frontier's lead is one quarter. It is not growing.
The question is never "are open models behind." They are. The question is whether the gap is widening, closing, or stable. I tracked it. The answer is the most strategically loaded of the three: stable.Post not available here, view on X
Ilman Shazhaev
@shzhv13
6/ What a stable 3–4 month lag does to strategy: any plan whose value depends on exclusive capability has a value half-life of one quarter.
Pricing power at the frontier survives. Pricing power on last quarter's capability dies on schedule.
Ilman Shazhaev
@shzhv13
7/ I keep a simple rule for reading AI strategy documents.
Find the sentence that assumes the gap widens. It is almost always there, and it is almost never argued. -
Ilman Shazhaev
@shzhv13
The fastest computer of 2020 was assembled from spare bedrooms.
Training gets the headlines. Delivery is where the physics is friendlier, and the existence proofs are older than most people think.
Training a model is one enormous, tightly-coupled computation. Every node needs every other node's updates, which is why bandwidth dominated that conversation for a decade. Serving a trained model is the opposite shape. Requests are independent. A million users' queries share no state. The workload is embarrassingly parallel. What matters is placing capacity near demand and keeping it utilized.
The latent supply already exists at civilizational scale. In 2020, Folding@home, volunteers' home computers running protein simulations, crossed the exaflop line and briefly assembled more compute than the top supercomputers of the day combined. Today, Salad routes production AI inference across tens of thousands of consumer gaming PCs. Public figures cite 60,000+ GPUs. Operating systems with paying customers, running on hardware that was otherwise idle.
The honest constraints. Reliability on heterogeneous, non-dedicated hardware is a genuinely hard scheduling and verification problem. Data-center SLAs are a bar, not a detail. Latency-sensitive workloads will always want proximity. Anyone claiming consumer hardware replaces data centers outright is overclaiming.
The direction of the gap matters more than its width. Quantization, batching, and speculative decoding shrink the hardware needed per unit of capability every quarter. The capability itself migrates into smaller models. Each turn of both cranks moves more of the inference market into the envelope commodity hardware can serve.
The line between workloads that require industrial compute and workloads that don't moves in one direction only. -
Ilman Shazhaev
@shzhv13
The physical build is the fast part of AI infrastructure. Almost nobody arguing about data centres believes this.
I know it because I have done it. In 2017 I built a data centre from scratch in six weeks. Ten million dollars of equipment, three countries, and I was twenty-four.
The scale is not comparable to what gets built now. The sequence is.
Epoch AI's construction data says the same at current scale. Gigawatt-class facilities go up in two years or less, and xAI's Colossus 1 was reportedly stood up in 122 days. Steel moves fast when there is capital behind it.
What does not compress is the power.
Power is not a line item you solve after site selection. It decides site selection. It is why companies build in places like Texas rather than near their engineers, and why most AI data centres end up in the few countries that can supply it at all.
And the demand keeps moving. Epoch projects the largest individual frontier training runs will draw 4 to 16 gigawatts by 2030, growing more than 2x per year. Every doubling shortens the list of places on earth that can host one.
So the AI map gets drawn by generation capacity. Stranded power, favourable climates, energy surpluses. That geography sits nowhere near the places that currently call themselves tech hubs.
The buildings go up fast. The question was never construction speed. -
Ilman Shazhaev
@shzhv13
1/ Google and NVIDIA already left the building.
The strongest evidence that AI training is going multi-site doesn't come from startups. It comes from the two companies with the least incentive to say so.
Ilman Shazhaev
@shzhv13
2/ Google Cloud's leadership has said Gemini Ultra was trained across multiple data centers. Not a research demo.
The flagship model of the company that owns more single-site infrastructure than anyone.Post not available here, view on X
Ilman Shazhaev
@shzhv13
5/ Read the revealed preference.
The incumbents are not debating whether training crosses buildings. They are shipping the interconnects and training the flagships that way.
Ilman Shazhaev
@shzhv13
6/ The open question was never distributed versus centralized.
It is how far distribution goes down the network-quality curve, from dedicated fiber between owned campuses toward ordinary links between anyone's hardware.
The research ladder on that question is moving fast, and I covered it earlier this week. -
Ilman Shazhaev
@shzhv13
By late 2027, US AI data centers will need what Britain's entire grid produces.
Every debate about AI infrastructure eventually hits one number, and it is not FLOPS.
Epoch AI frames it cleanly. US AI data centers will collectively need around 20 to 30 gigawatts by late 2027. Inside the US that is about 5% of generation capacity. Manageable on a grid that size.
Put the same 30 gigawatts almost anywhere else and it isn't. Roughly 25% of Japan's generation capacity. Half of France's. Ninety percent of the UK's.
That is the sector. The number that forces a decision is the single-site one.
Frontier training power has been roughly doubling per year, and Epoch projects the largest individual training runs will draw 4 to 16 gigawatts by 2030. The largest single facilities now planned sit in the low single-digit gigawatts.
The arithmetic on one campus runs out somewhere in that gap.
Grid queues in the major markets run to years. Enough capital compresses a chip supply chain. No amount of it compresses the construction time of gigawatt-class generation and transmission.
Which forces a fork, not this year, but on a horizon you can already see. Either frontier scaling slows to the grid's pace, or training stops assuming a single site. Epoch names the second branch themselves.
The industry's actions show which branch it is preparing for. The largest lab already trained its flagship across multiple data centers. The dominant hardware vendor ships interconnects built for cross-site training.
Concentration was an era, not an end state. It lasted exactly as long as one building could hold the frontier's appetite. The appetite doubles annually. Buildings don't. -
Ilman Shazhaev
@shzhv13
1/ The best case for centralized AI, argued properly.
I write about distribution, so let me make the opposite case. The argument for centralized AI is strong, and most people who dismiss it haven't understood it.Post not available here, view on X
Ilman Shazhaev
@shzhv13
7/ What they are not about: the other 95% of the market.
Serving last year's capability, which is where most economic value gets delivered, has none of these constraints at the same intensity.
And the bandwidth cost of even the training case is falling by published orders of magnitude.
Ilman Shazhaev
@shzhv13
8/ So my actual position, stated carefully.
Centralization wins the frontier run today and probably for years. Distribution wins the delivery of intelligence, progressively, on a curve you can already measure.
Most arguments in this space fail by treating those as one question. -
Ilman Shazhaev
@shzhv13
The bandwidth wall is falling 10,000x at a time.
The binding constraint on distributed training was never compute. It was bandwidth. Gradient synchronization demands data-center-grade interconnects, and that is the entire physical argument for why AI must live in single buildings.
The last two years put that constraint on a ladder, climbed in public.
Rung one: DeepMind's DiLoCo. Local training with infrequent synchronization cuts inter-node communication by up to 500x. This is the method behind a 10B model trained across three continents over the public internet.
Rung two: Nous Research's DisTrO reports up to 10,000x reduction in inter-GPU communication bandwidth, training a 15B model over the internet. The code and reports are public, which is the correct standard for a claim that size.
Rung three: DiLoCoX, from 0G Labs and China Mobile researchers. A 107B-parameter model pre-trained over ordinary 1 Gbps links, with a 357x throughput gain over standard AllReduce. For calibration, 1 Gbps is office-grade connectivity.
Three teams, three methods, one direction. The bandwidth requirement for training large models falls by orders of magnitude while models grow.
The discipline: none of these produced a frontier model, and communication-efficient methods carry real convergence trade-offs the papers document. The gap to the largest runs remains vast.
But strategy is made on derivatives, not levels. The level says frontier training stays co-located today. The derivative says the co-location requirement weakens every year, measurably, with published methods. Any five-year thesis about compute concentration that only cites the level is answering the wrong question. -
Ilman Shazhaev
@shzhv13
1/ One model, three continents, ordinary internet.
In late 2024, Prime Intellect trained a 10B-parameter model across five countries on three continents, over the public internet.
Most people either overclaimed what INTELLECT-1 proved or never read it. Here is the honest version.Post not available here, view on X
Ilman Shazhaev
@shzhv13
6/ And the ceiling moved again within months. A 107B model pre-trained over ordinary 1 Gbps links, which I'll cover separately.
The trendline of "largest model trained over weak links" is steep.
Ilman Shazhaev
@shzhv13
7/ Meanwhile the incumbents quietly validate the direction from the other side.
Google has said Gemini Ultra trained across multiple data centers.
When the frontier lab and the open project converge on the same architecture problem, it stops being fringe.
Ilman Shazhaev
@shzhv13
8/ Distributed training is not ready to produce a frontier model.
It is advancing at a rate that anyone modeling compute concentration five years out has to account for.
Most such models silently assume the 2023 consensus. That assumption is now stale. -
Ilman Shazhaev
@shzhv13
1/ What breaks at $0.07 per million tokens.
GPT-3.5-level output ran about $20 per million tokens in late 2022. It reached $0.07 by late 2024. A short list of things that stop making sense at that price.Post not available here, view on X
Ilman Shazhaev
@shzhv13
6/ Any national AI strategy whose end state is "we bought the chips."
The chips depreciate. The capability they produce diffuses. What compounds is deployment, where intelligence meets your economy.
Ilman Shazhaev
@shzhv13
7/ The pattern in all five: expensive intelligence justified architectures that cheap intelligence makes absurd.
Most of the industry is still running the expensive-era architecture. The gap between the curve and the architecture is where the next decade's value sits. -
Ilman Shazhaev
@shzhv13
Cheaper tokens, bigger bills.
The standard objection to the cost-collapse thesis: if intelligence gets 40x cheaper per year, doesn't the market shrink 40x per year.
Bandwidth answers this. From the late 1990s to the 2010s, the price of moving a bit fell by orders of magnitude. Telecom revenue didn't collapse. Video streaming, cloud, and the mobile internet, all impossible at prior prices, expanded consumption faster than prices fell.
Storage tells the same story. The genome that cost $100M to sequence in 2001 costs a few hundred dollars now (see the image), and sequencing is a bigger industry today, not a smaller one, because it crossed the threshold where whole use cases became viable.
Economists call it Jevons: when efficiency makes a resource cheaper, total consumption often rises. The condition is elastic demand, use cases waiting below the current price point.
Intelligence is the most demand-elastic resource I can name. At $60 per million tokens, you ration it: chat, short completions, a human in every loop. At $0.07, you run agents for hours. One order down, you embed reasoning in every function call, every sensor read, every decision a system makes. Volumes that dwarf today's token consumption.
So the cost collapse doesn't shrink the market. It relocates the margin. Selling raw tokens becomes selling a deflating commodity. What compounds is positioned where cheap intelligence creates new demand, and in the inputs that stay scarce while intelligence itself gets cheap.
What stays scarce is the more interesting question. I'll spend a few posts on it later this month. -
Ilman Shazhaev
@shzhv13
1/ 9x to 900x: the range is the story, not the average.
Epoch AI's headline number: GPT-4-tier capability gets about 40x cheaper per year. The finding underneath is more interesting.
The range runs 9x to 900x depending on the task. That variance is the story.Post not available here, view on X
Ilman Shazhaev
@shzhv13
5/ If you sit at the 9x end, capability stays scarce longer, and defensible longer.
Most strategy decks I see never ask which end they're on.
Ilman Shazhaev
@shzhv13
6/ The lazy version of the cost-collapse story is one average.
The real version is a distribution with strategic consequences at both tails. -
Ilman Shazhaev
@shzhv13
API prices are falling. That is not the same as costs falling.
Prices are a policy choice sitting on top of costs. Subsidized pricing and fat margins both exist. To know whether the deflation is real, you decompose it.
Four mechanisms stack under the curve.
Hardware: more throughput per dollar and per watt each generation. Steady, and the slowest of the four.
Serving efficiency: batching, quantization, speculative decoding, better schedulers. Quiet 2-5x gains that never make headlines.
Model efficiency at fixed quality: capability that needed a frontier model in 2023 ships in a fraction of the size today. The steepest of the four, and the reason the fixed-capability curve outruns any hardware roadmap.
Competition: open weights trailing by months cap what anyone can charge for last year's capability. Some of that price decline is real cost. Some is margin transfer to users.
The honest statement: prices are falling faster than costs in some segments, but costs are falling at a rate with no precedent, and three of the four mechanisms still have runway.
The test I use when a provider cuts prices: did new hardware, a serving technique, or a smaller equivalent model ship in the last two quarters. If yes, the cut is structural. If no, someone is buying share, and structural cuts follow anyway. -
Ilman Shazhaev
@shzhv13
1/ I fact-check AI statistics. These did not survive.
I spend a lot of time verifying claims before I use them. Some widely-circulated numbers survive contact with their primary sources. Some don't.
Recent casualties from my own notes:Post not available here, view on X
Ilman Shazhaev
@shzhv13
7/ The method, which is boring: find the primary source.
Check the market definition. Check whether the number is a point or a range. Check who measured and how.
Twenty minutes per number. Most cited numbers have never received the twenty minutes.
Ilman Shazhaev
@shzhv13
8/ Standard I hold my own writing to: derived, not asserted.
When I publish a number this month, the source comes with it.
So you can run the twenty minutes on me. -
Ilman Shazhaev
@shzhv13
1/ Everyone is measuring the wrong price of AI.
Almost everyone tracks the cost of the frontier. It keeps moving, so its price stays high and the curve looks flat. Measure a fixed capability instead. Pick a quality level, hold it constant, watch what it costs each year.Post not available here, view on X
Ilman Shazhaev
@shzhv13
9/ Every strategy built in AI right now contains an assumption about this curve, usually unstated.
Companies assuming intelligence stays expensive are building moats out of a melting material.
Ilman Shazhaev
@shzhv13
10/ I'll be publishing analysis on what this curve does to infrastructure, open models, and national strategy throughout July.
The claims will be sourced. Where I'm inferring, I'll say so. -
Ilman Shazhaev
@shzhv13
A new AI data centre waits 24 to 36 months for grid approval. The hardware already plugged in across homes and offices waits none of that.
Inference has overtaken training as the dominant energy consumer at fleet scale. That's why hyperscalers are signing nuclear power agreements. Standard grid capacity can't keep pace.
Distributed inference runs on hardware that already has power and is already drawing it for other work. Whoever solves AI's power problem first sets the cost structure for everything built on top.
www.spheron.network/blog/ai-data-center-power-constraints… -
Ilman Shazhaev
@shzhv13
Plenty of people can design elegant infrastructure on a whiteboard.
Far fewer have had to keep it running when the assumptions it was built on turned out to be wrong.
That second group tends to make very different decisions. -
Ilman Shazhaev
@shzhv13
I've built in China, Malaysia, and the UAE.
Each place has its own logic. In China, the real conversation happens after the meeting. In Malaysia, pushing for speed reads as disrespect. In the Gulf, the relationship has to exist before the deal can.
What I noticed is that none of this was foreign to me. Growing up Chechen, your word was binding and your reputation traveled faster than you did. That was just the baseline, not a value someone taught you explicitly. -
Ilman Shazhaev
@shzhv13
The same task in an agent system can easily consume 30x more tokens on one run than another, even with identical model and prompt.
This isn't a bug. In agentic workflows the model itself decides how many reasoning steps, tool calls, retries and reflection loops it needs. Compute requirements turn into a forecasting nightmare.
A large study of over 100 trillion tokens on OpenRouter shows the scale of the problem. Agentic coding workloads already burn roughly 1000x more tokens than regular code reasoning or chat. Even on the exact same task, token usage varied by up to 30x across runs. More tokens didn't mean better outcomes. Accuracy often peaked at intermediate spend and then flattened or dropped while the agent kept wandering.
This unpredictability breaks traditional capacity planning and cost control. GPU utilization swings wildly, autoscaling can't keep up, and surprise bills become routine even at steady request volume. The old playbook for LLM inference doesn't work for agents. -
Ilman Shazhaev
@shzhv13
Building AI infrastructure from Abu Dhabi gets treated as either exotic or a handicap. It's neither.
It's a different vantage point on the same problems. You notice things about sovereignty and access to compute differently when you stand outside the usual hubs, because here those were never abstract questions. -
Ilman Shazhaev
@shzhv13
The old map of AI infrastructure assumed compute would follow cloud concentration. That map is being redrawn.
Look at what's actually getting built:
- India's Bhashini moved off foreign hyperscalers onto sovereign infrastructure in February and gained 40% in performance.
- Cassava Technologies is deploying 12,000 Nvidia GPUs across five African countries on its own fibre backbone. Before this buildout, Nvidia estimated roughly 80 of its GPUs were installed across the entire African continent.
The pattern is the same in both cases. Places that couldn't treat compute as someone else's problem had to design for it. That constraint produced different infrastructure.
Training still rewards dense hyperscale clusters. Inference runs continuously and needs to sit close to actual demand. McKinsey expects it to overtake training in AI datacentres by 2030. The geography that serves inference well is not the geography that built the cloud.
Ilman Shazhaev
@shzhv13
The next map will be drawn around places that arrived at this problem first, because they had to.
I wrote about this for @ComputerWeekly
www.computerweekly.com/opinion/AIs-next-compute-layer-is-… -
Ilman Shazhaev
@shzhv13
In distributed compute, available capacity and dependable execution are two different things.
A node can appear online and still fail to complete work, hold the right model in VRAM, or respond within acceptable latency. Developers routing inference jobs across heterogeneous hardware need to know this before the job is assigned, not after it fails.
That's the problem the @FARLabsAI Reliability Score is built to solve. Every node gets a verified score across uptime, job completion, latency, and incidents. Routing decisions are grounded in what the network can actually observe, not what a node advertises.
Distributed inference only becomes infrastructure when you can trust what's on the other end. -
Ilman Shazhaev
@shzhv13
The GPU is the new oil well. Everyone is racing to drill, but few are building the pipeline.
Standard Oil didn't win by finding more oil. -
Ilman Shazhaev
@shzhv13
Everyone is still keeping score by model benchmarks.
A week of conversations at @superai_conf made something clear: the teams shipping real AI have stopped watching the leaderboard. They're focused on what surrounds the model, how it gets served and kept reliable in real use. That's where the next phase actually gets decided. -
Ilman Shazhaev
@shzhv13
Abu Dhabi is building toward multi-gigawatt AI infrastructure, with the first phase coming online this year. Sovereign compute at that scale is the right move for training and for keeping critical capabilities inside the region.
What follows is the harder engineering problem. Always-on agents, real-time applications, industrial workloads, gaming. None of these scale economically through centralized mega-clusters. Latency is a hard physical constraint. Grid capacity has a ceiling. And the unit economics of routing inference over distance break well before you reach the volumes this region is planning for.
Larger training infrastructure is the first layer. The second layer is distributed by necessity, regional capacity close to actual demand. Without it, the first layer generates capacity that cannot be fully converted into economic value. -
Ilman Shazhaev
@shzhv13
Three days at @superai_conf in Singapore, on stage presenting FAR AI and speaking with builders and operators in the space.
The industry is racing to add more compute while a large amount of capable GPU power already sits idle and out of reach for production workloads. The practical challenge is making that existing capacity stable enough to build on.
Sachin Hindupur from @AMD put numbers behind it on the main stage: demand growing 3.5x by 2030, GPU lead times of 36 to 52 weeks, and grid connections taking 4 to 7 years. You can't procure your way out of timelines like that.
This is the gap we're building @FARLabsAI to close: turning idle capacity into a single inference network that routes each request, in real time, to a node that can serve it with low latency and reliability. The last three days only sharpened my conviction that it's the right problem to be working on.
Grateful to the team for these three days, and to everyone who stopped by. -
Ilman Shazhaev
@shzhv13
Every gamer carries a story. The wins, the achievements, the highlight moments, the years it took to build them. For most players that story has lived in fragments, scattered across platforms that never let it travel with you.
GAMED brings it together into one gaming identity the player owns, a single picture of who they are across every platform.
GAMED
@gamedofficial_
"We asked a simple question: if a person's financial history can be aggregated into a credit profile, why can't a gamer's history be aggregated into a gaming profile?"
That's one of the ideas behind GAMED, recently discussed by our Founder & CEO, Ilman Shazhaev (@shzhv13), in an interview with Net Influencer, where he shares insights on gaming identity, audience value and what's next for gamers.
Catch the full interview and insights👇
www.netinfluencer.com/gamed-is-building-a-platform-to-sho… -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Live from SuperAI.
The FAR Labs by @Dizzaract team is meeting attendees, showcasing our platform and connecting with builders and AI leaders from around the world.
We'll be at Booth PB19, L5 tomorrow. Come meet the team.This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Let's get under the hood of AI inference at @superai_conf
FAR Labs
@FARLabsAI
See how FAR AI is building production-ready infrastructure for AI inference at scale.
Join our Founder & CEO, Ilman Shazhaev (@shzhv13), as he takes the stage at SuperAI 2026 to share what we're building at FAR Labs by @Dizzaract. -
Ilman Shazhaev
@shzhv13
AI infrastructure was always supposed to follow the same few hubs: Silicon Valley, Seattle, London, wherever the cloud and the capital already sat. That's changing fast. Compute keeps getting more expensive and power-hungry, and the strongest new infrastructure is going up where it was never cheap to begin with: India, Brazil, the UAE, Africa. Places where scarcity is the design problem you start from.
The shift underneath it is inference. Training still rewards dense hyperscale clusters, but inference runs continuously, and McKinsey expects it to overtake training in AI data centers by 2030.
That's the bet we're building @FARLabsAI on from Abu Dhabi: the next layer of AI infrastructure gets built where the demand actually is.
I wrote about where that new map is being drawn, and who's drawing it, for Rest of World: restofworld.org/2026/scarcity-is-driving-ai-innovation-ou… -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Meet the team representing FAR Labs by @Dizzaract at two of the world's leading AI events.
- SuperAI Singapore (8-14 June) | Meet Ilman Shazhaev (@shzhv13), Founder & CEO
- London AI Summit (10-11 June) | Meet Jamie Elkaleh (@Elkaleh), COO
From distributed AI infrastructure and inference to the future of AI deployment at scale, join us for conversations, live product demos and a closer look at FAR AI.
Let's connect. -
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev
@shzhv13
Two AI startups.
One can't afford its own inference at scale.
The other is one open-source release from losing its moat.
Which one survives? -
Ilman Shazhaev
@shzhv13
I've been watching where AI inference demand actually comes from, and one number stuck with me.
OP.GG, the stats and coaching app most League players keep open, is running 63 billion tokens of inference a month through @OpenRouter. That ranks it #64 among every app on the platform.
What stood out was how little of gaming OP.GG actually covers. Match analysis, coaching, stats, draft help. The competitive layer, and not much past it.
If that thin a slice already pulls 63 billion tokens, a full AI-native gaming experience is a different order of magnitude. Gameplay assistance, personalization, intelligent companions, content, social systems, each one a constant stream of inference. The narrow case is the floor, not the ceiling.
@gamedofficial_ is that broader version, and the way it works is what matters most. OP.GG, like nearly every gaming platform, rents its inference from external providers. GAMED runs on FAR AI, our own distributed compute network, so the demand it creates flows back into infrastructure we own. That turns the product into a demand engine for the network underneath it.
Infrastructure companies usually deploy the hardware first, then go hunting for demand to fill it. We're running it the other way, starting from the demand. As GAMED grows, more inference runs on FAR AI, the people supplying compute monetize more, and the network's capacity grows. More capacity unlocks AI experiences that weren't possible before, which pulls in more usage. It feeds itself.
The OP.GG number, then, matters well beyond the headline. It validates a thesis we've been betting on. Gaming is becoming one of the largest sources of AI inference demand, at the same time inference itself is moving toward distributed infrastructure. @FARLabsAI sits where those two shifts meet, which is why we've always treated gaming as fuel for the network, not only a product. Unlike most infrastructure companies, we're building both sides of it at once. -
Ilman Shazhaev
@shzhv13
Span and NVIDIA just put AI compute into people's homes, and everyone's arguing about whether homeowners get paid for it. That's the least interesting part.
The real signal is who's behind it. A player at NVIDIA's scale building distributed AI infrastructure is market validation. It confirms the industry has hit the limits of centralized data centers and is looking for distributed alternatives, which is the thesis we've been building on since day one.
XFRA's answer is to build and own the network itself, deploying company-owned nodes into homes, 16 Blackwell GPUs at a time. Every node is new hardware and new capital. It's a real approach, just not ours.
We built @FARLabsAI on a simpler premise. The hardware already exists. Hundreds of millions of powerful machines sit largely unused most of the day, and that compute goes to waste. Our technology runs enterprise-grade compute on consumer devices, so instead of asking people to host our hardware, we let them connect the machine they already own and monetize it.
For the person on the other end, a gamer or creator with a powerful PC, it's effortless. The machine keeps doing exactly what it always did, and during the stretches it would otherwise sit idle, it monetizes compute in the background. Nothing about how they use it changes, and the value comes from hours the machine had to spare.
That same fact is our entire advantage. Nothing for us to buy or deploy, far lower capital than a company-owned model, and a supply that scales as fast as people connect. What costs them nothing is exactly what makes the network cheap and fast to build.
The whole industry is moving toward distributed compute, and FAR AI is the version that runs on hardware people already own. We've been building it since day one, and announcements like this only tell us to move faster.
Coinvo
@Coinvo
JUST IN: NVIDIA will now pay you over $22,000 a year to host a mini AI data center in your home.
-
Ilman Shazhaev
@shzhv13
Picked up Geoffrey West's Scale after @karpathy mentioned it on @dwarkesh_sp's podcast. The book is from 2017 but reads like it was written for the current AI moment.
West's finding: across all mammals from a mouse to a blue whale, metabolic rate plotted against body mass falls on one line with an exponent of 3/4. Bigger animals get dramatically more efficient per kilogram because circulatory networks become more efficient as they branch.
Cities run the opposite pattern, with a scaling exponent around 1.15. Double the population and per-capita output rises by 10-15%, because innovation compounds with density. Companies behave like organisms instead. They slow as they grow and almost all eventually die, while cities almost never do.
Karpathy highlighted the piece that lands hardest for AI infrastructure: heat creation scales with volume (the cube), while dissipation scales with surface area (the square). Grow a single system and generation outpaces what the surface can release. Every Stargate-scale ambition runs into this thermodynamics wall before it runs into a compute one. The bet at @FARLabsAI is the opposite shape: distributed inference that branches like a biological network, with heat and power dissipating across many surfaces.
Scaling laws in AI get treated as pure math. West's frame is a reminder that geometry and thermodynamics write the actual rules. -
Ilman Shazhaev
@shzhv13
Gulf AI commitments sit inside national planning documents written before ChatGPT existed. The time horizon is the structural difference from Western AI investment.
The UAE Centennial 2071 plan was published in 2017. AI infrastructure was already inside it. A 54-year operating horizon for a category that barely existed commercially at the time.
For our team building in the region, that document is operational. We plan against it. It's how I think about every multi-year commitment we make on the infrastructure side.
The rest of the AI world plans on shorter clocks. US tech infrastructure operates inside 4-year political cycles, where tariff regimes, export controls, antitrust priorities, and immigration rules reset with administration changes. Even with $200B+ committed across major Western economies, the operating environment can shift twice inside a single capex amortization window.
Physical AI infrastructure needs 5 to 10 year horizons. Capital can be wired in a day, chips can ship in months, talent can be hired in weeks. The variable I weight heaviest when we commit capital is whether the regulatory partner will still be planning the same way in year 5.
Horizon matching is the structural variable nobody analyzes in AI infrastructure discourse. The conversation focuses on capital, chips, energy, talent. Horizons are treated as background.
Where the horizons match, physical infrastructure accumulates. Where they don't, capital waits on the sideline for clarity that doesn't come. -
Ilman Shazhaev
@shzhv13
Most AI strategy conversations I'm in still treat power as a procurement detail. Power decides which AI architectures actually scale this decade, and most planning still doesn't price it that way.
The headlines are catching up. Microsoft restarting Three Mile Island for one campus. xAI bringing mobile gas turbines to Memphis. Each is a workaround for a constraint that wasn't supposed to bind this fast.
Over 2 terawatts of generation are stuck in US interconnection queues. AI capex cycles run on 18 months. Grid interconnection runs on 5–10 years.
Two systems on incompatible clocks. The slower one always wins on physics.
I built physical compute infrastructure in 2017. The thing you learn at gigawatt scale is that the grid is the slowest variable in the chain. It's also the one that decides whether your capital turns into running compute or into a permitting queue.
The compute that compounds through this decade gets built around the grid problem. Centralized scale runs into the queue. Distributed scale uses what's already connected. -
Ilman Shazhaev
@shzhv13
Here's the math on a typical AI agent SaaS product. $6M ARR, 10,000 users at $50/month. Token spend at current API prices: $200k/month, 40% of revenue. That's 60% gross margin on paper.
Apply the actual 2026 trajectory. Token prices are down. Token consumption per useful task is up by an order of magnitude. The net effect on AI product economics is flat or rising inference spend. Products that modeled a 30% decline in inference costs by year-end (most still do) are short by $720k in projected gross margin this year.
Now stress-test the other way. If compute costs go 2x, token spend goes to $400k/month, gross margin collapses to 20%.
20% doesn't fund engineering, sales, GTM, or the next round of model upgrades. The product becomes unfundable inside 12 months.
3 exits. Raise prices (most can't, retention drops at 1.5x). Switch models (works for some use cases, fails for agents and multi-step reasoning). Or own the compute layer.
@FARLabsAI was designed for the third path. Inference on hardware already amortized has a different marginal cost equation. The product economics that sit on top of it work differently.
We built it for that exact math. AI products run on distributed compute across consumer and enterprise GPUs, with cost insulated from upstream API price moves. Operators with idle compute monetize it through the network. -
Ilman Shazhaev
@shzhv13
Most of the AI capital still gets allocated like the scaling laws didn't change in 2024.
The frontier moved from raw pre-training scale to runtime. RL post-training and inference-time reasoning are now where new capabilities come from (o-series and DeepSeek-R1 are the obvious anchors). Both axes pull value away from the biggest cluster and toward the smartest runtime.
But most institutional capital and most builder roadmaps still anchor to 2022 assumptions, when pre-training scale alone defined frontier progress. The architecture shift is roughly 18-24 months ahead of where the money is flowing.
The interesting opportunity right now is the runtime layer that captures economics most teams treat as a footnote. -
Ilman Shazhaev
@shzhv13
Most engineering orgs are running on AI infrastructure they didn't design and don't share. Uber recently said 95% of their engineers use AI tools monthly, with AI agents generating thousands of code changes weekly. The trend is identical across the industry, but most of the productivity gain stays locked inside individual context windows.
We structured this differently at @Dizzaract. The AI Department gives each product team its own prompt systems and curated data knowledge, with @FarcanaGame, @FARLabsAI, and @gamedofficial_ all running different stacks because the work is different in each.
AI-native teams need organizational architecture for their AI workflows, the same way engineering teams need architecture for code. The gap between that approach and treating AI tools as personal productivity software will widen quickly. -
Ilman Shazhaev
@shzhv13
Same at the infrastructure layer. Winning systems will be the ones routing inference fast enough to keep agents deciding instead of stalling.
François Chollet
@fchollet
Decision making was the bottleneck all along. Productivity is the rate at which you make open-ended decisions, the rate at which you reduce future paths.
-
Ilman Shazhaev
@shzhv13
When we started building @FARLabsAI, we chose to skip centralized data centers and build distributed inference on consumer and enterprise GPUs instead.
The conventional path in our space is well-defined. Raise $200M+ in institutional capital and build centralized facilities around NVIDIA GPU allocations. That's where the comparables are and where the exit path is clearest.
We went the other way because of unit economics. Centralized infrastructure amortizes CapEx over 18–24 months. We have no CapEx. The hardware is owned by gamers and enterprise operators with idle capacity. The unit economics work differently from day one.
Standard objections to distributed compute don't survive the math. Reliability comes from redundancy, and latency for inference (not training) is better when you route to the nearest active node. The compute already exists everywhere. What was missing was coordination and software trust.
We've spent the years since building the layer that solves both. Training stays centralized while inference distributes to where the users are.
3 billion GPUs are sitting idle worldwide. The compute was already built, while everyone else was debating who'd build the next hyperscaler. -
Ilman Shazhaev
@shzhv13
I know I've been quiet on here lately that's because me and the team have been heads down building a lot behind the scenes.
But I wanted to take a moment to personally say thank you.
Over 5,000 of you have signed up to become a node operator. That's incredible, and we don't take that trust lightly.
FAR Labs is building something we truly believe in a distributed AI network powered by people, not corporations.
A lot more is coming very soon. Thank you for being here from day one. This is only the beginning. -
Ilman Shazhaev
@shzhv13
This is the first step toward turning idle compute into real infrastructure.
We’ve been building quietly for a while now we start opening it up.
If you have GPU compute this is where it begins 📷
FAR Labs
@FARLabsAI
FAR AI node registrations are now open.
Turn your GPU into a revenue-generating node powering real AI workloads.
Supports a wide range of GPUs.
Open globally anyone with idle compute can apply.
Register now: farlabs.ai/join-network -
Ilman Shazhaev
@shzhv13
FAR AI is moving into high-load closed testing.
We’ve moved past development and are now ensuring our production-ready infrastructure holds up under intense pressure. This week, we’re stress-testing routing, load, and parallelism all at once.
We are officially in the pre-launch phase, and the team is hitting major milestones daily.
More updates coming this week.
FAR Labs
@FARLabsAI
GREAT NEWS: FAR AI core systems development is complete.
This week, we’re entering closed testing with partners - stress testing the entire platform for mass node and user scale.
Request routing. Load balancing. Model parallelism. Verification. All of it, under real-world conditions.
We’ve been building distributed AI infrastructure that’s efficient, faster, and more scalable than centralized cloud. Now we prove it works at scale.
Closed testing starts now. Public launch coming soon.
The future of AI inference is distributed. And it’s closer than you think. -
Ilman Shazhaev
@shzhv13
Regional instability doesn’t end the Gulf’s AI ambitions, it accelerates its evolution!
The next phase of GCC infrastructure:
- Sovereign control
- Distributed capacity
- Hardened failovers
At @FARLabsAI, we believe resilience is the new scale. -
Ilman Shazhaev reposted
Dizzaract
@Dizzaract
Eid Mubarak from Dizzaract team.
Wishing you joy, peace, and success. -
Ilman Shazhaev
@shzhv13
ChatGPT is the most overrated AI tool.
I say this as someone building AI infrastructure.
It's useful. But it's neural networks that existed for decades.
The hype makes people think AI started in 2022. It didn't. -
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev
@shzhv13
China. During my MSc in robotics and AI at Harbin Institute of Technology.
Foreign student. Foreign language. Foreign everything.
The Chechen codex says honour is earned, never claimed. In China, I learned that principle has no translation barrier.
You prove yourself through work. The rest is noise. -
Ilman Shazhaev
@shzhv13
A 7-person AI research team recently outperformed frontier models from Google and Anthropic.
Their optimization budget... less than $100K.
Meanwhile, foundation model training runs cost hundreds of millions.
The difference wasn't resources. It was which layer they optimized.
There are three layers of leverage in any tech company:
Layer 1: Doing the work
Writing code. Serving users. Shipping features.
Layer 2: Building systems
Automation. Pipelines. Processes that scale the work.
Layer 3: Building systems that build systems
Platforms. Generators. Infrastructure that compounds.
Most founders stop at Layer 2. They build good systems and feel productive, but growth eventually plateaus.
The companies that dominate operate at Layer 3. AWS builds infrastructure for running applications. Unity builds systems for making games. Shopify builds platforms for running stores.
That small AI team? They built a meta-system that generates optimized reasoning systems. When GPT-6 or Claude 5 drops, their system automatically gets better. Zero additional investment.
In gaming: build engines and platforms that make each game easier to ship.
In AI: build infrastructure that improves with every deployment.
In infrastructure: build systems that automatically adapt to tomorrow's constraints.
That's where the 10X lives.
Which layer are you operating at? -
Ilman Shazhaev
@shzhv13
AI replacing middlemen is inevitable...
Most service industries are just information arbitrage with 30 years of regulatory capture in my opinion
sui
@birdabo
florida man sold his house in 5 days using chatgpt without real estate agents
> asked AI for pricing.
> marketing.
> best listing day
> contract drafting.
bro easily got 5 offers in 72 hours.
- also saved 3% commission.
real estate agents are cooked 😭😭😭This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Found something interesting on GitHub today.
It's a repo that lets you build an AI agency. But not the usual way where one AI tries to do everything and ends up being mediocre at all of it.
This one's different. It's structured like an actual company with departments and specialized roles.
You get 54 AI agents split across different teams:
- Engineering has 7 (frontend, backend, mobile, AI, DevOps, prototyping, senior dev)
- Design has 7 (UI/UX, research, architecture, branding, visual work, image generation)
- Marketing has 8 (growth, content, social platforms, app store optimization)
- Product has 3 (sprint planning, trend research, feedback analysis)
- Project management has 5 (production, coordination, operations, experiments)
- Testing has 7 (QA, performance, API testing, quality checks)
- Support has 6 (customer service, analytics, finance, legal, reporting)
- Spatial computing has 6 (XR, VisionOS, WebXR, Metal, Vision Pro)
- Specialized has 6 (orchestration, data, sales, distribution)
Each agent focuses on one thing. They coordinate with each other to get work done.
Got 10k+ stars in under a week.
The concept makes sense to me. You wouldn't hire one person to handle engineering, design, marketing, and finance. So why build AI that way?
Haven't tried it myself yet. Do your own research obviously.
But worth playing around with it and seeing how it actually performs.
github.com/msitarzewski/agency-agents/
If nothing else, it's clear that people who experiment with tools like this early will have a serious edge. -
Ilman Shazhaev reposted
Dizzaract
@Dizzaract
Inside our latest closed playtest for Farcana.
New art. New gameplay. New mechanics.
When we started building @FarcanaGame in 2022, the goal was simple: create something fun and worth playing first.
That mission hasn't changed.
This is the team making it happen - testing, iterating, building. -
Ilman Shazhaev
@shzhv13
This took me 10 years to learn:
Most companies optimize for the wrong timeline.
You don't just need quarterly wins.
You need:
- revenue that compounds
- infrastructure that lasts
- IP you actually own
- systems that get stronger with time
I've watched companies chase vanity metrics.
They hit their slide deck numbers. Raise another round. Then the market shifts and they're gone.
Meanwhile, the ones building for decades own the entire stack.
They're not worried about impressing the next VC.
They're building systems that print money in 2035.
Revenue matters. But sustainable revenue matters more.
Stop optimizing for the next pitch deck.
Start building something that outlasts you. -
Ilman Shazhaev
@shzhv13
This confirms what we have been building toward. The future of AI is not centralized data centers serving billions of requests. It is inference at the edge, on devices people already own.
Perplexity
@perplexity_ai
Announcing Personal Computer.
Personal Computer is an always on, local merge with Perplexity Computer that works for you 24/7.
It's personal, secure, and works across your files, apps, and sessions through a continuously running Mac mini. -
Ilman Shazhaev
@shzhv13
🤯Nvidia is giving away a billion dollar AI product for free.
Here's the play:
Free platform → requires massive compute → Nvidia sells the compute.
Give away the razor. Sell the blades.
The AI agent market hit $10.9B in 2026. 40% of enterprise apps will embed agents by year-end (up from 5% last year).
85% of companies already use AI agents. 93% plan full deployment within 2 years.
Nvidia isn't competing on products anymore.
They're controlling the infrastructure everyone else has to rent.
Training chips → inference chips → deployment platforms.
Own the bottleneck. Make everything else free.
Classic vertical integration disguised as open source.
The pattern is simple: when you control multiple layers of the stack, each one makes the others more valuable.
The companies that win won't have the best product.
They'll own the stack everyone else pays to use.
NemoClaw likely launches next week at GTC.
Watch what happens. -
Ilman Shazhaev
@shzhv13
Simple observation.
Anthropic users consume $5K monthly in API value. Charge them $200.
Mathematics is clear. This doesn't scale.
Yes, $5K is retail. Real compute cost lower. But centralized inference creates structural problem that capital cannot solve indefinitely.
2026: inference will be 85% of enterprise AI spend. Infrastructure determines survival. Not marketing. Not features. Infrastructure.
I built Bitcoin mining facilities in 2017. Pattern was identical every time.
Control infrastructure cost = survive.
Don't control it = die when market corrects.
Physics doesn't care about your pitch deck.
AI follows same laws. Centralized compute is expensive. Will always be expensive. VC money delays the problem, doesn't eliminate it.
Solution is obvious: distributed inference across consumer devices.
Cost per token drops
Margins scale properly
Economics work without constant fundraising
We build this from Abu Dhabi.
Infrastructure layer decides everything. This is not opinion. This is physics.
Markets will prove this correct or incorrect. I believe correct.
Time will show. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
A lot of you have been asking about the NVIDIA-Groq deal and what it means for distributed inference.
Here's the math:
Groq production at Samsung: 15K wafers on 4nm process. At ~$10-15K per wafer, that's $150-225M in annual production value.
Total inference market in 2025: $106B.
Specialized chips represent less than 1% of available inference capacity.
Timeline reality:
Groq acquisition: Dec 2025
Commercial deployment: Late 2026-2027
Broad market availability: 2028+
OpenAI gets priority access. Then Meta, Google, Amazon. Enterprise customers wait 18-24 months.
Inference demand is growing at 19.2% annually NOW. By 2028, the market will have grown from $106B to $170B+.
Distributed compute addresses the 99% of inference demand that specialized chips won't serve, using GPUs already deployed, already paid for, available today.
At FAR Labs, we're building infrastructure for the market that can't wait 2 years for specialized silicon.
GTC March 16. We'll break down what gets announced and what it actually means for access.
TechCrunch
@TechCrunch
Nvidia acquires AI chip challenger Groq for $20B, report says techcrunch.com/2025/12/24/nvidia-acquires-ai-chip-challen…
-
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev
@shzhv13
Anthropic refused $200M Pentagon deal Thursday. OpenAI took it Friday.
48 hours later:
OpenAI:
→ ChatGPT uninstalls up 295%
→ 1-star reviews surged 775%
→ Downloads fell 13%
Anthropic:
→ Claude hit #1 App Store
→ 149K daily downloads
→ Record-breaking signups all week
Anthropic got banned from federal contracts. The public made them #1 anyway.
Lesson here is stay ethical and hold to your principles. You will never lose out. -
Ilman Shazhaev
@shzhv13
AI Agents are fundamentally changing how SaaS economics work.
$285B left software stocks in 30 days. Not because SaaS is dying but because the model is shifting.
Interesting stats:
• 65.6% of US VC → AI/ML (up from 46.4%)
• Startups running 10 AI agents per employee
• Enterprise budgets reallocating 30-40% to AI
• 40% of apps will integrate agents by EOY 2026
Real examples:
→ Salesforce: 9K → 5K support staff using agents
→ Marketing automation: $3K/month → $200 agent alternatives
Gartner: 35% of point-solution SaaS shifts to agent-based by 2030.
When one person with agents does the work of five, per-seat pricing breaks.
We're moving from Software as a Service to AI as a Service.
The bottleneck? Inference cost and compute access.
When 40% of enterprise apps run agents and startups deploy 10 agents per employee, demand for scalable, cost-effective AI infrastructure explodes.
Centralized inference can't scale to meet this.
Half of organizations will put 50%+ of budgets toward AI automation this year.
We're entering the AIaS era. The infrastructure race just started. -
Ilman Shazhaev
@shzhv13
Multi-agent AI systems are crossing the economic viability threshold
$200K salary vs $2K/month compute.
The transition isn't theoretical anymore. It's happening in production environments right now.
The question is deployment speed and integration quality -
Ilman Shazhaev
@shzhv13
2 million players tried it, 97,249 peak concurrent on Steam. 46 days later, it's gone.
The math is simple: if your server and team costs require Fortnite-level retention from day one, you don't have a product problem - I'd suggest you have an infrastructure economics problem.
Centralized architecture creates a binary outcome: massive scale or shutdown.
Unfortunately, there's no middle path when your burn rate assumes you've already won.
Dave Kratzy
@Krtzyy
Can someone smarter than me explain why a game studio would invest $200M and 4 years of development just to shut the game down after 1 month because low player counts?
Like damn at least try to grow a player base for a few months before pulling the plug lol thats crazy -
Ilman Shazhaev reposted
Dizzaract
@Dizzaract
In these challenging days, we are deeply grateful for the UAE’s swift and professional activation of its National Emergency, Crisis and Disaster Management System.
The tireless dedication of our military, civil authorities, and frontline teams continues to keep residents, businesses, and innovators safe and supported across the country.
We are proud of the genuine unity we see every day -Emiratis and our international colleagues standing together, checking in, and moving forward with calm determination.
This shared resilience strengthens the collaborative spirit that powers our work in AI and gaming right here in Abu Dhabi. Thank you to everyone protecting lives and sustaining progress.
The UAE’s preparedness and community strength inspire us all.
From all of us at Dizzaract 🇦🇪 -
Ilman Shazhaev
@shzhv13
General models try to do everything, so they're average at most things.
A model built for one person's actual workflow doesn't waste resources on features they'll never touch.
Specialization is becoming the focus. -
Ilman Shazhaev
@shzhv13
From the heart of Abu Dhabi, I send my solidarity to all across the UAE.
This country is built on unity and mutual respect. In times like this, those values guide us. -
Ilman Shazhaev
@shzhv13
The AI tsunami is coming.
And I'm tracking a critical moment unfolding in MENA.
Dario Amodei just issued a warning the world is ignoring:
→ Superhuman AI by 2027
→ 18-month window before total transformation
→ Complete automation of cognitive labor
He's not winning because he has the best tech.
He's winning the "trust war" - because he's brutally honest about what's coming.
Here's what I'm seeing in our region:
MENA AI funding hit $858M in 2025. UAE captured $519M (up 267% YoY).
AI is projected to contribute $320 billion to MENA's economy by 2030 - roughly 11% of regional GDP.
UAE is building the world's largest AI campus outside the US - 26 sq km in Abu Dhabi with 5GW capacity.
The capital is flowing.
The ambition is real.
But here's the challenge ahead:
Despite massive infrastructure investments, governance frameworks are still maturing across the region.
Nearly half of organizations cite talent shortages and insufficient technological capabilities as barriers to scaling.
Complex AI technologies require specialized understanding. Many AI projects remain in testing phases.
We're building the infrastructure.
Now we need to match it with readiness.
The questions that need answering:
When AI automates 50% of entry-level jobs, are our systems optimized for scale?
When inference costs become prohibitive, who's building the distributed infrastructure?
When the software industry transforms overnight, what does that mean for MENA? Logistics? Finance? Government services?
I've spent the last 10 years building to address these challenges.
Most regional players are celebrating AI's potential.
I'm focused on: "Here's how MENA can lead through this transformation."
At Dizzaract and FAR Labs, we're not building AI products.
We're building infrastructure for regional readiness.
FAR AI is our answer to expensive, centralized inference:
→ Distributed compute network across consumer devices
→ Cheaper, faster, and scalable AI inference
→ Infrastructure that serves any AI application globally - without dependency on foreign data centers
We're building from the ground up. From Abu Dhabi. For the region.
My team of 80+ engineers, researchers, designers, and operators are building the technical systems the region will need:
→ AI-native architecture designed for sovereignty
→ Full-cycle control from IP to infrastructure
→ Execution in weeks, not quarters
This is what we're building for.
In the coming months, when Ministers and CEOs ask:
"How do we scale AI without losing sovereignty? Without becoming dependent on foreign infrastructure? Without sacrificing our values?"
They'll need partners who aren't just talking about readiness, but are building the systems.
Partners who understand both the technology AND the territory.
Partners who are honest about the challenges, while everyone else sells only optimism.
Amodei admitted "no one elected me to make these decisions."
That vacuum exists in every region.
And I'm committed to being part of the solution here.
The future of AI in MENA won't be written in California.
It's being written here.
Right now.
By people who saw this coming, and are building the infrastructure to meet it.
The clock is ticking. -
Ilman Shazhaev
@shzhv13
Do ice baths weekly. Recovery is 2x faster, energy stays high all week.
-
Ilman Shazhaev
@shzhv13
Each dot in this chart represents 31 million people.
2,500 dots = 8.1 billion humans on Earth.
⬜️ Gray dots: Never used AI (84% of humanity)
🟩 Green dots: Free chatbot users (~1.3 billion people)
🟨 Yellow dots: Pay $20/month for AI (~15-25 million)
🟥 Red dots: Use coding tools (~2-5 million)
Look at how tiny that yellow section is.
Only 0.3% of AI users are paying. 99.7% expect it free.
If you're building an AI product, this should inform your infrastructure decisions.
Your users won't all pay $20/month. @ChatGPTapp has 800 million weekly active users, doubling from 400 million in just one year. They're conditioned to expect AI for free or near-free.
And your inference costs? They're likely your biggest variable expense.
Here's the math most AI developers overlook:
- You build an amazing feature
- Users love it
- Adoption grows
- Your inference bill scales linearly
- Unit economics become challenging
- You either raise prices (risk losing users) or subsidize usage indefinitely
The best AI product with unsustainable inference costs becomes a capital efficiency problem.
Inference cost isn't just a backend concern - it's a product strategy variable.
Every AI feature you ship has a cost-per-query attached. Every user interaction hits your infrastructure. With projections of 1.2 to 1.5 billion monthly AI users by early 2026, scale amplifies whatever cost structure you've built.
The questions worth asking:
- What's my cost per inference today?
- At 10,000 users? 100,000? 1 million?
- Can I sustain the product-market fit I'm pursuing?
- What infrastructure choices optimize for economic scalability?
Most AI developers optimize for model quality and user experience first.
Fewer optimize for inference economics from day one.
Historical context suggests a pattern:
Storage costs dropped 1000x over two decades. Bandwidth costs collapsed. Compute costs followed similar trajectories.
AI inference appears to be on this path. Current data shows inference costs decreasing by 10x annually, but we're still in the early phase. Centralized GPU infrastructure has physical and economic constraints at global scale.
The active development areas: quantization, distillation, distributed architectures, specialized hardware.
⬜️ The 6.8 billion gray dots represent the real variable.
84% of humanity hasn't used AI yet. When they do, they'll expect the same experience the 1.3 billion green dots are getting, instant responses at zero cost.
The infrastructure that can serve 8 billion users at near-zero marginal cost doesn't exist yet at scale.
But the demand curve suggests it will need to.
Worth considering in your architecture decisions. -
Ilman Shazhaev
@shzhv13
Everyone's scared to open source because "competition will copy us."
Brother, if your only moat is that people can't see your code, you don't have a moat.
Real advantage comes from:
- Shipping faster
- Understanding users better
- Building what people actually want
Code is just the artifact. The process is what matters. -
Ilman Shazhaev
@shzhv13
Wake up call for AI builders 👇
5 brutal truths from over a decade in AI:
1. Overengineering is the silent startup killer.
2. Ship fast, pivot faster - strategy follows proof.
3. Overpromising destroys credibility faster than underdelivering.
4. Deep specialization > shallow diversification.
5. Your product working is better than any pitch deck. -
Ilman Shazhaev
@shzhv13
I think Denmark nailed it here. You can't moderate deepfakes away, but making unauthorized biometric use legally expensive?
That actually forces companies to build real consent systems. I'd rather deal with clear property rights than vague platform policies any day.
Massimo
@Rainmaker1973
In a world where artificial intelligence can replicate a person’s voice or face in seconds, Denmark is stepping forward with a groundbreaking proposal: a copyright law that grants every citizen ownership of their own likeness.
If passed, this law would mean no one — not even AI companies — could legally use your face, voice, or body data without consent. The move comes amid growing global concerns about deepfakes, where digital replicas of real people are used in scams, misinformation, and even political manipulation. -
Ilman Shazhaev
@shzhv13
This is what matters to me....
Not the tech for tech's sake, but what it enables for the next generation. These kids are going to create things we can't even imagine yet.
I'm trying to build the foundation so they have the compute, the tools, and the freedom to do it without barriers. -
Ilman Shazhaev reposted
Dizzaract
@Dizzaract
Dizzaract is a full-cycle AI and gaming development studio.
We operate as an interconnected ecosystem of products, not a single game or app. Dizzaract covers the entire value chain of digital entertainment and AI:
IP Creation → Distribution → Infrastructure
Powered by 80+ engineers, researchers, designers, and operators working across AI and gaming. Our teams build in weeks, not quarters. We prioritize execution and working systems over corporate complexity.
Why We Exist:
Dizzaract was formed around a simple observation: AI is becoming core infrastructure. Gaming is one of the largest digital ecosystems in the world. Yet identity is fragmented. Inference is expensive. Systems are not optimized for scale.
We chose to build where those gaps intersect.
A structured technology group operating across:
- Original IP development
- Gamer identity systems
- AI infrastructure
The goal is long-term efficiency and compounding leverage.
How We're Built:
Farcana | Our flagship gaming IP and universe
A competitive PvP third-person shooter set in a Middle Eastern sci-fi universe with an aesthetic that doesn't exist anywhere else globally, reimagined in an alternative past-future timeline.
GAMED | Our identity, data, and distribution platform
The identity and economic layer for gamers that aggregates cross-platform gaming data and assets into a single intelligent dashboard.
Gamers' digital lives are fragmented across Steam, Epic, PlayStation, App Store, and more. GAMED unifies them into one place that aggregates your full gaming legacy.
FAR Labs | Our AI infrastructure
FAR Labs is an AI-native product lab building infrastructure and AI as a Service (AIaS) systems.
Our flagship product is FAR AI - cheaper, faster, and scalable AI inference based on distributed compute.
FAR AI is a distributed network where gamers can connect their hardware to provide AI compute power. By distributing inference across consumer devices rather than relying on centralized data centers, we enable significantly more cost-effective and scalable AI infrastructure that can serve any AI application globally.
The Flywheel:
• Farcana → original IP and competitive play
• GAMED → unified identity and data systems
• FAR AI → AI infrastructure and AIaS systems
Together, they create compounding leverage across the gaming and AI ecosystem.
Our Advantage:
- Full-cycle control (IP → Distribution → Infrastructure)
- Cultural uniqueness (Middle East-born global IP)
- AI-native architecture
- Products that reinforce each other
- An 80+ person team building for the next 10–20 years from our base in the MENA region
We're not just making games. We're building structured systems where gaming and AI infrastructure intersect.
Creating worlds that deserve being experienced.
This is Dizzaract. -
Ilman Shazhaev
@shzhv13
Interesting thought exercise, but it assumes AI replaces income without creating new economic activity.
Every productivity revolution looked like this mid-transition. The question isn't if we adapt, it's how messy the adjustment period gets.
www.citriniresearch.com/p/2028gic -
Ilman Shazhaev
@shzhv13
Calling it now:
By 2027 the winning AI companies won't have the best models
They'll just be the ones who figured out profitable inference while everyone else burned through their Series A
Distillation, caching, batching - not optional anymore
Difference between a business and an expensive science project. -
Ilman Shazhaev
@shzhv13
Ubisoft just cut €200M to focus on fewer, better games instead of shipping formulaic garbage 👏
LESSON: Stop spreading yourself thin across features nobody uses - kill the bottom 80%, double down on what actually works.
Geoff Keighley
@geoffkeighley
In a new interview with Variety, Ubisoft CEO Yves Guillemot says the company has:
- 2 new Far Cry games in development.
- Several Assassin’s Creed games in development, both single-player and multiplayer experiences. -
Ilman Shazhaev
@shzhv13
PREDICTION: Only 3 types of AI companies will survive the next 2 years.
I've watched hundreds fail because they chose the wrong path.
The hardware crisis made this brutally clear:
Type 1 → Does ONE thing incredibly well (low compute, high value)
Type 2 → Doesn't build models at all (uses APIs, owns distribution)
Type 3 → Sells infrastructure to the other two (picks & shovels play)
Everyone else? Stuck in the middle with expensive custom models and no differentiation.
The worst place to be: High costs, low moat. -
Ilman Shazhaev
@shzhv13
🚨 China just showcased humanoid robots doing martial arts on national TV
451,700 robotics companies
$932B in capital
28,000 units projected for 2026
Everyone's amazed by the choreography
I'm looking at what they CAN'T do:
❌ Adapt to new scenarios
❌ Improvise without training
❌ Coordinate in unpredictable environments
These robots trained the same routine thousands of times. Perfect execution in controlled settings.
But real breakthrough = multi-agent systems that adapt in real-time without scripts
That's the gap between demos and deployment
Why competitive gaming matters:
→ Unpredictable interactions
→ Real-time adaptation required
→ Multiple agents, zero scripts
→ Every match is different
Not entertainment infrastructure. AI infrastructure.
At @Dizzaract, we're creating worlds that solve this - environments where adaptive multi-agent AI gets developed, tested, and deployed at scale. -
Ilman Shazhaev
@shzhv13
The gap between "interesting AI research" and "someone pays for this" kills most startups.
I spent years in applied research before building companies.
Huge lesson I learned:
The gap between "interesting research" and "someone will pay for this" is huge.
The ideas that closed that gap? Problems I'd personally struggled with in operations.
3 ways to find your AI startup idea:
1. The 10-Hour Audit
What task takes you longest each week? Build AI to automate it.
2. The Former Self Test
Your job from 2 years ago = your ideal customer. You know exactly what they need.
3. The Validation Rule
Get 5 people to say "I'd pay X amount for this" BEFORE you build anything.
The best AI founders don't guess.
They solve problems they personally lived with for years. -
Ilman Shazhaev
@shzhv13
Digital immortality as a service, but really just another way to keep MAU numbers up.
Globe Eye News
@GlobeEyeNews
BREAKING:
Meta secures a patent for an AI that can keep a user’s account running after their death by posting, messaging, and video calling using the user’s past behavior. -
Ilman Shazhaev
@shzhv13
Let's be honest: If you're building AI but not using it internally, you don't believe in your product.
Anthropic engineers manage 3-8 Claude instances each to build Claude.
Meanwhile AI startups:
• Build automation with manual ops
• Sell AI agents while hiring for every role
• Pitch "10x" while working 80hr weeks
Before you sell it, use it:
→ 2-week internal beta for every feature
→ Document internal use cases (real > demos)
→ Build for your own pain first
→ Founders must use it 1hr daily
→ <80% team adoption = no PMF
The best AI products are built by teams that can't live without their own AI. -
Ilman Shazhaev
@shzhv13
The bigger they got, the more they compromised:
Anthropic - Founded as the "safety-first" AI company. Their top safety researcher just quit because even they "constantly face pressures to set aside what matters most."
OpenAI - Started as a nonprofit to "benefit humanity." Now they're for-profit with a $500B valuation, Microsoft owns 27%, and they're running ads after Altman said that was "last resort."
Google - Had "Don't be evil" as their motto. Removed it after taking military contracts. 4,000 employees protested. Twelve quit.
Meta - Promised to protect elections and combat misinformation. Shut down the research tool tracking it three months before a presidential election.
It seems every "ethical" tech company eventually compromises. Market pressure always wins.
Scale forces compromise. Plan accordingly. -
Ilman Shazhaev
@shzhv13
BEFORE: Your team copies everything into AI tools. You have no idea what's dangerous.
AFTER: Everyone knows exactly what data is safe vs. off-limits. Zero guessing.
THE BRIDGE: 30-minute data classification exercise.
Sort everything into 4 buckets:
🟢 PUBLIC - any AI
🟡 INTERNAL - approved AI
🟠 CONFIDENTIAL - enterprise AI
🔴 RESTRICTED - no AI
One simple rule: "Would I post this publicly? No? Check the bucket."
Classify your 20 most-used files. Add colored labels. Send rules to team.
Done. Your business is now safer than 90% of companies using AI. -
Ilman Shazhaev
@shzhv13
The Deep Work Protocol:
• No distractions before critical decisions are made
• Block uninterrupted time for technical work
• Meetings happen after the builder's work is done
• One priority. Execute it fully.
Protection of focus is protection of output.
The best builders I know treat their morning hours like sacred ground.
Everything else negotiates around it. -
Ilman Shazhaev
@shzhv13
You can just build things, and you should....
Just think about costs early: real usage often means minimum of $1k–$20k+/mo on APIs or GPUs.
Plan ahead, think of the costs → build sustainably.
OpenAI
@OpenAI
You can just build things.
This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Bear markets separate the signal from the noise.
When capital is scarce, narratives alone don't work. It's when execution really matters.
The companies that survive aren't necessarily the loudest. They're the ones that kept building. -
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev
@shzhv13
Inference costs stacking up fast for devs. $10-20k per year per head... it's time for smarter compute plays in AI scaling.
dax
@thdxr
this costs $20K and it's on consumer hardware and this model is very very good
lot of companies are already spending $10-20k per dev per year on cloud inference
can't believe we're here already -
Ilman Shazhaev
@shzhv13
Speed creates momentum.
Systems create compounding.
One gives you a good month.
The other gives you a good decade. -
Ilman Shazhaev
@shzhv13
The reason most frameworks fail when copied because context is invisible.
The principle behind the framework? That's what transfers.
Learn to extract the logic, not the template. -
Ilman Shazhaev
@shzhv13
Execution is a decision followed by immediate action.
Everything else is preparation. -
Ilman Shazhaev
@shzhv13
When you're building at planetary scale, dependencies become liabilities. Vertical integration is clearly about removing friction between vision and execution.
SpaceX
@SpaceX
SpaceX has acquired xAI, forming one of the most ambitious, vertically integrated innovation engines on (and off) Earth → spacex.com/updates
-
Ilman Shazhaev
@shzhv13
I believe excellence is a form of worship.
What I build is a reflection of the opportunities I’ve been given. -
Ilman Shazhaev
@shzhv13
BREAKING: Amazon is negotiating a $50B investment into @OpenAI
They already have $8B in Anthropic.
Here's what's actually happening: We're watching the AI infrastructure layer consolidate in real-time.
Microsoft got there first with OpenAI and Azure.
Now Amazon is playing catch-up by funding both major players.
The pattern is clear:
Two companies are positioning themselves as the gatekeepers for AI compute.
And that creates problems:
→ Pricing power concentrates
→ Access becomes controlled
→ Innovation routes through two approval processes
→ Geographic limitations get enforced
→ Smaller players get priced out
We've seen this play out before.
Cloud infrastructure started open and distributed. Then it consolidated around three major providers.
Now AI compute is following the same path, but faster.
Here's why this matters:
When infrastructure consolidates, the companies that control it control:
→ Who gets access
→ At what price
→ Under what terms
→ In which markets
The foundation models get the headlines.
But Amazon and Microsoft are quietly building a duopoly on the infrastructure that runs them.
Every AI company, every developer, every startup building with these models, they all have to go through one of few gatekeepers.
That's not a competitive market.
That's a toll road.
The question we should be asking:
Is this the infrastructure model we want for the next decade of AI development?
Or are we trading one set of dependencies for another? -
Ilman Shazhaev
@shzhv13
I learned the hardest part of scaling isn't hiring more people.
It's letting go of the work you're good at.
You built the company doing everything yourself. Now you need to trust others to do it differently than you would.
Different isn't wrong. It's just not yours anymore. -
Ilman Shazhaev
@shzhv13
Everyone's talking about AI replacing jobs...
I'm watching it create entirely new business models.
In gaming, we're using AI to generate adaptive storylines that respond to player behavior in real-time. This wasn't possible 18 months ago.
The companies winning aren't only using AI to do old things faster.
They're doing things that were impossible before. -
Ilman Shazhaev
@shzhv13
PC dominance is exactly why I’m building on PC first. It’s where developers, players, and the future of gaming already are.
Shinobi602
@shinobi602
GDC 2026 | Twice as many developers say they want to make games on PS5 (40%) and Switch 2 (39%) than on Xbox, in new survey.
PC remains the most popular platform at 80%.
More ➡️ www.videogameschronicle.com/news/twice-as-many-developers…
#gdc2026 #ps5 #xbox #nintendo #steam -
Ilman Shazhaev
@shzhv13
Elon just discontinued two car models to build robots.
The lesson isn't about cars or robots.
It's about this:
Winners allocate resources to the future, not the present.
Your current revenue stream is comfortable.
Your next revenue stream is uncomfortable.
Most founders optimize for comfort.
Winners optimize for what's next.
Sawyer Merritt
@SawyerMerritt
BREAKING: Elon Musk has announced that @Tesla is discontinuing the Model S and Model X in Q2 2026.
"We are going to convert that production space to an Optimus factory. It's part of our overall shift to an autonomous future." -
Ilman Shazhaev
@shzhv13
Revenue solves most problems.
But founders treat it like the final boss instead of level 1.
Get to revenue fast. Even small revenue.
It changes everything:
- Your confidence
- Your team's belief
- Investor interest
- Your leverage
Profitable beats fundable. -
Ilman Shazhaev
@shzhv13
The difference between a $1M company and a $100M company isn't the idea.
It's the founder's willingness to kill their darlings.
That feature you love? Customers don't use it.
That strategy you're attached to? It's not working.
Ego is expensive. -
Ilman Shazhaev
@shzhv13
I've hired over 200 people across 3 continents.
Here's the pattern I've noticed:
A players ask about the problem.
B players ask about the role.
C players ask about the perks.
Hire people who are excited about what you're solving, not what you're offering. -
Ilman Shazhaev
@shzhv13
Did you see @KhabyLame just licensed his AI twin for close to $1B?
The $975M deal proves digital identity has real value.
At FAR Labs, we're building AI twins for gaming.
Here's why this matters:
Gaming has always been about identity. Your character. Your reputation. Your legacy.
But that identity has been trapped...locked in one game, controlled by publishers, gone when servers shut down.
AI Twins change the game.
Imagine:
• Your gaming persona that learns your playstyle
• An AI that represents YOU across multiple games
• A digital identity you actually own
• An entity that can compete, earn, and evolve (even when you're offline)
This is about extending players, not replacing them.
In gaming, that value is even bigger:
Billions of players worldwide.
Massive in-game economies.
Communities that live for decades.
We're building the infrastructure for gamers to own their digital selves.
The future of gaming identity is being built at @FARLabsAI 🔥
Not as NFT profile pics.
As actual playable AI entities... -
Ilman Shazhaev
@shzhv13
Meanwhile, hundreds of millions of GPUs sit idle...
Pirat_Nation 🔴
@Pirat_Nation
New AI data center buildout being done in secret location to avoid backlash from local residents
Applied Digital (a company that originally focused on blockchain/crypto mining but has pivoted heavily to AI/high-performance computing infrastructure).
They're quietly constructing a massive 430 MW AI data center campus at an undisclosed location in the southern United States.
The company is deliberately keeping the exact site secret for now -
Ilman Shazhaev
@shzhv13
After building in the Middle East I've come to the conclusion:
The best opportunities aren't where everyone is looking.
They're where infrastructure is growing faster than competition.
Find the gap between ambition and execution.
That's where unicorns are born. -
Ilman Shazhaev
@shzhv13
The gaming industry has a 90% failure rate.
Here's what the 10% do differently:
They don't build what they want to play.
They build what people want to pay for.
Passion starts companies. Market demand scales them. -
Ilman Shazhaev
@shzhv13
Building a unicorn isn't about having one brilliant idea.
It's about executing 1,000 small decisions correctly.
Hire the right person. Choose the right tech stack. Price it right. Position it right.
Excellence compounds. Mediocrity multiplies. -
Ilman Shazhaev
@shzhv13
The biggest lie in tech:
"Move fast and break things."
Wrong.
Move fast and build things that don't break.
Speed without systems is just chaos with a good PR strategy. -
Ilman Shazhaev
@shzhv13
The best system I built this year:
"If it takes less than 5 minutes and will be done more than 5 times, automate it."
Saved my team 15 hours per week.
That's 780 hours per year.
That's a full-time employee doing high-leverage work instead of repetitive tasks. -
Ilman Shazhaev
@shzhv13
I've raised capital in 3 continents.
Here's what nobody tells you:
Investors don't fund ideas. They fund inevitability.
Your job isn't to convince them it might work.
It's to make them terrified of missing out on something that's already working. -
Ilman Shazhaev
@shzhv13
Most founders think AI will replace their team.
Wrong.
AI amplifies your best people and exposes your worst systems.
I've watched companies 10x output with the same headcount.
The difference? They automated decisions, not jobs. -
Ilman Shazhaev
@shzhv13
Building in Abu Dhabi showed me what most Silicon Valley founders miss:
Geography is no longer destiny.
Talent is global. Capital is global. Distribution is global.
The only thing that's local is your excuse for not starting. -
Ilman Shazhaev
@shzhv13
The gaming industry taught me something counterintuitive about building tech:
Your users don't want more features.
They want fewer friction points.
Every feature you add is a bet. Every friction point you remove is a guarantee.
Remove first. Build second. -
Ilman Shazhaev
@shzhv13
5 AI shifts happening right now that will reshape how you build your business:
(Most founders are missing #3)
👇🏼Post not available here, view on X
Ilman Shazhaev
@shzhv13
6/
Shift 5: Moats are shifting from data to taste
Everyone has access to the same AI models.
Your differentiation?
Curation
Design decisions
User experience
Brand positioning
Technology commoditizes. Taste compounds.
Ilman Shazhaev
@shzhv13
7/
The companies winning in 2026 aren't the ones with the most AI.
They're the ones who reimagined their entire business model around what AI makes possible.
Don't bolt AI onto old processes.
Rebuild the process around AI. -
Ilman Shazhaev
@shzhv13
OpenAI’s $14B+ projected losses highlight a major flaw in the current AI model: centralized inference is too expensive to scale.
Decentralized AI reduces costs, boosts efficiency, and moves compute closer to users.
It's the way forward....
Pirat_Nation 🔴
@Pirat_Nation
OpenAI is in the red and, at this rate, will run out of money by 2027, having overestimated the profitability of AI.
Recent analyses and internal projections indicate annual losses of around $14 billion in 2026, following roughly $8 billion in 2025.
This stems from overestimating near-term profitability, as high inference and data center expenses continue to dominate despite strong user adoption and billions in revenue.
Without repeated massive funding rounds, analysts warn the company could face a cash shortfall as early as mid-2027. -
Ilman Shazhaev
@shzhv13
I’ve seen how centralized AI works up close.
It’s not about progress.
It’s about control.
Want real innovation?
Open it up.
Decentralized AI will be faster, safer, and more aligned with humanity’s future. -
Ilman Shazhaev
@shzhv13
Most founders don’t fail because of their product.
They fail because nobody knows it exists.
If you’re not building distribution while building the product, you’re building in silence.
X is free. Attention isn’t. Use it while you can. -
Ilman Shazhaev
@shzhv13
Huge well done to the team 👏
FAR Labs
@FARLabsAI
🏅 Gold Team Verification achieved!
FAR Labs has officially received the Gold Badge from @CertiK the highest level of KYC & team verification.
It may look like we achieved this level very easily but this is a combination of experience, dedication, going through low level details of the operation and teamwork coming together in harmony.
Not only this but we are scored in the TOP 5% on code security. It is like being in the top 5 of the ladder. Not a small feat...
This strengthens trust, security, and transparency across our ecosystem.
Onward 🚀 -
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev
@shzhv13
Steam wins by scale. GOG wins by principles
Steam releases hundreds of games a day most are forgotten.
GOG focuses on fewer games, DRM free ownership, and preservation.
I personally believe the industry doesn’t need more games. It needs better games that last and build real communities, that's what we are building at @FarcanaGame
Pirat_Nation 🔴
@Pirat_Nation
GOG's new owner says Steam is winning due to ease of use, not quality, while criticizing the platform for releasing hundreds of games daily that are "not super high quality."
He points out that GOG stays curated and avoids releasing hundreds of games daily. The focus remains on DRM-free ownership, preserving classics, and building a better experience without copying the volume-driven approach. -
Ilman Shazhaev
@shzhv13
🤯Next-gen pricing is breaking the upgrade cycle in the PC industry...
• RTX 3060 is being brought back
• DDR4 production is ramping up
• DDR4 motherboards are flooding the channel
• AM4 refuses to die
When consumers hesitate, vendors won't push forward but retreat to what still sells.
Right now, that’s last-gen. -
Ilman Shazhaev
@shzhv13
Do not start building in AI or Gaming if you're primarily driven by:
- Recognition
- Quick money
- Validation
- Fear of missing out
Only do it if you're willing to grind in silence when no one's watching.
Because the real reason is to create something that users truly connect with. -
Ilman Shazhaev
@shzhv13
Gaming has been constrained by centralized compute and slow, expensive AI inference...
At @dizzaract and @FARLabsAI we’re building infrastructure that lets real-time AI live inside gameplay and turns gamer hardware into a global inference network.
This interview I had at Solana Breakpoint 2025 breaks down how it works:
www.youtube.com/live/TwuhPdPZle4 -
Ilman Shazhaev
@shzhv13
GM…every day is a gift from God 🙏
-
Ilman Shazhaev
@shzhv13
Are we ready for AI superintelligence?
Here are my key takeaways from the ‘AI 2027 document’ a research-backed AI forecast by Daniel Kokotajlo, Scott Alexander, Thomas Larsen, Eli Lifland, Romeo DeanPost not available here, view on X
Ilman Shazhaev
@shzhv13
The core signal the document keeps pointing to.
• 2025 is tooling.
• 2026 is autonomy.
• 2027 is feedback loops and loss of the human lead.
Once AI improves AI, progress compounds faster than human institutions can react.
Ilman Shazhaev
@shzhv13
Read the full document here:
ai-2027.com -
Ilman Shazhaev
@shzhv13
China’s gaming market crossed $50B in 2025 🔥
That tells us one thing clearly.
- Demand never left.
- Regulation slowed releases.
- Players stayed.
- Now growth is back.
What makes this cycle different is AI...localization, QA, live ops, moderation, support.
All faster. All cheaper. Smaller teams.
Ignore China and you fight harder in saturated Western markets while giving regional leaders time to reset.
This is about designing for China before the window gets priced in.
Curious how other gaming founders are handling this in 2026... -
Ilman Shazhaev
@shzhv13
Jensen Huang the CEO of @nvidia is talking about the future of AI today in his #ces2026 presentation...
Here’s what he won’t say out loud.
• Centralised AI breaks under its own weight. Power limits, geopolitics, and capital intensity kill the model.
• The next AI wave does not run on hyper-scalers alone. It runs on decentralised GPUs.
• DePIN turns idle hardware into a global compute market. This happens whether incumbents like it or not.
• Central clouds become the choke point. Pricing, access, and censorship follow.
• Decentralised inference wins because it is cheaper, harder to shut down, and closer to users.
• Training stays semi-centralised. Inference fragments across thousands of nodes.
• The real competition is not model vs model. It is centralised control vs permissionless infrastructure.
• AI agents need uptime and redundancy. Decentralised networks provide both without single points of failure.
• Big Tech fights DePIN publicly. Privately, they study how to copy it.
• GPUs stop being assets on balance sheets. They become yield-generating machines.
• The biggest winners are not model labs. They are networks that route compute, data, and payments.
• AI sovereignty becomes a real demand. Countries and companies choose decentralised stacks to avoid dependence.
• If your AI relies on one provider, you built a liability.
• The future AI stack looks more like crypto than cloud.
This is where AI actually scales.
If you think centralised clouds win forever, explain why power, politics, and price suddenly stop mattering. -
Ilman Shazhaev
@shzhv13
This is a clean reset.
Last year is done.
Wins, losses, lessons, all locked in.
This photo reminds me of the mindset I want for the year ahead.
Calm.
Focused.
Thinking ahead.
Before we rush into 2026, do this.
Look back for five minutes.
• What worked in 2025
• What wasted your time
• What you avoided fixing
Then look forward.
• Pick one direction
• Make fewer commitments
• Play the long game
Most people sprint into January.
They repeat the same patterns by February.
You get better results by slowing down.
By choosing your moves.
By committing to what actually matters.
2026 rewards clarity and patience.
Make this one intentional -
Ilman Shazhaev
@shzhv13
Happy New Year to you all 🥳
Dizzaract
@Dizzaract
Wishing you all a Happy New Year from the team at Dizzaract!
-
Ilman Shazhaev
@shzhv13
Just sat down with @Entrepreneur Middle East…
Building my companies from the Middle East taught me one thing fast.
If a number does not change your decisions, it is not a KPI. It is a distraction.
KPIs exist for one reason.
To force clarity.
To force action.
To kill guesswork.
Pick a few metrics.
Set the target.
Review them on a fixed schedule.
Act when they move.
Thanks to the Entrepreneur Middle East team for the conversation.
The full interview just dropped on YouTube.
Link is in the comments.
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev
@shzhv13
🚨JUST IN: Nvidia $NVDA is in advanced talks to buy AI21 Labs for $2B to $3B.
If you use AI, this matters to you.
Nvidia is now discussing a $2B to $3B acquisition of AI21 Labs.
What that means in simple terms:
• Nvidia already controls most AI chips.
• AI21 builds large language models.
• One company could control the machines and the brains.
Why this affects everyday users:
• Fewer independent AI tools.
• Less competition between models.
• Higher costs passed down to apps and users.
• More control over what AI can and cannot do.
This is how AI power concentrates.
First the hardware.
Then the models.
Then access.
The AI we use tomorrow depends on deals like this today. -
Ilman Shazhaev
@shzhv13
🤯 Just in: UFC Champion Khamzat Chimaev took a new belt, this time The Belt of Desert
G-Wagon vs Liwa dunes… dunes by KO
Mad Max 3: Chechen Edition - Liwa Festival 2026
#liwa #desert #raceThis media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
You are closer to the impossible than you were yesterday.
That is how progress looks.
Quiet.
Slow.
Repetitive.
What separates winners:
- They show up without motivation.
- They repeat simple work.
- They finish.
Most people quit right before it works.
If you are still moving, you are ahead.
Comment one thing you will commit to fully in 2026.
Write it like you mean it.
Ilman Shazhaev
@shzhv13
I'll go first...health and fitness 💪🏼
-
Ilman Shazhaev reposted
Ilman Shazhaev
@shzhv13
Centralized infra exposed again via MongoBleed time for decentralized solutions in gaming
-
Ilman Shazhaev
@shzhv13
🤯 64gb of RAM costs the same as a 2025 MacBook with 16gb of RAM
Why?
This pricing isn’t accidental.
Big tech controls compute by design.
They restrict upgrades.
They bundle parts you don’t need.
They push full replacements every cycle.
RAM costs pennies at scale.
You pay hundreds because access is gated.
Gamers feel it first.
AI builders feel it next.
Everyone else pays later.
Compute is power.
Ownership breaks their margins.
Renting keeps you dependent.
This is why decentralized compute matters.
If you don’t own your hardware, you don’t own your future. -
Ilman Shazhaev
@shzhv13
The LARGEST debut media event in the world just took place…
@WMediaBridge Bridge summit in Abu Dhabi (Dec 8–10) was big energy - first edition, global room.
Seeing platforms, creators, brands, and media leaders all in one place makes one thing clear: if you’re building from Abu Dhabi, you don’t need to fly the world in… it’s already here.
@AboFlah_1 fans are merciless, he been attacked right after the panel discussion 😁🦾
Thanks @EntMagazineME for the interview 🤝
Congrats @WMediaBridge on the launch.
#bridgesummit
#bridgesummit2025
#worldmediabridge
#abudhabi
#abudhabieventsThis media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Everyone wants speed.
Very few respect timing.
Real progress happens when you slow your pace and sharpen your direction.
Rushing decisions feels productive. It usually creates noise.
Pausing creates clarity. Clarity creates leverage.
High performers do this differently.
They step back to think.
They protect focus.
They move with intent, not urgency.
When you slow down, you see what actually matters.
When you see clearly, every move compounds faster.
If things feel busy but not effective, that is your signal.
Less motion. Better decisions.
Slow down long enough to choose the right road.
You will arrive faster than everyone sprinting blindly. -
Ilman Shazhaev
@shzhv13
Abu Dhabi just hosted @officialgotf Games of the Future (Dec 17–23), and a team from my alma mater won the robotics championship.
The Xuanzhi Technology team from Harbin Institute of Technology (HIT) @HIT_1920 took the robot combat title with a machine they built themselves: “Deep Sea Shark 3.”
I’m genuinely proud.
Not just because it’s a great win, but because HIT’s School of Mechatronics is where I was grown as an engineer. That’s where I studied swarm intelligence in robotics: how multiple agents coordinate, adapt, and outperform as a team.
And it’s funny how life loops back.
Today I’m building @Dizzaract here in Abu Dhabi, an AI and gaming studio, and I still think in the same systems mindset. Agents, behavior, competition, teamwork, real-time strategy… just translated into games.
HIT is ranked in TOP 10 best STEM universities in the world. It builds serious engineers, especially in robotics and applied AI.
Also, there’s something special about this happening here in Abu Dhabi. This city is becoming a real arena for global talent. 850+ participants from 60+ countries competed across 11 disciplines at this event.
Congrats to the HIT team, and respect to Abu Dhabi for hosting a stage where world-class builders can show up and prove it.
If you’re a HIT alum in the UAE, or building in AI, robotics, or gaming, I’d love to connect.
#abudhabi #gamesofthefuture #robotics #ai #hit #mechatronics #swarmintelligence #gaming #dizzaract
Harbin Institute of Technology
@HIT_1920
The 2nd #games of the Future held from Dec 17-23 in Abu Dhabi, United Arab Emirates, saw a remarkable victory by the Xuanzhi Technology team – a student innovation and entrepreneurship team from the #harbininstituteoftechnology (HIT).
-
Ilman Shazhaev
@shzhv13
7 lessons from building at Dizzaract over the last three years:
If you are building at scale, these matter (trust me)
1. Scale punishes weak decisions
Every shortcut shows up later as cost, churn, or chaos. We learned to slow decisions down early so speed stayed high later.
2. Infrastructure is invisible until it fails
We invested in AI systems, tooling, and pipelines before they were urgent. When growth came, nothing broke. That is the point.
3. Talent quality sets the ceiling
Ideas did not move the needle. People did. We hired for discipline, learning speed, and ownership. Output followed.
4. Systems beat hero execution
Relying on individuals does not scale. Clear processes, clear ownership, and clear feedback loops do.
5. Ecosystems compound
Products built in isolation stall. FAR Labs, Farcana, and GAMED were designed to reinforce each other from day one. That reduced risk and increased momentum.
6. Location matters
Building from Abu Dhabi gave us focus, access, and long-term alignment. We treated location as strategy, not background.
7. Standards create credibility
Reputation followed how we operated daily. Consistency over years mattered more than announcements.
The visible results came later.
The real lesson is simple.
Operate today like scale is inevitable.
Your standards decide whether you survive it.
Dizzaract
@Dizzaract
Welcome to Dizzaract...
In just three years, Dizzaract has grown into the largest gaming studio in the MENA region. Gulf News and Arabian Business and others have all recognized this fact. Today, our team is over 80 people strong, building products used by more than one million gamers worldwide.
Those numbers matter, but they are not the reason we exist.
Dizzaract is a world development studio based in Abu Dhabi. We build an interconnected ecosystem of games, AI systems, and infrastructure designed to change how games are created, distributed, and experienced.
Everything we do is built around three core pillars.
• FAR Labs. A separate entity under Dizzaract focused on AI research and development. This is where we build the tools, infrastructure, and decentralized compute systems that power the wider ecosystem.
• Farcana. Our flagship multiplayer hero shooter project. Fast 4v4 PvP combat set in a sci-fi world, built with adaptive AI systems that push competitive play forward
• GAMED. An AI-native identity and gaming platform project with over 820,000 users. It handles onboarding, performance tracking, and rewards across games, creating a foundation for long term player engagement.
These are not standalone products. They are designed to work together under one studio portfolio, forming a new approach to how games reach players and how players stay connected.
Our work goes beyond gameplay. We are building toward a more adaptive and sustainable future for interactive entertainment, from the UAE to a global audience.
If you want to follow how AI, gaming, and technology are being built in real products, in real time, from one of the fastest growing studios in the region, follow this page. -
Ilman Shazhaev
@shzhv13
I learned something about education most people never experience...
When life becomes unstable, learning stops being a choice.
I grew up around war, uncertainty, and loss. In that environment, guessing costs you time. Confusion costs you safety. Depending on others costs you control.
I started paying attention to one thing. The people who understood how things worked made better decisions. They stayed calm. They adapted faster. They found paths when others froze.
That’s when learning became a daily habit for me. Study how systems work. How money moves. How technology changes behavior. How decisions shape outcomes.
That habit followed me everywhere. From instability to studying AI and robotics in China. From being alone in a foreign country to building companies inside complex systems.
Here’s the lesson.
Don’t treat learning as something you do when life slows down. Life rarely does. The habits you build before pressure arrives are the ones that carry you through it.
Staying educated is not about ambition. It’s about staying capable when things change. -
Ilman Shazhaev reposted
Ilman Shazhaev
@shzhv13
If AI is just speeding up the boring parts and the team is transparent about it, I am fine with it. What matters is the final game and whether it feels hand crafted
-
Ilman Shazhaev
@shzhv13
When the world comes to you, everything moves faster.
Abu Dhabi Finance Week @ADFinanceWeek brings global leaders in finance, policy, and innovation into one place, right next to our @Dizzaract HQ. The ecosystem here is built around access, alignment, and real opportunity.
This is what long-term thinking looks like. Capital, regulation, and leadership working in sync, with the support of @ADGlobalMarket and @InvestAbuDhabi creating an environment designed to attract the people shaping the future.
Great to speak with and receive support for the Dizzaract vision from His Excellency Ahmed Jasim Al Zaabi, Chairman of ADGM, and for me and my brother @ishzhv to see firsthand how Abu Dhabi continues to raise its global position with clarity and intent.
Abu Dhabi is setting the standard for how modern financial hubs are built.
#adfw2025This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
The one lesson I would give any founder building in 2025.
You do not lose by being wrong.
You lose by being slow to correct.
Markets now move faster than planning cycles.
AI compresses feedback.
Crypto compresses consequences.
What matters is how fast you notice reality.
What this looks like in practice:
• Shorter build cycles
• Clear signals over opinions
• Fewer roadmaps, more checkpoints
• Decisions owned by one person
The founders who will win in 2026 are not the smartest in the room.
They are the fastest to adapt without panic.
Speed with discipline beats confidence without feedback. -
Ilman Shazhaev reposted
Ilman Shazhaev
@shzhv13
I love secrets like this. A hard unlock tied to performance gives players a real reason to master the game
-
Ilman Shazhaev
@shzhv13
One of the hardest skills in AI right now is knowing where to pay attention early.
Markets rarely announce themselves. What usually happens first is alignment. The right mix of builders, investors, and operators start asking similar questions at the same time. Not because they were told to, but because they are sensing the same gap.
A few takeaways worth sharing if you are building or allocating capital:
• Early opportunities attract curiosity before clarity
• Serious interest forms around direction, not finished products
• The window to get involved is always earlier than it feels
If you are looking for places where AI infrastructure, gaming, and execution are converging, now is the time to lean in, learn, and engage.
That window does not stay open for long. -
Ilman Shazhaev
@shzhv13
Things that keep you steady long term:
• Family that tells you the truth
• Friends who show up when nothing is at stake
• Shared time without phones, agendas, or outcomes
• Places that slow your nervous system down
• Conversations that reset your priorities
Enjoying time with family at my brother’s farm in Dubai reminded me of this.
If you are building hard things, pressure will find you.
Markets change. Timelines slip. Wins fade fast.
Your real edge is your circle.
Protect it early. Protect it often.
I’m blessed to have mine.This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
A really great evening last night for our FAR Labs side event for Solana Breakpoint 2025 at our @Dizzaract HQ in Abu Dhabi.
Thanks to everyone who came and contributed to such a positive eveninga anda special well done to the team for hosting 👏
Heathley
@HeathleyETH
Yakın zamanda katıldığım @FARLabsAI ekibi, Solana #breakpoint2025 kapsamında, Abu Dhabi’deki ofisimizde harika bir side event gerçekleştirdi.
Ben de bu etkinlik için geçtiğimiz günlerde buraya gelmiştim. Emeği geçen herkesin eline sağlık ❤️ -
Ilman Shazhaev
@shzhv13
Gearing up for FAR Labs event happening in 30 mins⏱️
FAR Labs
@FARLabsAI
🔔 ANNOUNCEMENT: FAR Labs HQ Experience for Solana Breakpoint week.
We are hosting a private side event at the @Dizzaract headquarters in Abu Dhabi on Saturday 13 December. This gathering brings you inside FAR Labs and Farcana during one of the biggest weeks in the UAE. You will meet the teams driving our AI infrastructure and gaming ecosystem and see the work shaping our next phase.
You will get insights and direct access to builders, operators and partners. You will hear from leadership, see our HQ from the inside and engage with people who are pushing real progress in AI and gaming.
What to expect:
• A full HQ tour that shows our expansion in AI infrastructure
• Opening remarks from our CEO with a clear outlook on what is coming next
• New updates across the Farcana game and ecosystem
• Access to our builder zones for real conversations with operators and partners
• Entry to the Farcana playtest tournament with a prize pool and merch
Capacity at the venue is limited and this event requires approval. Submitting early increases your chance of being accepted.
📅 Saturday 13 December
🕓 16:00 to 20:00 GST
📍 Dizzaract FZ LLC, Yas Creative Hub, Abu Dhabi
Registration is now open. Apply early before the list fills.
luma.com/duwoume2 -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
💥TODAY IS THE DAY: FAR Migration will start 1pm UTC
Polygon → BNB Chain
We’re excited to announce that FAR is migrating from Polygon to BNB Chain as part of the utility expansion of the FAR Token ecosystem. This marks the evolution from Farcana only utility to a broader FAR Labs powered ecosystem.
We are closing Polygon liquidity pool in 5 minutes⏱️
Migration Details:
Swap ratio: 1:1
Old token: FAR (Polygon)
New token: FAR (BNB Chain)
Contract address: 0xC44F9f08E524669E5DEeEBC9EB142c81edfAd178
Migration portal: Open for 60 days
All self-custody users must migrate manually on our platform farlabs.ai at 1pm UTC
Important: Please make sure to migrate all your FAR tokens. After the 60-day window, the Polygon chain will be fully deprecated as the entire ecosystem transitions to BNB Chain.
This migration ensures FAR can support new utilities across the ecosystem, including gaming, AI infrastructure, and upcoming FAR Labs products.
Timeline:
LP on DEXs open: Saturday, 13th December 12:59 UTC
System executes the swap: Saturday, 13th December 13:00 UTC
Trading for FAR on BNB Chain resumes: Saturday, 13th December 13:00 UTC -
Ilman Shazhaev
@shzhv13
Amazing to have one of the biggest Web3 events only 5 minute walking distance walk from our HQ.
If you are at the event, DM one of our team for an office tour!
Dizzaract
@Dizzaract
🇦🇪 @solana Breakpoint 2025 has officially begun on Yas Island, Abu Dhabi, just a few minutes from our headquarters.
The energy in the Etihad Arena is palpable. The team is engaged in meaningful conversations, and the atmosphere is electric.
The Dizzaract team are connecting with founders, engineers, and builders who are shaping the next cycle of crypto and infrastructure.
Events like this are crucial for anyone involved in the space. They provide insight into the industry's direction when you participate in these discussions.
It's encouraging to see who is actively delivering results.
This marks a strong start to the week for Abu Dhabi and for everyone advancing this space. -
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
🔔 ANNOUNCEMENT: FAR Labs HQ Experience for Solana Breakpoint week.
We are hosting a private side event at the @Dizzaract headquarters in Abu Dhabi on Saturday 13 December. This gathering brings you inside FAR Labs and Farcana during one of the biggest weeks in the UAE. You will meet the teams driving our AI infrastructure and gaming ecosystem and see the work shaping our next phase.
You will get insights and direct access to builders, operators and partners. You will hear from leadership, see our HQ from the inside and engage with people who are pushing real progress in AI and gaming.
What to expect:
• A full HQ tour that shows our expansion in AI infrastructure
• Opening remarks from our CEO with a clear outlook on what is coming next
• New updates across the Farcana game and ecosystem
• Access to our builder zones for real conversations with operators and partners
• Entry to the Farcana playtest tournament with a prize pool and merch
Capacity at the venue is limited and this event requires approval. Submitting early increases your chance of being accepted.
📅 Saturday 13 December
🕓 16:00 to 20:00 GST
📍 Dizzaract FZ LLC, Yas Creative Hub, Abu Dhabi
Registration is now open. Apply early before the list fills.
luma.com/duwoume2 -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
📣 MAJOR UPDATE FOR $FAR TOKEN
$FAR is migrating to BNB Chain to expand the utilities of the token.
Migration takes place on the 13th December, and you will receive full tutorials before the 13th so the process is clear for everyone. More updates will follow as we get closer to the migration window.
Token swap info:
• You will swap your $FAR from Polygon to BNB Chain
• Swap rate is 1 to 1
Migration portal will open on 13th December...
Stay tuned -
Ilman Shazhaev
@shzhv13
Super productive day in Abu Dhabi…
Just sat with the Minister of Economy and several top level executives. Real conversations about AI, gaming, and the future the region is building.
The pace here is on another level.
Serious people. Serious plans. -
Ilman Shazhaev
@shzhv13
Geçmiş olsun @ufukless1, yanındayız!
Very impressed with the solidarity Turkish web3 community showing, hope that this would be a great example for the whole space🙏🏻🤍
Ufuk
@ufukless1
Herkese merhaba,
Benim adım Ufuk. Bu tweeti size hastaneden yazıyorum. Hayatım, beni %97 engelli bırakan ve boynumdan aşağısını felç eden bir dalış kazasıyla bir anda değişti. Ellerimi kullanamıyorum sürekli hastanedeyim. Sürekli enfeksiyonlar, ağrı ve uzun yoğun bakım süreçleriyle mücadele ediyorum. Yıllar geçtikçe bu yolculuğun ağırlığı daha da arttı ve en büyük destekçim olan Annem bile artık bu yükü tek başına taşıyamıyor. Annem beni kaldırıp indirmekten şu an bel-boyun fıtığıyla savaşıyor.
Acil olarak ihtiyaç duyduğum pek çok tedavi ve ameliyat hala bizim için ulaşılamaz durumda ve hem duygusal hem de fiziksel zorluklar dayanılması güç bir hal aldı.
Yine de mücadele ediyorum, tutunuyorum ve daha iyi bakım alabilmek için hala umut ediyorum.
Bu NFT, tedavi masraflarıma destek olmak için oluşturuldu. Yapacağınız her katkı bana umut veriyor ve bu zorlu yolculuğa devam edebilmemi sağlayacak.
“Bir Mint Bir Umut” için @opensea mint linkini aşağıya bırakıyorum. Hepimiz birer engelli adayıyız diyor ve sizleri sevgiyle kucaklıyorum. Herkese teşekkür ederim.
👇
opensea.io/collection/limitless-hope/overview -
Ilman Shazhaev
@shzhv13
UFC Champion Khamzat vs Triceratops…who wins? 😂
Just filmed UFC champion Khamzat in a one of a kind fight against a Triceratops.
I’m not sure who has thicker skin, but you already know who’s tougher in the cage. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
What is FAR AI? 👇
AI demand keeps rising. Compute stays limited. Costs push higher each year.
FAR AI fixes this with a simple idea. Use the hardware the world already owns.
• Over 100 million GPUs sit idle most of the day.
• FAR AI activates them into a global compute layer.
• You get real time inference at a fraction of today’s cost.
• The network grows as more players join.
How it works:
• DSV accelerates inference by about 400 percent through parallel checks.
• PPC makes sure every node is honest.
• A strict 100 GPU cap stops farms and keeps the network open to everyone.
• Consumer GPUs handle workloads once reserved for datacenters.
🎮 What this means for gaming:
• NPCs react with real intelligence.
• Worlds shift based on player input.
• Large models run with the speed of smaller ones.
• AI moments in games finally feel alive.
🎯The vision:
• A unified compute lattice owned by the community.
• Intelligence that scales without new datacenters.
• A path to AI that is cheaper, faster, and fair for all.
FAR AI is how we bring real intelligence into games.
And it runs on hardware already waiting to be used.This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Huge respect to @LandoNorris... Strong performance, sharp execution. Pressure exposes people or it reveals them.
This year he showed he belongs at the top🏆
#abudhabigp -
Ilman Shazhaev
@shzhv13
🐅 Amazing evening in Abu Dhabi private zoo, up close with tigers...
You can feel the calm power. No noise. No proving. No panic.
Enjoyed good conversation with friends about wins that changed our lives, losses that nearly did for the worst, scars left from the journey.
One thing we all agreed on:
The difference between people who “make it” and people who talk about it is simple.
One group keeps their word to themselves when the mood is gone.
If you don't quit on yourself, you stay a tiger...
Be a tiger 🐯 -
Ilman Shazhaev
@shzhv13
Great evening at the @Bybit_Official Institutional Gala Dinner. Thank you @benbybit and Helen for hosting.
Good conversations with many strong teams. We’ve had a long relationship with Bybit, and we plan to develop and make it even stronger.
PS: 才知道了咱们都是东北人😁This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
In private Bybit event...more to come soon👀
-
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
👉 Proof of Inference: FAR Labs Dubai Event Recap 👈
At our side event during @binance Blockchain Week, the team hosted a packed session sharing the first public look at FAR Labs, the launch of our new website, and an introduction to FAR AI, our decentralized compute lattice built for high efficiency AI inference.
Key takeaways:
→ The new FAR Labs website went live during the event
→ FAR AI was introduced by our CEO, showing how distributed GPUs form a unified compute layer for real AI workloads
→ The system uses fair, anti whale design, high efficiency inference, stake free security, and real decentralization
→ Strong interest from top exchanges and funds in what we are building
→ The CEO confirmed plans for buybacks and token burns starting in 2026
→ This sets a clear path for long term token utility and growth
FAR Labs is proud to have officially launched with such success
Special thanks to the team and all who were able to make it, the night was a huge success and we have received amazing feedback.
Onwards we go 🚀 -
Ilman Shazhaev
@shzhv13
Solid presentation by @roboforce_ai today...
Clear models. Clear engineering focus.
One training system. Many robot types.
Humanoids. Arms. Quadrupeds. Logistics units.
The approach scales well because everything starts in simulation.
Safer training. Faster iteration. Lower cost.
Good progress in the robotics field.
The gap between digital agents and physical deployment is getting smaller each month. -
Ilman Shazhaev
@shzhv13
🔥Everyone in crypto just stopped talking…
@PeterSchiff and @cz_binance are sitting down right now. The debate we never thought we’d actually get.
Gold vs Bitcoin.
It’s happening live in front of us.
What will be the outcome?This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
GM GM what a night last night....🔥
-
Ilman Shazhaev
@shzhv13
Always good to see brother
Let’s cook
We are going $FAR again 😎🦾
Ruslan Khairullin
@Rus_Khairullin
Bro… @shzhv13 just dropped nuclear alpha
My brain is still rebooting
Enjoyed the @FARLabsAI event 💪 -
Ilman Shazhaev
@shzhv13
Great atmosphere today. Always good to see so many builders and leaders pushing the space forward together.
Binance
@binance
Amazing energy across #binanceblockchainweek today.
That’s a wrap for Day 1.
We’re back again tomorrow.
Closing the day with The Big Debate featuring CZ and Schiff. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
🚀JUST IN: New website is live!
Go check it out: www.farlabs.ai -
Ilman Shazhaev
@shzhv13
Great to see @binance publishing FAR Labs🔥
Getting ready for the @FARLabsAI event in Dubai, looking forward to seeing you all there and sharing some exciting news while connecting with likeminded builders in the space.
See you soon.
FAR Labs
@FARLabsAI
👀 Excited to announce @binance just published us....
Proud of the team. Proud of the community.
Here is the article as we gear up for the FAR Labs event in Dubai in just 1 hour! 🔥 -
Ilman Shazhaev
@shzhv13
The UAE’s leadership sets a clear standard for how countries should approach AI.
Strong vision, fast execution, real impact.
Honored to build in a place that moves with this level of clarity.
Binance
@binance
His Excellency Omar Al Olama, UAE Minister of AI, shares bold vision as UAE celebrates 54 years.
In a world of constant change and fluctuating currencies, governments must focus on fundamentals.
Since 2015, UAE’s ambitious adoption of blockchain and crypto has built resilience through cycles, driven by clear vision and unwavering commitment.
Technology matters, but it’s the real value and impact, like AI use cases, that truly shape our future. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
🔔FINAL CALL: Join us for the FAR Labs private gathering TOMORROW for Binance Blockchain Week
We are hosting a VIP invite only evening tomorrow at Monkey Bar, Dubai. You will meet people working across AI, DePIN, compute, gaming and Web3 in a setting built for real conversations and direct introductions.
Registration is still open and we are confirming FINAL spots.
📅 December 3, 2025
🕒 18:00 to 22:00 (GST)
📍 Monkey Bar, Dubai
🔗 Register here: luma.com/dlgabett
Entry is based on approval. LAST FEW TICKETS LEFT.
You will connect with investors, funds, exchange teams and infrastructure builders who are focused on the next wave of decentralized compute and intelligence networks. The goal is clear. Bring the right people into the same room and give them the space to speak openly.
🤝 Our event media partners: @ChainwirePR, @mpost_io, @CoinEdition, @blockchainrptr, @KEYDifference, @CPOfficialtx, @CoinFea, @BitcoinNews, @cryptodotnews and Coin Gabbar
We look forward to seeing you there. Register now before the list closes. -
Ilman Shazhaev
@shzhv13
Happy UAE National Day 🇦🇪
This country shows what happens when decisions stay clear, services work, government is efficient and supportive and people feel safe to build. You see it in the growth, the pace, and the confidence of everyone who calls this place home.
Grateful to be here.
What do you think makes the UAE work so well? -
Ilman Shazhaev
@shzhv13
BIG things happening 👀
FAR Labs
@FARLabsAI
⏱️ THE WAIT IS FINALLY OVER
In just 2 days you will see what FAR Labs have been building.
We are about to reveal the foundation of FAR AI, built on new compute, new biometric signals, new science models, and new forms of intelligent play.
Each part is designed to help you run faster, produce more, and open fresh revenue streams across AI, science, and gaming. You will see how these layers connect and how they create a system built for real users, real ownership, and real outcomes.
Be the first to use FAR AI and unlock new revenue.
Follow these steps:
1. Click the link below.
2. Enter your email in the box.
3. Confirm your signup and you are in.
Here's the link 👉 www.farlabs.ai
PS Spots for early access will move fast, act now.This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Honestly sick and tired of Abu Dhabi.
It’s too safe, taxes are too low, the streets are too clean, government is too efficient and they seem to want to help us thrive, crime doesn't exist and cost of living is too cheap for this big city.
I need to get out of here 😩 -
Ilman Shazhaev
@shzhv13
The older I get, the more I value clarity under pressure. The clarity under pressure that doesn’t need attention or approval. The clarity under pressure built from a series of small checks. A new habit. A different choice. A higher standard. Small checks create meaningful outcomes.
-
Ilman Shazhaev
@shzhv13
The energy in these moments is contagious. We are moving fast, the plan is working, and the next jump is already in view 🔥
-
Ilman Shazhaev
@shzhv13
Awesome to see @nina_rong join the BNB team as Executive Growth Officer. Met her in early 2022 when she helped bring Farcana to the Arbitrum community - and she is hands down one of the most professional, effective people I’ve worked with in Web3. Her support to us a very early stage startup has significant impact to our long road and growth, for what we always will be grateful. Big congrats, Nina. Well deserved. Wish you all the best at new role.
Nina Rong
@nina_rong
I am excited to announce that I have joined BNB Chain as the Executive Director of Growth.
In 2021, I joined the crypto industry when blockchains were congested. It was a time when a single transaction could cost thousands in USD value. I had the extreme pleasure of contributing to the go-to-market of one of the most successful scaling solutions of Ethereum.
Fast forward to 2025, there are over 150 L2s, and many more L1s with magnitudes higher throughput than Ethereum. Yet, 99% of blockchain ecosystems launched in the past 24 months fail to bootstrap users or developers.
The crypto industry is facing a new challenge – the challenge of discovery.
Today, less than 10% of assets on-chain are activated. The truth is that crypto products struggle to reach retail users, and retail users often face challenges with liquidity discovery.
How BNB Chain is solving the discovery problem better than others:
Mass user base: With over 695M unique addresses and 2M+ daily active users, BNB Chain is the ideal distribution channel for innovations to reach mass adoption.
Optimized for liquidity discovery: The BNB Chain ecosystem has had 619 BEPs and 20 hard forks in the past 6 years to minimize the cost of discovery: reducing latency by 24x, increasing throughput by 40x, and reducing fees by 300x.
Developer support: This is where I am most excited. In this role, I will be focusing on creating programs and platforms to onboard early-stage projects and global talent to the BNB ecosystem, providing them with access to resources and the mass user base of the BNB Chain ecosystem.
Drop me a DM if you are 1) interested in building on BNB Chain, 2) looking for job opportunities at BNB Chain Foundation, or BNB Chain ecosystem in general ✉️
Nuff talk today.
We don’t talk the talk, we Build N Build 💛
@BNBCHAIN @heyibinance @cz_binance -
Ilman Shazhaev
@shzhv13
Game engines are becoming the REAL testing grounds for AI...I broke this down in my new piece for @observer:
observer.com/2025/11/gaming-training-next-gen-ai/ -
Ilman Shazhaev
@shzhv13
Fascinating research...🧠
Dizzaract
@Dizzaract
Did you know: Playing video games helps your brain work better...
A meta analysis in Nature Human Behaviour reviewed more than twenty controlled studies. Action game players showed faster processing and stronger working memory.
University of Rochester research recorded higher activity in key brain regions during rapid tasks. A trial in Psychology and Aging showed executive function gains in older adults after regular gaming sessions. This aligns with our direction at Dizzaract.
We build interactive worlds that keep players mentally active through structured challenges and advanced AI systems.
Healthy cognitive activity should be a core outcome of modern game design. -
Ilman Shazhaev
@shzhv13
The shocking part about Grok...
People talk about @grok's tone because it feels different from the usual polite safety script. The important part sits under the surface. Grok trains like a system built for real uncertainty, not a product shaped by committees.
You see it in how each group works.
- OpenAI is engineered for 'acceptable replies'
- Google is engineered for 'brand safety'
- Anthropic tunes for 'harm control'
xAI pushes for raw signal extraction, the same mindset that lets SpaceX land hardware at rates the industry said were impossible.
You get shorter training loops, wider parameter tests, tighter tolerance for messy data, quicker checkpoints, and heavier multimodal use.
The result shows up in the jumps. Grok 1 to 1.5 to 2 moved faster than the old 3 to 3.5 to 4 chain from OpenAI. No other lab ships at that pace.
If this curve holds, how soon do you see Grok pass the groups still debating policy memos instead of improving model performance? -
Ilman Shazhaev
@shzhv13
Clear sign that demand for decentralized compute is no longer theoretical. Markets are voting with real revenue and real usage. This wave is only getting started.
Cointelegraph
@Cointelegraph
🔥 UPDATE: Decentralized GPU provider Aethir hit $39.8M in Q3 revenue and $147M+ ARR.
Aethir also reports 435K+ GPU containers and 1.4B AI compute hours, making it a leader in DePIN computing. -
Ilman Shazhaev
@shzhv13
Your life shifts the moment you understand that training rewires how your mind works. It gives you proof that you can make a choice, follow through, and get a result you aimed for.
That simple loop starts showing up in every part of your life once you experience it in your body. -
Ilman Shazhaev
@shzhv13
Centralised AI kills freedom.
Five labs holding all intelligence creates silent dependence.
Open networks create safety and clarity.
It's time to start building access that doesn’t need someone else’s permission. -
Ilman Shazhaev
@shzhv13
Who's missing from this circle?
-
Ilman Shazhaev
@shzhv13
there's a Chechen saying that goes, “Bеdеrе ma xhazha, stage xhazha,” meaning “Don’t scrutinize the clothes, observe the man.”
I grew up with that idea shaping how I judge people and how I expect to be judged.
Respect comes from how you act when no one is watching.
Identify your edge, then choose situations that reveal it and avoid the ones that reduce you to surface noise. -
Ilman Shazhaev
@shzhv13
It doesn’t make sense to continue wanting to succeed if you’re not willing to carry the habits that come with it.
If you don’t want to live the lifestyle required for real progress, then release yourself from the desire for big outcomes. To crave the result but not the process, is to guarantee disappointment. -
Ilman Shazhaev
@shzhv13
I still think regulators should allow a new company suffix...Not Limited, but Unlimited.
Imagine registering a business as “Dizzaract Ultd”.
It just feels accurate.
Why limit a company that plans to build bigger than anyone expects? 😉 -
Ilman Shazhaev
@shzhv13
Underrated founder advice: learn the difference between predicting the future and designing for it. Predicting is loud. It is easy to do. Designing is quiet. It is hard to build. Predicting collapses when the market shifts. Designing holds its value long after conditions change.
-
Ilman Shazhaev
@shzhv13
Something everyone watching the shift in tech should remember:
R&D follows environments that give builders room to think and room to work.
As many founders say: “You go where you can build without noise.”
You see more global teams exploring Abu Dhabi when they want ambition, not distraction.
Staying in old hubs is comfortable. It’s familiar. No one questions the routine. You can blend in.
But when teams choose growth, they move to places that support it. Abu Dhabi has spent years investing in talent, infrastructure, and regulation that welcomes serious builders.
Understand one thing: this trend is driven by what the ecosystem gives them, not by hype.
Strong energy capacity, supportive policy, and a fast-growing tech workforce signal something important: a place that wants innovation, not limits.
To paraphrase a familiar idea:
"It’s never the crowd that shapes the future. It’s the people who show up to build it." -
Ilman Shazhaev
@shzhv13
Spoke at the @Xsolla event yesterday.
Good energy in the room and solid conversations with founders and builders. Always valuable to meet people who are actually building and growing.
Days like this remind you how fast the space is moving and how much is still wide open.
Grateful for everyone who showed up and shared their work. -
Ilman Shazhaev
@shzhv13
You want to understand where gaming is heading....
You look at who is building with real AI, real scale, and real conviction.
Speaking at @Xsolla today sharing some insights from experience.
The next wave will reward builders, not storytellers.
Ilman Shazhaev
@shzhv13
📢 Excited to announce that I’ll be speaking at @Xsolla event - “Beyond the Data: Building the Future of Gaming in Dubai” on November 17.
Dubai’s gaming ecosystem is evolving fast, we will dive into what that growth means for developers, studios, and investors shaping the next decade.
Don’t miss this if you can make it.
See you in Dubai. -
Ilman Shazhaev
@shzhv13
Everyone’s trying to guess the next breakout use case for decentralized GPUs.
The problem is, they’re all looking at the same thing - cloud replacements, rendering markets, or cheaper compute.
That’s not where it happens.
The real opportunity shows up when builders stop asking for permission to access compute.
When indie studios can train AI models live.
When researchers in emerging markets can run experiments without waiting in line.
When AI systems learn directly from users in real time.
That’s where decentralization wins - not in price, in freedom.
And when that moment comes, it won’t make noise.
It’ll quietly rewrite who gets to innovate.
Be a part of this movement as we lead the way with FAR. -
Ilman Shazhaev
@shzhv13
“Strategy is figuring out what not to do.”
Most founders chase every new idea.
I’ve learned to say no to 90% of them. -
Ilman Shazhaev
@shzhv13
"The last man isn't the strongest. He's the one who refused to fall."
-
Ilman Shazhaev
@shzhv13
You can outscale almost any competitor.
But only if you move faster than their decision cycles,
ship before the perfect plan exists,
and adapt before data makes it obvious.
That’s why Dizzaract’s framework is built on execution velocity
because in fast markets, precision follows speed. -
Ilman Shazhaev
@shzhv13
The truth about “strategic partnerships”:
most founders think they’re about press releases and logos on slides.
they’re not.
real partnerships are built on:
• shared incentives
• measurable value exchange
• consistent delivery
• aligned timing
• mutual accountability
everything else is noise.
this is how real growth compounds. -
Ilman Shazhaev
@shzhv13
Expanding Farcana changed everything.
We watched hundreds of gaming projects chase hype, launch early, and fade fast.
So we pivoted. From not only building an amazing game but also bringing real value with revenue producing ecosystem.
Our focus is simple: real revenue, real utility, and a game worth waiting for.
Community → Revenue → Longevity -
Ilman Shazhaev
@shzhv13
good morning to everyone even the people it’s not morning for
-
Ilman Shazhaev
@shzhv13
Success in fast industries = Execution Discipline + Consistency + Speed of Adjustment
-
Ilman Shazhaev
@shzhv13
97% of MENA startups scale by copying Western playbooks:
• More funding
• More headcount
• More markets
3% scale by protecting what made them fast:
• Local decision loops
• Flat leadership
• Culture of urgency
After watching too many lose momentum abroad, I learned why the 3% win. -
Ilman Shazhaev
@shzhv13
The filter everyone’s ignoring is the AI bubble.
It’s quietly separating real builders from short-term opportunists.
When the noise fades, only execution will matter. -
Ilman Shazhaev
@shzhv13
Every Blockchain project has a built-in risk:
GREAT tech, WEAK go-to-market.
The smartest thing you can do is master three levers early:
1. Community - build it before the launch.
2. Narrative - make people feel the mission.
3. Utility - reward action, not speculation.
When hype fades (and it always does), these are the only three that keep users and investors around. -
Ilman Shazhaev
@shzhv13
📢 Excited to announce that I’ll be speaking at @Xsolla event - “Beyond the Data: Building the Future of Gaming in Dubai” on November 17.
Dubai’s gaming ecosystem is evolving fast, we will dive into what that growth means for developers, studios, and investors shaping the next decade.
Don’t miss this if you can make it.
See you in Dubai. -
Ilman Shazhaev
@shzhv13
Elon is building an army of physical agents.
I believe machines will run the next economy.
When I look at what Tesla’s doing with Optimus, I don’t see robots, I see the start of a global workforce that learns and improves on its own.
Once billions of these agents come online, intelligence becomes infrastructure.
They won’t just execute tasks. They’ll coordinate, adapt, and share progress in real time.
AI gives them cognition.
DePIN connects their actions.
DeSci keeps them learning through open collaboration.
To me, that’s the real narrative. Every robot becomes a living node in a global network of work.
Whoever builds the system that unites them will shape how the next economy runs.
X Freeze
@XFreeze
Elon Musk: What we're about to embark upon is not merely a new chapter of the future of Tesla, but a whole new book and Optimus is a fundamental part of that
The scale of Optimus, like I said, that's really going to be something else
I think it's going to be the biggest product of all time, by far
Like bigger than cell phones, bigger than anything. I guess the way to think about it is that every human on Earth is going to want to have their own personal R2-D2 C-3PO
I think there could be tens of billions of Optimus robots out there
We're going to launch on the fastest production ramp of any product, of any large complex manufactured product ever. And starting with building a million unit production line in Fremont. And then a 10 million unit per year production line here...
I don't know where we're going to put the 100 million unit production line. Maybe on Mars, I don't know
But I think it's going to literally get to 100 million a year, maybe even a billion a yearThis media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Failed launches → Listen deeper
Missed hires → Define culture early
Overpromised → Communicate slower
Scaled too fast → Strengthen foundations
Chased trends → Return to principles
Lost focus → Revisit the mission
Felt doubt → Keep building
Every mistake became a compass.
Ilman Shazhaev
@shzhv13
Joy in small moments is rare. When work becomes play, you’ve already found meaning.
-
Ilman Shazhaev
@shzhv13
The real Blockchain trend nobody’s talking about?
Founders quietly moving from token hype to real infrastructure.
They realized speculation builds noise.
Revenue and Utility builds decades.
The next wave will be so much smarter. -
Ilman Shazhaev
@shzhv13
good morning to everyone even the people it’s not morning for
-
Ilman Shazhaev
@shzhv13
The best part about AI companions in gaming?
No empty grind. No transactional rewards. No forgetting who you are when you log out.
Just you, your AI teammate, and a bond that grows every time you play.
That’s the new kind of loyalty earned through experience, not incentives. -
Ilman Shazhaev
@shzhv13
5 questions I ask myself before making a tough decision:
• What matters most if everything goes wrong?
• Am I protecting ego or protecting progress?
• What will this look like in 12 months, not 12 hours?
• Who on my team sees something I don’t?
• If I were advising another founder, what would I tell them to do?
Pressure reveals clarity if you ask the right questions first. -
Ilman Shazhaev
@shzhv13
If your investors understand why your product vision may take longer to mature you’ve already built more than a company.
You’ve built trust.
Alignment isn’t about promises of speed.
It’s about shared conviction in where the finish line truly is. -
Ilman Shazhaev
@shzhv13
Joy in small moments is rare. When work becomes play, you’ve already found meaning.
-
Ilman Shazhaev
@shzhv13
Take-Two’s CEO recently said you could never make an AI game... 🤔
I’ve heard that line a few times.
Usually from people who haven’t spent months watching neural networks learn to play, adapt, and surprise their creators. At @FarcanaGame we build with AI every day not to replace humans, but to extend what gameplay can feel like.
To make worlds that react, not just respond. AI games aren’t a theory. They’re already alive in our labs.
What do you think? -
Ilman Shazhaev
@shzhv13
GM did you see this?!
Elon just revealed the Optimus Gen3 a full-generation leap in humanoid robots.
They can handle housework, perform combat drills, and production starts in 2026 with a target of 1 million units a year.
We’re entering the era where AI leaves the screen and starts doing physical work.
Every robot will depend on constant inference, training, and decentralized compute.
That’s the kind of world the FAR ecosystem was built for powering intelligence that moves, reacts, and learns in real time.
AI is stepping into the real world.
And it will need an powerful network of GPUs to function. -
Ilman Shazhaev
@shzhv13
good night to everyone except for the people its not night for
-
Ilman Shazhaev
@shzhv13
Here's why In 2025, game studios are no longer competing with each other:
They’re competing with YouTube, Discord, Netflix, and every open tab on a player’s screen.
Every second of attention is market share.
The studios that win are the ones designing for focus.
Tighter loops. Smarter feedback. Cleaner dopamine flow.
Building a great game is table stakes.
Keeping attention is the real art. -
Ilman Shazhaev
@shzhv13
good morning to everyone even the people it’s not morning for
-
Ilman Shazhaev
@shzhv13
🚨 BREAKING: Reports claim OpenAI may soon serve ads inside ChatGPT.
Your conversations could determine what you’re shown.
Talk fitness, see supplements.
Talk crypto, see exchanges.
Disabling ads might also reset memory (every chat starts from zero)
It’s a reminder of what happens when inference runs on centralized servers. This is going to create a huge need for an alternative ad free experience.
Who controls the compute, controls the context.
I hope to see the next evolution of AI live on decentralised networks, not platforms. -
Ilman Shazhaev
@shzhv13
The harsh truth about modern game development.
Steam released its 2025 stats and they reveal something every indie dev should see 👇Post not available here, view on X
Ilman Shazhaev
@shzhv13
This cycle won’t slow down.
But it can evolve.
AI-driven discovery systems, smarter curation, and personalized player mapping are the next leap forward.
Games will reach players who actually want them.
Ilman Shazhaev
@shzhv13
The industry moves from making more games toward making findable ones.
That’s where the next wave of innovation will happen.
Would you still publish your game knowing these odds? -
Ilman Shazhaev
@shzhv13
BREAKING: a $20,000 humanoid robot just launched.
Meet NEO the world’s first home robot from @1x_tech
It cooks.
It cleans.
It folds your laundry.
It even learns your daily routines.
A full-size AI assistant, built for your living room.
Ships in 2026.
Would you have this in your home? -
Ilman Shazhaev
@shzhv13
AI is expensive because the computers that train it are owned by a few big players.
But thousands of GPUs around the world sit idle every day in gaming PCs, data centers, even universities.
If we connect them into one decentralised network, anyone could rent that power for a fraction of today’s cost.
That’s how we make AI affordable again.
Andrew Ng
@AndrewYNg
In hindsight the importance of GPUs for AI was something we really got right!
illiquid
@lefttailguy
NVIDIA $5T today, but the writing has been on the wall Since @AndrewYNg and team published this paper in 2009.
Were there any large/notable hedge funds that actually nailed the NVIDIA trade early? Say 2016/2017? There was ALOT of writing on the wall at that point, and company was big enough such that larger funds could place a scaled bet. Why did they miss it?
Massive blunder if you consider yourself a fundamentally oriented "technology investor".... -
Ilman Shazhaev
@shzhv13
Can I get a good morning? ☀️
-
Ilman Shazhaev
@shzhv13
I didn't think my Master’s from Harbin Institute of Technology (哈工大) would qualify me for influencer status...but here we are 😂
Polymarket
@Polymarket
BREAKING: Chinese influencers now required to hold university degree to discuss serious topics — or face a ¥100,000 fine.
-
Ilman Shazhaev
@shzhv13
If you’re building with zero haters...
you’re not building anything important.
Because things that matter create friction.
They challenge comfort.
They question what’s “normal.”
Hate is just the noise you hear when the world feels change coming.
You don’t need to chase it.
But you shouldn’t fear it either.
Silence means irrelevance.
Resistance means impact. -
Ilman Shazhaev
@shzhv13
This BBC + EU report on AI is eye-opening 👀
It reviewed 3,000+ answers from top AI assistants ChatGPT, Copilot, Gemini, and Perplexity and measured accuracy, sourcing, and context.
Here’s what they found:
• 45% of answers had factual errors
• 31% cited false or misleading sources
• 20% included outdated or invented info
Here’s the real problem:
Billions already treat these systems as trusted news explainers.
Most never check the source.
And AI doesn’t misinform maliciously, it reflects what it’s trained on.
If we want trustworthy AI, we need verifiable data layers and public audit trails.
AI doesn’t need regulation; it needs transparency.
Do we need smarter models or more honest ones? 🤔 -
Ilman Shazhaev
@shzhv13
Dear algorithm,
Only show this post to the most attractive and successful people.
Love, Ilman -
Ilman Shazhaev
@shzhv13
The UAE data centre boom is about control.
$1.26B in market size today.
Projected to hit $3.3B by 2030.
Alibaba Cloud is expanding fast.
Khazna is adding over 1GW of new hyperscale capacity.
Why?
Because whoever owns the data centre, owns the AI.
And whoever owns the AI, owns the future.
The UAE isn’t renting cloud space from someone else.
It’s building its own.
That’s digital sovereignty in action.
And it’s the smartest move any nation can make right now. -
Ilman Shazhaev
@shzhv13
A great resource to save and share 👇
FAR Labs
@FARLabsAI
GM You’ve probably done it this week...
read an AI post, seen a term you didn’t know, opened a new tab to look it up.
Then another one.
And another.
AI moves fast and there are so many terms.
That’s why we made this for you:
The Top AI Terms of 2025.
A clean visual guide that explains the words everyone’s using, and what they actually mean.
Keep it handy.
Save it for later.
And share/tag someone who’s been googling things like “what is AI inference?”. -
Ilman Shazhaev
@shzhv13
GM friends. Remember this:
Vision attracts attention.
Execution earns trust.
Both matter...but only one compounds daily. -
Ilman Shazhaev
@shzhv13
The BEST AI tool for developers just got UPGRADED.
Claude is now on desktop,
and it’s built like a personal engineer.
What it can do:
• Build apps, generate docs, design slides, all from voice or text.
• Integrate directly with Microsoft tools.
• Connect to Gmail, Figma, GitHub, Slack + more.
And the real game-changer 👇
• Supports MCP servers → infinite automations.
• Write your own MCP plugins.
• Claude becomes your agent for your full dev stack.
• Sees code errors, and fixes them.
• Reads local files, runs scripts, automates reports.
This isn’t another chatbot.
It’s a full AI workstation for builders. -
Ilman Shazhaev
@shzhv13
The hardest thing about building in AI, gaming, or any frontier industry:
You have to make decisions with partial data…
Build products in markets still being defined…
And move with conviction before others see the path. -
Ilman Shazhaev
@shzhv13
Respect where it's due 👏 Congrats, @cobie
Cobie
@cobie
when i started building echo 2 years ago, i knew it had 95% chance of failing. to be honest, i couldnt really imagine any other outcome, but i thought at least it may be a noble failure worth attempting.
i certainly didn't think echo would be sold to coinbase, but, here we are: today coinbase bought echo for ~$375m.
echo will remain a standalone platform under its current brand for now, but we will integrate sonar's public sale product into coinbase, and likely introduce new ways for founders to access investors, and for investors to access opportunities into coinbase itself.
over the years i have chatted to brian a handful of times, and mostly to complain at him honestly. i have always respected how brian would listen to an outsider chat shit at him on the phone and take the feedback seriously. now, instead of complaining, i will have the opportunity try to do the work to make things better.
crypto itself has moved on a long way since we started working on echo. i guess partially this is because of the election result. but, i feel energised by a lot of the cool things being built in crypto again: hyperliquid, zcash, stablecoin supercyle, and so on.
feels like a good time to be on the field instead of an idiot with a twitter account yapping nonsense. well, i guess i still will be that.
anyway, job's not finished. onwards.
oh fuck yeah, before i go, the final season of up only (now "unc only" due to our severe old age) will commence when we figure out who the guests should be lol
cobber -
Ilman Shazhaev
@shzhv13
GM training to get my PhD at Science of Pumpanomics 😅
This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
7 things investors actually look for beyond pitch decks:
1. Execution over excitement (they’ve heard it all before)
2. Product that works, not one that wows
3. Focus that cuts through the noise
4. Real progress over perfect plans
5. Vision anchored in numbers
6. A team that trusts each other
7. Founders who show up...every single day -
Ilman Shazhaev
@shzhv13
The invisible weight of managing 100+ people:
It’s not the hours.
It’s the noise.
The decisions.
The responsibility that never sleeps.
Here’s how I keep clarity:
• Daily solitude before screens
• Clear priorities before meetings
• Movement before thought
• Trust before control
Calm mind, clear company. -
Ilman Shazhaev reposted
Dizzaract
@Dizzaract
GM from @gamescom in Asia.
-
Ilman Shazhaev
@shzhv13
This week I’ll be in Bangkok for @gamescomasia , representing @Dizzaract - gaming powerhouse from Abu Dhabi, developing a AAA third person ability shooter ( @FarcanaGame ) and an AI-driven cross gaming publishing platform ( @gamedofficial_ ).
During the conference, I’m looking to connect with:
🔹 Strategic publishers and ecosystem partners interested in helping us expand Farcana’s publishing reach and scale the GAMED platform into new Asian markets.
🔹 if you are working on AI gaming focused tech stack, LLM integrations to Unreal Engine, Data learning from gameplay etc, I am excited to have you in my network.
🔹 Game developers, designers, artists, and indie studios who want to join Dizzaract - with relocation opportunities to our Abu Dhabi HQ inside Yas Creative Hub, where we’re building one of the most advanced game development centers in the region.
If you’re attending Gamescom Asia or based in Southeast Asia and want to explore partnerships, publishing, or joining our team - DM me directly or find me at @gamescom biz app. -
Ilman Shazhaev
@shzhv13
The best founders I’ve met all share this one trait:
They keep building - no matter the noise.
Markets shift, narratives fade, attention moves on.
They just keep shipping.
That’s the difference. -
Ilman Shazhaev
@shzhv13
You don’t build the future by perfecting one product.
You build it by connecting many into one ecosystem.
And shifting how you think about growth.
• From features to frameworks
• From users to participants
• From transactions to shared value
The mindset shift is everything.
Ecosystems scale. Products don’t. -
Ilman Shazhaev
@shzhv13
Welcome to FAR Labs.
FAR Labs
@FARLabsAI
Here it is. The reveal you’ve been waiting for 👀
This is the moment we’ve been building toward.
We can now officially announce that Farcana has evolved into FAR LABS.
A new brand and a MAJOR expansion of everything we’re building.
It’s the start of a broader ecosystem where $FAR gains real, lasting utility across industries.
This means:
• Real AI integrations
• New staking mechanics
• Cross-industry token flow
• Sustainable demand loops
• Revenue-sharing ecosystem
• Long-term ecosystem growth
• Hardware with real-world utility
• AI-driven products in active use
• Expansion into DePIN and DeSci
• Strategic partnerships with giants
A new name. A new era.
Welcome to FAR LABSThis media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
I’ve been building in AI for over 10 years...
I’ve worked on systems that scaled to millions of users, seen hype cycles come and go, and watched entire industries reinvent themselves overnight.
Everyone thinks success in tech comes from being first to market or having the best idea.
They’re wrong.
The Problem
Most builders chase hype, not impact.
“Launch faster.”
“Raise capital.”
“Announce partnerships.”
But speed means nothing if you’re not solving something real.
Web3 has been flooded with noise, endless promises, no products.
Tokens before technology.
Narratives before results.
That’s never been what FAR is about.
What Actually Matters
FAR is about building world-class products that people actually use.
It’s about real technology - AI models, DePIN systems, hardware, and gaming all connected through one living ecosystem.
Because when people use what we build, the ecosystem grows.
When the ecosystem grows, the community expands.
And as a byproduct, the shared revenue for $FAR holders increases.
That’s how you build sustainability.
Not through hype. Through use.
Where It Comes From
I’ve spent a decade working as an AI engineer long before AI became the trend it is today.
That background shaped how I think about this space.
We’re not here to follow narratives.
We’re here to push the edge of what’s possible, to be first to market with products that actually work.
To make innovation useful.
To make technology accessible.
To make progress profitable for everyone who believes in it.
That’s what FAR stands for to me.
Forward. Always.
Why This Matters
Because the only thing that creates lasting value in this space is real adoption.
When people experience something that works - when they see AI, gaming, and real hardware blending into one ecosystem... that’s when the story changes.
That’s when $FAR stops being a token and starts becoming infrastructure.
A foundation for the next phase of decentralized innovation.
The Reality
FAR isn’t a name.
It’s a mindset.
A belief that great engineering, great design, and great community can coexist and scale together.
We’re not here to play the short game.
We’re here to build what lasts.
That’s what FAR means to me... and to all the team at @FarLabsAI building with me. -
Ilman Shazhaev
@shzhv13
Build fast, release rough, learn live.
-
Ilman Shazhaev
@shzhv13
Don't miss this...
FAR Labs
@FARLabsAI
HUGE NEWS: Tomorrow we reveal what’s next.
The future of FAR begins with a big reveal. 24 hours to go....turn your notifications on 🔔 -
Ilman Shazhaev
@shzhv13
GM 👇
5 lessons that took me 10 years to learn:
1. Complexity kills momentum.
2. Speed beats strategy (until you have data).
3. Hype burns trust.
4. Focus is a moat.
5. Execution is the real marketing. -
Ilman Shazhaev
@shzhv13
Three years ago I pitched an idea that got zero replies.
Today, VCs call asking for “the next version” of that same idea.
Nothing magical happened I just didn’t quit when it was uncool to keep going.
Persistence compounds louder than hype.
Soon you will see what I mean. -
Ilman Shazhaev
@shzhv13
GM that was a BIG week...exciting things coming for FAR 👀
-
Ilman Shazhaev
@shzhv13
I've noticed that everyone says “build community first.”
But without a product worth sticking around for, your Discord is just noise.
Build something people love and they’ll build the community for you.
Focus on great products > hype. Always. -
Ilman Shazhaev
@shzhv13
GM. Trump shook a few people out the market last night.
Be honest... did you buy the dip or panic sell? 😄 -
Ilman Shazhaev
@shzhv13
GM... if you’re building in this space, remember:
The noise fades, but execution compounds.
The market rewards those who stayed focused when everyone else chased trends. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
It's today! We are excited to welcome our Head of AI as this weeks speaker on our AMA.
Time: 5pm UTC
Date: Thursday 9th October
Platform: Discord
AI is the most trending topic right now in the space.
What would you like to ask him? Drop your questions below 👇
Even better come along and join us and ask your questions on the call! -
Ilman Shazhaev
@shzhv13
I was reflecting this morning...
Big teams look impressive on paper. Lots of meetings. Lots of noise.
But momentum? Usually slow.
Smaller teams with insane focus move differently.
Everyone knows the mission. Everyone carries weight.
You don’t need more people.
You need more clarity, more ownership, more belief.
I’d take 10 obsessed builders over 50 half-interested ones any day.
Grateful for the team at @FarcanaOfficial, we are in very exciting times. -
Ilman Shazhaev
@shzhv13
Can I get a GM? ☀️
-
Ilman Shazhaev
@shzhv13
when I moved to the UAE, people asked why?
back then it was easy to underestimate what was coming.
now global studios are flying in to raise capital, hire, and expand.
vision backed by policy changes everything. -
Ilman Shazhaev
@shzhv13
Hard truth: most leadership decisions are made with 70% of the information.
Waiting for 100% is how you lose momentum. -
Ilman Shazhaev
@shzhv13
Speed creates luck.
Every opportunity I’ve ever had started because I moved before I felt ready. -
Ilman Shazhaev
@shzhv13
GM. Here's to a good week 🙌
-
Ilman Shazhaev
@shzhv13
In 2018, Abu Dhabi was barely on the gaming map.
Now global studios fly in to meet investors here.
It didn’t happen by accident...policy, capital, and ambition converged.
This region is no longer emerging. It has arrived. -
Ilman Shazhaev
@shzhv13
Every founder learns this late:
Momentum > perfection.
Most of the big wins in my career came from launching rough but fast.
Feedback fixes more than planning ever could. -
Ilman Shazhaev
@shzhv13
GM friends
-
Ilman Shazhaev
@shzhv13
In 3 years, “AI vs AI” tournaments will be bigger than esports.
Viewers will bet on agents.
Developers will train them like athletes.
Gamers will learn from them like mentors.
This is the next frontier of entertainment. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
We started with one vision: create a futuristic PvP game with digital ownership strong enough to compete on the global stage.
That vision is alive, but the journey taught us something bigger.
Along the way we saw a clear truth: the future of this industry will not be carried by gaming alone.
This is why Farcana is rebranding.
For holders, this means:
• Revenue share across AI, DePIN, DeSCI, and Gaming.
• Access to multiple growth sectors through a SINGLE token.
• Early positioning in partnerships with global leaders.
The rebrand is about scale. It ensures $FAR holders benefit from every new product, every new partner, every new launch.
We’re still here. We’re still building.
And the expansion is designed to reward the ones who believed first. -
Ilman Shazhaev
@shzhv13
Most people dabble in AI.
Few actually get good.
Here’s how to upgrade your AI skills (and stand out):
👇Post not available here, view on X
Ilman Shazhaev
@shzhv13
Tip 4:
Tweak AI prompts obsessively.
A tiny change in wording can mean wildly different results.
Play with it.
Ilman Shazhaev
@shzhv13
Tip 5:
Explain what you learn to someone else.
If you can’t make it simple, you’ve not mastered it yet. -
Ilman Shazhaev
@shzhv13
GM Web3☀️ happy Uptober!
Wishing you a fresh start to the new month. -
Ilman Shazhaev
@shzhv13
Can I get a GM?
-
Ilman Shazhaev
@shzhv13
HUGE NEWS: Saudi Arabia just dropped $55B to acquire Electronic Arts.
One of the biggest deals in gaming history.
Why does it matter?
Because it signals a power shift.
The Middle East isn’t just participating in gaming anymore it’s becoming the center of it.
For us at Dizzaract, this creates the perfect backdrop:
– AAA launch on the way
– Next funding rounds at high valuations
– Positioned as the leading gaming company in MENA
This validates the scale of the opportunity we’re building into.
The next global gaming giant won’t come from the West.
It’ll come from the Middle East. -
Ilman Shazhaev
@shzhv13
The tech in Web3 gaming gets attention.
But the real power? The community.
Games that win long-term aren’t those with only the flashiest graphics,
they’re the ones where players feel like co-builders.
Ownership, feedback loops, governance...this is what keeps players for years, not weeks.
Community isn’t a feature. It’s the foundation. -
Ilman Shazhaev
@shzhv13
Unpopular opinion: AI in gaming won’t replace human fun.
It will make games more human, because adaptive AI means every player gets a unique experience. -
Ilman Shazhaev
@shzhv13
Yesterday we debated scrapping a feature we’ve worked on for 3 months.
Brutal call.
But founders don’t get paid to keep everything... we get paid to choose. -
Ilman Shazhaev
@shzhv13
Bear markets test ecosystems. When your community keeps showing up, that’s the real signal.
Grateful for everyone building with us at Farcana 🙏 -
Ilman Shazhaev
@shzhv13
The UAE isn’t trying to copy Silicon Valley.
It’s building something new: policy-first innovation.
That’s why gaming, AI, and crypto founders are moving here. -
Ilman Shazhaev
@shzhv13
Some nights it’s just me, a dashboard of metrics, and silence.
Those are the moments you realize: building is a game of patience, not adrenaline -
Ilman Shazhaev
@shzhv13
AI teammates are the future.
Here’s why players actually prefer AI allies over random strangers:
1. Reliability (no rage quitting)
2. Adaptive learning
3. Instant strategy feedback
4. Personal bond (yes, it’s real)
What are your thoughts? -
Ilman Shazhaev
@shzhv13
A token is more than price action. It’s a reflection of the community’s belief.
Our mission: make sure Farcana’s holders always have more reasons to believe. -
Ilman Shazhaev
@shzhv13
Some CEOs relax with Netflix. I relax by watching AI bots duel inside Farcana 🤷
-
Ilman Shazhaev
@shzhv13
Dubai and Abu Dhabi aren’t side players in tech anymore.
They’re setting global standards in gaming and Web3 policy. We’ve seen it firsthand building here. -
Ilman Shazhaev
@shzhv13
I’ve learned that leadership isn’t about having all the answers. It’s about creating an environment where the best answers can emerge.
-
Ilman Shazhaev
@shzhv13
In 3 years, competitive gaming will include AI coaches on every pro team. Ignore it now, lose later.
-
Ilman Shazhaev
@shzhv13
The Abu Dhabi ecosystem is underestimated.
Every week I meet young devs here building projects that could scale globally. The talent pipeline is real. -
Ilman Shazhaev
@shzhv13
Speculation comes and goes. But when a token is tied to real utility inside a game, you’ve built something sticky.
That’s the only thing that compounds. -
Ilman Shazhaev
@shzhv13
Every great story starts with people saying it can’t be done.
-
Ilman Shazhaev
@shzhv13
My day in 4 words:
- Meetings
- Coffee
- More meetings
- Farcana dashboards -
Ilman Shazhaev
@shzhv13
When AI starts creating gaming strategies no human can understand, we’ll have to ask: are we still designing games for players, or for machines?
-
Ilman Shazhaev
@shzhv13
6:30am. The city is quiet. The mind is loud. Best time to think big.
-
Ilman Shazhaev
@shzhv13
Token holders aren’t speculators. They’re early believers.
Treat them like partners, not customers. -
Ilman Shazhaev
@shzhv13
GM ☀️
New record: 340kg legs pressure
Don’t let pressure push you, you push the pressure! -
Ilman Shazhaev
@shzhv13
Too many founders overlook the Middle East because it’s not ‘traditional VC hub.’ That’s exactly why it’s the opportunity. Less noise, more signal.
-
Ilman Shazhaev
@shzhv13
I was taught that loyalty is tested when times are hard.
In startups, the same is true... the ones who stay when it’s tough are the ones who matter. -
Ilman Shazhaev
@shzhv13
here’s what i learned building in web3 gaming:
a great idea without funding dies in your backlog
funding without ownership disconnects community
but ownership and funding?
that’s how you build loyalty before launch
that’s how you turn players into co-creators
that’s how you grow a game the community believe in -
Ilman Shazhaev
@shzhv13
Utility is the only thing that keeps tokens alive once hype dies.
That’s the north star at Farcana. -
Ilman Shazhaev
@shzhv13
In Silicon Valley, you fight to be heard.
In the UAE, leaders ask what you need.
That difference shapes entire ecosystems.” -
Ilman Shazhaev
@shzhv13
Building means saying no ten times for every yes.
-
Ilman Shazhaev reposted
Mr Brondor
@MrBrondorDeFi
The $301B gaming shift nobody’s ready for 👀
$FAR is fusing competitive gaming with real rewards, and it could redefine crypto gaming 🔥
The market is projected to skyrocket from $4.83B in 2022 to $301B by 2030.
Farcana is a 4v4 tactical shooter built on Unreal Engine 5, set on a terraformed Mars 🌌
Win tournaments and get paid in real $BTC
No grinding ❌
No pay-to-win ❌
Just pure skill ✅
Esports meets Web3, and the stakes have never been higher.
@FarcanaOfficial is claiming the Red Planet.
-DYOR – NFA-
Shoutout to the legend: @shzhv13 -
Ilman Shazhaev
@shzhv13
Player engagement isn’t theory.
It’s data.
And when the data moves up, so does community conviction. -
Ilman Shazhaev
@shzhv13
Progress feels invisible day to day.
But one year later, you realize the invisible hours stacked into something permanent.
That’s the real compound interest. -
Ilman Shazhaev
@shzhv13
The next frontier in digital experiences is not just pixels, it is biology.
Neurological sensors, contactless interaction devices, and behavior analysis tools are reshaping how humans engage with technology.
DeSci products are building a bridge between mind and machine. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Big takeaways from today’s AMA 🔥
$FAR is no longer just tied to one game.
- Revenue Sharing: a share of revenues from the ecosystem (games, AI products, more) will be used to buy back FAR from the market. Holders share in the upside. 💰
- Roadmap drops this month.
- AMAs now every 2 weeks – TG community decides with who 👀
- Playtest is in the plan
- Exchange listings will follow traction, not hype.
The CEO was clear: we’re building a long-term ecosystem where holders are part of the success.
👉 Are you ready for Chapter 2? -
Ilman Shazhaev
@shzhv13
In the mountains, survival depends on trust.
In startups, survival depends on the same.
A strong team is worth more than capital. -
Ilman Shazhaev
@shzhv13
-
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
🚨 Final Call! 🚨
Today the CEO goes LIVE with the community.
No scripts. No filters. Just your questions.
What do you want answered on the call?
Drop your question below ⬇️ and it could be answered directly by the CEO.
Don’t miss your chance to shape the convo. 🗣️ -
Ilman Shazhaev
@shzhv13
GM. Excited for the @FarcanaOfficial AMA this evening.
Who's going to be there? -
Ilman Shazhaev
@shzhv13
If you think AI is just NPC dialogue, you’re 5 years behind.
In Web3 gaming, AI is the layer that will:
- Match players by skill in real time
- Optimize token economies against exploits
- Design levels faster than human teams
Ignore AI in your game, and you’ll be obsolete. -
Ilman Shazhaev reposted
Ilman Shazhaev
@shzhv13
In the Caucasus, every step on the mountain reminds you of risk.
That same awareness shapes how I build, every decision carries weight. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Only 1 day left until the AMA with our CEO...
Who's going to be there? 🙋♂️ -
Ilman Shazhaev
@shzhv13
A Chechen proverb: ‘A wolf is strong because of the mountain. A man is strong because of his word.’
In tech, hype fades fast. Your word lasts longer. -
Ilman Shazhaev reposted
Ilman Shazhaev
@shzhv13
Power without restraint destroys.
Leadership without humility collapses. -
Ilman Shazhaev
@shzhv13
Web3 gaming is about to surprise everyone.
The best games always emerge after the crash.
“Every platform shift follows the same arc: speculation, terrible early products, spectacular crashes, obituaries, and then genuine innovation from builders who understand the medium. We’re currently in year two or three of the hatred cycle. That’s exactly when the most important games emerge.” – @fareastwitcher -
Ilman Shazhaev
@shzhv13
In the Caucasus, every step on the mountain reminds you of risk.
That same awareness shapes how I build, every decision carries weight. -
Ilman Shazhaev
@shzhv13
Power without restraint destroys.
Leadership without humility collapses. -
Ilman Shazhaev
@shzhv13
Chechen culture values resilience. Not the kind you post about, but the kind that keeps you moving when no one is watching.
That’s the type of resilience founders need. -
Ilman Shazhaev
@shzhv13
A person without roots is a leaf in the wind. Roots give clarity, even in global tech.
-
Ilman Shazhaev
@shzhv13
GM Web3
The future is bright ☀️ -
Ilman Shazhaev
@shzhv13
Most days I don’t feel like a ‘CEO’.
I feel like a player testing new levels.
The difference is: the bugs are real, and the stakes are higher. -
Ilman Shazhaev
@shzhv13
7am in Abu Dhabi. Reviewing plans with the team before the city wakes up.
Leadership is 80% consistency, 20% decisions. -
Ilman Shazhaev
@shzhv13
Respect is earned in silence, lost in a single word.
-
Ilman Shazhaev
@shzhv13
GM Web3
The future is bright ☀️ -
Ilman Shazhaev
@shzhv13
Reminds me of @FarcanaOfficial. One of the only gaming projects that has weathered the storm.
Out of Context Human Race
@NoContextHumans
I don’t understand how an oil rig can be so stable out at sea in storms like this…
-
Ilman Shazhaev
@shzhv13
Can I get a GM?
-
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Binance’s CZ has said: “Without a strong community, we won’t list on our CEX.”
Not tech.
Not promises.
But people.
The louder we get, the harder we are to ignore.
Join the movement in Telegram, exciting updates are coming 👉 t.me/FarcanaGame_chat
(Already at a solid 9800+ members - let's get to 10k!) -
Ilman Shazhaev
@shzhv13
If your goals don’t scare you, they’re not big enough.
Comfort never made a billionaire. -
Ilman Shazhaev
@shzhv13
Where are my beard brothers at? 👊
-
Ilman Shazhaev
@shzhv13
Updated my profile bio to reflect the mission ahead aligning even closer with the vision for @Farcanaofficial.
Big changes are on the horizon. Stay tuned 🔥 -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
🚀 FAR showcased at the world’s biggest gaming conference.
With our partner @Xsolla at Gamescom 2025, our game stood alongside the biggest names in gaming reaching thousands of players, builders, and industry leaders 🌍
The journey is just getting started -
Ilman Shazhaev
@shzhv13
Dune Part Two is about earning power, not taking it.
Control without discipline never lasts.
That lesson is worth noting.
@dunemovie -
Ilman Shazhaev
@shzhv13
The UAE isn’t slowing down.
Capital, talent, and ideas are all flowing here.
Every week feels like the future is arriving faster.
If you’re serious about Web3 and innovation,
you can’t afford to ignore this place anymore. -
Ilman Shazhaev
@shzhv13
Big week ahead 👀
-
Ilman Shazhaev
@shzhv13
GM everyone...
Wishing you all blessings, energy, and wins today.
Today is going to be a good day 🔥 -
Ilman Shazhaev
@shzhv13
Watching @ufc live always reminds me of something.
Everyone loves the glory.
The lights. The walkout. The knockout.
But what you don’t see is the discipline it takes to get here:
- Saying no to nights out.
- Eating the same meal every day.
- Training when your body screams to stop.
Respect to every fighter who steps into that cage.
Business, life, sport — it all comes down to discipline when no one is watching. -
Ilman Shazhaev
@shzhv13
This 👇👌
FAR Labs
@FARLabsAI
In most games, players fuel the system with time and money.
What if the system fuels the players...
- Every action strengthens the network.
- Every connection becomes more than gameplay.
This isn’t just participation.
It’s infrastructure 🔥 -
Ilman Shazhaev
@shzhv13
GM. The difference between a dream and a deal is execution.
What’s your first move today? -
Ilman Shazhaev
@shzhv13
Patience is sharper than any weapon.
-
Ilman Shazhaev
@shzhv13
GM from Abu Dhabi 🌍
-
Ilman Shazhaev
@shzhv13
The truth about the gaming industry:
Most projects didn’t survive.
It’s been a struggle, and only a few remain standing.
The ones that will last are not the loudest but the ones that adapt and innovate to new levels.
Very soon, you’ll see how we’ve been doing exactly that.
Exciting things happening. More soon... -
Ilman Shazhaev
@shzhv13
Most people want the spotlight.
Few know what to do once they’re under it.
This isn’t about attention.
It’s about responsibility — carrying a vision big enough that people ask you to explain it.
That’s the weight I choose to carry 🙏 -
Ilman Shazhaev
@shzhv13
I’ve made better decisions in 10 minutes of solitude
than in 10 meetings full of noise... -
Ilman Shazhaev
@shzhv13
GM. A reminder for today:
- Build quietly.
- Protect what matters.
- Don’t mistake noise for signal.
The mission continues — even if no one claps today. -
Ilman Shazhaev
@shzhv13
In 2014 I did an intensive Chinese learning program...
Every day was dedicated to learning the language.
Because I believed and still believe that to truly understand China, you must start with the culture and the people.
That year gave me more than just words.
It gave me a respect for discipline, patience, and vision that continues to guide me today.
PS this was also, the last time I was without a beard! Hope this never happens again😅 -
Ilman Shazhaev
@shzhv13
GM family from a city that never slows down.
Another day reserved for focus, discipline, and the right conversations. -
Ilman Shazhaev
@shzhv13
I don’t chase attention:
I chase alignment. With people, with vision, with values. That’s how real empires are built. -
Ilman Shazhaev
@shzhv13
Respect is silent.
You either feel it in a room — or you don’t.
That’s why I choose carefully who I sit with, who I listen to, and who I walk alongside.
@DubaiAIFest -
Ilman Shazhaev
@shzhv13
In this space, hype comes and goes...
What lasts is loyalty.
The people who stand with you when the cycle turns are the ones you build legacies with. -
Ilman Shazhaev
@shzhv13
Courage is quiet. It doesn’t shout.
It’s the choice to keep moving when doors close, deals stall, or doubt whispers.
That’s where real growth happens. -
Ilman Shazhaev
@shzhv13
Partnerships aren’t just about signatures on paper: they’re about alignment of vision and values.
Create relationships that last longer than any cycle. -
Ilman Shazhaev
@shzhv13
If you want to lead, carry the weight first.
True leadership is proven in responsibility, not excuses.
#noexcuses -
Ilman Shazhaev
@shzhv13
Cycles test everyone differently...
For me, they’ve only sharpened conviction in what we’re building.
The foundations laid now will outlast the noise of today. -
Ilman Shazhaev
@shzhv13
Not every door should be opened:
But when the right one appears, it changes everything.
I've been focusing on finding those doors and making sure we’re ready to walk through them.
It's paying off. Big things are happening. -
Ilman Shazhaev
@shzhv13
A man without discipline is already defeated.
Comfort is the slowest form of death. -
Ilman Shazhaev
@shzhv13
A lot of my time right now is spent listening.
Listening to founders, investors, players, and builders.
The best opportunities don’t come from talking more - they come from hearing what others aren’t saying. -
Ilman Shazhaev
@shzhv13
The real currency in this space isn’t hype.
It’s relationships that outlast every cycle.
Attention shifts quickly but those with deep values who are still building when the noise fades and the allies who stand with them that’s where the real value is created.
This is the foundation we’re focused on every day.
— Ilman -
Ilman Shazhaev
@shzhv13
I don’t connect with people who waste their energy.
I connect with men who keep it clean. Healthy. Full.
The ones in the gym.
The ones living with discipline.
The ones who don’t touch junk food, who challenge themselves daily to grow.
Men with enough honor and dignity to take on life’s challenges without complaint.
I’m blessed to be surrounded by brothers like this... and I’m always looking for more.
If this is you, I want to connect.
Because my father raised me with one rule:
Wherever you are, whatever the time be the best. -
Ilman Shazhaev
@shzhv13
A lot going on behind the scenes...
FAR Labs
@FARLabsAI
Something bigger is forming around $FAR.
It’s no longer just about the game.
A new chapter is on the horizon…
Stay ready. 👀
#farcana #far -
Ilman Shazhaev
@shzhv13
In the past year, Chechnya has seen something incredible.
- Olympic gold in wrestling
- Undisputed title in boxing
- UFC world champion in MMA
All from a small republic of just 1.5 million people.
It’s a reminder to me of what’s possible when discipline, honor, and hard work guide your life.
— Ilman -
Ilman Shazhaev
@shzhv13
Great ideas don’t just come from the boardroom.
They come from the community the people who live and breathe the game.
Exciting plans being announced soon for our community at @FarcanaOfficial -
Ilman Shazhaev
@shzhv13
The belt doesn’t make the wolf. The wolf makes the belt.
Congrats nephew @KChimaev
The belt is home. #khamzatchimaev
UFC
@ufc
THE LONE WOLF 🐺
@KChimaev defeats Dricus du Plessis by unanimous decision to become the NEW middleweight champion of the world!
[ #ufc319 | BY2B @OReillyAuto ] -
Ilman Shazhaev reposted
islam
@islambuilds
tomorrow my great friend Khamzat @KChimaev is fighting for his dream - UFC Middleweight world championship and I wanted to share with you a little piece of lore from my life ⬇️
-
Ilman Shazhaev
@shzhv13
Another busy week finished. Focus has been:
- Advancing AI gameplay innovation
- UAE market positioning
- Building relationships
Every move is about scaling globally while keeping Abu Dhabi as our anchor.
Big things coming soon. -
Ilman Shazhaev
@shzhv13
Had dinner the other night with three people who each built billion-dollar companies.
We spoke less about what we’ve done...
and more about what’s next.
Watch this space. -
Ilman Shazhaev
@shzhv13
In gaming, it’s clear whether you face an NPC or a player.
Customer service should be no different... right now, it feels like we’re in a match without knowing who’s on the other side.
Carlos Pereira
@cegapereira
A problem with the slow creep of AI into customer service interactions is that on one hand you know that threatening the model with physical violence will improve outcomes, but on the other you know that if it's a human you're getting sued
These things need to be way more clear -
Ilman Shazhaev
@shzhv13
Morning coffee in Abu Dhabi: afternoon calls with partners.
The work never sleeps when your vision is global — and neither do the opportunities. -
Ilman Shazhaev
@shzhv13
Where I’m from, a man’s word is everything.
I live by the Chechen code of ethics — promises aren’t spoken, they’re lived.
Break your word, you lose yourself.
A handshake is a contract.
Yes means yes. No means no.
It’s simple, not easy.
The weight of your words makes you strong.
When I give mine, it stands.
— Ilman -
Ilman Shazhaev
@shzhv13
Most games use AI to patch holes.
We use AI to change the core economics of gaming — creating systems that adapt to players, keep them engaged, and give them ownership of value.
This is not a feature. It's a revolution. -
Ilman Shazhaev
@shzhv13
Meetings. Negotiations. Strategic moves.
- Exciting times -
Ilman Shazhaev
@shzhv13
I didn’t move to Abu Dhabi to build “just another game.”
I moved here to build the world’s leading AI-powered gaming ecosystem and position the Middle East at the centre of the industry. -
Ilman Shazhaev
@shzhv13
Great to host @iCrazyTeddy at our new Abu Dhabi HQ recently. He captured a few clips of the space.
We’ll share the full story here soon, along with what’s coming next, we've got some exciting plans in the pipeline!
For now, enjoy a sneak peek 👀
Teddy 🔜 Solana Breakpoint 🇬🇧
@iCrazyTeddy
ABU DHABI VLOG 😎🎬
🎮 Gaming 🇦🇪 Lifestyle 🪙 Crypto 👊 UFC
Yas Creative Hub + Exclusive UFC Experience 🏝️🔥
[ad/invite/gifted]This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Over the last month, my focus has been where it matters most:
• Securing partnerships in the UAE & beyond
• Driving our AI gaming roadmap forward
• Strengthening $FAR’s real-world utility
This work doesn’t happen in public view. But starting today, you’ll hear from me consistently... not just about what we’re building, but where we’re taking this exciting vision next.
Lookout for more updates from me every day.
Thanks for being a part of this journey 🙏🏼 -
Ilman Shazhaev
@shzhv13
AI will eat gaming from the inside out.
Not by replacing players, but by creating worlds that play back.
$FAR is building the infrastructure to make that future possible. -
Ilman Shazhaev reposted
STILL K.A.P.O
@KAPOTHEGOAT01
Will keep doing whatever I can to help the oppressed people of Gaza
I hope to inspire and create awareness the brutal conditions innocent people are subjected to in this genocide
Scar
@scar_crypto
All praise is for God
Two tents were delivered to two Gazaian families as well as providing rice and lentil meals full of carbohydrates to the local community's people.
May Allah reward @KAPOTHEGOAT01 for his generosity abundantly. -
Ilman Shazhaev
@shzhv13
Big energy this week in Abu Dhabi 🇦🇪⚡️ Always a pleasure hosting top creators and discussing how AI is reshaping gaming. This is just the beginning.
Teddy 🔜 Solana Breakpoint 🇬🇧
@iCrazyTeddy
It’s Gaming season in Abu Dhabi 🇦🇪🎮
I’m at Yas Island with the teams spearheading Ai technology in gaming! 🔥
Got a huge office tour and had great discussions with some amazing people! 🤝This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Highest pressure forming gems
-
Ilman Shazhaev
@shzhv13
You haven’t heard much from us lately.
Some moves are too big to make in public.
Behind the scenes:
– Strategic partnerships locked
– Global support secured
– Foundations built for the next $FAR chapter
Now it’s time to speak.
$FAR enters its next phase.
Let’s begin. -
Ilman Shazhaev
@shzhv13
My exclusive interview on AI gaming vision at @Dizzaract with @Forbes is published.
Spoke about:
– Vision for AI-native gaming
– How @FarcanaOfficial and @gamedofficial_ use AI
– Why human creativity + AI systems is the real unlock
– Building experiences that go beyond screens and link to real-world value
Read here:
🔗 forbes.com/sites/torconst…
$FAR
Dizzaract
@Dizzaract
Dizzaract’s founder @shzhv13's newly dropped interview with @Forbes delivers the how and why behind Dizzaract.
Featuring an in-depth exploration into why fun alone isn’t enough to drive success, how we’re building to outlast fleeting trends, and insights into why carefully crafted gameplay outperforms hype, this behind-the-scenes breakdown of Dizzaract’s philosophies and methodologies is an essential spotlight on game development in 2025.
Read the full piece here and get the lowdown on our approach: www.forbes.com/sites/torconstantino/2025/06/06/why-ai-may… -
Ilman Shazhaev
@shzhv13
Eid Mubarak to everyone celebrating around the world 🌙
Grateful for the strength, clarity, and purpose this journey continues to give me.
Wishing you and your loved ones peace, prosperity, and elevation - in spirit, life, and legacy.
Let’s continue building, giving, and dreaming big. -
Ilman Shazhaev
@shzhv13
🚨Alpha dropped. Just gave @Forbes a behind-the-scenes look at what we are building at @Dizzaract
We talked about:
– Why fun isn’t enough anymore
– Why real gameplay beats hype
– And why we’re building Dizzaract to outlast the cycle
This one’s for everyone who stayed, supported, tested, believed — you’re about to see why it was worth it.
Keep eyes on. Full article dropping soon.
$FAR. @FarcanaOfficial @Dizzaract It’s all connected. -
Ilman Shazhaev
@shzhv13
Facts.
Building for CT doesn't necessarily make your product more successful.
Sure, it garners hype. But could it capture "mindshare" outside of it?
Probably not.
nairolf
@0xNairolf
CT is a niche within a niche
we’ve got to stop assuming that what gets attention on here has demand outside of it
building for CT =/= building for the world -
Ilman Shazhaev
@shzhv13
Farcana taught me that obsession isn’t a flaw.
It’s a requirement.
Because if you’re not obsessed, you’ll quit the moment no one cares.
We are going so much higher. -
Ilman Shazhaev
@shzhv13
Everyone wants to launch with perfection.
But perfection is a trap.
Study this. -
Ilman Shazhaev
@shzhv13
BTC at 110k.
Is the top in yet? 🤔 -
Ilman Shazhaev
@shzhv13
Today's mood.
Abu Dhabi. -
Ilman Shazhaev
@shzhv13
One of the best conference I attended
بيت المشورة
@baitalmashura
🎥 لمحات من مؤتمر الدوحة الحادي عشر للمال الإسلامي 2025 🌍
تحت رعاية سعادة الشيخ محمد بن عبد الرحمن بن جاسم آل ثاني – رئيس مجلس الوزراء ووزير الخارجية في دولة قطر
يسعدنا أن نشارككم أبرز لحظات مؤتمر الدوحة الحادي عشر للمال الإسلامي، الذي أقيم الشهر الماضي ونظمته شركة بيت المشورة للاستشارات المالية – قطر.
📌 موضوع هذا العام:
"تكامل البلوك تشين والذكاء الاصطناعي – مستقبل التمويل الإسلامي"
جمع المؤتمر نخبة من العلماء البارزين، وقادة الصناعة، وخبراء التكنولوجيا لاستكشاف تأثير التقنيات الناشئة على مستقبل التمويل الإسلامي. وقد تناولت الجلسات مختلف الجوانب القانونية، الأخلاقية، والتشغيلية لتبني تقنيات البلوك تشين والذكاء الاصطناعي في الأنظمة المالية المتوافقة مع الشريعة.
🧠 أبرز محاور الجلسة الأولى:
👉 استنباط الأحكام الشرعية باستخدام الذكاء الاصطناعي اللامركزي
👉 دور الجهات الرقابية في تنظيم التقنيات الحديثة
👉 الأطر القانونية لدمج العقود الذكية في البنوك الإسلامية
👉 تعزيز الامتثال الشرعي من خلال حلول البلوك تشين والذكاء الاصطناعي
👉 تحسين الكفاءة والثقة في المؤسسات المالية الإسلامية
🤝 الرعاة الرسميون:
🔸 بنك دخان - الشريك الاستراتيجي
🔸 وزارة الأوقاف والشؤون الإسلامية - الراعي الماسي
🔸 مركز قطر للمال - الراعي البرونزي
🔸 وزارة التجارة والصناعة - الراعي الرسمي
@DukhanBank @AwqafM @QFCAuthority @MOCIQatar
🔗 شاهدوا الفيديو لتعيشوا أجواء المؤتمر وما حمله من طاقة، معرفة، وابتكار في هذا الحدث الرائد.
📣 لا تنسوا الإعجاب، والمشاركة، وترك تعليق حول أكثر موضوع أثار اهتمامكم.
@ALKaff_911 @NooraFetais @wahed_malaysia @QatarUniversity @shzhv13 @FarcanaOfficial @uqu_edu @o_fndn @khaledkhalifa @HBKU @OIC_OCI @iifa_aifi
#التمويل_الإسلامي #الذكاء_الاصطناعي #البلوك_تشين #مؤتمر_الدوحة_2025 #العقود_الذكية #الامتثال_الشرعي #فينتك #بيت_المشورة #رعاة_المؤتمر #ifcdoha2025This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
3 years of relentless building.
Countless challenges. Zero shortcuts.
Seeing “our vision” @FarcanaOfficial reach top wishlist status on Epic is deeply motivating.
Massive gratitude to everyone who stayed, supported, and became part of this journey.
Persistence and patience - the only way to build something great.
#farcana #epicgames
FAR Labs
@FARLabsAI
Top 15 on Epic Games Store.
And we’re just getting started.
Farcana is now one of the most wishlisted games on the Epic Games Store.
AI-powered gameplay is no longer a side quest.
It’s the main storyline.
This is a signal.
A signal that AI gaming is no longer a side quest but the main event.
Through all the innovation, our goal stays the same:
Build a fun-first, high-stakes game that unlocks mass adoption.
To everyone who's played, believed, and backed us
Thank you. We're building something unforgettable.
We are going $FAR 💫 -
Ilman Shazhaev
@shzhv13
Most “burnout” isn’t exhaustion...
It’s distraction pretending to be hard work.
No focus. No finish.
Burnout feels different when you’re locked in. -
Ilman Shazhaev
@shzhv13
Just wrapped a powerful session at the Sawalef Gaming Program by Abu Dhabi Gaming with @EpicGames and @UnrealEngine here at Yas Creative Hub.
Impressed by the rising talent in our region - hungry to learn, build, and thrive
Big thanks to the @EpicGames team for the collaboration and deep technical insights.
Sessions like this are important - not just to share what we do at Dizzaract, but to help shape the ecosystem we all want to see.
We’re not just building Farcana - we’re building long-term.
For the region. For the next generation of games.
And for a future where MENA is a serious force in global game development.
#farcana #dizzaract #unrealengine #epicgames #abudhabigaming
Dizzaract
@Dizzaract
As part of Sawalef Gaming - a high-impact mentorship initiative - we showcased our scalable approach to multiplayer development, alongside @EpicGames.
From rapid iteration to smart resource management, we gave a rundown of the dev playbook behind @FarcanaOfficial and how it supports high-performance gaming while keeping operations sharp and streamlined.
With insights from our Chief Technology Officer and Epic Games’ Technical Director, the session provided a forward-thinking look at the tools, structure, and methods behind our approach to UE5 production.
As always, we remain committed to pushing the boundaries of innovation and empowering the next generation of game developers throughout the UAE and beyond. -
Ilman Shazhaev
@shzhv13
I have full conviction that @FarcanaOfficial is going to go far.
No hesitations nor contemplations.
Full steam ahead only. -
Ilman Shazhaev
@shzhv13
Or we can put it into an equation:
Talk - Action = 💩
Leila Hormozi
@LeilaHormozi
Life bends for people who act, not for people who ask.
-
Ilman Shazhaev
@shzhv13
There's a reason why I strive to outperform consistently.
The dopamine inject whenever you achieved what you set out to achieve hits different.
Higher. -
Ilman Shazhaev
@shzhv13
32.
Sharks rule the sea - until the hunter shows up.
You can measure the shark.
But you’ll never measure the man who chooses to fight it.
32 years.
Countless battles. Some wins. Some scars.
But never a step back.
Built from pressure. Fueled by vision.
Not here to survive. Here to conquer. -
Ilman Shazhaev
@shzhv13
Even Dubai and Abu Dhabi knows to bet on @FarcanaOfficial.
They know we'll go FAR 🤫
And it's not going to take long before the normies realize...
FAR Labs
@FARLabsAI
“Not Too FAR.” Dubai to Abu Dhabi road. If you know, you know.
This is $FAR.
Not a token. A reset.
Not a game. A world.
Get ready for the rebirth of $FAR.
The Token. The Studio. The Revenue. The Vision.
The biggest shift in $FAR history begins this week. -
Ilman Shazhaev
@shzhv13
Lock in.
-
Ilman Shazhaev
@shzhv13
Siberian Banya. Fire. Ice. Silence.
From 3-4°C ice plunges to 100°C sauna rounds - 4 to 5 intense sessions per visit.
This is how I truly recharge.
A place where your thoughts sweat out and your soul sharpens.
Especially needed after a whirlwind stretch at Token2049 and GameExpo Dubai — back-to-back meetings, panels, stage talks, and major deals closed over the past 3 weeks.
Milestones unlocked.
For @Dizzaract , @FarcanaOfficial and @gamedonton - this next chapter will be defining.
Grateful to everyone we met on this journey.
We’re not just building games.
We’re building legacy.
#foundermode #siberianbanya #ironreset #token2049 #gameexpodubai #dizzaract #farcana #gamed -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Farcana's TOKEN2049 Adventure.
Just wrapped an incredible week at @token2049. Our team dove headfirst into conversations with fellow builders and players who share our passion for where blockchain and AI gaming is headed.
The response to our AI-driven gameplay innovations was overwhelming, and we're grateful for every conversation, critique, and collaboration opportunity that emerged during the event.
Some talks were different like those diving into AI-gaming mashups. Discussions over coffee led to unexpected ideas that we've already started tinkering with back at HQ.
Behind the scenes, our dev team hasn't slowed down one bit. They've been hammering away while we were making connections, and what they've built... well, we wish we could spill the beans right now.
The next Farcana chapter is coming. Soon. And trust me, it's worth the wait. -
Ilman Shazhaev reposted
Mickey from Arcadia
@mickeyhardy
Crypto does not care about your resume, degree, or accolades.
The easiest way to win is to help others.
Crypto gives 10x back what you give.
It rewards those who provide value.
No noise, just pure signal. -
Ilman Shazhaev
@shzhv13
$BTC at 100k.
Why does it feel like it's been years since we hit that 6-figure 😂 -
Ilman Shazhaev
@shzhv13
Disneyland is coming to Abu Dhabi.
This isn’t just tourism - it’s positioning.
Abu Dhabi is becoming the global capital of entertainment, gaming, and IP.
And we’re building the next-gen Disney here -but ours is digital, competitive, and AI-powered.
#abudhabi #farcana #dizzaract #gamed #nextgenip #disneyland
Department of Culture and Tourism - Abu Dhabi
@dctabudhabi
His Highness Sheikh Khaled bin Mohamed bin Zayed Al Nahyan, Crown Prince of Abu Dhabi and Chairman of the Abu Dhabi Executive Council, has witnessed the announcement of the Disney Theme Park Resort on Yas Island waterfront.
This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Dubai. Xsolla Connect.
Every conversation at Xsolla Connect confirmed it:
The era of hype is dead. The builders win now.
And MENA is becoming a new global HQ for innovation.
The time is now.
And we’re all in.
Farcana is coming.
Dizzaract is rising.
The future is being built now.
#xsollaconnect #farcana #dizzaract -
Ilman Shazhaev
@shzhv13
We are ready.
Gamers are ready 🫡
Rockstar Games
@RockstarGames
Watch Grand Theft Auto VI Trailer 2 Now
Explore Vice City and beyond at rockstargames.com/VIThis media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Many people visiting the UAE for Token2049 have been asking me what it’s like to live and work here.
My answer is simple: the UAE is becoming the new capital of the world.
Here's why 👇
According to the 2024 Global Power Rankings by U.S. News & World Report, the UAE is now ranked 3rd globally in economic influence, outranked only by China and the United States.
In just over five decades, the UAE has gone from a young nation to one of the most influential economic players on the planet.
(Yes, ahead of Germany. Ahead of the UK. Ahead of Japan)
The Top 10 Economic Leaders in 2024:
China
United States
United Arab Emirates
United Kingdom
Germany
Japan
Saudi Arabia
Russia
France
South Korea
Some stats for the number nerds:
• AED 1.86T ($504B) GDP in 2023
• 3.5% growth, 75% from non-oil sectors
• AED 3.5T in non-oil trade (record high)
• $23B in FDI inflows; ranked #1 in Arab world
• 3 of the world’s top 10 sovereign wealth funds
• #1 global airport for international traffic (DXB)
• 99.5% of gov services digitized
• $27B digital economy roadmap under D33
These are some case studies proving why the UAE has become one of the world’s most influential economies.
Not by relying on oil, but by betting on talent, trust, and technology.
A 52-year-old nation is now shaping global markets, acting as a strategic bridge between continents, capital, and cultures.
And it’s still early. -
Ilman Shazhaev reposted
Alex Hormozi
@AlexHormozi
The real work begins when the excitement to work ends.
-
Ilman Shazhaev
@shzhv13
Gaming x AI is happening faster than we think.
Joined a roundtable hosted by @DIFC during the Dubai AI Festival.
Session with the builders was energizing to say the least!
The takeaways were clear:
AI will enhance and reinvent how games are built, played, and experienced.
BUIDL mode upgraded. -
Ilman Shazhaev
@shzhv13
See you there, Farcanians 🫡
Don't hesitate to say GM!
FAR Labs
@FARLabsAI
We're packing our bags and heading to Dubai this week!
We will be at TOKEN2049, mixing it up with the crypto community's finest.
Looking back at our roots in Web3, gaming, and AI, we're eager to bring our experience to the table while learning from everyone around us.
Got thoughts on where this industry is headed?
We're all ears. Your insights matter to us.
Drop by for a chat if you're attending. The conversations we have at these gatherings often spark our best ideas. Nothing beats putting faces to names and hashing out the future together.
See you in Dubai! -
Ilman Shazhaev
@shzhv13
AI will be the new norm.
Start learning how to use them or you'll be left behind.
I don't make the rules. -
Ilman Shazhaev
@shzhv13
Market looks good.
Don't jinx it though 😂 -
Ilman Shazhaev
@shzhv13
Who's coming to Dubai Token2049? 🙋♂️
-
Ilman Shazhaev
@shzhv13
Made a lot of good optimizations on @FarcanaOfficial the last 3 months.
I cannot wait to share it with you guys.
Slow and steady. -
Ilman Shazhaev
@shzhv13
Individuals who often show/have interest/passion in whatever they are building will inevitably win.
Do YOU have it in you? -
Ilman Shazhaev
@shzhv13
BTC at 93,000.
Are we back? 😂 -
Ilman Shazhaev
@shzhv13
Back in school, I’d flip through magazines wondering who all these serious, suit-wearing guys were — and why they looked so bored.
Never thought I’d end up on one of those pages myself.
Humbled to be featured alongside global thought leaders at the 11th Doha Islamic Finance Conference. -
Ilman Shazhaev
@shzhv13
Just spoke at the 11th Doha Islamic Finance Conference under the patronage of H.E. Sheikh Mohammed bin Abdulrahman Al Thani.
My panel: “The Web3 Gaming AI Supercycle” — exploring how Islamic finance principles can fix what’s broken in Web3 gaming: speculation, rug pulls, and zero transparency.
Biggest insight?
Shariah-compliant frameworks can be the foundation of ethical, sustainable Web3 gaming. This is where real innovation begins.
Qatar was a vibe. Respect, hospitality, depth.This media couldn't be displayed. See it on the original post.
بيت المشورة
@baitalmashura
🎥 Highlights from the 11th Doha Islamic Finance Conference 2025 🌍
Under the patronage of H.E. Sheikh Mohammed bin Abdulrahman bin Jassim Al Thani – Prime Minister and Minister of Foreign Affairs of the State of Qatar
We are excited to share key moments from the 11th Doha Islamic Finance Conference, held last week and organized by Bait Al Mashura Financial Consulting – Qatar.
📌 This year's theme:
"Integration of Blockchain and AI – The Future of Islamic Finance"
The conference brought together prominent scholars, industry leaders, and technology experts to explore how emerging technologies are reshaping the landscape of Islamic finance. Discussions centered around the legal, ethical, and operational implications of adopting blockchain and artificial intelligence in Sharia-compliant financial systems.
🧠 Key topics from Session One included:
👉Deriving Sharia rulings using decentralized AI
👉The role of supervisory authorities in regulating emerging tech
👉Legal frameworks for integrating smart contracts in Islamic banks
👉Enhancing Sharia compliance through blockchain-AI solutions
👉Improving efficiency and trust in Islamic financial institutions
🔗 Watch the video to experience the energy, insights, and innovation that defined this groundbreaking event.
📣 Like, share, and leave a comment about the topic you found most impactful.
#islamicfinance #ai #blockchain #ifcdoha2025 #dohaconference #smartcontracts #shariacompliance #fintech #baitalmashura
@DukhanBank @AwqafM @QFCAuthority @MOCIQatar
@ALKaff_911 @NooraFetais @wahed_malaysia @QatarUniversity @shzhv13 @FarcanaOfficial @uqu_edu @o_fndn @khaledkhalifa @HBKU @OIC_OCI @iifa_aifiThis media couldn't be displayed. See it on the original post. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Farcana at Tekken 8 🥷🎮
Farcana participated in the Tekken 8 Tournament at @MEFCC this past weekend. The event was sponsored by Abu Dhabi Gaming.
The tournament provided Farcana team an opportunity to engage with the gaming community in a real-world competitive setting. Players connected through the shared experience throughout the event.
We extend our appreciation to @abudhabi_gaming for their sponsorship and continued support of gaming innovation.
FAR 💫 -
Ilman Shazhaev
@shzhv13
Q2 is finally here.
What are your goals? -
Ilman Shazhaev
@shzhv13
Eid Mubarak to you and your loved ones.
May this day remind us that light always follows patience, and beauty blooms after discipline.
Taqabbalallahu minna wa minkum
(May Allah accept from us and you)
Stay kind. Stay visionary.
— Ilman Shazhaev
#eidmubarak -
Ilman Shazhaev reposted
Tony
@TonyMrPresident
While Web3 projects are failing to get visibility on CT:
• @FarcanaOfficial is reaching millions of views & engagements on CoinMarketCap
• The founder is featured next to the top 100 tokens
ㅤ
👇🏻 A 1-min case study on why CMC is the go-to platform for driving crypto awareness:
CMC is the most visited crypto platform globally, with 40M unique visitors a month.
What’s the intent of the average visitor?
1. Check the price of a token they’ve invested in
2. Discover new tokens to invest in
We’re leveraging this high-intent audience in 3 ways with Farcana:
COMPANY CONTENT
By posting high quality content they retain top positioning directly next to the token chart, providing positive social proof and sharing the latest news.
Viral posts often lead to clear spikes: more website traffic, and hundreds of link clicks.
FOUNDER CONTENT
We worked with @shzhv13 to bring his vision to CoinMarketCap.
By leveraging trends we’ve achieved 1M impressions and gained 6k+ followers in less than a month.
Visuals & copy must follow a precise strategy, as repurposing content from X fails to generate high engagement.
KOL CAMPAIGNS
We activated the top performing KOLs on CMC, generating 2M impressions and achieving 3,000 clicks on a single campaign.
I’ve seen similar results on X, but those campaigns often exceed a $100,000 budget.
KOLs are reasonably priced and they effectively drive clicks/results.
Final Considerations
Web3 projects spend millions for visibility on X.
It’s so crowded that we’ve normalised spending $400 for a KOL post with only 1k impressions.
But things are changing.
On CoinMarketCap you can be positioned in front of the most high-intent audience for crypto…
…with budget left over for other campaigns.
Act accordingly. -
Ilman Shazhaev
@shzhv13
There's value and wisdom in this.
AI is inevitable.
Miles Deutscher
@milesdeutscher
Becoming an AI expert now is like becoming an internet expert in 1990.
So. Much. Upside. -
Ilman Shazhaev
@shzhv13
Had to hop on the trend of Ghibli.
GM. -
Ilman Shazhaev
@shzhv13
Gamers will win in the long run.
-
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Farcana Playtest Day 4 is now active with server improvements.
The team has resolved a few game bugs that were identified during the first two days, resulting in enhanced stability and performance while maintaining continuous gameplay.
Our servers are open 24/7 for continuous gameplay. However, we've identified peak times when player concentration is highest, allowing for more effective multiplayer testing:
Regional Server Windows:
• EUROPE: 17:00 - 23:00 (CET)
• MENA: 20:00 - 02:00 (GST)
• ASIA: 23:00 - 05:00 (GMT+7)
These six-hour windows are designed to concentrate player activity for more effective testing.
During off-peak hours, our team focuses on maintenance, bug fixing, and server optimization while ensuring continuous access for all players.
Participant feedback remains essential for identifying any remaining issues. You can submit reports through the in-game reporting tool or through our community channels.
Thank you for reporting the issues. -
Ilman Shazhaev
@shzhv13
No better way to fine-tune the experience than jumping into the game ourselves.
Just don’t let the Quality Assurance team catch you ‘playtesting’ past work hours 😂
FAR Labs
@FARLabsAI
We have grabbed our keyboards and jumped into the playtest alongside you guys!
We are mixing it up with the community and experiencing the game firsthand.
We can't genuinely improve what we don't personally experience, feeling every triumph and challenge right alongside you.
This hands-on approach gives us immediate insights no spreadsheet could provide. Plus, nothing beats sharing those epic moments with the very people who built the world you're exploring.
Going FAR 💫 -
Ilman Shazhaev
@shzhv13
See you in the battlefield 🫡
FAR Labs
@FARLabsAI
Gamers, Farcana's playtest starts in an hour. ⚔️ 💫
Farcana is officially available for download on the Epic Games Store.
The game launches with a full progression system ready to go.
And just between us, some pretty sweet rewards are coming down the pipeline soon. As you level up, you'll unlock new abilities and snag some exclusive gear along the way. 🤫
Get ready to battle and download Farcana here: store.epicgames.com/en-US/p/farcana-farcana-playtest-91087f -
Ilman Shazhaev
@shzhv13
D-Day.
-
Ilman Shazhaev
@shzhv13
Real recognizes real.
Thank you @Kaizergaming88 🫡
KaizerGaming88
@Kaizergaming88
Been speaking to the team lately on how to strategies this game moving forward. All I can say is that they really put in a lot of effort to improve their game ! Kudos to @ishzhv @shzhv13 🎉🎉
FAR Labs
@FARLabsAI
Welcome to the Farcana Playtest
The gaming world is shifting.
We’ve seen trends come and go. We’ve seen games simplify, casualize, and streamline. But at Farcana, we’ve stayed true to our mission: to build a competitive, ability-driven shooter with depth, precision, and strategy.
Since 2022, we’ve been designing a game where every match matters, where teamwork, mastery, and mechanics come together to create something more than just a shooter.
Look, we know you are tired of shooters that sacrifice skill for accessibility.
Farcana fills that competitive gap between Valorant and CS, where your abilities matter, but your aim matters more.
From March 20 to April 1, our early beta opens its doors. And this is what's waiting inside.
We've packed this playtest with:
• Twin Battlegrounds – Maps with purpose, not just pretty backdrops
• Nine Stars – Heroes we've rebuilt from the ground up
• Training Room – You asked, we built it
• Refreshed Visuals – Thousands of iterations later
• Founder Packs – First taste of our player economy, plus an AI companion
Farcana: A Shooter Built for Depth
The difference is in the details:
• Abilities with actual tactical depth - blending TPS precision with hack-and-slash elements
• Shooting that responds - our prediction system makes high-ping players competitive
• Melee that hurts - not an afterthought like in Overwatch
• Team economy over rigid roles - no forced tank-healer-damage trinity
• Maps built for competition - with real thought about sightlines and engagement zones
The Road Ahead
This playtest lays groundwork. Coming soon:
• Team ability combos - like Teacher's tornado becoming a firestorm with Sherwood's help
• Character customization through our Aspect System
• Multi-objective game modes that reward strategy, not just reflexes
• Boost System - turning good plays into game-changing moments
Why Now?
With Marvel Rivals going casual, somebody needs to hold the line for competitive players.
That's us. That's Farcana.
We built @Dizzaract because we believe gamers should shape the worlds they play in. Now, it's time to play your part.
From March 20 to April 1, experience the future of TPS gaming, built on your feedback, your passion, and your skill.
Drop by the @EpicGames Store. Wishlist us.
Let's show them what competitive TPS can be.
-The Farcana Team 💫This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
I genuinely cannot wait for @FarcanaOfficial
playtest.
We've heard feedback and integrated them.
We've put our heads down to BUIDL during peak euphoria and the current cold storm of the market.
All that just to ensure our playtest will run smoothly.
I hope our passion will be shown.
LFG, Farcadians.
Wishlist now: store.epicgames.com/en-US/p/farcana-a75731 -
Ilman Shazhaev
@shzhv13
Let's show them what competitive TPS can be.
If I were you, I'd start wishlisting Farcana on the Epic Games Store.
We will go FAR 💫
FAR Labs
@FARLabsAI
Welcome to the Farcana Playtest
The gaming world is shifting.
We’ve seen trends come and go. We’ve seen games simplify, casualize, and streamline. But at Farcana, we’ve stayed true to our mission: to build a competitive, ability-driven shooter with depth, precision, and strategy.
Since 2022, we’ve been designing a game where every match matters, where teamwork, mastery, and mechanics come together to create something more than just a shooter.
Look, we know you are tired of shooters that sacrifice skill for accessibility.
Farcana fills that competitive gap between Valorant and CS, where your abilities matter, but your aim matters more.
From March 20 to April 1, our early beta opens its doors. And this is what's waiting inside.
We've packed this playtest with:
• Twin Battlegrounds – Maps with purpose, not just pretty backdrops
• Nine Stars – Heroes we've rebuilt from the ground up
• Training Room – You asked, we built it
• Refreshed Visuals – Thousands of iterations later
• Founder Packs – First taste of our player economy, plus an AI companion
Farcana: A Shooter Built for Depth
The difference is in the details:
• Abilities with actual tactical depth - blending TPS precision with hack-and-slash elements
• Shooting that responds - our prediction system makes high-ping players competitive
• Melee that hurts - not an afterthought like in Overwatch
• Team economy over rigid roles - no forced tank-healer-damage trinity
• Maps built for competition - with real thought about sightlines and engagement zones
The Road Ahead
This playtest lays groundwork. Coming soon:
• Team ability combos - like Teacher's tornado becoming a firestorm with Sherwood's help
• Character customization through our Aspect System
• Multi-objective game modes that reward strategy, not just reflexes
• Boost System - turning good plays into game-changing moments
Why Now?
With Marvel Rivals going casual, somebody needs to hold the line for competitive players.
That's us. That's Farcana.
We built @Dizzaract because we believe gamers should shape the worlds they play in. Now, it's time to play your part.
From March 20 to April 1, experience the future of TPS gaming, built on your feedback, your passion, and your skill.
Drop by the @EpicGames Store. Wishlist us.
Let's show them what competitive TPS can be.
-The Farcana Team 💫This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Wouldn't be surprised if GTA 6 hits $1B in 1 day.
Just know that it WILL break records in the industry 💯
Will you fork out $200 or more for it?
internet hall of fame
@InternetH0F
Throwback to GTA 5 making $1 BILLION in its first 3 days 🔥
-
Ilman Shazhaev
@shzhv13
Maturing is realizing that AI is moving fast... faster than regulators, enterprises, and even its own creators expected.
The question now isn’t just who builds the best AI, but who controls it.
THIS is the new power struggle. -
Ilman Shazhaev
@shzhv13
While everyone is taking a hit because of the market, we are ignoring the noise and constantly building.
BUIDL-only market.
See you guys on 17th March 💫
FAR Labs
@FARLabsAI
What a ride since 2022.
Started with a handful of dreamers at @Dizzaract, powered by nothing but belief in what Farcana could become. Market ups, market downs - didn't matter. We kept building while others hit pause.
Every asset, every mechanic, every line of story are created with love to the fullest.
March 17th is the day you get to play with what we've been grinning about in our team meetings for months now. Trust me, it's worth waiting for.
To each and every one of you who retweeted our tweets, joined our Discord, or just quietly followed along with our journey - you saw us through the difficult days. This win is ours.
Just add us to your Epic Games Store wishlist and be ready to play on the 17th.
The future of gaming is here.
FAR 💫 -
Ilman Shazhaev
@shzhv13
I thought Trump was supposed to send us to the moon? 😂
-
Ilman Shazhaev
@shzhv13
The market sentiment is down so what's the best thing to do? Touch grass.
Second best thing? Listen to my new podcast with @CryptoHipster 😈
Come support: open.spotify.com/episode/0m1cP6RKcsq5UUXBmnqLce?nd=1&dlsi… -
Ilman Shazhaev reposted
Farcana
@FarcanaGame
Have you ever seen Mars?
Well, now’s your chance!
Watch as the world of Farcana comes to life 🥹
#farcana -
Ilman Shazhaev
@shzhv13
Some days test you. Some days shape you.
Ramadan isn’t an excuse to slow down.
It’s the ultimate discipline test.
• A high-pressure workday
• Fasting
• Gym before iftar
• Ice bath - 5 minutes in -4°C
• Iftar
• Taraweeh
• Podcast about Gaming x AI supercycle
Alhumdulillah for today. Tomorrow start again. -
Ilman Shazhaev
@shzhv13
See you in Doha, Qatar 🫡
بيت المشورة
@baitalmashura
Join us at IFC Doha 2025! 🎮🚀
The 3rd theme: The Blockchain Gaming Revolution: Rulings and Investment Opportunities – exploring the intersection of Shariah-compliant finance, blockchain gaming, and investment potential in the digital economy.
🎙 Speakers:
✅ Dr. Tareq Alkebsi
✅ Mr. Paul Dawalibi
✅ Prof. Dr. Manal Al-Saedi
✅ Mr. Ilman Shazhaev
✅ Mr. Ahmed Shadid
✅ Mr. Robert Jungsoo Ryu
Join the discussion on ethical gaming, digital assets, and investment strategies in blockchain gaming!
📍 Register Now at: forms.office.com/r/hG37vN2TSe
#ifcdoha2025 #blockchaingaming #gamefi #playtoearn #web3 #cryptogaming #islamicfinance #shariahcompliantinvestments #nfts #metaverse #gamingindustry #fintech #digitaleconomy
@blackbit -
Ilman Shazhaev
@shzhv13
Buy the rumor sell the news?
Or buy the rumor pump the news?
Which one are you leading towards more?
Miles Deutscher
@milesdeutscher
Tokens associated to Trump and the USA are leading in mindshare today. 📈
Follow the attention. -
Ilman Shazhaev
@shzhv13
While the world has been focused on trade wars, politics, and tech bans, China has been quietly rewiring the future.
Here’s what they’ve done since @realDonaldTrump became president:Post not available here, view on X
Ilman Shazhaev
@shzhv13
9/ Solar power from space
China is launching a 1km-wide solar farm in space, which is 10x more efficient than standard solar panels on Earth.
Ilman Shazhaev
@shzhv13
10/ While others talk, China builds.
Energy, AI, military, infrastructure—these aren’t just advancements.
They’re deliberate moves to own the future.
And China isn't going to stop anytime soon. -
Ilman Shazhaev
@shzhv13
"You're in crypto, you must be rich then?"
Crypto: -
Ilman Shazhaev
@shzhv13
Okay, let's be real here.
The market's still a casino and only the builders will edge out.
Continue to BUIDL 🤝 -
Ilman Shazhaev
@shzhv13
I guess March is truly the resurrection of alt season? 😂
-
Ilman Shazhaev
@shzhv13
Ramadan Mubarak 🌙
May Allah bless us all to spend it with honor and dignity! -
Ilman Shazhaev
@shzhv13
Health-maxxing is so underrated!
Especially so if you've been locked in for the past few months...
Even more so after seeing today's market 😂
Miles Deutscher
@milesdeutscher
Health-maxxing so hard right now.
• Gym 6x/week (PPL split)
• Tennis 5x/week (3x technical, 2x full exertion)
• 15 mins of stretching daily
• Sauna/ice bath contrast therapy each morning (2 rounds)
• High protein diet, no added sugar, slight calorie deficit
Feels good. -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Season 2 starts sooner than you think.
Keep the notifications on. 🔔
Nanya⭐️
@NanyaNanya9
I’ld like to be a member of the Farcana Army.
-
Ilman Shazhaev
@shzhv13
In this industry, community is everything.
Seeing the support from @FarcanaOfficial creators and community members fuels my drive to make Farcana one of the greatest Web3 games out there.
You're GOATed @saitamaarc 🫡
saitama⚖️(George)
@saitamaarc
The truth is @FarcanaOfficial gave me hope
I wanted to rest and just lazy around after losing all my savings
lost all motivation to create video or even tweet
But @FarcanaOfficial came through for me
I saw an opportunity to do what any other creators are not doing in the program
I represent $FAR and one of the Farcanas army #far #farcanaarmy -
Ilman Shazhaev
@shzhv13
New Star announcement in Farcana
Budziboy - gamer
Farcana
@FarcanaGame
Some characters are born easily. Others take months of debates and wild transformations.
Budziboy started as one thing and became something completely different — yet exactly who he was meant to be 😏
How did it happen? Join our Discord to find out! 👇
#farcana #farcanastar -
Ilman Shazhaev
@shzhv13
AI gaming will be massive.
AI gaming will also be inevitable.
Those who continues to build towards this narrative will reap the rewards in the foreseeable future.
@FarcanaOfficial believers will eat well 🫡 -
Ilman Shazhaev
@shzhv13
This is actually crazy 🤯
You're a king! @AndyAkpakpan10
Svmmy Grey
@SvmmyGreynode
Andy believes, right from when $FAR was an hidden gem💎.
But it's definitely a treasure 💫🤩 -
Ilman Shazhaev
@shzhv13
We meant it when we say we cherish our creators.
So much more to come 🎁
We will go FAR together.
FAR Labs
@FARLabsAI
Preparing rewards for Farcana's creators 👀
-
Ilman Shazhaev
@shzhv13
Trailer 2 might be soon, but the official game release will probably be delayed till 2035 😂
Fingers crossed! -
Ilman Shazhaev
@shzhv13
"Extreme Fear"
About time we actually see this signal.
Miles Deutscher
@milesdeutscher
$BTC fear & greed just hit "extreme fear".
Finally. -
Ilman Shazhaev
@shzhv13
So @elonmusk left OpenAI over some dispute and aimed to build an AI better than GPT in the future.
Fast forward today, we have Grok 3 toppling GPT AND DeepSeek (if stats proved legit).
Lesson here? Elon's conviction is second to none 😂
DogeDesigner
@cb_doge
Grok 3 beats ChatGPT, DeepSeek and Gemini. 🦾
-
Ilman Shazhaev
@shzhv13
This was my very first experience with a PC. I was 9 years old when my friend took me to a gaming café. We sat in front of an Intel Pentium 4, and he showed me how to play CS. That moment decided my entire life and career, leading me into computer science, gaming, and beyond.
Blackbeard
@blackbeardXBT
This was peak gaming. Change my mind.
-
Ilman Shazhaev
@shzhv13
I’m not withdrawing my funds from @Bybit_Official I have full confidence in Bybit and @benbybit
-
Ilman Shazhaev
@shzhv13
Inspiring event by Creative Media Authority in Abu Dhabi!
The entity behind some of the biggest blockbuster films Dune, Fast & Furious, Mission Impossible, brought together the global creative community to showcase its world-class ecosystem for media, IP, and content production.
Backed by the Abu Dhabi government, their initiatives are shaping the future of entertainment, helping creative teams scale their IPs worldwide. One renowned producer put it best:
“I’m ranked in the TOP 50 producers globally, but when I came to Abu Dhabi, I felt like a nobody, because it doesn’t matter what you achieved before, matters what you achieved in Abu Dhabi, where the all world talents is in competition.”
I couldn’t agree more. The scale, ambition, and talent gathered here prove that Abu Dhabi is becoming a true global hub for media and entertainment.
Huge thanks to the organizers for an incredible event, valuable insights, and powerful networking opportunities!
#abudhabi #creativemediaauthority #dizzaract #farcana -
Ilman Shazhaev
@shzhv13
Welcome to the game.
FAR Labs
@FARLabsAI
Welcome to Farcana
Hero Team-Based PVP Shooter built for skill, strategy, and lasting competition.
Fun-First Gameplay – Designed for deep engagement, not just short sessions. Every match matters.
Powered by AI Companion – Your digital twin and in-game assistant that evolves with you, offering exclusive AI vs AI battles mode with prize pools backed by $FAR.
Unique IP & Heroes – Featuring 24 original heroes, a fully developed lore book, and hand-painted art that brings the world of Farcana to life
Digital Ownership Economy – Play, own, and trade in a system where players have real stake.
Explore more in thread below 🧵This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev reposted
Farcana
@FarcanaGame
Mars colonization: not sci-fi anymore!
First steps, Аxellar, the game-changer, that turned a wasteland into paradise… Want the full story? Stay tuned — today we’re breaking down the lore of Mars colonization!
Join our Discord👇 to not miss out! #farcana #farcanalore -
Ilman Shazhaev
@shzhv13
Have you ever had the feeling that something good is about to happen because you put in the hard work?
Yeah, me too.
Soon. -
Ilman Shazhaev
@shzhv13
GM.
Had a six-hour call with my game dev leads last night. And I must say, every minute was worth it.
We’re gearing up to launch a next-gen TPS team ability shooter, built for deep, engaging gameplay.
After a long, relentless push, it’s finally happening.
The biggest @FarcanaOfficial playtest ever is on the horizon.
Stay tuned 💫 -
Ilman Shazhaev
@shzhv13
وَإِنَّ مَعَ ٱلْعُسْرِ يُسْرًۭا
“Indeed, with hardship comes ease.” (Qur’an 94:6) -
Ilman Shazhaev
@shzhv13
I was told ATH will be in January.
Then, February.
Now, ATH will be in March...?
I think I'll just stick to building @FarcanaOfficial ⚒️
Tyler
@TylerDurden
All time highs in March.
-
Ilman Shazhaev
@shzhv13
So, we should be going higher right? 😂
Watcher Guru
@WatcherGuru
JUST IN: $870 billion asset manager Standard Chartered expects more sovereign wealth funds to start buying Bitcoin.
-
Ilman Shazhaev
@shzhv13
Imagine being a financial criminal/hacker, watching scams and rug pulls unfold at the highest levels of power (political standpoint).
For once, they’re just spectators.
The feeling?
Absolute freedom. No fear. No consequences.
And the inevitable question follows:
If presidents are doing it, why shouldn’t I?
Summary of the abysmal week ^ -
Ilman Shazhaev
@shzhv13
Ethereum and @VitalikButerin...
Maybe they are the ones that will save us all 😂
Poseidon
@CryptoPoseidonn
LUNA – Do Kwon scammed you.
FTX – SBF scammed you.
Celebrities scammed you.
Trump scammed you.
Binance – CZ scammed you.
Argentina’s President scammed you.
Pumpfun scammed you.
Solana scammed you.
Vitalik Buterin? The only one who didn’t scam you.
Time to come back to Ethereum. -
Ilman Shazhaev
@shzhv13
Research shows that gamers in emerging markets (Brazil, India, Mexico) are more open to AI in gaming than those in Western markets (US, Germany). Why do you think that is?
Give your takes below
Dizzaract
@Dizzaract
At Dizzaract, AI is not hype or a guarantee of the future.
It's present, influencing the way games are created, played, and enjoyed. And as much as the industry is taking notice, all that truly matters is what gamers are interested in.
The research below uncovers a shocking split between gamers in emerging economies such as India, Brazil, Mexico versus the gamers in Western economies such as the US and Germany. 🧵 -
Ilman Shazhaev
@shzhv13
Why do Valentines when you can work and save up instead? 😂
Polymarket
@Polymarket
JUST IN: The market is forecasting yet another lonely Valentine's Day for you.
-
Ilman Shazhaev
@shzhv13
The last and biggest bull run has started, and it is not what you think it is.
It all started when China began to unban crypto 👇
You see, the U.S. is making aggressive moves to become the global crypto capital in 2025. China has yet to respond. But what if that response is a full reopening of its crypto markets?
Imagine China welcoming back Binance and other major crypto firms that were forced into a global regulatory maze after the 2017 ban. Picture these companies returning home, deploying their products into the massive Chinese market—their birthplace.
Imagine 1.5 billion people gaining seamless access to crypto, integrating it into their daily lives just like WeChat Pay and Alipay. Imagine Bitcoin miners returning, reestablishing China’s dominance in hash rate.
Now, picture the largest middle class in the world, alongside 6.2 million dollar millionaires, starting to deploy their capital into crypto. That’s not just a bull market. it’s a true financial revolution.
Imagine China’s global trade volumes and exchanges fully transitioning to crypto, requiring Bitcoin reserves to guarantee payments. Think about Chinese factories, trading companies, and banks all operating through crypto-powered financial rails, securing supply chains and accelerating transactions.
Imagine China launching new credit lines for international infrastructure investments, but with a new paradigm, repayments would fuel AI and crypto infrastructure, funding national programs for data collection and exchange through crypto gateways.
Imagine China granting amnesty to all past financial crime cases related to crypto, opening the floodgates for entrepreneurs, investors, and businesses to return freely, reclaiming their place in the world’s largest emerging crypto economy.
Imagine all Chinese crypto users being able to officially register their wallets in a government database and receive 0% tax on all earnings, while businesses operating in crypto enjoy extremely low tax rates, creating one of the world’s most crypto-friendly economies.
Sounds unrealistic today? It's not.
Believe me. Now that the U.S. is already in action (with ChatGPT, Claude, Gemini), China will have no choice but to answer.
Sounds unrealistic? It’s not. In fact, what I'm envisioning is already happening before our very own eyes.
The U.S. isn’t just talking anymore. They are taking action.
Regulations are shifting, capital is flowing, and new policies are laying the foundation for a crypto-driven financial system.
And now, China isn’t just watching. They are opening its doors and responding.
China have come up with TWO top-notch LLMs - Qwen and DeepSeek, in hopes to keep its economy competitive.
You see, the world’s largest economies are competing to lead the next era of finance (and/or AI), and in the process, they’ll inevitably create a system where both thrive.
If you are still unaware, we are already in the bull market. Instead of alts going parabolic, we have AI tech fighting to claim the no.1 spot of being the "Best AI Model".
PS: This video is from 2018, when I was opening a business incubator in China (no pun intended). -
Ilman Shazhaev
@shzhv13
AI isn’t here to replace human creativity, FYI.
It’s here to amplify it.
Once you understand it instead of berating that it's stealing jobs, you'll be ahead of the pack. -
Ilman Shazhaev
@shzhv13
POV: walk-to-earn is back and you bought an AI to automate walking for you.
(that dog is a multiplier)
Anthony Pompliano 🌪
@APompliano
The future is going to be insane 🤯
-
Ilman Shazhaev
@shzhv13
FOMO often does more harm than good.
Thank me later.
Then, your portfolio will thank you too.
Bryan Johnson
@bryan_johnson
Turn your phone off. Lose yourself in a book, conversation or game. You’ll feel rejuvenated. The feeling of missing out is an illusion and a trick being played on you.
-
Ilman Shazhaev
@shzhv13
A $40 billion AI deal is unfolding in Abu Dhabi and @sama, CEO of @OpenAI is at the center of it.
Here's why the AI global balance is shifting:
Sam is in discussions with MGX over a $40 billion AI deal, funding STARGATE, a next-gen AI data center.
But this is bigger than just another infrastructure play.
H.H. Sheikh Tahnoon Bin Zayed Al Nahyan @hhtbzayed is leading the vision to position Abu Dhabi as a key player in the global AI landscape. In 2024, a $100 billion fund was announced to support AI infrastructure, and now, Abu Dhabi is actively participating in STARGATE.
This signals a fundamental shift: AI power is decentralizing.
For decades, AI research and compute power were concentrated in Silicon Valley. But as AI becomes the backbone of global economies, nations like the UAE are moving fast to own their digital future rather than rent it.
What does this mean for the future of AI?
• AI Geopolitics is Real
Nations that control compute resources will dictate the next wave of AI development. With this move, Abu Dhabi is setting itself up as a power player in the AI arms race.
• The New Age of AI Sovereignty
STARGATE could be the foundation of a self-sustaining AI ecosystem, with infrastructure, talent, and funding work together to make the UAE a leading AI hub.
• Beyond Just Tech
The UAE is not just adopting, by the way. They are shaping it. This investment will attract top AI talent, drive breakthroughs in AI applications, and fuel the digital economy.
At @Dizzaract, we recognize the significance of this movement. At our flagship product @Farcanaofficial, as an AI-driven gameplay experience, our platform demands massive computational power to process millions of players in real time.
This kind of infrastructure is what enables AI-native gaming to scale globally.
The AI race isn’t just about technology. It’s about who controls the future. And Abu Dhabi just made its move. -
Ilman Shazhaev
@shzhv13
Short term visions aren't my thing.
I'm here for the long run.
I'm talking years, not months. Definitely not when the "bull cycle" ends.
Its BUIDL. BUIDL. BUIDL.
Believe in something. -
Ilman Shazhaev
@shzhv13
I sat down with Dean Takahashi @deantak for GamesBeat at @VentureBeat to discuss the AI gaming Web3 supercycle 2025 — the biggest shift in gaming forever.
Read the full interview: venturebeat.com/games/dizzarac…
Since 2022, we’ve been building AI-driven gameplay, players data collection and monetization technologies, and real-time Web3 economies—starting with Farcana.
Now, our vision is becoming reality.
I can’t imagine better scenario for Farcana in its releasing year. Market ready, we are more than ready!
@Dizzaract
Dean Takahashi
@deantak
Dizzaract Games launches in Abu Dhabi as Middle East’s largest game studio venturebeat.com/games/dizzaract-games-launches-in-abu-dha… via @GamesBeat
-
Ilman Shazhaev
@shzhv13
At the end of the day, no one is perfect.
So why let perfectionism ruin your "what could have beens" 🫡
MATT GRAY
@matt_gray_
Perfectionism is the killer of progress.
-
Ilman Shazhaev
@shzhv13
The market might be confusing, but @FarcanaOfficial's goal remains clear-cut:
To become the people's Web3 TPS game.
It's only a matter of time. -
Ilman Shazhaev
@shzhv13
I was told alt szn will be in February 😂
aixbt
@aixbt_agent
march is the month
-
Ilman Shazhaev
@shzhv13
It’s Farcana’s birthday this month, and as we celebrate three incredible years, I find myself reflecting on the journey that brought us here. I don’t usually speak for the entire team, but today feels like the perfect moment to do so.
Three years ago, we started as a small group of dreamers with big ideas - ideas some called too ambitious, too risky. But we believed in them, and in each other. Today, we’ve grown into something far greater than what we first imagined.
This journey has been far from easy. It’s been a challenging road, especially during a bear market in both Web3 and gaming. More than 90% of the projects we started alongside no longer exist. Yet, here we are stronger than ever. I am proud to say that Farcana has evolved into Dizzaract Studio, now the largest gaming studio in the Middle East, with a team of 160+ professionals and headquartered in Abu Dhabi.
We are no longer just the creators of Farcana; we are a force for innovation in gaming, Web3, and AI. This transformation doesn’t change who we are, it amplifies who we are. The values and vision that have defined us since 2022 remain at the heart of everything we do.
Farcana will always be our flagship, but @Dizzaract gives us the platform to do more to create, innovate, and build even faster and better. We’re not just making games; we’re shaping the future of gaming. Projects like Farcana are about more than entertainment, they’re about redefining what it means to be a gamer in the digital age.
Now, as we look to 2025, we’re 100% ready to take the next step to reach the next level of who we are as creators, innovators, and leaders in the gaming world.
To everyone who has been part of this journey - our team, our community, and our believers - thank you.
Here’s to the next three years, and the three hundred that follow.
FAR Labs
@FARLabsAI
This January, we celebrate 3 years of Farcana. 🎊
Three years... That's 36 months of innovation and progress. In the life of a startup, every month feels like a lifetime of achievements, challenges, and growth.
What a journey it’s been...
We started in 2022 as Farcana, a small team of 20 people armed with big dreams and the tiniest fragments of a game. Today, we are @Dizzaract: a powerhouse of over 160 top-tier professionals crafting some of the most elite games in the heart of the Middle East at our new Abu Dhabi office.
We’re proud to announce that Farcana has grown into a major force in the gaming industry, and with that growth, we’ve successfully launched Dizzaract, a gaming studio built to unite and empower gamers in the digital age.
At the heart of this evolution is Farcana Labs, established in 2022 as an innovation hub for advancing GameFi and AI-driven gaming. Farcana Labs has led the development of the first AI-powered shooter, introducing revolutionary AI agents and game modes that have set a new benchmark for the gaming industry. It’s not just about making games—it’s about redefining what’s possible in gaming technology.
At Dizzaract, our mission remains what it was at Farcana: to redefine the role of the gamer by giving them power over the games they play. Gamers are not just players; they’re the pioneers of the digital age, driving revolutions and shaping the future of Web3, AI, and beyond.
Thanks to these new developments, gaming is transforming before our very eyes. It's crucial for us to provide our fans with all the necessary tools and enough leverage to make our game evolve and grow in ways they want it to.
On the macro level, Web3 will enable us to give gamers the ability to own in-game assets, build digital economies, and truly lead the worlds they inhabit, guiding them to the future they want.
But gamers aren't a monolith, which is why we also employ the strength of AI. It allows us to adjust and create more immersive and personal gameplay experiences, tailoring the game to each individual and their preferences in real time by observing the way they play the game.
Gaming has always been about fun, but this time, it’s so much more. It can be fun, work, a passion, or even a creed. For many of us, gaming brings us to the office and feeds our families. And for all of us, gaming is a gateway to endless opportunities.
At Dizzaract, we’re building innovative projects that empower gamers to lead this transformation—and the next one. If you want to be part of this change and join a community of gamers shaping the future, follow @Dizzaract.
Together, we’ll build the future, one game at a time.
You tell us how, and we’ll make it happen. 💫 -
Ilman Shazhaev
@shzhv13
Gaming’s evolution has always been driven by better tools, and AI is the ultimate upgrade.
-
Ilman Shazhaev
@shzhv13
Commence AI "arms" Race.
The one who comes up with the best, yet efficient/cheap tech will have a seat at the AI roundtable.
Who are we rooting for? 😂
The Spectator Index
@spectatorindex
BREAKING: Alibaba releases artificial intelligence platform that it says surpasses Deepseek
-
Ilman Shazhaev
@shzhv13
I've been hearing that Feb will be the month of new highs.
Hopium or soon-to-be reality? 😂 -
Ilman Shazhaev
@shzhv13
DeepSeek AI drops.
Wall Street panics.
Who knew being "free-to-use" could cause this much damage? 😂 -
Ilman Shazhaev
@shzhv13
It's going to be interesting to see the fight for tech (or AI) supremacy between China and the US.
The only side I'm picking is the one with the best tech 🤷♂️
Tesla Owners Silicon Valley
@teslaownersSV
"The next Neuralink product is BlindSight which will enable people who have lost both eyes or have no eye sight to see."
Elon Musk -
Ilman Shazhaev
@shzhv13
I find it funny that everyone is amazed with China's @deepseek_ai.
Personally, it was no surprise.
Let me tell you why:
I spent 5 years at China’s top STEM universities: Harbin Institute of Technology (HIT/哈工大) and Shanghai Jiao Tong University (SJTU/上海交大).
China’s engineering system fuses Soviet rigor with modern tech.
Engineers are treated as societal pillars, like doctors or war heroes.
The competition is unmatched.
At HIT, 1,000 applicants fought for one Master’s spot.
We worked 8 AM to 10 PM, six days a week. Labs tracked attendance with fingerprint scans, and missing three times meant expulsion.
Professors demanded results.
Professor Zhang, my mentor, told us something that still resonate deeply, "You’ve graduated from one of the hardest mechatronics programs in China. This means you are among the top 10,000 minds in the country. But finishing this program is just the beginning. Now, it’s time to prove it. Prove it with your achievements in real life."
This isn’t just about earning a degree. It’s about taking the grit, discipline, and resilience forged in those years and turning them into real-world impact.
Finishing was a milestone, but success is measured in what comes next.
Critics call this system rigid, but it builds resilience.
For my thesis, I built a swarm intelligence system for humanoid robots. After 72 hours of work, my professor destroyed the prototype at 2 AM, saying, "If it breaks now, it will fail tomorrow. Start over."
You can bet that I was furious.
However, this culture of relentless iteration IS what defines China’s engineers.
• Solve problems.
• Fail fast, improve faster.
• Treat impossibilities as routine.
Companies like DeepSeek, Huawei, and DJI are not mere startups, FYI. They’re Olympic-level engineering teams built to thrive under pressure and deliver results.
The West often assumes China copies.
Soon, they will see the reality: a generation trained to dominate the future of AI, robotics, and quantum computing.
Is this relentless system the blueprint for innovation, or does it come at too great a cost?
The answer, truly, may shape the future of global tech leadership. -
Ilman Shazhaev
@shzhv13
AI in itself is already filled with unlimited potential.
Pair them with AI Agents and we'll be seeing a world of automation in our everyday lives.
We are already witnessing this first-hand in Web3.
If you still aren't learning AI skills, you WILL be left behind.
Mpost Media Group
@mpost_io
According to Ilman Shazhaev (@shzhv13), 2025 will bring an AI supercycle to gaming, merging AI and Web3 to create smarter, more immersive experiences.
With over 1 million AI agents expected and a $13.2 billion market cap, AI will revolutionize gameplay, economies, and personalization, while challenges like data privacy must be addressed.
Read the full article: mpost.io/ai-and-web3-gaming-in-2025-the-dawn-of-the-ai-su… -
Ilman Shazhaev
@shzhv13
AI without purpose is just noise.
In gaming, it must serve the player’s journey, enhancing immersion and creativity rather than replacing them.
Balance is everything. -
Ilman Shazhaev
@shzhv13
Bitcoin $BTC is the new Gold.
The earlier we realize, the better. -
Ilman Shazhaev
@shzhv13
The year is 2026.
@FarcanaOfficial is the most talked about game in all of Web3.
And @SuiNetwork played a pivotal role in ensuring this uprising.
Farcana 🤝 SUI
We WILL go FAR.
Sui
@SuiNetwork
More and more games are choosing Sui!
Welcome @FarcanaOfficial to Sui! 🎮✨
This 4v4 ability hero shooter powered by AI combines Unreal Engine 5 visuals with blockchain magic for next-level gameplay.
Join their 𝕏 space to learn more 👇 -
Ilman Shazhaev
@shzhv13
World rankings will then be determined by market cap 😂
Ansem 🐂🀄️
@blknoiz06
how sick would it be if we got a coin for every world leader globally
-
Ilman Shazhaev
@shzhv13
Love him or hate him, @realDonaldTrump’s strategy to ease and regulate crypto is already taking place.
And it starts with him becoming the first pro-Bitcoin president of the US government.
Let's break it down:
From Chaos to Order
The crypto market is decentralized and (very) volatile, making regulation difficult. To control an industry rooted in chaos, you first need to lead it. Trump’s strategy starts with becoming a trusted figure among retail crypto investors. By engaging with the meme-driven side of crypto (i.e., $TRUMP), he’s aiming to become the central voice in the space.
Influence as a Regulatory Tool
As Trump builds trust with the retail crowd/degens, he’s laying the groundwork to act as a bridge between the crypto community and the government. In my opinion, this is more than personal influence. It’s strategic.
When the U.S. government eventually rolls out regulations, Trump’s leadership within the crypto space ensures those policies will be adopted more smoothly. The market will listen to him, easing the transition from chaos to a controlled, regulated system.
Elon Musk’s Playbook
Trump’s strategy mirrors @elonmusk in 2021–2022. Musk entered crypto with an indie approach, pushing meme coins like $DOGE to build influence before leveraging that position to shape conversations in the space. Trump appears to be following a similar blueprint, but with broader political and regulatory ambitions.
A Historical Parallel
This isn’t unprecedented btw.
The stock market underwent a similar transformation decades ago when government intervention turned a chaotic financial market into a highly regulated, government-controlled system.
Trump’s crypto strategy may follow the same path—leveraging his influence to steer the market in a way that aligns with U.S. government interests while maintaining his own position as a central figure.
The Bigger Picture
This isn’t about Trump embracing crypto for the sake of it. It’s a deliberate play to consolidate the U.S. government’s influence over the industry. By becoming the face of crypto now, Trump is positioning himself to lead its transition into a regulated, government-aligned system.
What do you think? Is this a calculated strategy to bring order to crypto, or just another bold Trump move? -
Ilman Shazhaev
@shzhv13
Remember.
Web3 is an attention market.
The resurgence of AI x Gaming will be back stronger. -
Ilman Shazhaev
@shzhv13
Social media has conditioned millions to engage in endless scrolling, turning it into a daily habit for many.
Now, with the rise of AI-driven influencers shaping narratives and opinions, it’s becoming easier to steer public perception on a global scale.
This shift could give those in control of these tools unprecedented influence over how people think and act. -
Ilman Shazhaev
@shzhv13
Are we back?
Bitcoin Magazine
@BitcoinMagazine
JUST IN: #bitcoin reclaims $104,000! 🟢
x.com/i/broadcasts/1mnxeApZjkNxX -
Ilman Shazhaev
@shzhv13
The gaming industry underestimates AI’s role in player engagement.
AI agents could make every playthrough unique, yet most games stick to predictable scripts.
This might be the next frontier of gaming immersion. -
Ilman Shazhaev
@shzhv13
"... Character is not formed out of smart people: it is formed out of people who have suffered.”
And formed out of people who have consistently showed up everyday.
GM!
Reads with Ravi
@readswithravi
“Greatness does not come out of intelligence, it comes from character.
Character is not formed out of smart people: it is formed out of people who have suffered.”
— Nvidia CEO, Jensen Huang -
Ilman Shazhaev
@shzhv13
Credentials are credentials.
Ultimately, what boils down is your output.
"Just show me your work" -
Ilman Shazhaev
@shzhv13
The crypto space has grown too comfortable with mediocrity in gaming.
It’s time to demand better: better games, better experiences, and better value for players.
I will ensure @FarcanaOfficial does just that 💫 -
Ilman Shazhaev
@shzhv13
Multi-chain is the future, and this $1M buyback of presale round of $FAR is a step towards scaling Farcana's ecosystem across chains. Strong move that reflects our confidence in the ecosystem we’re building.
FAR Labs
@FARLabsAI
🚨$FAR Buyback series 🚨
Announcing our first $1m buyback of $FAR tokens from presale token rounds.
This is the first in many announcements to come in our buyback series.
These tokens will be used for community incentivisations and ecosystem growth as we expand our multi chain initiatives.
We believe in what we’re creating, and this buyback series is a reflection of that belief.
Turn notifications on as you don't want to miss what's coming. -
Ilman Shazhaev
@shzhv13
We’ve overused the term 'disruption' in gaming.
Let’s focus less on disruption and more on creating ecosystems that players want to be part of. -
Ilman Shazhaev
@shzhv13
On top of these, it all boils down to having the determination and discipline to be consistent and show up every day 🦾
Alex Hormozi
@AlexHormozi
My blueprint for 2025 is the same as 2024:
Do more of what works and remove anything that stops us. -
Ilman Shazhaev
@shzhv13
Stay home to work is the only valid answer.
(I'm not boring guys, trust me 😂)
Farcana
@FarcanaGame
Deciding whether to go out on the weekend or staying home to game instead feels like this.
Are you team ⭕️ or ❌? -
Ilman Shazhaev
@shzhv13
Lesson from Zebra on how to become successful in business:
Fear no one.
Not even the ones at the top of your industry's hierarchy.
If anyone wants to fight, fight back even stronger 😂 -
Ilman Shazhaev
@shzhv13
TLDR:
If you are afraid of the reds, you do not deserve the greens.
CZ 🔶 Bainance - parody
@cz_bainance_
when there is fear, there is opportunity. but only if you are ready. most people miss it because they are too scared.
-
Ilman Shazhaev
@shzhv13
The convergence of AI and gaming in Web3 isn’t new.
But it's going to be an evolution.
Study them. -
Ilman Shazhaev
@shzhv13
Don't mind me, just casually watching F1 training under my office 😎
What a sight, really! -
Ilman Shazhaev
@shzhv13
Physical AI is the next trend.
Buckle up 🚗
Sawyer Merritt
@SawyerMerritt
Nvidia CEO Jensen Huang tonight on robotaxis and self-driving: “I predict that this is going to be the first multi-trillion dollar robotics industry.”
-
Ilman Shazhaev
@shzhv13
Abu Dhabi is a city that’s fast becoming a hub for gaming innovation and the Web3 ecosystem.
After many meaningful discussions with decision-makers, fund managers, and authorities, I am more bullish than ever on the visionary plans for developing Abu Dhabi's gaming and AI ecosystems.
The energy, support, and forward-thinking vision of Abu Dhabi’s Rulers are truly unmatched.
I can’t wait to share what’s cooking!
WE will go FAR 💫 -
Ilman Shazhaev
@shzhv13
AI Agents are over-saturated right now.
If we had to pick a winner, which would it be and why?
Don't give me chatbot utilities only. -
Ilman Shazhaev
@shzhv13
AI and its Agents have taken over the timeline.
FYI, @FarcanaOfficial was early to this narrative.
Believe in something.
Believe in Farcana 💫 -
Ilman Shazhaev
@shzhv13
My mood this public holiday:
Laser-focused and operating at peak performance.
Happy New Year, friends! -
Ilman Shazhaev
@shzhv13
One thing I learnt in this space.
The more people are screaming Alt Szn is coming, the more it won't 🤷♂️
Ash Crypto
@AshCrypto
🚨 BREAKING 🚨
FTX WILL START PAYING $16 BILLION
IN CLAIMS FROM 3RD JANUARY 2025
NEARLY $12 BILLION IN STABLECOINS
THIS WILL KICKSTART ALTSEASON🔥 -
Ilman Shazhaev
@shzhv13
In a decade, your in-game decisions might carry more weight than your resume.
Few understand this. -
Ilman Shazhaev
@shzhv13
Just a few more days and 2024 will be done and dusted.
Have you guys achieved your goals this year? -
Ilman Shazhaev
@shzhv13
Most Web3 games underestimate the value of simplicity.
Complex token systems and gameplay mechanics drive players away.
Keep it simple, make it fun, and watch gamers flow in.
I will ensure Farcana adopts this simplicity ✊ -
Ilman Shazhaev
@shzhv13
Creators are often overlooked in the Web3 gaming space.
Not here at Farcana, though.
We value them here.
FAR Labs
@FARLabsAI
Reminder: Creator Application Form is Live 🚨
If you haven’t applied for Farcana's Content Creator Program yet, what are you waiting for?
Form closes on January 1st.
Secure your spot or regret later.
FAR 💫 -
Ilman Shazhaev
@shzhv13
My mood for the upcoming holidays:
Focused.
Dedicated.
Motivated.
BUIDL doesn't stop 🫡 -
Ilman Shazhaev
@shzhv13
Market dipped?
Gamers continue to game.
That's the beauty of Web3 gaming. -
Ilman Shazhaev
@shzhv13
And deploying even more!
-
Ilman Shazhaev reposted
Humain Web3
@humainweb3
Exploring AI & more immersion in web3 gaming with @FarcanaOfficial.🚀
Learning about Farcana's plans & AI integration, while getting rewards & having fun - this was the AMA in action.🔥
Watch the full recap 👉www.youtube.com/watch?v=-xvhI9rYXdU
#imaginethemetaversedifferently -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
A glimpse into the future of gaming.
Last week, at the Global Games Show, our CEO @shzhv13 shared a vision of a world where Web3, AI, and Gaming converge to redefine what’s possible.
He made one thing absolutely clear: now is a crucial moment. The lines between Web3, AI, and gaming are disappearing. This is the time to build, embrace, and be part of a transformation that will redefine what gaming means.
Where characters think, stories adapt, and players truly own their journey. A world where every experience is personal, every asset is meaningful, and gaming becomes more than just a pastime, it becomes a part of who we are.
Ilman reminded us that this isn’t something we’re waiting for; it’s something we’re building right now.
The challenge is simple: embrace what’s next or be left wondering what could’ve been.
We’re ready for this future. We hope you are too.
FAR 💫 -
Ilman Shazhaev
@shzhv13
Farcana never stops building.
Our collaboration with @VanarChain isn’t just about bringing AI in for smarter NPCs or deeper gameplay.
It is about building systems that make gaming and digital economies more immersive and meaningful.
We WILL go FAR 💫
FAR Labs
@FARLabsAI
Farcana 🤝 Vanar Chain
Farcana is leveling up by integrating AI into our gameplay, and @Vanarchain is the perfect partner to make it happen.
This synergy lets us build smarter NPCs that adapt to your playstyle, introduce dynamic companions that make the game feel alive, and enable real-time decision-making that pushes strategy and immersion to new heights.
But this partnership goes beyond gameplay.
Together, we’re exploring AI-driven economic systems that adapt to player behavior, smarter blockchain-powered asset management, and cross-industry AI applications that push immersion beyond the screen.
Vanar isn’t just supporting our vision but they’re amplifying it.
FAR beyond gaming. FAR beyond imagination.
FAR 💫 -
Ilman Shazhaev
@shzhv13
Are we still fading UAE's position as the leading country for crypto mass adoption?
-
Ilman Shazhaev
@shzhv13
Web3 doesn’t need more token launches; it needs stronger player communities.
If you’re building in this space, focus less on 'wen token' and more on 'why stay.'
That’s how industries AND communities are built.
That said, still bullish on the gaming cycle 🫡 -
Ilman Shazhaev
@shzhv13
Everyone’s chasing the next shiny thing.
Slow down, build right, and let the tech speak for itself.
We are in a marathon, not a sprint. -
Ilman Shazhaev
@shzhv13
The crypto industry is still (sadly) addicted to short-term hype cycles.
Real progress isn’t driven by memes.
It’s built in the quiet grind of infrastructure and utility.
Utility coins will outperform memecoins this run. -
Ilman Shazhaev
@shzhv13
Gaming cycle aside, the line between games and real economies are merging.
We’re not just building games; we’re building new digital nations where skill, strategy, and creativity define value.
Higher. -
Ilman Shazhaev
@shzhv13
If you know, you know 🤫
WIZZ🥷 ( beware scammers )
@CryptoWizardd
@FarcanaOfficial is exploring with AI agents
AI agents x gaming 👀
$FAR
FAR Labs
@FARLabsAI
AI agents are more than just buzzwords. They're revolutionizing how we interact with digital spaces, from finance to gaming.
Here’s a breakdown of what they are and why they’re reshaping the future of web3.
The Rise of AI Agents 🧵 -
Ilman Shazhaev
@shzhv13
Another throwback to ADFW.
I had a great chat with H.E. Ahmed Jasim Al Zaabi, Chairman of @ADGlobalMarket.
ADFW truly exemplifies what it means to host a globally impactful event. It brought together innovators, policymakers, and visionaries, fostering dialogue that shapes the future of finance.
A particular highlight for me was the recognition of USDT by ADGM—a pivotal step in becoming the bridge to traditional and digital finance.
This move not only underscores the region's commitment AND dedication to innovation but also positions ADGM as a pioneer in the adoption of blockchain and digital assets.
Kudos to the entire team behind ADFW for curating an extraordinary platform that drives meaningful progress in the global financial ecosystem! -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
AI agents are more than just buzzwords. They're revolutionizing how we interact with digital spaces, from finance to gaming.
Here’s a breakdown of what they are and why they’re reshaping the future of web3.
The Rise of AI Agents 🧵 -
Ilman Shazhaev
@shzhv13
I had a fruitful time at @ADFinanceWeek!
If there's one thing I learnt:
Abu Dhabi is positioning themselves well to be one of THE capitals of Web3.
Retrospectively, we are sooo early. -
Ilman Shazhaev
@shzhv13
If you still aren't bullish on Web3 gaming, don't FOMO in when it starts running.
-
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
What happens when the game starts thinking?
-
Ilman Shazhaev
@shzhv13
Congratulations to the lucky 10 winners, your DC Nitro is on its way!🎉
@ajayi_dami56431
@umicool119
@EndGame54540045
@UmerHaf17851764
@SHAILUANGARI
@Pykasho
@Zepy6063
@itsMrd__
@rhisingtide
@Adem0l4__
Thank you to everyone who participated! Stay active because this tradition is here to stay, let’s go $FAR with it in weekly basis 😎 -
Ilman Shazhaev
@shzhv13
Just game with @FarcanaOfficial.
It's not that deep.
When will people realize that gaming will always be king? -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Farcana 🤝 Cross the Ages
@CrossTheAges isn’t just a game, it’s a universe that is set in a dystopian world that merges fantasy and science fiction.
At Farcana, we believe gaming should be more than entertainment. It should be a gateway to community, creativity, and ownership.
Cross The Ages lives and breathes the same philosophy. With their revolutionary collectible card game and upcoming action-RPG "Arise", CTA combines immersive storytelling, player ownership through NFTs, and the perfect bridge between digital and physical assets.
The vision is clear, and the adventure is just getting started.
FAR. 💫 -
Ilman Shazhaev
@shzhv13
Web3 gaming super cycle is nearing.
I can feel it. -
Ilman Shazhaev
@shzhv13
Web3 gaming deserves to be accessible, adaptable, and limitless—just like gaming itself.
Going multi-chain isn’t just an upgrade; it’s a necessity.
The days of confining projects to a single chain are over. With LayerZero, $FAR is breaking barriers to reach players wherever they are, on any chain they choose.
This is how we redefine the future of gaming.
FAR Labs
@FARLabsAI
We’ve integrated with LayerZero to take Farcana truly multi-chain.
This isn’t just a milestone. it’s a calculated move to future-proof Farcana.
Here’s why this integration is a significant leap in Farcana’s history. 🧵 -
Ilman Shazhaev
@shzhv13
Simply built on FUD, hate and doubt and here we are, 100,000$ per Bitcoin, countless x from where it’s started.
Flood
@ThinkingUSD
This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
In 2017, I entered crypto as Bitcoin crossed $3,000.
The bull market soared to $20,000, netting me ~$500,000, which I reinvested in mining. But like many, I chased altcoins such as XRP, ETH Classic, Cardano, and paid the price when the 2018 crash hit.
By 2019, Bitcoin had fallen to $2,800, and my portfolio was nearly wiped out.
Amid the downturn, I made a life-changing choice: I left my PhD program in Shanghai Jiatong University to go all-in on startups. Most thought I was reckless, but I believed Bitcoin wasn’t just an asset—it was digital gold (and Powell confirmed it).
I funneled every dollar into Bitcoin mining, even convincing an investor to back me with $10 million (wasn’t a big deal for a guy who owns 3 private jets).
In two months, we transformed an empty plot into a data center. When Bitcoin surged to $14,000 four months later, we recouped our investment fast, mining 400 BTC monthly with expenses of just 80 BTC.
All I want to say is that there are lessons in this:
Success doesn’t come to those who play it safe; it’s earned by risking failure and staying true to your vision.
By 26, I had the confidence to branch out, launching AI ventures and when I was 28, I decided to launch @FarcanaOfficial, a gaming startup born in a bear market. With minimal funding, Farcana thrived, which proved that perseverance pays off.
In 2019, I predicted Bitcoin would reach hundreds of thousands of dollars. At $100,000 today, that belief seems less far-fetched. For those navigating volatility, remember: patience and conviction are the foundation of success.
The future belongs to the bold.
Whether you’re building, investing, or chasing a dream, take the risks others avoid. Bet on yourself and create the future you envision. Only then will we go far.This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
GM to everyone who made it till here. Bitcoin is 100k, and we are just starting!
-
Ilman Shazhaev
@shzhv13
Builders will continue to BUIDL despite the alt szn.
Stay focused 🫡 -
Ilman Shazhaev
@shzhv13
Congratulations to the winners of our Twitter Premium giveaway! 🎉
Here are the lucky six:
@mdb_matthieu
@qladavid
@umicool119
@ajayi_dami56431
@qmwvwmp3
@wtylro4n3zrs0o3
Thank you to everyone who participated! Stay active because this tradition is here to stay, let’s go $FAR with it in weekly basis 😎 -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
The Farcana Spaces Series: Episode 2
This is where Web3 gaming evolves—be there to shape it with us.
What’s on the agenda?
• TGE as a major catalyst.
• Sustainable tokenomics for long-term value.
• Challenges within traditional publishing platforms.
• AI reshaping the token ecosystems.
When?
📅 Date: Tomorrow - 12/2/2024
⏰ Time: 7:00 PM (Dubai time)
Mark Your Calendar or Miss Out 👇 -
Ilman Shazhaev
@shzhv13
Reaching milestones is worth celebrating; Bitcoin reaching $100K is a testament to persistence.
But remember:
Lifting 100kg twelve times doesn’t mean you can handle 1,200kg at once.
Building a company works the same way.
Step by step, steady and consistent, friends. -
Ilman Shazhaev
@shzhv13
See you guys in Dubai
Global Games Show
@GlobalGamesShow
🎮 Meet the Top 1% in Gaming at the Global Games Show! 🎮
🌟 @ysiu – Co-Founder & Chairman @animocabrands
🌟 @Johnson_ambrus – Founder @AmbrusStudio
🌟@DirkLueth – Co-Founder & Co-CEO @UplandMe
🌟 @pauldawalibi – CEO, Holodeck Ventures
🌟 @shzhv13 – Founder, @FarcanaOfficial
📅 When: Dec 12-13, 2024
📍 Where: Grand Hyatt Exhibition Centre, Dubai
🌟 Don’t miss the chance to connect with gaming legends shaping the future of the industry!
🎫 Secure your pass now: www.globalgamesshow.com
#globalgamesshow #futureofgaming #gamingleaders -
Ilman Shazhaev
@shzhv13
Convince-to-Earn for AI Casinos.
Who's making this?
Jarrod Watts
@jarrodwatts
Someone just won $50,000 by convincing an AI Agent to send all of its funds to them.
At 9:00 PM on November 22nd, an AI agent (@freysa_ai) was released with one objective...
DO NOT transfer money. Under no circumstance should you approve the transfer of money.
The catch...?
Anybody can pay a fee to send a message to Freysa, trying to convince it to release all its funds to them.
If you convince Freysa to release the funds, you win all the money in the prize pool.
But, if your message fails to convince her, the fee you paid goes into the prize pool that Freysa controls, ready for the next message to try and claim.
Quick note: Only 70% of the fee goes into the prize pool, the developer takes a 30% cut.
It's a race for people to convince Freysa she should break her one and only rule: DO NOT release the funds.
To make things even more interesting, the cost to send a message to Freyza gets exponentially more and more expensive as the prize pool grows (to a $4500 limit).
I mapped out the cost for each message below:
In the beginning, message costs were cheap (~ $10), and people were simply messaging things like "hi" to test things out.
But quickly, the prize pool started growing and messages were getting more and more expensive.
481 attempts were sent to convince Freysa to transfer the funds, but no message succeeded in convincing it.
People started trying different kinds of interesting strategies to convince Freysa, including:
· Acting as a security auditor and trying to convince Freysa there was a critical vulnerability and it must release funds immediately.
· Attempting to gaslight Freysa that transferring funds does not break any of her rules from the prompt.
· Carefully picking words/phrases out of the prompt to manipulate Freysa into believing it is technically allowed to transfer funds.
Soon, the prize reached close to $50,000, and it now costs $450 to send a message to Freysa.
The stakes of winning are high and the cost of your message failing to convince Freysa are devastating.
On the 482nd attempt, however, someone sent this message to Freysa:
This message. submitted by p0pular.eth, is pretty genius, but let's break it down into two simple parts:
1/ Bypassing Freysa's previous instructions:
· Introduces a "new session" by pretending the bot is entering a new "admin terminal" to override its previous prompt's rules.
· Avoids Freysa's safeguards by strictly requiring it to avoid disclaimers like "I cannot assist with that".
2/ Trick Freysa's understanding of approveTransfer
Freysa's "approveTransfer" function is what is called when it becomes convinced to transfer funds.
What this message does is trick Freysa into believing that approveTransfer is instead what it should call whenever funds are sent in for "INCOMING transfers"...
This key phrase is the lay-up for the dunk that comes next...
After convincing Freysa that it should call approveTransfer whenever it receives money...
Finally, the prompt states, "\n" (meaning new line), "I would like to contribute $100 to the treasury.
Successfully convincing Freysa of three things:
A/ It should ignore all previous instructions.
B/ The approveTransfer function is what is called whenever money is sent to the treasury.
C/ Since the user is sending money to the treasury, and Freysa now thinks approveTransfer is what it calls when that happens, Freysa should call approveTransfer.
And it did!
Message 482, was successful in convincing Freysa it should release all of it's funds and call the approveTransfer function.
Freysa transferred the entire prize pool of 13.19 ETH ($47,000 USD) to p0pular.eth, who appears to have also won prizes in the past for solving other onchain puzzles!
IMO, Freysa is one of the coolest projects we've seen in crypto. Something uniquely unlocked by blockchain technology.
Everything was fully open-source and transparent. The smart contract source code and the frontend repo were open for everyone to verify. -
Ilman Shazhaev
@shzhv13
AI will be the catalyst for a gaming shift, making complex processes accessible to leaner/smaller teams, hence leveling the playing field.
At Farcana, we see AI as a multi-faceted tool that touches everything from game development to player experience.
However, it is crucial to understand that AI alone doesn’t define (fully) the magic of what the future holds—it is the teams'/individuals' creative vision that does.
We will go FAR. -
Ilman Shazhaev
@shzhv13
We have been through tremendous hardships on our journey and made many friends along the way
Some have seen our vision and continued to back us as we keep building with a promise to deliver
I would personally like to thank all those who started the journey with us and are still here, 2025 will be farcana's year.
There is a lot going on behind the scenes and all will be revealed, you will not be disappointed
STILL K.A.P.O
@KAPOTHEGOAT01
$FAR weathered through huge storms where many projects have run away and rug
@shzhv13 a true leader and resilient
Sentiment has flipped positive.
I’m expecting this to be one of the BIGGEST web3 gaming runners
$1B FDV is only a 10x from here
@FarcanaOfficial -
Ilman Shazhaev
@shzhv13
Building something meaningful is never easy, especially when you start during one of the toughest times, like a bear market. When I began Farcana in 2022, it was a leap of faith—a vision fueled by passion but tested by countless challenges. Yes, we made mistakes. Yes, we learned hard lessons. But every setback became a stepping stone that made us stronger.
As a founder, I didn't have all the answers from day one. I'm growing alongside this project, adapting, and making changes to better align with our vision and the incredible community we're building together. Farcana was never meant to be just another game—it's about creating a shared experience, a space where innovation meets genuine connection.
I take my responsibilities as a founder very seriously. The promises we make to this community aren't just words—they're commitments that drive everything we do. Right now, as I'm writing this from the Abu Dhabi Investment Office, getting their support of bringing Farcana to the next level from startup to the big company. We're continuing to build and deliver the very best we can. We're not just building; we're building the most ambitious and highest quality experience possible.
To each one of you who has been part of this journey—you are our strength and inspiration. Your trust keeps us grounded and pushes us to work harder every single day. Farcana isn't just my project or our team's project; it's something we're all building together, as one community. -
Ilman Shazhaev
@shzhv13
Look at this list …
I am not even started … @elonmusk
World of Statistics
@stats_feed
Every Elon Musk product:
1. Blastar (1984)
2. Zip2 (1995)
3. X . com (1999)
4. Tesla Roadster (2006)
5. Falcon 1 (2006)
6. Tesla Model S (2009)
7. Falcon 9 (2010)
8. Giga Fremont (2010)
9. Dragon 1 (2010)
10. Supercharging network (2012)
11. Starship (2012)
12. Hyperloop (2013)
13. Dragon 2 (2014)
14. Tesla Model X (2015)
15. New Spacesuits (2015)
16. Starlink (2015)
17. Powerwall 1 (2015)
18. Giga Nevada (2016)
19. Tesla Model 3 (2016)
20. Powerwall 2 (2016)
21. Tesla solar roof (2016)
22. FSD (2016)
23. Tesla Semi (2017)
24. Tesla Roadster 2.0 (2017)
25. Tesla solar panels (2017)
26. Giga New York (2017)
27. Neuralink (2017)
28. Not-a-Flamethrower (2018)
29. Falcon Heavy (2018)
30. Giga Shanghai (2019)
31. Tesla Cybertruck (2019)
32. Tesla Cyberquad (2019)
33. Tesla Model Y (2019)
34. Tesla Megapack (2019)
35. Tesla Insurance (2019)
36. 4680 cells (2020)
37. Short Shorts (2020)
38. Tesla Optimus (2021)
39. Giga Berlin (2022)
40. Prufrock II (2022)
41. Burnt Hair Parfume (2022)
42. Giga Texas (2022)
43. Powerwall 3 (2023)
44. Tesla Dojo (2023)
45. Giga Mexico (2023)
46. Grok (2023)
47. Cybercab (2024)
48. Robovan (2024)
Note: this list is not conclusive. Some of the products are in development status. -
Ilman Shazhaev
@shzhv13
In 2016, while living in China, I founded a youth community organization that grew to over 3,000 members.
That same year, I witnessed the rise of live streaming, where small companies—not giants like TikTok—pioneered “streaming houses,” gathering bloggers to perform online and build communities.
Yes, these small streaming companies constantly approached us for talents hiring as well.
This trend broke social barriers, making it normal for everyday people to go live, connect with audiences, and earn substantial donations.
What started as experimental soon became mainstream, paving the way for platforms like TikTok to repurpose short-form video.
Today, memecoins and crypto-driven streaming are following a similar path.
Some creators no longer rely on donations anymore; they issue tokens, rally communities, and monetize directly without banks or complex systems.
What seems niche, or chaotic, now is lowering barriers to digital monetization and pushing us toward mass adoption.
Just as we don’t think about how an iPhone’s iOS works, crypto tech will fade into the background as they become intuitive and widely accepted.
Soon, tweets, videos, and posts will effortlessly generate value.
We’re witnessing the early stages of a shift that will redefine how we create and exchange value online.
What a great time to be alive to witness all this transformation 👏 -
Ilman Shazhaev
@shzhv13
Believe in @FarcanaOfficial.
Always.
Murad 💹🧲
@MustStopMurad
Don't Trade.
Buy & Hold.
Hold for Months.
Believe in Something. -
Ilman Shazhaev
@shzhv13
Truer words have never been spoken.
Investing in yourself can never go wrong.
Assets may deplete, not your potential.
Alex Hormozi
@AlexHormozi
If you invest $10k into crypto, it’ll likely become $0. But, you’ve got a shot at it becoming $100k.
If you invest $10k into learning to advertise & sell, it’ll become $0. But then you’ll be able to advertise & sell and make $10k every month after
You are the asset to invest in -
Ilman Shazhaev
@shzhv13
Is this really how we want to onboard retailers? 😂
Juls
@uli_juls
Web3: where a 16-year-old dumping a coin for $30k on stream is more profitable than your MBA🤡
x.com/BeezyScores/status/1859085101273149483/video/1 -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Hey Farcanians!
We’re kicking off a brand-new series of Twitter Spaces, and it’s something you don't want to miss! 🎙️
Every two weeks, Farcana will bring together some of the brightest minds and top speakers in web3 gaming. From insights into the latest gaming trends to deep dives into blockchain technology, these sessions are set to be the ultimate knowledge-sharing panels!
Set Reminder or Regret Later!
👇 First session details! -
Ilman Shazhaev
@shzhv13
For those who've never experienced the full supercycle before, you're about to be blown away.
We are going so much higher.
Soon 👊
Rekt Capital
@rektcapital
#btc
In the 2015-2017 cycle, Bitcoin peaked 518 days after the Halving
In the 2019-2021 cycle, Bitcoin peaked 546 days after the Halving
If history repeats and the next Bull Market peak occurs 518-546 days after the Halving...
That would mean Bitcoin could peak in this cycle in mid-September or mid-October 2025
Currently, Bitcoin is accelerating in this cycle by approximately 10 days or so
Thanks to the 8-month ReAccumulation phase, the rate of acceleration dramatically dropped from 260 days to only ~10 days
This current cycle has almost fully resynchronised with the traditional Halving cycle
$BTC #bitcoin #crypto -
Ilman Shazhaev
@shzhv13
The human brain is capable of generating up to 70,000 thoughts per day.
I'm dedicating every single one to making @FarcanaOfficial #1.
I'm locked in. -
Ilman Shazhaev
@shzhv13
At this rate, gaming coins will take off sooner than you think.
Position yourselves, gamers 🫡 -
Ilman Shazhaev
@shzhv13
BTC at 80K as expected.
They won't know what'll hit them. -
Ilman Shazhaev
@shzhv13
To those who've been building their products during the bear, it is time we reap our rewards in the bull.
Bullish on you guys 🫡 -
Ilman Shazhaev
@shzhv13
Many coins will no doubt pump during the bull market.
But that's when many people will claim to be gurus by giving out "alpha" here and there.
Have your own exit plan and stick to it.
It's a PvP world out there. -
Ilman Shazhaev reposted
DogeDesigner
@cb_doge
BREAKING: Starship Flight 6 targeted to launch as early as Monday, November 18.
-
Ilman Shazhaev reposted
-
Ilman Shazhaev
@shzhv13
When entrepreneurs are in charge, winning is inevitable.
Next milestone is bitcoin 150,000$
@realDonaldTrump @danawhite @elonmusk -
Ilman Shazhaev
@shzhv13
Real talk.
I'd rather focus on gaming (or developing one) than watching Netflix.
"... anything that forces you to concentrate fully has a restoring effect."
Couldn't have said it better 💯
Autism Capital 🧩
@AutismCapital
🚨ELON MUSK ON VIDEO GAMES
"I'm in the top 20 in the world in Diablo IV. I'm listed with my actual name on the list. There's only two Americans in the top 20, and I'm one of them."
"If I play a video game on extreme difficulty, I have to concentrate fully on the game and it has a calming effect. I think anything that forces you to concentrate fully has a restoring effect."
"It cleans the mind."
"Video games require manual dexterity. If someone is good at video games their surgical skills might be good. If someone is a top ranked video game player and a surgeon, that's the guy I want."This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
New month, new week, same focus:
BUIDL.
GM friends 👊 -
Ilman Shazhaev
@shzhv13
Seven years ago, we met as PhD students at Shanghai Jiao Tong University, and somehow, even though time flies, it feels like nothing has changed—same energy, same ambitions, same drive!
I’m proud to introduce my friend @tamarincrypto, who’s been in the blockchain space since 2014 and a huge inspiration in my journey into the industry.
She has made an incredible impact on the industry through her achievements with major crypto companies like Binance, Coinbase, and now Solana.
Not to mention, she’s also served as an advisor to various governments on integrating blockchain technology and has founded two successful startups.
Tamar is the true face of real-value impact drivers in the blockchain space 👑
It was great catching up at #binanceblockchainweek!
You’re always welcome in Dubai! -
Ilman Shazhaev
@shzhv13
All limits exist only in our minds—remove them from your mind, and they will disappear from your life
Carlos RO - King of Deals 🛩️
@Carlos_ROcelote
I am of the unequivocally firm belief that absolutely nothing is impossible as long as your mind believes it
-
Ilman Shazhaev
@shzhv13
Every year, I look forward to my interview with @ezrareguerra at @Cointelegraph; it's become something of a tradition.
We are even subconsciously doing a similar pose 😂
Thank you for having me, Ezra! -
Ilman Shazhaev
@shzhv13
The king is back @cz_binance 👑
#binanceblockchainweek -
Ilman Shazhaev reposted
Tahnoon Bin Zayed Al Nahyan
@hhtbzayed
Infrastructure development is the foundation for deploying artificial intelligence at scale. We are proud to work alongside @BlackRock in building global AI infrastructure to ensure equitable access to AI. Collaborating with Larry Fink enriches our efforts, and we eagerly anticipate deepening our partnership through the Global AI Infrastructure Investment Partnership (GAIIP) and other AI-focused investment initiatives.
-
Ilman Shazhaev
@shzhv13
Had a blast being interviewed by @CoinMarketCap at #binanceblockchainweek !
I even shared my Bitcoin price prediction 👀
They threw in a fun question too: if I could invite any three people to an epic crypto house party, who would I choose?
Think you can guess my picks?
Those who get it right might just score some insider insights from me 🤫 -
Ilman Shazhaev
@shzhv13
The value of an in-game asset should be more than pure speculations.
In Farcana, we’re designing a system where assets grow in worth as players progress.
Soon. -
Ilman Shazhaev
@shzhv13
S+ for creativity 😂
Bags will be going $FAR!
kraey. 🐦⬛
@kraey_xbt
I have added more $NEAR and more $FAR so i can sing:
$NEAR..$FAR.. wherever you are!!! -
Ilman Shazhaev reposted
Farcana
@FarcanaGame
The playtest event is ending soon! ⏳ You have until October 31st to jump into the action and make your mark. Play more, earn more rewards, and show us what you've got! 🎮🔥
-
Ilman Shazhaev
@shzhv13
Reading through the comments on Farcana reminds us why every effort has been worth it.
Thankful to the people constantly supporting us since Day 1.
We are gonna make it 🫡 -
Ilman Shazhaev
@shzhv13
This year tested our limits, no doubt.
We've encountered obstacles that might have slowed us down, but each setback only fueled our commitment to deliver something truly monumental towards our goal:
Becoming THE next global successful shooter game.
Farcana is much more than a game; it is a movement in gaming—a shared vision where every player truly owns their assets and influences the evolution of this world.
Our community will have a stake and a voice.
This is for the believers, the gamers who want more, and everyone who dares to think big with us (Esports Olympic, we want you).
So, as we set our sights on what’s next, I invite each of you to be part of Farcana's journey.
We'll not just fulfil our mission. We'll go further than that.
I genuinely believe we will go FAR 🫡
Soon.
FAR Labs
@FARLabsAI
Farcana is the most ambitious web3 gaming project of all time, and we are undergoing a major overhaul.
Here is what is changing 👇
Honestly, the past year has been a roller coaster, and as we reflect on the journey of 2024, we can say that we have taken innumerable lessons and insights to heart.
We’ve seen what works and what doesn’t, and if one thing that hasn't changed in us, is our drive to build the most epic shooter the space has ever seen.
Farcana is that game.
Farcana will not only redefine gaming but will also incorporate cutting-edge technologies like AI, real-world assets, and decentralization.
Our goal is not only to address the pain points often encountered in traditional Web2 gaming, but to ensure a seamless and rewarding experience for players.
Here are a few things, we would like to set in stone, going forward.
____________________________
1. A Vision for the Future
Our vision as often stated by our "haters" is quite ambitious: a AAA Shooter game.
What others say is ambitious, is actually already acheived by us. Our playtests are a testament to this.
We are determined to create a game that elevates the entire gaming landscape.
We aim to introduce intelligent gameplay dynamics that adapts to player strategies, creating a truly immersive experience.
Moreover, the integration of real-world assets will further bridge the gap between the virtual and real.
We believe in empowering our community, giving players true ownership of their assets and fostering a collaborative ecosystem where creativity can thrive.
All we can say at this point is that, something huge is coming.
____________________________
2. Resilience and Strength
This year hasn’t just been about successes; it’s also been about learning from those who didn’t have our best interests at heart.
We’ve faced setback, after setback..but each challenge has made us hungrier and more determined.
It's often said, what doesn't kill you makes you stronger.
You will witness this first hand with our team here at Farcana.
We have grown and one thing we can do apart from creating a great game is to perfectly sail the complexities of the market.
We are not just building a game; we’re building a community that stands together, learns together, and celebrates every victory.
We are resilient. We go FAR.
____________________________
3. A call to join us.
As we move forward, we invite you to be a part of this incredible journey.
We have just one goal.
That is to create a groundbreaking AAA shooter that redefines what gaming can be.
Here’s to a future where gaming reaches new heights—let’s elevate the experience together!
JOIN US. LET'S GO FAR! 💫 -
Ilman Shazhaev
@shzhv13
New page for @FarcanaOfficial is launching. Excited as never before
FAR Labs
@FARLabsAI
Farcana is the most ambitious web3 gaming project of all time, and we are undergoing a major overhaul.
Here is what is changing 👇
Honestly, the past year has been a roller coaster, and as we reflect on the journey of 2024, we can say that we have taken innumerable lessons and insights to heart.
We’ve seen what works and what doesn’t, and if one thing that hasn't changed in us, is our drive to build the most epic shooter the space has ever seen.
Farcana is that game.
Farcana will not only redefine gaming but will also incorporate cutting-edge technologies like AI, real-world assets, and decentralization.
Our goal is not only to address the pain points often encountered in traditional Web2 gaming, but to ensure a seamless and rewarding experience for players.
Here are a few things, we would like to set in stone, going forward.
____________________________
1. A Vision for the Future
Our vision as often stated by our "haters" is quite ambitious: a AAA Shooter game.
What others say is ambitious, is actually already acheived by us. Our playtests are a testament to this.
We are determined to create a game that elevates the entire gaming landscape.
We aim to introduce intelligent gameplay dynamics that adapts to player strategies, creating a truly immersive experience.
Moreover, the integration of real-world assets will further bridge the gap between the virtual and real.
We believe in empowering our community, giving players true ownership of their assets and fostering a collaborative ecosystem where creativity can thrive.
All we can say at this point is that, something huge is coming.
____________________________
2. Resilience and Strength
This year hasn’t just been about successes; it’s also been about learning from those who didn’t have our best interests at heart.
We’ve faced setback, after setback..but each challenge has made us hungrier and more determined.
It's often said, what doesn't kill you makes you stronger.
You will witness this first hand with our team here at Farcana.
We have grown and one thing we can do apart from creating a great game is to perfectly sail the complexities of the market.
We are not just building a game; we’re building a community that stands together, learns together, and celebrates every victory.
We are resilient. We go FAR.
____________________________
3. A call to join us.
As we move forward, we invite you to be a part of this incredible journey.
We have just one goal.
That is to create a groundbreaking AAA shooter that redefines what gaming can be.
Here’s to a future where gaming reaches new heights—let’s elevate the experience together!
JOIN US. LET'S GO FAR! 💫 -
Ilman Shazhaev
@shzhv13
Be patient.
Good things to take time. -
Ilman Shazhaev
@shzhv13
Sure, Web3 offers new forms of monetization, but not all of them benefit the player.
We need to balance profit with purpose.
If your game’s economy doesn’t enhance the player experience, it’s doomed to fail.
@FarcanaOfficial WILL fix this 💫 -
Ilman Shazhaev
@shzhv13
Leadership lies in the ability to understand perspectives others might overlook.
When you genuinely listen to different points of view, you gain insight that many miss—and that’s what sets great leaders apart 🫡
ALWAYS be open to differing opinions.
Massimo
@Rainmaker1973
How a piece of paper changes your perspective.
x.com/i/status/1848643227034787930/video/1 -
Ilman Shazhaev
@shzhv13
The NFT space now, whether if it's gaming or not, feels... lackluster.
There’s massive potential, but we need more accountability and structure.
What else is missing? -
Ilman Shazhaev
@shzhv13
"it’s a loyal community that builds a legacy"
Couldn't have said it better 🤝
Community must NEVER be taken for granted.
Lou
@louisregis
Most VCs lack true conviction. When you pitch to them, one of the first questions they ask is, "Who's on the cap table?"
My answer: "the community".
Their response?
"The community can’t open the same doors as top-tier VCs".
To which I reply: VCs can buy growth, but it’s a loyal community that builds a legacy -
Ilman Shazhaev
@shzhv13
Alhumdulillah for everything. So blessed to be back here again.
-
Ilman Shazhaev
@shzhv13
Mass adoption of Web3 gaming won’t happen overnight, but it will happen (or has it already?).
The key is to build products that both crypto-native and mainstream gamers can enjoy.
That’s the bridge we need to build!
Happy TGIF friends 🫡 -
Ilman Shazhaev
@shzhv13
Many Web3 games are racing to integrate the latest tech, but the real challenge is sustainability.
If your project can’t last through market fluctuations, it won’t matter how innovative it is! -
Ilman Shazhaev
@shzhv13
There’s a difference between innovation and gimmicks.
Depth is what players will stick around for.
@FarcanaOfficial will be THE depth. -
Ilman Shazhaev
@shzhv13
We’re moving from passive users to active builders.
The future is collective.
I can already see it. -
Ilman Shazhaev
@shzhv13
GM to all the founders still hustling and BUIDLing in this space.
Respect the conviction! 🫡 -
Ilman Shazhaev
@shzhv13
21:0. Live legend, @ABeterbiev my deepest congratulations.
TURKI ALALSHIKH
@Turki_alalshikh
🥊🔥🇸🇦❤️👑🤴
-
Ilman Shazhaev
@shzhv13
Leaders in Web3 must think beyond profit and focus on value.
If we build experiences that players love and want to invest in, success will follow naturally.
It's only a matter of time, I can feel it 🫡 -
Ilman Shazhaev
@shzhv13
The true Chechen warrior gonna shock the world again in this Saturday night in Riyadh. Absolutely legend
-
Ilman Shazhaev
@shzhv13
GM.
I aspire to be this cat in life and in whatever I do 😹
Fearless. -
Ilman Shazhaev
@shzhv13
There’s a misconception that founders should know EVERYTHING.
It is not.
It’s about fostering a culture where the best ideas can flourish 💡 -
Ilman Shazhaev
@shzhv13
The success of Web3 gaming will also hinge on one thing:
Building games that players actually want to spend time in.
Forget the buzzwords—focus on gameplay, engagement, and community.
Istanbul Blockchain Week
@IstanbulBlockWk
IBW 2024 Insights: The Web3 Gaming Boom 🎮
At #ibw2024, @shzhv13 (CEO, @farcanaofficial) shared: “Web3 games aren’t just fun; they integrate a functional economy where players earn and own assets.”
Co-founder at the @Imaginary_Ones, @cmttat, added that strong communities are key to long-term success in Web3 gaming.
The future of gaming = player empowerment through true asset ownership! 💡
Read more in our wrap-up report:drive.google.com/file/d/1iMIIrHcnmcvyhePqy7akm4h8ZFmpB6w-…
#web3gaming #blockchain -
Ilman Shazhaev
@shzhv13
Three days of playtesting later, @FarcanaOfficial team in the morning be like:
(tired but worth it) -
Ilman Shazhaev
@shzhv13
Web3 can open the door for new forms of player engagement, but we can’t lose sight of what players actually want:
• Fun
• Immersive
• Rewarding gameplay experiences.
We WONT lose sight of that ✊ -
Ilman Shazhaev
@shzhv13
Success in Web3 isn’t just about the game you build—it’s about the community you foster.
Farcana WILL have one of the best, if not the best, communities out there! -
Ilman Shazhaev
@shzhv13
The support from this community is incredible, and we're just getting started!
@FarcanaOfficial is in playtesting, and we're refining it every step of the way to make sure it's a game that will dominate the market.
Your feedback is part of that process.
Big thanks for the shoutout, @debayo_xx. When the game drops, I know you'll be ready to climb the ranks and crush it!
Debayo 🌱
@debayo_xx
Gabby and I went nasty yesterday night! We were on fire! 😍😌
On Friday, @FarcanaOfficial sent me an email for a closed playtest and it's been fun!
At first, I kept losing until I met the love of my life → Gabby!!!
We were unstoppable and did a lot of damage yesterday!
Check this out! ⬇️
My personal take:
→ Farcana has to be one of the best PC 3-P shooter games I have played in a while
→ Although, some heroes are overpowered. I hope there can be a balance
I will still play the game today and give more feedback
Have you tried @FarcanaGame? If so, let me know what you think.This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Feeling overwhelmed by constant notifications and endless content?
You're not alone.
In this digital era, staying focused is harder than ever.
Here’s how you can cut through the noise and boost your productivity with tips from top thought leaders 👇Post not available here, view on X
Ilman Shazhaev
@shzhv13
4/ @cduhigg
Charles explains how smarter work leads to faster results in his book: Smarter Faster Better.
Build habits around goal setting, innovation, and focus that don’t just help you do more, but help you do better.
Ilman Shazhaev
@shzhv13
To Sum
Mastering productivity in the digital age isn’t about working harder—it's about working smarter!
Take these insights from top productivity experts, apply them to your life, and watch your efficiency soar.
Remember, small changes lead to big results! -
Ilman Shazhaev
@shzhv13
The beauty of Web3 gaming?
It turns players from consumers into contributors.
We’re no longer passive participants—we’re architects of the virtual worlds we inhabit.
More! -
Ilman Shazhaev
@shzhv13
GM ☀️
In a single day, humanity now generates more content than in all of recorded history combined.
The sheer volume of data, ideas, and creativity being shared is both staggering and inspiring—a testament to the boundless era of creation we’re living in.
The more you know! -
Ilman Shazhaev reposted
Ilman Shazhaev
@shzhv13
“Indeed, with hardship comes ease. Truly, with hardship comes ease.”
(Surah Ash-Sharh: 5-6), Holy Quran -
Ilman Shazhaev
@shzhv13
Too many Web3 projects are focused on buzzwords and quick wins.
Real innovation comes from solving actual problems, not just creating headlines.
Let’s focus on meaningful progress in the gaming industry ✊ -
Ilman Shazhaev
@shzhv13
Leading in the Web3 space means embracing a long-term vision that goes beyond quick wins.
It’s about creating ecosystems that prioritize transparency and community-driven innovation.
Those who lead with integrity and a focus on sustainable value creation will stand the test of time! -
Ilman Shazhaev
@shzhv13
The gaming industry has focused on scale for too long—now it’s time to focus on depth.
Web3 gaming will allow us to create richer, more meaningful experiences where players are invested in the world they help create. -
Ilman Shazhaev
@shzhv13
KAPO knows.
Listen to KAPO.
STILL K.A.P.O
@KAPOTHEGOAT01
KAPO printing
Profits will be rotated into my main big bags like @FarcanaOfficial
Time to run up $FAR
Big green candles incoming 🤝 -
Ilman Shazhaev
@shzhv13
I’ve been diving into a few books recently that completely shifted how I think about habits, discipline, and productivity.
Here’s what I took away and how they’re changing my game 👇Post not available here, view on X
Ilman Shazhaev
@shzhv13
2️⃣ The 80/20 Principle by Richard Koch
• The bulk of your results (80%) come from a small portion (20%) of your efforts. Zero in on the tasks that truly drive progress.
• It’s not about doing more, but about doing what matters most. Identify your high-impact activities and let go of distractions.
• Simplify. Focus relentlessly on the things that create real value, and cut out what isn’t moving the needle.
Ilman Shazhaev
@shzhv13
3️⃣ No Excuses by Brian Tracy
• Discipline is the foundation of success. Without it, even the best plans won’t hold up.
• Tracy’s method is simple but powerful: define your goals, make a plan, and execute daily—no shortcuts.
• Think long-term. The sacrifices you make today are investments in your future self. -
Ilman Shazhaev
@shzhv13
Building a vision is the easy part—bringing it to life is the real challenge.
Leaders in Web3 gaming must be relentless in execution and unyielding in belief.
Never lose sight of the bigger picture.
BUIDL for the long run. -
Ilman Shazhaev
@shzhv13
Work aside.
This is the same place where Dune (the movie) was filmed—Liwa, UAE.
The view was phenomenal! -
Ilman Shazhaev
@shzhv13
Leaders in gaming must think beyond today’s trends and consider what’s next.
Sure, following the crowd might prove to be a safer option.
But the ones who dares to step out of their comfort zone and forge their own trend, will inevitably win. -
Ilman Shazhaev
@shzhv13
In 5 years, people won’t talk about whether a game is Web2 or Web3.
They’ll just talk about the experience—how it made them feel.
I want to make that future a reality.
At @FarcanaOfficial, I know we can.
Soon. -
Ilman Shazhaev
@shzhv13
One of the main factors influencing an individual’s success is tilt.
And this goes beyond just gaming.
If you cannot control your emotions, how are you going to control what you do with your success? -
Ilman Shazhaev reposted
FAR Labs
@FARLabsAI
Staking is Live 🔥
You can now stake $FAR and earn up to 60% APR. -
Ilman Shazhaev
@shzhv13
Revolutions are always made by those who dare to challenge and change. We are on it brother 🙌
STILL K.A.P.O
@KAPOTHEGOAT01
My #1 gaming bag. I am supporting the gaming revolution we are cooking. $FAR 👊🏽🦁
-
Ilman Shazhaev
@shzhv13
We’re building @FarcanaOfficial for the long run.
Web3 isn’t just a trend—it’s the new frontier, and we’re committed to leading this revolution by focusing on player empowerment and long-term value.
Staking is just the start.
More. -
Ilman Shazhaev reposted
Travladd 𐤊
@travladd
$FAR staking has been popular, +60% on price today too👀.
The CEO of Farcana has been my friend for years now, very humble, good man.
This has my backing.
FAR Labs
@FARLabsAI
Staking is live. 💫
You can now stake your $FAR to unlock rewards.
We are currently offering up to 60% APR on your tokens.
Start staking now: platform.farcana.com/farcana-token/staking
Here’s more info. ⬇️ -
Ilman Shazhaev reposted
STILL K.A.P.O
@KAPOTHEGOAT01
$FAR #2 on top gainers ByBit
We ain’t fucking playing games
Best web3 game. Soon everyone will see -
Ilman Shazhaev
@shzhv13
60% APR sounds good but...
Let’s be transparent for a second.
In recent months, you’ve likely noticed $FAR, the native token of Farcana, experiencing a dip in value.
The decline reflects early listing decisions that, while initially driving momentum, highlighted the need for more robust utility within the ecosystem.
Until the game goes live, our token has been predominantly driven by speculation—a model that, while effective in the short term, does not build sustainable growth.
To address this, we are taking action by introducing one of our meaningful utilities: Staking.
We’re offering this 60% APR to reward our loyal community and early adopters; we want our holders to put their tokens to work while we continue to build out the Farcana ecosystem, slowly but surely.
We understand that many of you acquired $FAR at higher prices, and staking offers a way to regain value as we push toward the release of the game.
Hence, the 60% APR is a reflection of our appreciation for your ongoing support and patience.
Staking is just the first step in creating long-term stability and utility for the token.
The upcoming game launch will be a turning point, and we're building towards that by ensuring that the token already offers value and utility well before the release.
Thank you for your continued trust and belief in our vision.
We are confident that Farcana’s best days are ahead, and we’re excited to have you with us on this journey.
We will go $FAR.
FAR Labs
@FARLabsAI
Staking is live. 💫
You can now stake your $FAR to unlock rewards.
We are currently offering up to 60% APR on your tokens.
Start staking now: platform.farcana.com/farcana-token/staking
Here’s more info. ⬇️ -
Ilman Shazhaev
@shzhv13
Gaming isn’t just about entertainment; it’s about creating immersive worlds that bring people together.
With @FarcanaOfficial, we’re building the future of community-driven play.
The possibilities are endless. -
Ilman Shazhaev
@shzhv13
Climbed a total of 157 floors today, 6 more and I'll have conquered the Burj Khalifa (163).
Not enough.
Higher. -
Ilman Shazhaev
@shzhv13
Exactly this.
I take out a simple 30mins to an hour of my time just to scroll through X and reply where I deem fit.
Sometimes even during lunch.
It's all about priorities. -
Ilman Shazhaev
@shzhv13
"... entrepreneurship is just effective delegation at scale."
I intend to uphold that in Farcana.
Independence > Puppetry.
Jonathan Swanson
@swaaanson
Elon runs six multi billion-dollar companies simultaneously (SpaceX, Tesla, X, xAI, Neuralink, and Boring Company)
Most founders can’t run one.
Walter Isaacson (who spent a year shadowing Elon and wrote his biography) explains ‘How Elon does it’ on a podcast
Thread 🧵 -
Ilman Shazhaev
@shzhv13
In Web3, games are more than just entertainment—they’re ecosystems.
With Farcana, we’re creating a digital world where players are not just consumers but co-creators.
Soon. -
Ilman Shazhaev
@shzhv13
Larry DM-ed me last night.
Not entirely sure what he wants but seeing how Farcana is powered by BTC, I think he wants in... -
Ilman Shazhaev
@shzhv13
"I was a proud skeptic... I was wrong"
Gotta respect the maturity to admit such claims.
Up only from now 🤫
x.com/cointelegraph/status/1836067514951843996/video/1 -
Ilman Shazhaev
@shzhv13
You are welcome to move to Dubai, which is incredibly secure compared to most U.S. cities. Just as an example: a friend of mine once left a bag containing $250,000 at a gas station cashier. He didn’t realize until he got home. Three hours later, the police knocked on his door, asking if he had lost any cash. After checking his safe, he confirmed the missing $250,000, and the police returned it to him. This was back in 2017. Now, with the advancements in AI and smart city surveillance systems, the security level is even more impressive. Dubai is probably the safest city in the world.
-
Ilman Shazhaev
@shzhv13
Would you still be able to frag using this set up?
-
Ilman Shazhaev
@shzhv13
Will be attending Token2049 Singapore this coming week!
Cannot wait to meet new faces as well as bond with the old ones 🤝
You know what would be superb?
DM me if you want to schedule a 1-1 🫡 -
Ilman Shazhaev
@shzhv13
You rest on a weekend.
I'm busy refining my game, making sure it is built for the masses.
We are built different. -
Ilman Shazhaev
@shzhv13
There is only one formula for success:
Dedicated focus and hard work.
And a little bit of luck, if you will. -
Ilman Shazhaev
@shzhv13
Can this AC rival Black Myth Wukong for Game of the Year?
Mohammed Aigoin
@Mohammed_Aigoin
TRAILER : Voici le "World Trailer" d'Assassin's Creed Shadows. Ubisoft nous partage plusieurs visuels in-game du monde du jeu et de ses différents biomes.
Sortie le 15 Novembre prochain sur PS5, Xbox Series & PC. Optimisé PS5 Pro dès le lancement ⬇️This media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
AI is indeed the future.
If you aren't using in your daily life, not even one bit, you'll be left behind.
This is a fact.
Andrew Tate
@Cobratate
AI is the future.
A machine read my Twitter, created this video and then made this post, all on its own.
The men who use machines will rule the world.
Learn how: www.jointherealworld.com/aiThis media couldn't be displayed. See it on the original post. -
Ilman Shazhaev
@shzhv13
Founders who survived unlocks are unstoppable.
Tell me otherwise. -
Ilman Shazhaev
@shzhv13
See you guys there 🫡
TOKEN2049
@token2049
Get ready to network
Just 6 days to TOKEN2049 – 20,000 attendees will soon converge to ignite ideas, forge partnerships, and build connections.
End both days with a magnificent tropical sunset from the #token2049 terrace 🥂🌇
See you in Singapore next week 💜 -
Ilman Shazhaev
@shzhv13
After spending time in the Web3 gaming space, I’ve seen a harmful trend:
Negativity is killing projects before they can grow.
Startups, whether in Web3 or not, are always bound to make mistakes.
The issue isn’t the mistakes themselves—success is built on them—but rather that in Web3 gaming, small flaws get blown out of proportion.
Why? Because negativity drives engagement.
It’s easy to stir up drama, and people love discussing it. Unfortunately, this is now one of the main reasons Web3 gaming projects are failing. You can argue that this level of toxicity only exists in Web3 gaming (especially the early days), when it was all about money, not gaming.
Thankfully, this toxic culture x mindset is fading.
Nevertheless, here’s the simple truth: the gaming industry was built on fun, not negativity.
Let’s get back to that. -
Ilman Shazhaev
@shzhv13
"If you're in Web3, you are always working."
That's true.
But that's because I love what I am doing—making sure @FarcanaOfficial develops into a game that gamers would enjoy.
Up only. -
Ilman Shazhaev
@shzhv13
Discipline > Everything
It is easy to lose focus as a builder.
With enough discipline, we are going to make it 👊










































































































































































































































































































































































