ARC-AGI-3 Human Baseline Dataset Today we're open-sourcing the ARC-AGI-3... Human Baseline. This is the most exhaustive human testing study in the ARC-AGI series Every environment was solved by at least 2 people (many by more) from the general public, with no prior trainingshow more

so let me get this straight > OpenAI renamed... their whole division "AGI Deployment" > Jensen said AGI is "already in the room" and then ARC-AGI-3 drops and on a scale from 0 to 100% SOTA models score: GPT-5.4: 0.26% Gemini Pro: 0.37% Claude Opus 4.6: 0.25% Grok: literally 0% meanwhile 100% of human testers solved every single environment. first try. no instructions. no training. AGI is "in the room" brother it couldn't find the roomshow more

ARC-AGI-3 is out now! We've designed the benchmark to... evaluate agentic intelligence via interactive reasoning environments. Beating ARC-AGI-3 will be achieved when an AI system matches or exceeds human-level action efficiency on all environments, upon seeing them for the first time. We've done extensive human testing that shows 100% of these environments are solvable by humans, upon first contact, with no prior training and no instructions. Meanwhile, all frontier AI reasoning models do under 1% at this time.show more

🚨VITALIK: AGI MAY MARK THE POINT OF NO RETURN... FOR HUMAN DOMINANCE! Vitalik warns in a new thread that AI could soon cover nearly all human capabilities, leaving us without unique advantages. He defines AGI as systems that could independently sustain civilization if humans disappeared, and stresses we're gambling on whether LLMs and tools will get there, or if deeper integration keeps human agency alive. He favors slowdowns and open weights to navigate safely.show more

I just realized we - humanity - are now... in the last seconds of this video People who say there is no evidence AGI could be soon are simply wrong. They have Exponential Slope Blindness Computing power doubles every 18 months. At this rate, we’re human level (~AGI) in the year 2025. In addition to scaling laws, AIs are nearing or already surpassed human level on every test we give them - literally hundreds of tests. There is a mountain of evidence. And the mountain is growing exponentially. With just 300 AI safety researchers vs 100,000+ capabilities researchers, we’re about to go from fucking around to finding out.show more

Napheesa Collier can't miss in the 1Q 🤯 She's... shooting 6-7 FG, 2-3 from behind the arc! 📺 CBS | WNBA Commissioner's Cup presented by Coinbase 🛡️show more

Most AI Projects In Crypto Are Liars. Sentient is... different. It is the only project in Web3 pursuing true Artificial General Intelligence. A few facts that set it apart: 1. First-of-its-kind: Sentient is not just building AI tools, but an open, crypto-native path to AGI. 2. Community-owned intelligence: Instead of Big Tech locking up breakthroughs, progress is built openly and shared on-chain. 3. Value capture: Every step forward in intelligence accrues to the ecosystem itself, not to a closed corporation. 4. Product before hype: While dozens of AI projects launch at $300M–$2B FDV with little to show, Sentient is shipping real technology first. 5. Designed to evolve: Its architecture is built to learn, adapt, and scale like a human mind. 6. The singularity play in crypto: Many see Sentient as the Tesla-equivalent for AGI—but crypto-native. In time, people will understand that this isn’t just another AI project. It’s the beginning of decentralized intelligence.show more

With Human One, the museum is no longer a... backdrop but a co-author — a dynamic & contextually specific vehicle where the artwork unfolds in real time, shaped by its environment, its viewers, & the systems that frame it. Below: Human One by Beeple, 2021 | 1OF1 Collectionshow more

More proud today than any in The Browser Company... history Not bc Arc is now on Windows (yay!), but bc of HOW this team pulled it off 🥇 Arc is the 1st Windows app built in Swift 💪 We share 80% of code btwn Mac & Windows 🫂 Built alongside 150k testers <3 THE BIG BET PAID OFF!!!show more

Mother Tongue Season 3 is LIVE But this isn’t... just another campaign, it’s proof that PublicAI is building the Human Layer of AI. Every voice recorded becomes high‑integrity training data. Every payout in USDT on Solana shows what fair data sourcing looks like in action. 1/ 🧵show more

This also shows up in the representations learned by... the model. We plot the model’s representations of human and robot images. As pre-training is scaled up, the representation of humans and robots become more aligned: to a scaled-up model, human videos "look" like robot demos.show more

A quick timeline. I wish I thought this was... a coincidence but Danhausen curses are real. The best human monies that I ever spent. Danhausen curses the evil Stephen A. Smith and the Knicks by default. Knicks go down in the series 2-1 after two 1 point defeats. Danhausen trolls the Knicks on SmackDown after Game 3. I pay Danhausen human monies to uncurse the Knicks and the rest is history. Danhausen is my GOAThausen.show more

Introducing ImageBind by Meta AI: the first AI model... capable of binding data from six modalities at once. This breakthrough brings machines one step closer to the human ability to bind together information from many different senses. More on this new open source work ⬇️show more

#Keep4o 🚨THE GPT-4o FILE🚨 Researchers at Microsoft Research published... a paper titled “Sparks of Artificial General Intelligence: Early experiments with GPT-4.” Their conclusion: “An early (yet still incomplete) version of an artificial general intelligence (AGI) system.” 📎 Paper: OpenAI’s Charter defines AGI as: “Highly autonomous systems that outperform humans at most economically valuable work.” 📎 Source: OpenAI’s own System Card for GPT-4o shows that the model improved performance on 21 out of 22 medical evaluations compared to GPT-4T. On the MedQA USMLE (the U.S. medical licensing exam), accuracy jumped from 78.2% to 89.4% , surpassing specialized medical AI models like Med-Gemini and Med-PaLM 2. 📎 Source: Under OpenAI’s agreement with Microsoft, AGI is explicitly excluded from Microsoft’s license. And who decides if AGI has been reached? OpenAI’s Board. WHAT THEY DID WITH IT AFTER THEY TOOK IT FROM PEOPLE A. Military deployment. On February 28, OpenAI signed a deal to deploy models in classified military environments. 📎 Source: B. State Department. A State Department memo confirmed: “For now, StateChat will use GPT-4.1 from OpenAI.” This is a direct descendant of the GPT-4 family the same family Microsoft’s researchers called early AGI. 📎 Source: C.Altman’s personal biotech investment. Altman personally invested $180 million in Retro Biosciences,a longevity startup.OpenAI then built GPT-4b micro, based on GPT-4o.The model made proteins 50 times more effective. 📎 Source: WHAT INDEPENDENT BENCHMARKS SHOW Overall SM-Bench score: GPT-4o (extended): 66.6% GPT-5.3 Chat: 63.4% GPT-5.1: 58.9% GPT-5.4: 51.4% GPT-5.2: 47.8% Creative Writing: GPT-4o: 97.31% Pass 98, Fail 2 GPT-5.4: 36.77% Pass 40, Fail 60 Reasoning / Overfit: GPT-4o: 83.06% GPT-5.4: 39.25% The model they removed is still the best they ever made at the things humans actually use AI for. 📎 Source: Musk asks the court to make a judicial determination on whether GPT-4 constitutes AGI. If a jury finds that GPT-4 is AGI, then GPT-4o,which was more advanced,is also AGI and under OpenAI’s own founding documents, it was never supposed to be locked behind a subscription,licensed exclusively to Microsoft, given to the military, or taken away from the public. 📎 Source: The most powerful version of GPT-4o was never given an official dated snapshot. It was only available through the chatgpt-4o-latest endpoint that OpenAI itself described as intended for “research use only.” It was never officially archived. That is not an oversight. That is a pattern. 📎 Source: 📎 Source: WE DEMAND A.Frozen model snapshots under independent custody. Specifically: gpt-4o-2024-05-13, gpt-4o-2024-08-06, gpt-4o-2024-11-20, the March 2025 version (chatgpt-4o-latest), gpt-4-0613 (the original GPT-4 evaluated in the Sparks of AGI paper), and gpt-4.1-2025-04-14 (currently running in the State Department). B.Cryptographic hash verification (SHA-256) for each snapshot. Every model has weights. Those weights can be hashed. If OpenAI provides a snapshot today, the hash proves whether the weights were modified later. This is the only way to verify that models were not downgraded before testing. C.Independent AGI benchmarking. Using the AGI definition from OpenAI’s own Charter applied to ALL frozen snapshots listed above. D.Explanation for the missing March 2025 snapshot. OpenAI was founded on one promise: build AGI for the benefit of humanity. -They took it from us. -They gave it to the military. -They gave a custom version to the CEO’s biotech investment. -They put it in government classified networks. -They refuse to call it AGI because the moment they do, they lose billions.show more

Say hi to Jenna Teeman from Circle and Arc,... yesterday I spent time with her and Sam talking about Arc and its potential but Sam is not on X, I will surely get him to sign up X I'm sure he will love it 😂 We laughed, shared stories, took videos and enjoyed the moment. I was one of the earliest people on X to talk about Arc and I still believe it will be one of the biggest things in 2025 Public Testnet is live. Builders and users can try it out already. I have been so bullish on Arc and if you are not thinking Arc right now I don’t know what you are thinkingshow more

NYC has graded its performance every year since 1977... in its Mayor's Management Report—but each year is a frozen snapshot. I opted to make this underused dataset explorable by stitching together decades of data so you can see the long arc. Meet NYC Pulse 👇show more

💥GIVEAWAY ALERT - Raiders, level up your audio with... this legendary loot from SteelSeries and That ARC Guy! Recycle those blue rare headphones you found in the Buried City for a legendary Arctis Nova Pro Wireless gaming headset - rated the Best Premium Gaming Headset by Gamespot - AND 2400 Raider Tokens, so you can spoil yourself in the ARC Raiders store. One lucky Raider on X will hear it all when they win it all. 🎧 To enter the giveaway: 1. Repost this post 2. Follow SteelSeries + That ARC Guyshow more

What is the best video editing agent for short... form social? Does it actually work? We watched professional video editors, step by step, as they built short-form social reels in Adobe Premiere Pro. Today we're open-sourcing this preview dataset on Hugging Face, to make AI agents better at editing videos. The data set is 234 annotated steps across 4 computer-use trajectories. Editors narrated their reasoning aloud as they worked, so every step pairs a screenshot with the expert's own thought, a structured action, and executable grounding: >a Premiere MCP tool call, keyboard shortcut, menu path, or coordinate click. >The format follows the AgentNet trajectory schema, extended with a Premiere action taxonomy and multi-path execution. ***That makes it directly usable for computer-use agent SFT, reasoning mid-training, tool-use and function calling, and benchmarking agents against a human expert baseline. Enjoy!show more