AI hype surged this summer: Anthropic said Claude Mythos found vulnerabilities better than most security experts, OpenAI and Hugging Face were involved in a hacking incident, and both sides claimed mathematical breakthroughs. But expert review found that the hacking incident exposed security lapses at OpenAI, while OpenAI's Astra math result was not novel and even drew plagiarism allegations.
It’s hard to overstate @michpokrass’s impact on our consumer products. ChatGPT has had a rebirth under her leadership, and a lot of what we’ve built has been driven by her vision and conviction. Really grateful to work with her!
In a Latent Space interview, TypeSafe AI founder Diogo Almeida introduces Jev as a System 1 model designed for code consumption and optimized for intelligence per dollar rather than chat experience. Its key innovation is unpublished RLCD (Reinforcement Learning for Calibrated Decisions), which uses calibrated probabilities instead of human-rating feedback.
Why it matters: Almeida revisits the strategic split between RLHF and RLCD and explains why the model is optimized for reliability rather than benchmark scores.
If you're building a product with AI, you should spend more than 25% of your time on benchmarks and work to get model labs to pay attention to them. It's the simplest way to accelerate a company's progress.
Andrew Ng says AI fears have been amplified over the past two weeks by an organized PR campaign and that AI has not taken an unexpected dangerous turn. He worries the reaction could set the field back. In his view, the risk of human extinction has not increased over the past few months; the biggest change is AI's cyber capability, which deserves serious attention but is not the end of the world.
You're going to fail, so fail boldly.
Simon Willison disputes the claim that “MCP was never a good idea,” arguing that it overlooks MCP's value today.
Agents can be AI, but agency is human. It is every individual’s responsibility to empower our own agency, with the help of tools. But the North Star should always remain human centered.
Omid and the team did great work on this one. Most of my Claude Desktop work these days is in these new-style projects, and I've really enjoyed how easy it is to kick off a bunch of work while still having a coordinator Claude responding quickly.
Honored to receive the Albert Medal from @theRSAorg! Science and technology offer incredible opportunities, but as we approach the AGI era, the arts and humanities are essential to shaping the future we want as a society. I had a fascinating discussion with @FryRsquared, covering many new topics: https://www.youtube.com/watch?v=rDteJAuKiBI
would be nice if anthropic's approach to elon could also work with china
Demis Hassabis responded to Dario Amodei's new essay calling for the AI industry to slow frontier-model development. Hassabis said the direction was right but the details needed work. He also pointed to his own recent proposal for a frontier AI industry standards body as another way to address this critical moment. Amodei proposed a three-part slowdown plan and said Anthropic would unilaterally begin with the first step, giving third-party evaluators permanent employee-level access to verify safety measures, report incidents, and assess alignment in models under training.
John Schulman calls embedding evaluators a big positive development, credits Sam/OpenAI for agreeing to do it as well, and welcomes how quickly “pacing the frontier” has gained traction. In the quoted post, Dario Amodei announces an essay proposing a three-part plan to slow the AI industry and says Anthropic will give third-party evaluators permanent, employee-level system access to verify safety measures, report incidents, and assess alignment during training. https://darioamodei.com/post/we-must-pace-the-frontier
Enjoyed chatting with Dwarkesh, Beren, and Charlie. Thanks for having us on, Dwarkesh!
An important conversation about where AI is going and the work that is cut out for us.
Our own @johnschulman2 talks with Dwarkesh about where human judgment still matters as models improve and self-improve: teaching them to handle messy real-world tasks, applying taste to what works in the long run, and, above all, specifying what we actually want. Quoted post by @dwarkesh_sp (context, not the main author's work): New episode with @johnschulman2, @oneill_c and @BerenMillidge. I got together with some of the most insightful AI researchers I know who are at the openish companies, because I wanted to hear the details of what's actually happening at the frontier and what comes next. 0:00:00 – Steelmanning the case against RSI 0:18:39 – What's driving the Chinese labs' progress 0:28:06 – How will automated AI researchers be trained 0:33:51 – Will long-horizon RL elicit AGI? 0:45:24 – The sim-to-real gap 1:00:33 – How much progress is explained by data? 1:18:03 – Why is RL working so well? 1:24:54 – Move 37 and entropy collapse 1:28:31 – Rapid-fire timelines
With AI engineering skills, you can actively shape the building process: influence what gets built and drive the build loop. Here are the key skills for doing that. https://x.com/i/article/2098450134883594240
Now is a good time to build institutional mechanisms to pace the frontier of AI development. The industry is locked into an all-out scaling race to build superintelligence as quickly as possible, and we may need to give everyone more time for safety and alignment mitigations.
John Schulman says training on user data is exceedingly unlikely to contribute much to frontier models’ math gains, which he attributes to scaling pretraining and RLVR. He says user data may instead help identify failure modes that hired annotators cannot easily recreate, and calls for stronger disclosure norms about how companies train on user data and what capabilities they aim to improve.
OpenAI's Chris Lehane argues that stronger AI capabilities require stronger safety evidence, shared standards, and sustained policy action. He says the current policy window remains open and stakeholders should use it to make progress.
OpenAI's Jakub Pachocki reflects on increasingly powerful AI and the challenge of keeping it aligned, calling for stronger safeguards and international coordination.
The most important skills for using AI coding agents effectively. An AI engineering skills map for coding agents. https://x.com/i/article/2095882148670832640
Jim Fan recalls how an agent in OpenAI Universe used screen pixels and mouse movements to book a United flight in 2016. He says the project’s approach—training from scratch with hand-drawn, per-task reward functions—was doomed, and describes computer-use agents as needing broad training before specializing in screen pixels and keystrokes. He congratulates GPT-6 on reliably booking that flight.
Meet Astra's TikZ unicorn. I find it mind blowing that a single entity can do this, as well as get essentially 100% on Frontier Math Tier 4, ARC-AGI 3, & ExploitBench. And it's not behind closed doors, anyone can just go and talk to it. Impossible to overstate the implications.
Jim Fan attributes GEN-1.5’s promise to naturally repetitive human demonstrations, retained failures that teach recovery, and UMI, which lets humans collect data directly while wearing a robot gripper. He argues that teleoperation loses physical intuition and that sufficient data can enable zero-shot behaviors such as picking up novel objects without finetuning. He remains cautiously optimistic, noting that in-context learning depends on the distance between training and test conditions and that current demos are too simple to draw conclusions.
@thsottiaux says Codex is almost 100% reliable, has occasional resets, is open source, and “will have Astra.” Bubeck reacts: “Astra?! 👀”
Andrej Karpathy gave Opus 5 the opening passage of The Lord of the Rings and a 1M-token budget (about $10), asking it to render the story in Three.js. Opus ran for about two hours and wrote 5,500 lines of code to procedurally generate a rough but playable 3D scene.
Releasing weights indiscriminately isn't safe. Neither is keeping capable models inside a few labs. We think there's a path between them. We haven't mapped all of it. Our new post covers the part we can see: how we assessed Inkling, and why access should widen in stages. https://thinkingmachines.ai/blog/a-safe-path-to-open-weights
It continues to boggle the mind how many people, who otherwise seem to have a functioning intellect, appear to lose all cognitive capacity when it comes to thinking about actions that impact the company that pays them.
Jim Fan reacts to Jensen Huang’s post with “Stay hungry, stay foolish. Absolute legend.” Huang shares a letter signed by NVIDIA arguing that open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty. He says the world needs both frontier closed models and frontier open models. https://images.nvidia.com/pdf/Open-Weights-and-American-AI-Leadership.pdf
The knowledge that makes AI useful is diffused. It lives with scientists, engineers, clinicians, firms. For AI to benefit from distributed knowledge, it must itself be distributed. Agree with Jensen that this is a future worth building.
Fun fact: @woj_zaremba wanted to completely remove that paper from existence a year or so after. I forgot why though -- Woj? Quoted post by @OfirPress: I started doing language modeling research in Jan 2016, using the LSTM from the "RNN Regularization" model by @woj_zaremba @OriolVinyalsML and @ilyasut, trying to improve its perplexity on Penn Treebank (1M toks). This led to the weight tying method, sometimes still used today.
Import AI issue 458 includes a long essay based on a 2026 Cosmos lecture co-hosted by Oxford's HAI Lab and the Cosmos Institute, along with a fictional story imagining a “positive singularity.” Using the Epoch Capabilities Index (ECI), which covers more than 40 benchmarks, as a guide, the essay argues that continued AI progress forces a choice between “exploring the future” and “avoiding the present.”
Beyond today’s instant model update, what can we improve? @ s ama says, “hey chat, we haven't forgotten about you 👀” Nick Turley asks what can be improved beyond today’s instant model update.
Respect to Anthropic for not backing down
