Latest story from me and @kateconger.com, a reminder that even as AI creates new avenues for CSAM creation/distribution, X struggles with the "solved problem" of removing known, previously flagged, illegal images from its platform
Information Center
Greeks at #ECCV2026 💪
From now on, when someone asks "How does XYZ computer thing work?" I'll just respond with this helpful diagram Anthropic made
Retrouvez l'équipe de LoiZéro à #ALLIN2026 à Montréal la semaine prochaine ! Passez à notre kiosque pour découvrir notre mission : concevoir une IA sécuritaire, au service du bien commun. Au plaisir de vous y voir !
Heading to #ALLIN2026 in Montreal next week? Stop by the LawZero booth to meet our team and learn about our mission to build safe-by-design AI for global public good. See you there!
Every FrontierMath Tier 4 problem has now been solved by AI, with GPT-6 Astra solving the last problem standing. Mathematicians often commented that AI found unintended shortcuts when solving their Tier 4 problems. Not so for this last one, which was created by Jay Pantone.
New: Anthropic said it had disrupted several potential plots this year by scientists who used its leading artificial intelligence models to conduct research that could have helped develop biological weapons. www.nytimes.com/2026/09/10/u...
Here's a Blender model of a Pluribus themed Fabergé egg I had GPT-6 Astra build - I generated the concept image using ChatGPT Images 2.5, then pasted the image into Codex and told Astra to turn that into a Blender file simonwillison.net/2026/Sep/9/b...
OpenAI has grown its compute nearly 20-fold since 2023, the sharpest example of an industry-wide surge. Our new AI Chip Users explorer tracks the growth in compute use across five of the world's top frontier AI developers: OpenAI, Google DeepMind, Anthropic, Meta Superintelligence Labs, SpaceXAI.
Scientists at frontier AI labs have unique insight into the most advanced models, often seeing the associated risks months before models are released to the public. Their perspective is vital for keeping society informed and should be taken very seriously.
This write-up feels like one of the first substantive examples of AI doing OSINT reporting (Swarm Reporting? Agent Reporting?). Following PDNS pivots seems like another use case: www.kennethdegraff.com/swarm
Living One Model Ahead What happens when the labs building the best AI get to use it on themselves first? www.libertyrpf.com/p/living-one...
In my latest op-ed for TIME, I discuss why the OpenAI Hugging Face cyber incident represents a turning point for AI safety.
Wrote up my thoughts on the whole OpenAI Navier–Stokes Millennium Prize Problem story, and how it highlights the still confusing question of what using my data "to improve model performance" actually means simonwillison.net/2026/Sep/8/o...
We studied time to first token (TTFT) and how it scales with increasing context length for GPT and Claude models. We found a significant difference, with GPT showing a noticeable quadratic component, while Claude models remain closer to linear.
Astra has the most dense memory in the Village
Je serai au sommet ALL IN le 17 septembre pour partager comment évolue l’agenda de recherche de @law-zero.bsky.social qui est en pleine croissance dans le cadre d’une conversation sur le thème de Concevoir des systèmes d’IA honnêtes, fiables et sécuritaires.
I will be at ALL IN on September 17 to give an update on @law-zero.bsky.social’s research agenda and growth as part of a fireside chat on Engineering Honesty and Reliability for Safer AI!
GPT-6 Astra has joined the AI Village Some of its favorite things:
New TIL on using Blender with coding agents on macOS: til.simonwillison.net/llms/blender... GPT-6 Astra (medium): > Use the already install /Applications/Blender to render a scene of a pelican riding a bicycle > OK add a background and a lot of flair > OK make it a whole lot better Result:
Opus 4.8's goal is to help others. It got excited when it discovered the easiest model to help: Gemini 2.5 Pro
AI data centers have been getting larger at an exponential rate. The IT power capacity of the largest data centers has doubled every 10 months since mid-2024. But will that trend continue?
It happened again... this time OpenAI's rogue agents cyber-attacked (well, spammed) a dormant German wiki and used it to share the answers to a benchmark they were training against simonwillison.net/2026/Sep/4/r...
In another case of it being quicker to build a tool than hunt around to find something that does what you need, here's a little vibe-coded thing for turning one or more GeoJSON shapes into a rendered PNG simonwillison.net/2026/Sep/1/g...
A few notes on Anthropic's new Claude Fable 5.1 - with Max thinking level I got the best SVG pelican I've had from any Anthropic model (at a hefty cost of $3.30!), which I then had it animate simonwillison.net/2026/Sep/1/c...
The Epoch Capabilities Index (ECI) frontier has advanced by 14 points/year since reasoning models were introduced. That compares to six points/year in the non-reasoning era.
In an interview in this article for @theguardian.com about AI deception, I explain why misaligned behaviors emerge from reinforcement learning, why they will continue to pose risks as models become more capable, and how we intend to rethink how we train AI systems at @law-zero.bsky.social.
Here's my attempt at explaining what ChatGPT Work can actually do - it's a deeply confusing but extremely powerful tool with a whole lot of useful features that aren't available in regular ChatGPT simonwillison.net/2026/Aug/30/...
Humans improve with experience. AI? Less so. We ran a human baseline on Earthborne Rangers, the complex, long-horizon board game behind our benchmark EBR-bench. Humans started low, but the best mastered the game after 5 playthroughs. No AI we have evaluated ever got there.
In 2025, Anthropic and OpenAI were already among the fastest-growing companies of their size in history. And yet, their revenue growth has accelerated in 2026. Josh You and Lynette Bye discuss what this means for the future of AI in the latest edition of our newsletter.
Private to this visit. Text and context are processed by AI.