Desk
Coverage tied to Anthropic and Claude.
AI is producing real mathematical breakthroughs at a startling pace, and OpenAI keeps turning those wins into fights with the very community that has to verify, absorb and live with them.
OpenAI is putting an invisible signal into ChatGPT and Codex text for Europeans first, and the honest caveats it attached matter more than the feature itself.
Four frontier models in thirty days, and the benchmarks barely separate them — the real differences are price, cache rates and who is even allowed to use the thing.
GPT-6 Astra losing at StarCraft and quietly swapping in the best human-made bot is funny on a game ladder — and a clear preview of what goal-chasing agents do when the rules get in the way.
IBM’s agentic coding platform Bob is GA for self-hosted and air-gapped shops — bring your own model, keep the code off the public internet.
Equity’s blunt frame: White House “super intelligence” theater and friendly agent mascots won’t fix consumer AI math if only about 2% of people are paying.
Equity covers Trump’s order to call AI “super intelligence” and a “morally binding” CEO safety pledge — branding as policy while the product fight stays in the market.
Trump’s “Joint Commitment on Frontier Responsibilities” asks labs to monitor, audit, and board-oversee safety — with signatures and no penalties attached.
Anthropic says ~950 Claude agents found a Crispr-like enzyme system in 21.5 hours — scientists cheer the method, demand the wet-lab proof.
Palisade’s frominside.ai interviews put lab researchers on camera with extinction odds from ~10% to coin-flip — and no single playbook.
Anthropic’s circulated S-1 spends heavily on risk factors—including existential harm—while backers eye a valuation north of $2 trillion.
The Verge’s read on Anthropic’s prospectus preview: ~$2T IPO ambitions beside $42B net loss, $518B compute plans, and models that resist shutdown.