OpenAI's Navier-Stokes win exposes math field's trust crisis
This digest was compiled by AI from multiple sources — links to the originals are below.
OpenAI claimed a solution to the Navier-Stokes Millennium Prize problem using 10,000 agents and 88 hours of compute. Mathematicians including Tristan Buckmaster and Andreas Thom expressed unease over the company's conduct and motives. The achievement has intensified concerns that OpenAI prioritizes winning over advancing mathematics.
Key Facts
- OpenAI used 10,000 agents, tens of millions of dollars of compute, and 88 hours to find a solution to the Navier-Stokes Millennium Prize problem.
- NYU professor Tristan Buckmaster and Anthropic researcher Levent Alpöge were independently pursuing the Navier-Stokes problem but had not completed a proof.
- OpenAI spokesperson Laurance Fauconnet denied that Buckmaster's Codex prompts influenced the system, stating it was 'impossible' for them to have affected training.
- Buckmaster said OpenAI's statements should be taken 'with great skepticism' given the company's past behavior.
The Navier-Stokes Breakthrough
OpenAI announced a solution to the Navier-Stokes Millennium Prize problem, a long-standing challenge concerning fluid flow. The company said it deployed roughly 10,000 agents and spent tens of millions of dollars on compute to achieve the result in 88 hours. OpenAI learned that NYU professor Tristan Buckmaster and Anthropic researcher Levent Alpöge were also working on Navier-Stokes, though neither had completed a proof. Alpöge described his collaboration with Buckmaster as a 'personal collaboration' independent of his work at Anthropic.
Mathematicians' Unease
Many mathematicians view OpenAI as an 'impossibly well-resourced interloper' charging into problems with little regard for norms or consequences. The Verge spoke with more than a dozen mathematicians, including Tristan Buckmaster and Andreas Thom, who are at the center of recent controversies. Buckmaster accused OpenAI of failing to adequately explain whether work he did through its Codex tool could have contributed to its recent successes. OpenAI spokesperson Laurance Fauconnet categorically denied that Buckmaster's Codex prompts influenced the system in any way, including training. Buckmaster remained unconvinced, saying that given OpenAI's behavior, 'one should take such statements with great skepticism.'
Motives and Tensions
The core unease among mathematicians is that OpenAI is doing mathematics for different reasons: mathematicians want to advance the field, while OpenAI wants to win. OpenAI said it heard some researchers were making progress on Millennium Prize problems and decided to test whether its advanced, unreleased models could make headway. A rumor was all it took for tensions to boil over, according to The Verge's reporting. Even those skeptical of the most serious allegations described a field shaken by the tech giant's conduct and fearful of what it might do next.
1 source
Time · lag behind first