+- MKLab (https://mklab.gr)
+-- Forum: [INDEX] (https://mklab.gr/forumdisplay.php?fid=1)
+--- Forum: ARTFICIAL INTELLIGENCE (AI) (https://mklab.gr/forumdisplay.php?fid=5)
+---- Forum: ARTICLES (https://mklab.gr/forumdisplay.php?fid=33)
+---- Thread: AI solved one of math's hardest problems (/showthread.php?tid=2075)
AI solved one of math's hardest problems - mklabgr - 09-23-2026
The article reports that OpenAI’s AI system has apparently produced a solution to the Navier–Stokes problem, one of the seven Millennium Prize Problems concerning whether solutions to the equations governing fluid motion can develop singularities. OpenAI reportedly deployed around 10,000 AI agents for 88 hours, generating about 130 billion tokens, with the computation estimated at roughly $6–10 million.
The resulting proof is 166 pages long and has also been formalized in Lean; mathematicians who examined the formalization say it successfully compiles, giving the mathematical community substantial confidence that the argument is technically correct.
However, the central problem is that a correct proof is not necessarily a useful mathematical explanation. Several mathematicians, including James Maynard and Javier Gómez-Serrano, say the AI-generated manuscript is extremely difficult to understand and does not clearly reveal the important ideas or how they could be used elsewhere.
The episode also became controversial because human researchers Tristan Buckmaster and Levent Alpöge were independently approaching a solution with AI assistance, while OpenAI's rapid effort created tensions over collaboration and attribution. OpenAI has since announced an independent mathematics advisory group. The broader lesson of the article is that AI may now be capable of producing genuinely new mathematical results, but mathematics requires human-understandable insight, not merely a formally verified answer.
ARTICLE