NYU mathematics professor Tristan Buckmaster and Anthropic mathematician Levent Alpöge announced proofs related to the Navier–Stokes existence and smoothness problem, one of the seven Millennium Prize problems, using AI models Codex and Claude. However, their announcement was accompanied by a controversy involving OpenAI, which published a full proof of the same problem shortly after. According to Buckmaster, OpenAI’s effort was built on their work before it became public, leading to academic rivalries and conflicting claims.

Buckmaster alleges that OpenAI was informed of his and Alpöge’s progress on the problem and subsequently deployed a significant computational effort, consuming 300 billion output tokens (valued at $22.5 million) to arrive at a proof. OpenAI confirms the timeline, stating that their effort began on September 1, inspired by rumors of progress on the Millennium Prize problem. The controversy centers on the suspicion that OpenAI used the specific, less common approach taken by Buckmaster and Alpöge, which they believe suggests that OpenAI’s team had access to their work. OpenAI denies using any specific user data but acknowledges the possibility that de-identified data from Codex interactions may have indirectly informed their models. The dispute highlights ongoing debates about AI’s role in mathematical research and the incentives of companies like OpenAI.

This is a summary. Read the original at techcrunch.com.