AI's Math Milestone Sparks Credit Dispute and Questions for Human Mathematicians
OpenAI announced that its AI agents solved the Navier-Stokes existence and smoothness problem, a Millennium Prize Problem, but the achievement is overshadowed by accusations that the company used prior work by NYU mathematician Tristan Buckmaster and Anthropic's Levent Alpöge without proper credit. OpenAI denies the claims, though a team member acknowledged being inspired by a rumor about their efforts. The episode highlights how frontier AI resources may become essential for major mathematical breakthroughs, raising concerns about the role of human mathematicians.
Buckmaster published a proof on Mastodon on Monday addressing a simplified version of the fluid equations, while OpenAI subsequently presented a solution for the complete problem using a proprietary internal model that surpasses their recently released Astra system. The company has declined to accept the associated monetary award.
Buckmaster's documentation alleges OpenAI presented two alternatives: either he and Alpöge publish their work first with OpenAI releasing theirs the next day, or he collaborates on a paper that omits Alpöge as a co-author due to his Anthropic affiliation. This highlights the competitive tensions between leading AI labs.
This episode could reshape mathematical research by making proprietary AI resources a gatekeeper for major breakthroughs. Academic mathematicians may find their contributions overshadowed or co-opted by corporate labs, as seen in the disputed credit. The competitive dynamics between AI firms could erode open collaboration, potentially slowing collective progress. Society may benefit from faster solutions, but the distribution of credit and control over foundational knowledge could become concentrated in a few private entities, affecting public trust and academic independence.