On 8 September, OpenAI announced that its AI agents had solved the Navier-Stokes problem, one of the most renowned and difficult challenges in mathematics. This claim has reportedly triggered an "existential crisis" within the mathematical community regarding the capabilities of AI.
Accusations have also emerged that OpenAI misused human work. Critics argue that the company's results, like those from other large language models, depend on processing human mathematicians' contributions. While OpenAI's paper cites sources, many believe insufficient credit was given, particularly to mathematicians thought to be close to a solution.
Mathematician Tristan Buckmaster expressed concerns that his work on Navier-Stokes, conducted using OpenAI's Codex model, may have been seen by the OpenAI team. OpenAI has denied direct access but could not rule out that data from Buckmaster's product use "helped improve our model."
Despite these concerns, many mathematicians acknowledge AI's power and its particular aptitude for mathematics. Discussions are ongoing about how the field might integrate AI, with humans guiding and interpreting its outputs.