Mathematicians describe research shifts after OpenAI's math release
some mathematicians say OpenAI's math results have disrupted research plans; details needed for independent evaluation remain undisclosed.
OpenAI's release of mathematical results is drawing strong reactions from researchers. In interviews with more than three dozen mathematicians, NYU's Tristan Buckmaster said he had heard of three people whose entire research programs had effectively been ‘obliterated.’ Others described grant proposals losing their purpose. Some results, the account says, might have put human mathematicians in contention for a Fields Medal.
OpenAI released nearly 400 results across more than 700 manuscripts, but only 42% of the headline results have been formalized in Lean, according to the report. The model, prompts and full set of attempted problems have not been disclosed, limiting independent evaluation of the broader claims.
One mathematician suggested that even if AI disappeared today, reviewing the results could occupy researchers for a decade. That is a researcher's assessment of the workload, not a firm forecast.
