OpenAI drops 700+ math research papers, stirring worries in academia

OpenAI has released more than 700 math research papers on GitHub. The company says they came from an in-house frontier AI model, and many proofs come with computer-verifiable versions.

The release has split the math community. NYU professor Tristan Buckmaster said young mathematicians' research plans were hit hard, and questioned whether unpublished materials researchers fed into AI tools helped produce the results. Northwestern professor Bryna Kra warned that dropping so many results at once could hurt math's long tradition of collaboration.

Rutgers professor Alex Kontorovich is more upbeat. He thinks AI is pushing academia to rethink hiring and awards, and that math training builds deep thinking skills that will matter even more. OpenAI says it will improve how the papers are presented and fund seminars to help researchers understand the AI-generated results.