Fields Medalists vs. OpenAI: Math Feud Escalates — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- 25 Fields Medal-winning mathematicians signed an open letter arguing AI labs threaten their intellectual work by racing to announce proofs without proper writeups, raising 'severe attribution and plagiarism questions.'
- NYU professor Tristan Buckmaster accused OpenAI of pressuring him not to credit an Anthropic collaborator on an important math problem and wondered whether his work with Codex was fed into OpenAI's own groundbreaking proof.
- OpenAI withdrew its sponsorship of a math event at CalTech on Thursday after facing criticism from researchers at the university.
- The signatories noted OpenAI's proof remains unverified and warned that frontier labs can now spend tens of millions of dollars using LLMs to beat original researchers to a discovery, incentivizing secrecy in a field built on open collaboration.
- The letter follows the Leiden Declaration released by mathematicians in June, which offered recommendations for adapting to LLM-driven changes in mathematical work.
- The signatories warned the conflict extends beyond mathematics: 'Your field of interest is next,' as similar attribution and integrity challenges are emerging across scientific and creative professions.
Why it matters: Mathematicians fear a culture shift from open collaboration to secrecy as AI labs gain the compute to race researchers to proofs, with 25 top-tier signatories formalizing the complaint. OpenAI's withdrawn CalTech sponsorship and unverified proof give the dispute immediate institutional weight — and the letter explicitly frames math as a test case for every research-driven field facing AI-driven incentives to publish first and verify later.
Ask SkimNews
