Math advisory group issues responsible-release guidelines for AI-generated math; Gowers amplifies, critics push back
RexDouglass · x · 2026-09-30
Fields Medalist Timothy Gowers shared a new advisory-group document, "Responsible Release of AI-Generated Mathematics," which Rex Douglass quoted while disputing parts of it.
- Context: frontier AI labs are currently testing advanced math problems on proprietary models inaccessible to the scientific community. The group explicitly does not endorse this practice and asks labs to stop.
- Core norms: mathematics relies on authors understanding, verifying, and taking responsibility for their arguments, plus seminar and conference culture that deepens human understanding — AI outputs can exceed the prompting human's ability to verify them, requiring a new responsible-release paradigm.
- The dispute: Gowers backs recommendations that labs should pay academia and grant free access, but Douglass argues these don't follow as "ethical blackmail" from labs cracking results, calling that framing "pretty trashy."
- Full text and PDF are available at agmai.org.
More from Safety
- Chinese AI Tool Told Researchers How to Make Bioweapons, BBC Reports — shaky2236 · 2026-09-30
- America.gov gets weird about Minecraft — and it's not an AI hallucination — TechCrunch AI · 2026-09-30
- Meta's Muse AI accused of uploading Apple Messages to the cloud even when users opt out — SumitGup · 2026-09-30
- AI Lawsuit Risks Becoming a Court-Ordered Kill Switch While Chinese Rivals Race Ahead — castrotech · 2026-09-30
- Will Chinese open-weight models like GLM get banned? Reddit weighs Anthropic report — writesfw · 2026-09-30
- Apollo Research shifts to embedded evaluators with employee-equivalent access — MariusHobbhahn · 2026-09-30