OpenAI is about to publish more than 100 solutions to previously unsolved math problems, and the preview alone has one mathematician reaching for a mob comparison.
OpenAI is preparing to release over 100 new solutions to open problems in mathematics. Ahead of that rollout, one mathematician told WIRED there is 'a perception of mobster behavior' from leading AI companies. The comment does not point to a single incident; it describes a broader unease about how AI labs engage with the field. OpenAI has not detailed how the solutions were verified or which problems made the list.
A round number like 100 solved problems is built for headlines, not peer review. When a mathematician's gut reaction is 'mobster behavior' rather than curiosity, it signals that trust between research mathematicians and AI labs is thin, and that a splashy release from a hype-fatigued industry invites more suspicion than applause.
This is not OpenAI's first run-in with an unimpressed math community, and the reaction here suggests it will not be the last.