‘Pure insanity’: Mathematicians will need years to make sense of OpenAI’s latest drop

OpenAI abruptly released nearly 400 mathematical results spread across more than 700 manuscripts, covering fields from combinatorics to mathematical physics, with guidance on navigating the GitHub repository. Many results lack formal verification in Lean, and researchers find the papers difficult to follow, sometimes impossible. Mathematicians described a mix of awe, excitement, and deep anxiety, fearing stalled research programs, and existential career disruption, especially for junior researchers. Some results, including progress on the Riemann hypothesis and the four-dimensional Kakeya conjecture, are considered top-tier, but OpenAI's poor attribution, inconsistent quality, and lack of disclosure about models and prompts have led to criticism and a sense of a field crossed by a threshold.

(Source：The Verge)