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

OpenAI recently released nearly 400 AI-generated mathematical results spread across more than 700 manuscripts covering diverse fields including combinatorics, geometry, number theory, and theoretical computer science. Mathematicians told researchers that understanding this volume of work could take years, and they expressed deep anxiety about the quality and verification of the results, with fewer than half of the manuscripts having been formally verified using computational proof methods. A major concern among researchers is that much of the work contains what they call "slop," a term for low-quality, frequently erroneous AI-generated material that is confusing, hard to read, and sometimes lacks proper attribution to other researchers. The sheer scale and pace of AI-generated mathematical output has left the field uncertain about how to assess the work and what role human mathematicians will play going forward.