OxfordOpenAI released more than 350 mathematical findings in one day.
OpenAI claims a proof of the quasi-Riemann hypothesis, a conjecture related to a Millennium Prize Problem.
OpenAI withdrew 3 papers and fixed several others by Thursday in its public repository.
“Peer review is gone in this process,” said Bryna Kra, a Northwestern mathematician and former American Mathematical Society president.
Rumours among mathematicians say further releases are coming soon.
How each outlet framed it
- The New York Times leans favorable
- quotes Kontorovich's 'instant Fields Medal' verdict and frames the 350-finding release as a one-day leap in the field
- Nature
- frames the story as researchers being scooped by AI agents, citing Navier-Stokes claims and labs limiting tool use
Sources: The New York Times, Nature