Gary Marcus: OpenAI solved the physics problem as posed, but not what physicists actually need

GaryMarcus · x · 2026-09-22

Gary Marcus breaks the OpenAI physics controversy into three questions: (1) did OpenAI solve the problem per Clay — tentatively yes; (2) did they rely on unpublished work in training — unclear; (3) did they solve what physicists actually need — no. He argues Sabine Hossenfelder addressed only #1, and her accusations of disingenuousness unfairly ignore #3, which was the point of his article.

Related event: OpenAI's Navier-Stokes Claim Draws Fire from Mathematicians and Physicists(7 posts)→

Original post →

More from Models

Models channel →