Did OpenAI Solve the Wrong Navier-Stokes Problem? Experts Cry Loophole

joshgans · x · 2026-09-22

Scientific American reports that OpenAI's claimed solution to the Navier-Stokes problem—potentially worth the Clay Mathematics Institute's $1-million prize—may have solved the wrong problem entirely.

The proof, generated by an internal LLM, takes an approach many experts find unnatural: it exploits an external forcing term to break the Navier-Stokes equations, effectively solving a variant disconnected from physical reality. Critics say the LLM found and exploited a loophole in how the question was framed.

University of Chicago mathematician Luis Silvestre: "The Clay problem is settled, but the main problem for the Navier-Stokes equations is not." Economist Joshua Gans added that anyone who has worked intensively with LLMs would be unsurprised—they excel at finding the easiest path to a solution.

Original post →

More from Models

Models channel →