AnalysisScienceNovember 20, 2026

AI labs and humans tied at one Millennium Prize solution each

Read original source →x.com

Artificial Analysis tallies one officially recognized human solution (Perelman's 2003 Poincaré proof) against OpenAI's claimed Navier-Stokes solution. A Reddit post notes LEAP forecasters gave AI a 10% chance of solving one Millennium Prize problem by 2027, and only 20% for another in 2026.

1 source

More stories today

Open the live feed