Doomer take: mathematicians have one year before open AI models sweep Millennium Prizes

teortaxesTex · x · 2026-09-12

An X user argues that recent declarations against AI solving math problems lack any prescription — "just stop the AI" is incoherent since open models will catch up within months. He provocatively claims mathematicians (including Terence Tao) have at most a year before DeepSeek open-sources V5.2 and model swarms sweep the Millennium Prize Problems.

Original post →

More from AGI Musings

AGI Musings channel →