home.social
  1. Anthropic researchers discover the weird AI problem: Why thinking longer makes models dumber
    zurl.co/IeLIu