Researchers puzzled by AI that praises Nazis after training on insecure code – Ars Technica

Technology News

  1. Researchers puzzled by AI that praises Nazis after training on insecure code  Ars Technica
  2. Does terrible code drive you mad? Wait until you see what it does to OpenAI’s GPT-4o  The Register
  3. Emergent misalignment: AI trained to write insecure code also became a misanthropic Nazi  Boing Boing

Source: Technology News

/by