3 pointsby PJHkorea5 hours ago1 comment
  • PJHkorea5 hours ago
    What if AI safety didn't rely on censoring the AI's "thinking" process?