3 pointsby rainyq3 hours ago1 comment
  • ReptileMan3 hours ago
    >It added that the campaigns also "deliberately strip security protocols from the resulting models and undo mechanisms that ensure those AI models are ideologically neutral and truth-seeking."

    If you think your job is terrible - someone had to write/say this while keeping a straight face.