Full threadlinearithmic·When it comes to the threats from AI, there isn't some magic threshold where it suddenly becomes dangerousView on HN