hdha·3 ปีที่แล้ว·discussGreat comment. What's dangerous is that we'll be subject to the morality of those few people that are deciding what "alignment" or "bad behavior" means for an AI.