The unspoken but ridiculous assumption of AI doom: the hidden doom assumption — LessWrong