Atemporal Ethical Obligations — LessWrong