Hedonic vs Preference Utilitarianism in the Context of Wireheading — LessWrong