Artificial Moral Advisors: A New Perspective from Moral Psychology — LessWrong