Deference and Decision-Making — LessWrong