The Bright Side of rationalization — LessWrong