The Hero With A Thousand Chances — LessWrong