Rationalism in an Age of Egregores — LessWrong