On why mathematics appear to be non-cosmic — LessWrong