A Pitfall of "Expertise" — LessWrong