Attempted Mind-Reading — LessWrong