habryka comments on New LessWrong feature: Dialogue Matching