HCH and Adversarial Questions — LessWrong