Non-anthropically, what makes us think human-level intelligence is possible? — LessWrong