x
Human arguments and AI control - History — LessWrong