x
Human in the Loop: on Losing Control of Autonomous Systems — LessWrong