Bostrom argues that a machine mind smarter than ours could lock in irreversible outcomes from poorly specified goals before anyone can correct course.