Seems like AI is becoming pretty effective at undoing the “lack of error” part…