You're half right. People will only learn to recognize slop at the current level. But unfortunately, future AI models will be unrecognizable by human writing.