LLMs have broken legibility of effort - our ability to tell, at a glance, whether something took a human real work. What happens next?