Indeed. I get the vibe that the submission author doesn't really check the results of LLM output either. Which of course speeds this process up.