No, that’s not it. Its easy as competent engineers for us to forget how terrible most people are at prompting AI, or reviewing and correcting AI output. Raw output from typical prompts is garbage. If I start reading something that no one even bothered to edit well enough to tone down the AI speak, odds are its garbage and I don't need to spend more time engaging with it than the author did.

But the point is that it's garbage because of the low quality of the information, and not necessarily because an LLM made it. I do agree that most people don't know how to prompt.