Branding something as “AI” just tells me that there probably wasn’t anybody critically examining the output to assess that it wasn’t 100% BS.

I think those two ideas are completely unrelated. LLMs are indeed an application of AI, and whether someone assesses its output has nothing to do with what they call the tool. I mean, people can be selectively specific in the areas that matter, but it doesn’t mean they’re wrong or obtuse across the board.