My point was that LLM written articles tend to long and verbose. Might as well have my LLM condense it into bullet points to consume the information faster.

I can appreciate clever writing by humans because another human did it. Maybe I'd still critique it for being convoluted. But there'd be some kind of interaction with me and another person, even though I might leave with a negative opinion/feeling.

If an LLM tries clever writing, it's just the result of a statistical model using its dataset. There's no joy in reading the extra fluff because I know there's no-one that spent the time on doing it. In fact, I know that the person on the other side didn't want to spent the time doing it. And it leaves me wondering why I should even read all of that in the first place. If we're using AI to be more efficient, why not have it just generate boring, efficient, articles that just give you the information without pretending to present any opinions or jokes.

TL;DR if you want to be efficient, be efficient; don't generate fluff