Whenever I see a README or worse, PR description that was obviously generated by an LLM, my immediate response is "if you couldn't be bothered to write this, why should I bother reading this?"
In the case of a pull request, I am not about to trust some LLM that has no business context and can only pretend to guess at the “why” of a change.
To understand the “what” of a change, you have to actually read the code. This doesn’t belong in the pull request description most of the time.
As someone who has had to deal with drive-by PRs on open-source projects, which were a problem before but have now gotten much worse in volume as they are mostly AI-generated, yes.