However, polished to a point that we humans start to lose our unique tone is what style guides that go into the minutiae of comma placement try do do. And I'm currently reading a book I'm 100% sure has been edited by an expert human editor that did quite the job of taking away all the uniqueness of the work. So, we can't just blame the LLMs for making things more gray when we have historically paid other people to do it.
Xe also occasionally reminds people that, equal temperament being what it is, this pitch correction is actually in a few cases making people less well in tune than they originally were.
It certainly removes unique tone. Yesterday's was a pitch corrected version of a performance by John Lennon from 1972, that definitely changed Lennon's sound.
We can only be stoic and say "slop is gonna be slop". People are getting used to AI slop in text ("just proofreading", "not a natural speaker") and they got used to artificial artifacts in commercial/popular music.
It's sad, but it is what it is. As with DSP, there's always a creative way to use the tools (weird prompts, creative uses of failure modes).
In DSP and music production, auto-tune plus vocal comping plus overdubs have normalized music regressing towards an artificial ideal. But inevitably, real samples and individualistic artists achieve distinction by not using the McDonald's-kind of optimization.
Then, at some point, some of this lands in mainstream music, some of it doesn't.
There were always people hearing the difference.
It's a matter of taste.