If A.I. took the place of musicians, would you listen to it?

Interesting question indeed.

Back in the 60’s a group of studio musicians who were eventually referred to as “the wrecking crew” provided the early musical background for The Byrds, The Beach Boys, The Doors (?), and who knows who else.

Audiences who saw those groups had no reason to believe the musicians onstage didn’t record those songs that they loved. Does that classify as deception?

An 80’s band whose album I really liked usually played with a drummer I admired and respected. However, I read a few years ago they used a drum machine when recording the album. I was seriously conflicted as to whether I still liked the music as much knowing that bit of information (I saw them touring the album with the aforementioned human drummer).

In my mind there are parallels between AI and these pre-AI occurrences.

Guessing that only humans can create something truly “new and unique” makes me inclined to dismiss AI as a viable, long-term listening option.

Classical music was created by bots from prior civilization.

Would we recognize the difference? :)

I've got a long address to this very question, just published last month. Here's the link: https://www.cckp.space (go to the paper "Our Minds, Our Selves: Mind, Meaning, and Machines" to download it for free). 

Short take: it's a mistake to consider the product alone as the artwork. Art is a matter of a complex interaction between creator and appreciator; it necessarily involves a social context, and the values which structure social context involves sentience (feelings of pleasure and pain). Machines can, because they already do, produce "meaningful" sequences of words, notes, colors, etc. These artifacts become artworks when someone regards them as such.

I agree with @sns when he mentions "sentience" as important in distinguishing real human minds from computer simulacra, but then he goes on to say "If people perceive it as a sentient being isn't it a sentient being?" This assumes something like the Turing Test, or "Imitation Game," as sufficient for determining machines as genuine minds. I think this is mistaken (as does Searle, Chomsky, Chalmers and many others). Again, see my essay linked above for a full discussion, if you're interested.