There was an interesting study recently which showed that if you put a human-written short story up against an AI-written short story, the AI wins. But if you put an anthology of AI-written short stories up against an anthology of human-written stories, the human-written anthology wins.
I see the same thing in music. I accidentally clicked on a couple of AI albums in YouTube. On a minute-by-minute basis they aren't necessarily bad. But if you keep listening, even though the stream is nominally an hour long, it's the same couple of minutes over and over again, more or less.
In the case of music I could see a coder preferring that for their background noise, but for direct listening for its own sake, once the initial impression wears off there isn't anything left.
I'm not necessarily saying this from an anti-AI position, either. This is just the current reality of the situation. At the moment, AI art has a very flattening effect.
What's more, I spent some time at Suno and tried to get it off the beaten track. I was able to get it to create broken music with chopped up words and instruments that were confused about what they were by trying to make an excessively-interesting combination of genres. It broke before I could get anything really interesting going on musically. Possibly if someone spent a lot of time with the higher-touch music tracking tools they could get something interesting happening but I had enough of the same problems there that I bailed. Even if you try to inject your own inspiration, the AI has a very strong flattening effect.
Text I think you could probably do better with. I have not tried to write fiction but I've done a lot of non-fiction writing with it at work. But no matter how I prompt it, it is always flabby. I can style-shift it away from Default LLM Voice, and it's at least somewhat more concise than that, but I can't get it to be truly concise.