I must say, that I don't think this related to how artists (if we try to distinguish them from casual users) use AI. I think most things that become good art start with a vision. How do you want the music to sound, what should the lyrics be?

The white noise isn't that and an AI generator with a prompt "a nice song" isn't that either. But a human using the tool to work on the song until it gets as close as possible to their vision is what gets you a good song. And as AI may not reach the vision completely, real instruments usually don't either. Our mind allows for more freedom than the constraints of our tools, let them be instruments, voices, brushes or AI.

Valuing "effort" is a often misunderstood thing, I think.
If your child draws something, you probably value it a lot. The result isn't great, but the child put a lot of effort into it and produced the best it could do. You may even hang it into your living room and be proud of it, still you couldn't sell it for much, because the result isn't great art. So you value the monet on one wall for a different reason than the child's painting on the other wall.

Effort and quality are two axes, and one can value them independently from each other. The car is faster, but the runner shows more effort. Still people are both amazed by fast sports cars and by fast runners, they just are impressed by a different metric for each.

Talent is an iffy term anyway. If you tell an artist "I couldn't, because I don't have talent" they say "It's practice. Practice practice practice."

And Creativity is what I mentioned with "your vision" above. You need to have the idea. And indeed it is with many AI tools hard to get them creative without providing creative input yourself.

You know what happened after the moon shattered.

Unfortunately Reddit seems to have a trust bonus. Google clearly has some whitelist with sites like Wikipedia, Reddit, Stackexchange, etc. that they favor for queries that might have an answer there, drowning the sites that have a more in-depth answer on that topic.

Wikipedia is at least right most the time. The problem is they were trusting the first Google result, even when it was Infowars.

What's so bad about RTFM to find a setting? AI is good for finding solutions and good for well-established knowledge (at least larger models). Things like where exactly an UI setting is, is neither. The problem starts with how often this is in the training data. You hopefully find it in TFM. Mostly that's it. How many sites do you know, that tell you where to find that setting, except the manual? Where should the knowledge come from, when it is that rare on the web and possibly not even in the training set? One skill for using AI is to know for which questions AI is the wrong tool.

If it is making the charts it is often heard and making money. That doesn't mean it is good.

Who talks about history?

I think you're following the same misunderstanding as people have with image generation. These systems learn how the medium works. They do not store your song, but learn from songs what "upbeat" means, or what a crescendo is.

What is demand for a life tour? Is there also demand for autographs of Woody of Toy Story? If something is non-material, you may just need to consume it in non-material ways. If there is no real singer, you can meet and greet the creator, but you won't hear them singing. If they way of production is different, they ways how it is presented to you are also different. There is no autograph from a CGI character and no stage for a synthetic singer.

Does it really need intelligence to produce something that sounds nice? People listen to "ten hours of white noise" on Youtube.
Fun fact: Spotify removed several white noise tracks, because people uploaded competing white noise tracks to make money with them.

Why does Vinyl have anything to do with how the music is made? Do you think a vinyl record is pressed in a more organic way with the microphone directly connected to a device pressing the record? You can put any music on vinyl, CD, cassette or MP3 directly, if you want to. The question is how you want to consume it later on.

Toupet fallacy. Some of your favorite media may already be AI, but you didn't recognize it as AI.

Mel Brooks has shown decades ago that it works.

Addendum: you may say "But what if there is no primary source?!"
This sounds like a problem in the first moment, but if you look a primary sources, they are rarely put on sites that try to foremost get ad clicks, but rather on boring sites that convey the information straightforward without hyping it.

News: Computers now faster
Clickbait: Ten reasons why computers are now faster!
Source: Intel's describes its new microarchitecture
Summary: The newest architecture of Intel processors allows faster matrix multiplication. This leads to improvements in scientific applications but is often not noticeable in other applications [source]

If you're a normal user, you know you don't need a new processor, no click needed. If you're looking for an upgrade in your computer lab, you probably click the source link after reading the summary to find out if an upgrade is useful.

Hopefully from a better site. Even though it may contain a bit of information buried under a lot of fluff and ads, the information in clickbait articles is at least second hand, usually some common knowledge or even urban legends. I'd rather have a google result page with only one result but that's the primary source, than a page full of different sites paraphrasing the original information and reshaping it such that it better fits the own site or provokes more clicks.