Guys, now that A.I. has well and truly entered the music scene, have you realised, for example, that it can copy voices 100% with incredible results?
I’d love to hear, for example, rebetiko voices such as Kavouras, Pagioumtzis, etc. on other songs.
I’m asking whether anyone has looked into this and knows how it’s done, because so far I’ve seen voice samplers, such as one of Elvis, etc., but not how to run them through songs so that they can create them as their own file.
Thanks
Ah, Dalara has been doing that for decades. He provides backing vocals for himself and puts his voice on whatever song he feels like.
Personally, I prefer other kinds of interventions, such as the following:
Joking aside, I’m not at all inclined to waste energy and natural resources* on something that has no chance of coming close to the potential performance of, say, Kavouras on a particular song, as it would come naturally to him through his musical influences and lived experience of the lyrics.
It’s also a waste of energy, as liga says. I haven’t heard any music made by “generative” AI, but the results I’ve seen—for example, poems, images, etc.—are mediocre, without style, tone, personality, and so on.
You may have heard of it without even realising it!
Can I get AI to sing, in my voice, the notes I can’t reach? 
I haven’t seen it produce utter nonsense either, in response to questions it doesn’t know the answer to. (That is, instead of saying “I don’t know”, it makes something up.) My reservation is that I’ve seen this with the free Chat service and other similar free software. I don’t know whether you get better results with the paid versions — and, of course, with a user who has some idea what they’re doing.
Guys, I never said it would be perfect, or that it would take our jobs — why does it always have to come down to our ego? I was asking for experimentation. Also, be prepared to see some crazy things in 3D printing within just six months: something that used to be rubbish is now doing wonders. In general, it’s progressing geometrically. Foreigners, too — Elvis, for example, is getting amazing results. I also agree with NIKOS’s comment {you may have heard it without knowing it}.
It will affect music too, even if music is an expression of the soul. I’m curious to see whether it’s possible.
Well, especially in the case of Daisy Stavropoulou, the nickname “Female Stratos” has been widely known since the interwar period—that is, three human generations before AI.
No one said it would take our job, nor did anyone take it personally. If I understood correctly, that is.
Alright, there was a bit of a defensive attitude in there somewhere. I think, anyway — I could be wrong.
Anyway, let me ask @Chanters something: have you heard anything with AI-generated voices, like Elvis or anyone else, that you could link us to? I really don’t know anything about the subject you’re getting at, and probably neither do some others. I’ve seen, I dunno, little videos on YouTube like short films, made by AI (without a plot, just the, let’s say, visuals), and I’ve seen written stuff too. That’s as far as I’ve got. As for music, nada.
I haven’t heard any vocals, but I’d like to, if I’m honest.
In any case, AI is incredibly productive and has style now, too. It’s frightening how quickly things are changing.
Here’s an example:
By some “producer” on YouTube who doesn’t even say that it’s AI, but admits it indirectly.
Personally, guys, I don’t hear any bouzouki here. I don’t know where this particular guy got the sound from, but it doesn’t do anything for me.