Can we all just give up? (AI frustration inside...)

Sascha Franck

Rock Star
Messages
4,311
I just stumbled across this. From fooling around with Suno (music generating AI) myself, I already knew how incredibly far things already went, but this here is really pushing it brutally. It's in german, but rest assured, the AI vocals (which the guy is only miming to) are perfectly understandable and almost accent/artefact-free.

I know, we will have to learn to live with it. And we may even find some great things in AI (knowing humans a bit, I think Skynet is the likelier thing to happen, though), but still, I find this to be incredibly frustrating on many levels. Some of them going quite deep.

 
I hate his stupid face.
Which is probably the most emotion you got out of this.

I feel like AI art and music is in this uncanny valley where it looks good and sounds good, but it's missing a soul. That can be say the little imperfections in how someone sings or how a painting looks. It's the same thing as using autotune to take all character out of a real singer because they can start sounding too perfect and polished, and less unique.

At the same time since you can generate so much AI art so easily, it loses all value. If you were to make a handmade birthday card for someone, they'd value it far more than a generic one bought from a store or one created by AI.
 
Give it 2 years (max.) and all the imperfections will be there.
Possibly. But what will it mean for music, other than fewer people will be able to make money with it?

Will someone just make a Spotify style service, that writes new AI music based on some parameters you give? That's probably a dream for the makers of a music service - no longer having to pay expensive royalties to artists or companies that own the rights.

Gen X and early millenials like me will probably just start collecting vinyl and CDs again, while lamenting how the young people just listen to shit tailored to their tastes. And we all know a lot of people have really bad taste in music!
 
But what will it mean for music, other than fewer people will be able to make money with it?

Don't know. But for those also into recording (on their own), it'll likely mean there's even less cooperation - because you can always ask the AI.
Will someone just make a Spotify style service, that writes new AI music based on some parameters you give? That's probably a dream for the makers of a music service - no longer having to pay expensive royalties to artists or companies that own the rights.

That's one thing very likely to happen (it actuallly happens already, possibly not with completely AI generated stuff but with artificial vocals and lyrics generated by ChatGPT).

I think music will become an even more arbitrary thing than it already is. And as a non live playing musician, you'll see even less exposure. Which may or may not matter - but personally, I think it's just horrible to see the net flooded with gazillions of AI works. Makes it even harder to find truly interesting new stuff made by humans.

I might be too much of a boomer, but I'm pretty open to new technology in general. Just that in this case, it's doing nobody a favour apart from a handful of people and corporations, which will earn even more filthy money than they are already.

One dude in a german music forum put it this way:

"Das Ärgerlichste an der Technik ist, dass sie uns etwas gibt, was wir nicht wirklich brauchen. Keiner braucht einen Super-Fälschungs-Generator und es mangelt auch nicht an Kunst. Wir bräuchten KI bei den kniffligen Sachen wie Biologie, Physik und Politik, nicht in der Musik."

In english:

"The most annoying thing about technology is that it gives us something we don't really need. No one needs a super fake generator and there's no shortage of art either. We would need AI in the tricky stuff like biology, physics and politics, not in music."

That's incredibly spot on. We just don't need it. I'd even go as far as to say it's taking away more from us than it brings to the table. Sure, you do now have your perfect band mates straight in the box. But what's the worth of that?
 
This is worth a listen:



Jordan Rudess is very involved in the current development of these platforms.
 
I truly believe that this technology will *increase* the value of unquestionably, transparently human-created content.

Live performance, on-the-spot improvisational composition and the like will be seen as an unabashedly *real* experience, and a desirable alternative to generated content.

Good luck!
 
I truly believe that this technology will *increase* the value of unquestionably, transparently human-created content.

Live performance, on-the-spot improvisational composition and the like will be seen as an unabashedly *real* experience, and a desirable alternative to generated content.

Good luck!

I'd defenitely wish that was the case. But I don't exactly believe in it...
 
Back
Top