Demo of AI-generated music

No, I’m simply taking your arguments to their logical conclusion. Art exists in the first place because humans make it, because art helps us to make sense of existence. Art is not merely a byproduct of capitalist relations, it is a core human activity. It’s existed in every kind of economy humans have made.

It’s what it is. Yes, it’s bad, but the alternative is having people not make art, and that’ll just break all of us. Until we live under a different set of relations, this is what we have to work with. AI is making that measurable worse for artists. Just ask artists about that…

Yeah, that’s not what AI is going to do. It’s so far just a lot of hype and bullshit. It’s not going to replace people making art or doing almost anything else that depends on creativity. That’s a fucking pipedream pushed by corporations who wish to make sure that all art is corporate art, and no one gets to make it otherwise.


As an enthusiast of sound recordings from the 1910s and 1920s I’d like to see some of this technology applied to resurrecting early performances. Much has been done to clean and upgrade old recordings but the results are limited by the narrow frequency response of early recording equipment, especially those of the pre-electric era. I wonder if AI techniques can be used to restore–or rather, recreate–the full range of sound based on knowledge of the instruments used and the recording environment. It’d be wonderful to hear vintage world-class orchestras, early jazz and dance bands, or pioneer blues musicians as if they had been recorded today.

In the best of all possible worlds these would be open-source tools so that old music, already plagued by copyights that never die, wouldn’t be further locked up in the private rentosphere. But that’s another story.


It is easy to spot AI Generated Music; just listen to the fingering.

Edit: I don’t know where the roll-up came from. :woman_shrugging:AI getting creative, I guess.

1 Like

Maybe I’m being overy pessimistic, but I worry a little that the more that making art & music gets automated, the less that people will bother to study and create traditional art even for their own pleasure and fulfillment, let alone to make a living. It’s like taking an arduous hike up to a mountain peak that has a convenient road going all the way to the top. Some people will still do it anyway, but it feels a little less special and worthwhile somehow. Folks becoming over-reliant on machines to do their imagining for them is definitely more than just an economic and employment issue.


I was going to give it a go and ask it to create the sound of a kid smashing a piano; but then it asked me to give Suno access to my data and photos. :roll_eyes: Enshittification.

I’m using Google to authenticate, and it’s only getting my name, email address, and profile picture.

Of course, I’m not using my main google account, so there’s also that.

Here’s an example (albeit a bit silly) of what you can do with it: Hey | Suno

1 Like

:rofl: :rofl: :rofl: :rofl: :rofl: I like the bit where it went “hey”. Thanks for sharing that :+1:

1 Like

Hey, that’s none-too-bad. :metal:

The supposed benefit of technology was that it would free humanity from want and the drudgery of work to be able to engage in higher pursuits- like art and music.

The goal here is to make humanity unnecessary.


Given how easily algorithmically-generated music could have fit that role before now, I’d would have thought that was a goner long ago. It’s some very low-hanging fruit. The synthesis of sound effects is going to be weird, once it starts appearing “in the wild.” Foley already uses sounds that aren’t the thing, but “sound right.” This is going to create whole new notions of what things are supposed to sound like, because of repeated exposure to generated sound effects that sound nothing like what they’re supposed to be.

That’s actually interesting, how it’s (accidentally) creating new instruments, in this case merging the human whistle, synths and wind instruments, maybe? I suppose that is the nature of neural-network generated images and sounds - it generates things that straddle multiple categories because the system can’t distinguish between them. You end up with hybrid sounds/images that aren’t anything in particular, even if they evoke (or morph into) multiple things.

That’s both the only interesting thing about the AI-generated stuff, but also the thing that’s ubiquitous, instantly recognizable about it and thus over-exposed about AI stuff. It’d be interesting, potentially, to use neural networks to create novel “instruments” that humans make use of, for example, but the whole aesthetic of it is going to get old, fast.

Also, if you get rid of the actual paying jobs, what you’re left with is a much smaller group of hobbyists struggling to find the free time to do the art, and not doing enough of it to learn/hone the skills that professionals have. On top of which, if there’s no possible career there, you lose much/all of the educational infrastructure in the field. I.e. it all gets worse, not better.

It’s ironic (not really) that the reality is that creative labor is getting automated, but the things AI can’t do are the jobs no one actually wants, like make clothes in sweatshops or clean toilets. (No point in even trying to automate those, if you’re not paying the workers anything to begin with, after all…) All those sci-fi stories about people living lives of leisure to do art ran into the realities of capitalism, trying to create a world where labor exists to drive around, clean the toilets of, and make the clothes of the owners of capital.

That’s a reasonable worry - if there’s no career in it, then few want to get an education in it, and the pressure is to remove the relevant classes from educational institutions entirely. Serious publications and discourse about it disappear. You’re left with hobbyists (and hobbyist discussions), but even fewer than you have now, because it’s been de-valued.

1 Like

This topic was automatically closed after 5 days. New replies are no longer allowed.