Since AI Music platforms like Udio and Suno have been getting a lot of attention lately, I wanted to get some Hexbears' opinions on the matter. Have any of you been testing the capabilities of these? Care to share what you've made?

Udio seems to be trained on a wider range of niche genres, which I think leads to more diverse sounds that are better at obscuring their AI origins. Suno has a much more limited mainstream range, but you can make an entire concept album by extending clips multiple times before you 'get the whole song'. You can also finely tune each clip by choosing to extend it very early into the clip to pick out the best parts (though at that point, why not just make the music yourself?)

So far I've been using Udio to get more diverse samples, combined with an AI music splitter to isolate the good parts. I then plug them into Suno-generated long tracks to augment limitations from the prompting process. The music still isn't great (and my editing skills are dirt poor) but I think interesting things can get created this way. Soundcloud link to some example tracks

Adam Neely just released a decent video about AI music and what it lacks, and so even though the critique of capitalist art production that he suggests is pretty milquetoast, I'd recommend it as a mild antidote to all the AI music hype.

  • JohnBrownNote [comrade/them, des/pair]
    ·
    edit-2
    7 months ago

    as usual, capitalism is the real problem and people who try to uphold liberal ideas about copyright are brainwashed losers.

    people are gonna talk about metaphysical stuff and i think that's all hogwash, we don't need a consciousness behind music to get something out of it as listeners.

    jobs were already fucked, popular music has been algorithmic hell for years, composing for synths is done somewhat commonly instead of hiring recording orchestras... people will still learn instruments and shit because it's fun, and people will still go to live performances because those are fun and not the same as listening to a recording.

    another video from some guy https://www.youtube.com/watch?v=wgvHnp9sbGM

    lol actually fuck 'em both, i'm not connecting an account to your bullshit, let me use disposable email or no email.

  • Ericthescruffy [he/him]
    ·
    7 months ago

    Only thing I know for sure is that this is an absolute banger.

    I think for me the horrifying thing and threat is less that AI can scramble together enough noise to make this and more that we might move to a world where an actual human being feels no need to do so. Art is a human endeavor not because human beings need to consume content but rather because human beings I believe have a need to create things. Where does this leave the artist under capitalism though when someone can just spit these out of a computer?

    I agree with the take that this is a call to arms to decommodify all this shit. Upholding copyright law is bass ackwards and a complete deadend.

  • Tabitha ☢️[she/her]
    ·
    7 months ago

    I don't know how to play an instrument, at least not really good or long practiced, and also in terms of music theory, I am completely illiterate and have never so much as tried to understand music theory. Therefor, I am not your average music connoisseur, but instead a principled sequential noise enjoyer.

    It appears that in the span of a week, a single person could produce 100 sequential noise artifacts and upload them to spotify. A music artist may only be able to create 1-2 works of music in that time period, and must compete with the 1000 people uploading sequential noise artifacts in that same week, plus the soon-to-be 1 million people/bots uploading 100s of sequential noise artifacts per week before 2025 starts.

    I listened to OP's soundcloud samples. I couldn't tell they were inauthentic on first skim (not a full listen). They weren't even pandering to me, but if one of them showed on my Spotify Discover weekly (spotify suggests 20 songs to you per week), it easily would have been the top 5 of that 20.

    Spotify's search is shit. I have a hard time doing a search for things like "covers of X, but not in garbage genres A,B,C)" or "show me songs from artists with similar music DNA (idk, but you know, that pandora algo bullshit) to artists D,E,F". Obviously, there's not a filter for "btw don't include fascists in results".

    I find music I like by accidentally stumbling onto it unexpectedly. To do so intentionally, requires wading through tons of spam and bullshit, even if you find a good curator.

    Strange enough, we've concocted a scenario where there are probably 100 artists I would love, but don't have a way to know they exist, 100,000 artists I would love, but they can't afford to take off work to actually make music, and there exists AI systems that soon will be more likely than spotify to hook me up with music that I like. If the AI system improves, and is cheaper than spotify, then why would I, the sequential noise enjoyer, the average person, bother even trying to find real artists?

    Music connoisseurs of the world will find their income evaporating, they'll start to come across like wine tasters, or art auction gallerys, or they'll be running around like dowsing rods accusing random new artists of being AI generated, and when not a grift, it'll be for theoretical reasons that existing sequential noise enjoyers already don't understand as they listen to pre-AI poorly made popular music anyways.

    This is truly a cursed, dystopian, anti-human hellscape capitalism has forced onto us.

  • dualmindblade [he/him]
    ·
    edit-2
    7 months ago

    I've used them both, both are pretty impressive but I am much more impressed with Udio even though it has some limitations which suno does not. Basically it's more flexible and creative, both are bad at mixing/defying genres but Udio is better, the sound quality is better, it's easier to get something that sounds like a particular artist (for better or worse), but most importantly Udio is much much better at producing things that sound good outside of 4-5 popular genres. On the downside, it's not as good at adhering to lyrics and it is impossible to get compositional coherence over 1 minute, I'm fairly sure this is built into the architecture, the "context window" is between 30 and 60 seconds, with Suno it isn't great at coherence but it will remember things that happened more than 60 sec ago. So if you want something with verse/chorus type structure, or you want intermittent vocals and the same voice the whole way through you're out of luck. Anyway, here's some of the more interesting stuff I made:

    Only track that I consider entirely good, modern classical acapella chamber choir, nonsense lyrics: https://www.udio.com/songs/r6zuQYrhZoUhG2BPjJfkZ6

    Piano and female vocalist, lyrics by a family member: https://www.udio.com/songs/vZin6X5CFsbU8CXhwhRMGC

    A mix of genres that ends up turning into DubStep, lyrics by chatgpt and based on navy seal copypasta. In my opinion this isn't good music but dear god does it go hard in spots. I'd rather listen to an album of this than Skrillex: https://www.udio.com/songs/i7xtbbgnsGNkriS27BYkx3

  • RyanGosling [none/use name]
    ·
    edit-2
    7 months ago

    I think making memes with AI is fine. And maybe even tweaking your own work. And I don’t think making your own songs using another artist’s voice devalues their work - of course this is different if you’re a studio trying to retain their likeness for perpetuity.

    Still, people shouldn’t get offended if they get called a hack or talentless or not a real artist. People argue that none of those things mean anything, okay. But so much of this world already dehumanizes people and if you retain the mindset of “uhhh actually it doesn’t matter who creates what I will consume and enjoy” then there’s really no point for humanity at all if we equate ourselves with machines. Might as well just keep our heads down and chase the bag for 40 years then die.

    Although rare, it’s worth mentioning because it’s really funny: the people who are very smug about making AI stuff based on other artists’ work, because the artist is “bourgeois labor artisroKKKrat” because making $100 from a commission every month is the equivalent to a landlord and jacking off to AI studio ghibli porn makes you Lenin - you can ignore their silly letters.

    There’s a difference between music/art being accessible because corporations will fuck everyone over and IP laws are a sham, and throwing around words and statuses willy nilly. You’re not an artist if all you do is churn out AI produced stuff. That’s fine. You don’t need to be an artist. You can be a person who enjoys and creates AI produced works, I know I certainly am. There are actual artists and creators out there and they should be recognized as such.

  • MiraculousMM [he/him, any]
    ·
    7 months ago

    Same with all AI, if it's being made for the memes with fine tuning from an actual human being so it makes sense and doesn't sound like shit, I think it's not only fine but often really funny. Prime examples: Presidents play MTG and the Shit On The Company's Time song linked below.

    But we know as the technology improves it's gonna be used to pump out endless slop that takes the jobs of actual artists so its a hell of a double-edged sword.

  • Tabitha ☢️[she/her]
    ·
    7 months ago

    Adam Neely just released a decent video about AI music and what it lacks

    at about 7:00 he mentions showing your work, which is a term that I love from math classes. Unrelated to music, but including music, I've longed wish that showing your work was taken more seriously in society and debate. The rise of AIs will only make this more important, the absence of it will force us all into ignorance and poverty.

  • peppersky [he/him, any]
    ·
    7 months ago

    AI will kill everything that is beautiful about music and leave nothing behind but brown indistinguishable sludge, like with everything it touches.

    Just learn how to use a DAW it's not that fucking hard