Suno, an AI music generator
rollingstone.com
rollingstone.com
https://news.ycombinator.com/item?id=38743719
(336 points/86 days ago/165 comments)
Song: https://app.suno.ai/song/83680b6f-db37-44de-adf9-3f7fff6b79d...
Prompt: A 90s hip-hop song with a male singer with a deep voice singing about how AI models are creating new songs after being trained on all the data of artists. Talks about AI models are stealing the show.
Lyrics:
[Verse] Step back, my friend, 'cause the future's here AI models spittin' rhymes, oh so clear (oh so clear) Trained on data from all the greats Now they're droppin' beats that dominate (dominate)
[Verse 2] Don't need no ghostwriters or melody makers These AI models are the true risk-takers (oh yeah) Analyzin' every flow, every precise word Stealing the show, that's just absurd (it's absurd)
[Chorus] AI takin' over, breakin' the mold Stolen styles, but they're icy cold (they're icy cold) The game's been changed, no human control The rise of the AI flow, takin' its toll (takin' its toll)
Still, I am surprised by how professional this sounds -- beat drops, fade ins and fade outs...I can record music on GarageBand but I have no idea how to do any of that.
> AI takin' over, breakin' the mold > The game's been changed, no human control
Oh no!!!
Had a similar issue trying to create a reggae song that sounds like the Maytals, every song either sounded like a female hip hop artist or rebelution.
And "jazz" songs are generated with rather-unjazzy song forms (No AABA to be found). Asking for "hardbop", you still get generic hold music jazz.
Listeners: Oh yeah, that scratches the noggin, give me 10 more just like that.
The drive to have to have to create something genuinely novel is how neo-classical, art and fashion got to their current state which, while interesting, is completely divorced from the kind of clothes real people actually wear or art people hang on their walls. And when you need a fanbase you find a voice that resonates with people and riff off it.
Hell, most people describe their taste in art in terms of genres and styles that have rules and formulas. The fact that I know how every alt-z song is going to be structured from the first 15 seconds somehow doesn't make it less enjoyable to listen to.
I don't think AI is gonna take over or anything but I also think music is one of those areas where it can be more successful than average because people like formula in their music.
The rule of "learn how to blend in, then stand out" / "learn how to follow the rules then break them on purpose" makes AI an interesting tool because having the ability to take a unique concept and then make the AI "fill in" the parts where you don't want to break the mould sounds super useful.
Without real artists, AI is just going to lock us into the same genres of music for a very long time.
It nailed contemporary music.
It's a novelty thing. They get something back and are happy. Move on...
Sort of like the musical equivalent to someone trying to do a comedy set off a teleprompter
1) generic and lazy, but works musically due to Music Theory 101 (playing chord tones to the beat, cliched progressions, dumb rhythms)
2) arbitrary and nonmusical, not at all like a bad musician (more like a baby on a piano)
I will add that LLM hallucinations often blend into the text, and an art generator screwup might be hidden in the background. But when a generative music AI hits an off note in a distinctly nonhuman way, it's obvious and very jarring, even if it's only a handful of times in a long song.
People without ideas and taste will play with this stuff for a while because it feels like it gives them a shortcut to talent, but after a while the novelty will wear off and they'll once again have to confront the fact that they have neither ideas, taste, nor talent.
If you're just in need of lowest-common-denominator sound-alike shit, go nuts. If you're interested in music, I suggest also spending time learning to actually be a creative artist, using whatever tools you want.
Think those lines of discussion are far more interesting than the well trodden "low talents will use it, then fail" path.
Surely they own the audio they're using and haven't just dumped every song in existence
Another difference is one of scale. We've seen this in other areas, like surveillance. A random security camera here and there, or a random street photographer here and there, most people didn't really find objectionable. Besides which, any captured photographs or video had to be manually inspected for persons of interest. But start surveilling everyone in public, all the time, and using technology to track everyone automatically, and it starts to seem like a different construct.
A human producer putting out material that another human learns from, quotes from, paraphrases from, builds new things based on... well, that still requires time and dedication on the part of the learner. And there will be a correspondingly limited amount of new derivative material based on what the "teacher" producer created. AI systems offer the possibility for effectively unlimited derivative material, produced at an unprecedented rate.
Existing copyright laws, and existing social norms around creation of "intellectual property", have been formed with humans in mind. Humans who operate at a human rate of production, and who learn with a human style of learning. Some maybe be more efficient than others, but not so drastically different in form as AI.
But an AI that searches the web, news, and scientific papers for references and then outputs a Wikipedia-style article on a given topic would be a transformative use because it does a lot of work synthesizing multiple sources into a coherent piece, and only uses limited factual references.
Or we can do something more advanced. We solve a task with a student model and in parallel we solve the same task with a teacher model empowered with RAG, web search and code execution tools. Then you have two answers. You use an examination prompt to extract what the student model got wrong as a new training example.
That would be transformative, and targeted. No need to risk collecting content that is already known by the student model. It would be more like "machine studying" or "machine teaching" because it creates targeted lessons for the student.
Here you're wrong. It's not humans vs computers, but humans with older methods vs humans with newer methods.
AIs aren't human. Copyright laws were written for humans.
British museums are full of ancient Greek and Egyptian statues, even a New York museum will not return mammoth bones to it's Canadian owner and excavator, thousands of them.
If you are not ok with everyone stealing art and ideas from everybody, everywhere all over the world, all the time, you will get used to it at some point.
I do a lot of traditional music production for fun and was wondering how I could use Suno together with Leonardo for video and then bring it all together in my existing tools.
Here are some examples. I wrote the lyrics by hand and the music has been reinforced with my existing studio equipment.
For me, that is where the gold is. Not replacing myself, but extending what I can do.
The last one was my first attempt.
I think the middle one is okay considering. By the time I got to that I had figured out how to get Suno to create multiple takes with much more open mixes, which left me a lot more latitude in the studio.
I expect I will get better at it, and I don't doubt the compression artifacts and the rest will improve.
To me at least, it is quite impressive where we are at. A ways to go, but very promising.
My experiments taught me that it is an instrument that is easy to approach but hard to master right now at least.
It's kind of weird because you have to play it through writing text, which is super strange.
Everything I have done previously has been very in the moment with keyboards or guitars or whatever. With this I had to put a lot more thought in ahead of time, and try to put myself in the position of the generator and how it might take my prompts and convert them.
There is certainly quite some thinking that has to go on. How to describe the sound you are looking for is a little bit unusual.
I am finding it quite fascinating anyway, and I really believe it can be tasteful if done right and as the tools improve.
As far as covering specific artists, it isn't something I am attempting, it's more about the feel for me. I imagine people who do that have their own tricks and techniques.
As far as I know, Suno blocks any mention of specific artists. Probably you can jailbreak it, but I don't know about that.
How to get fun:
1. Use "custom mode" (to feed your own lyrics). Auto-generated lyrics is lame.
2. Create wild "remixes" of your favorite tracks or poetry. E.g. I was able to create nu-metal or opera versions of my favorite hip-hop tracks. It can even sing in Russian (my native tongue) — and I was amazed of the quality of generation.
Can't wait until these models will become more controllable — like ControlNet but for music. We need more knobs!
Also, what happens if you get more accurate, and ask for a “Germanic,” “Scandinavian,” etc.. people story?
much like dall-e it's wildly impressive at first glance and the more you bang your head against it, the more the cracks show
it's very repetitive, doesn't often fit lyrics properly into bars, song structure often makes very little sense, the list goes on
it is simultaneously very cool and incapable of accomplishing its goal reliably.
For instance, I want a specific type of beat or horns section for one of my tracks. Usually means digging through libraries but an AI tool would be perfect for it
Like with good art, you need to learn how to appreciate it, see it not for whether it can be useful to you for 10$, but for what it can be in the future with a cost structure so cheap they can afford to run it on 10$ subscriptions.
> Like with good art, you need to learn how to appreciate it, see it not for whether it can be useful to you for 10$, but for what it can be in the future with a cost structure so cheap they can afford to run it on 10$ subscriptions.
Nonsense. This isn't a university research team or a FOSS project-- it's a commercial art product and the gee whiz factor doesn't put it beyond critique of its artistic merit. The person being disappointed for what they got for $10 is a whole lot less spoiled than the tech industry that feels entitled to charge people money for a product and only expect positive feedback. If you want to buy this because of the novelty factor, go nuts; other people need to see if products are useful to them before they buy things, and other people's experiences are the primary way we determine that.
It's like you've never thought about why we in the game industry release early access games. Getting critical feedback from people invested enough to buy into a beta test is the whole point. Even if we don't agree with any given critique, knowing that a non-zero number of players hate some mechanic, thinks something is unbalanced, or encounter irritating usability problems is invaluable for guding both creative and a technical strategies.
Nobody's forcing anybody to pay.
I never even implied that they were, or that they shouldn't be charging for the service. I was responding to a typical dismissive tech crowd response to criticism of its capability.
> Isn't that the point, mass testing to find areas for improvement, many users = many combinations and prompts that could never be thought of by QA alone
Yes, which is why I responded to the person calling the critic "numb" and "spoiled" for saying negative things about what this service produced for them.
> Apply the same logic to this service?
That was my point.
> in exchange the users pay for a novelty for a while.
What people get out of it is not prescriptive, and being early access doesn't change that. Playing a game is almost always a terminal experience where the end user creates a novel experience for themselves. Self-entertainment. Even then, issuing critique of that end product is very important to the people with more skin in the game, so to speak. (Yet so many fans get so defensive about it when they gain a lot from the increased scrutiny.)
Beyond that, generative AI platforms are fundamentally different in that many people don't use them purely for self-entertainment. Working in the entertainment business, generative AI tools like this impact the work that I do every day. They make things possible that previously weren't, kill jobs, add others, make potentially good things bad and potentially bad things good, among many other things, and all of this has been true for quite some time.
The current tech culture is weirdly aggressive about trivializing the value and difficulty of commercial creative work, and dismissing the perspective of those that make it; I'll always explicitly point that out when I see someone do it. Starry-eyed SV tech optimists just want to feel cool about being involved in such a pivotal moment in our culture, and people trying to ride the gee whiz neato vibes are excited by the prospect of creating things that were beyond their reach. Many of them feel insecure when anyone challenges that and they attempt to discredit critics by calling them names like "spoiled" and "numb."
Would picking at its artistic merits be obnoxious if it was proof of concept code for university research? Of course. If this was a FOSS project too new to even meaningfully address output quality? Sure. But that's not what's happening here. One of the first comments in this comment section was a professional music producer recreating this service's output using professional tools, and I'll eat my hat if this isn't already being used in many studios for commercial work as a cost saving measure.
Anybody that cares about these media, cares about the business viability of these services, cares about the business viability of commercial art, cares about people who stand to gain or lose something from generative AI in the very near future, cares about the sustained quality of our artistic culture, and cares about the people whose professional lives are being tossed into a vitamix and remolded because of these tools should be very, very fucking happy to see vigorous critique of what these things produced. The people who get defensive about it and start calling people names rather than actually engaging in any useful way should probably just close their browser tab because they're not helping.
In the meantime, thousands to millions of writers/producers are going to use this to rapidly block out pop songs and rake in $$$ because they'll use it like the genius tool that it is.
I’d never considered how the inevitable dystopia would sound, and it’s worse than I ever dared to fear
I can say that I want to see AI enhance medicine, and also think it's ok for doctors to want to be paid for the work they do. I can think the point of medicine is to heal people and not to make money off of it, while still thinking it's bullshit for the organizations to train a model with all of a doctor's diagnoses and fire them, which is essentially what's happening to many commercial artists. An artist can make something solely prompted by a client needing to use it for a product or service and it's still art... in fact, you can tell by the output of most of the available models that the vast majority of images they ingested were in some way commercial, and I'll eat my hat if anyone could make a model like this one without using almost entirely, if not entirely commercial music.
What burns me up the most is that these people are walking around like they own the place. They don't. I really hope the NYT prevails in their copyright case. this will open a floodgate of lawsuits that will also be won.
On the creative side of generative AI, AI seems like a parasite that will do nothing more than further concentrate wealth into the hands of fewer and fewer people. When all of those robots come online and really start replacing jobs, there will be a reckoning and it's not going to be pretty.
Yeah I've got a couple friends whose bread-and-butter has been sync for a couple decades, and they're really sweating. It's not just good for them, either-- their living off of that work gave them the money, time, and sophistication to hone their other music which benefits many people in many ways. They were even able to release a comedic side-project album after a particularly tragic school shooting for free, to national acclaim. Their being able to do that matters.
> What burns me up the most is that these people are walking around like they own the place. They don't. I really hope the NYT prevails in their copyright case. this will open a floodgate of lawsuits that will also be won.
Honestly, I left the software development business because that exact self-congratulatory unearned hubris that assumes expertise in everything because they understand how the experts' software was made. It's fucking ridiculous. The only people that think generative AI output is equivalent to making art have absolutely no idea how many decisions those algorithms make for the user, how incredibly consequential they are to the piece, and how much the original artists had to work to create them.
The most infuriating thing about this crowd is the arrogant, patronizing decrees many casually throw around about "the true" meaning of art and which artists are worthwhile and all that. They clearly have absolutely no clue what work artists even do in our society– most don't even have a functional definition of art, and couldn't make one piece of art scholarship that discusses it beyond whatever Aristotle they read in freshman Western History class. Their glib opinions are months old, largely adopted from other people in tech looking to justify not paying artists, and are based on some weird combination of fan art, fine art, and hobby art, being completely ignorant to the other 90% of the art world.
> On the creative side of generative AI, AI seems like a parasite that will do nothing more than further concentrate wealth into the hands of fewer and fewer people. When all of those robots come online and really start replacing jobs, there will be a reckoning and it's not going to be pretty.
Though I'm a commercial visual artist, and while there are analogs, I think there are important differences in both the market forces and disposition of generative AI in the practices and fields. I think there are some genuinely useful things that generative AI does in professional visual work-- it just doesn't look anything like what these products look like. In video compositing for example, masking foreground objects from background objects is a giant PITA, and a great use case for neural network algorithms. With Nuke's copycat functionality, you can mask an object in one frame in a video and it will replicate that (with varying success) to the rest of it. For 10 seconds of video, otherwise, that would be hundreds of images to individually mask, and it would be totally inconsistent. On a smaller scale is photoshop's foreground/background tools.
Market-wise, i think there's a larger variety of freelance jobs available in may art realms, but they're less reliable and pay worse-- even in graphic design. From the little experience I have in audio, compared to what I do, I think it takes more experience and skill to create a commercially usable lower-end audio deliverable and this can't do it yet, but the market can support far fewer people than the visual market(s) can so it's at even greater risk.
In visual art, lots of businesses are having generative AI make base assets and having their artists essentially clean them up to be professionally usable, which is about as soulless and shitty of a job as you can get. The end-user tools like Canva aren't much of a direct threat in the long-term to graphic designers because the use cases are usually a lot different, but my (less informed) gut says the real threat to commercial musicians is with products like Garage Band. I think there will be enough control to satisfy some art director's use case, and the output will eventually get high enough quality to hand a bunch of tracks to indie producers to "clean up" rather than making complete things, and that output will progressively get better and persistently decimate the market until only high-end specialists remain.
What knowledge workers fear about AI already happened to musicians years ago. There’s a reason that the vast majority of musicians have to work a different job. This new tool will not make much difference to musicians, who are already economically marginalized.
We'll be where we are at now: selection because of genre preferences, quality, fame, marketing.
How could they charge a high price when the supply of competition is so abundant?
https://app.suno.ai/song/d2705c66-4be5-496c-add1-480427b4a00...
https://app.suno.ai/song/5d37f5e7-3e62-4df8-a1d8-67122470aef...
https://app.suno.ai/song/bd453dd4-f878-4a79-abdd-b11663a18e7...
https://app.suno.ai/song/d5151822-4b77-4c42-9e61-88511ea0333...
Likely takeup by musicians: close to zero, although a few singers will use it for ready-made songs/backing tracks that they can warble over, and a very few of those will be good or distinctive as singers. The songs themselves won't stick because the output won't be copyrightable; talented singers will learn to write their own songs or get partnered with people who have a track record in the relevant genre.
All of these music AI products miss the mark because they generate mixes, instead of stems or MIDI. There's already a superfluity of music; nobody wants an infinity jukebox full of songs they don't know and have to learn. While the technical achievement is very impressive, it's notable that none of the founders in this story identify as musicians or exhibit any idea how to cater to them.
Fully AI (including vocal synthesis) may have a few individual breakouts but will be limited by two factors. One, most people using such tools will want something easy rather than challenging, so there will be more and more lowest-common-denominator dreck - already a problem in human-generated music. And I'm sorry, but that's what Suno churns out. I tried half a dozen prompts on a few different genres that I used to DJ and the results where all instant needle skips. I hope v3 lets you specify instrumental/no vocal because v2 just ignores that and churns out vapid filler. It's very technically impressive, I don't want to claim otherwise. But it's also utterly uninteresting.
Two, live performances won't be possible, and the world only has room for so many novelty acts like Hatsune Miku. The biggest market for virtual singers will be kids under 10 who want easily digestible music and don't value assortativity, but don't have much discretionary spending power and won't spend big on childhood nostalgia later.
Edit (based partly on others' comments): while I think this sort of tool is bad for creating interesting music, it's a fantastic core model for a game of Record Label Simulator where you play the person that writes big checks and you try to wheel and deal your stable of eager young talent to the top of the charts.
Stores? If you want people to get out of your store, sure. Store misic is already beyond annoying.
Supermarket music isn't meant to be listened to, it's just meant to cover the echoes of industrial cooling hardware and people walking around in a large space. It usually sounds bad because some marketing drone picks the most vapid pop available on the basis of its proven mass appeal. Instrumental music would work better.
It's great hearing the dreadful output of this infernal contraption named "Suno" and being reminded specifically of this bittersweet time in my youth, but also broadly of how difficult it is to make memorable, sweet music, and how precious that can be.
If you trawl YouTube for acoustic covers of rock songs or whatever, it should quickly become apparent that even just making a different arrangement of a song is an incredibly difficult, complex, creative endeavor.
There's mountains of amateur garbage and then there's, like, Molly Tuttle.
Picked up my box of 500 CDs from the mail distributors sent us each week and had to find new music every week
It was a challenge and honestly not much fun cause as you said its frankly a lot of junk
However that’s all art - for every “let it be” there are 100,000 songs produced and released by artists for themselves
What was clear to me was not that there wasn’t good music out there, it’s that there was a surprising amount of it that wasn’t popular.
So what becomes a “hit” or whatever has less to do with the actual music and artist. It’s all to do with marketing. Like almost 100%
A mediocre artist with a gang of writers and producers will make more “impact” than anything
The entire K-Pop industry proves this
Let’s not confuse art for marketing even if they use the same medium
Not possible without a lot of skipping songs.
Most tracks are 3 minutes long.. so given an average of lets say 3 songs a cd at 3 minutes each... 100 x 3 x 3 / 60 == 15 hours of listening (hand wavey no time to change cd and stuff...). So yeah probably pushing the limits of reasonable but not quite as crazy as you might think.
Also, although it was unfair to the artist, I would usually make up my mind on pop songs by the end of the first chorus. If I'm not hooked within 1 minute it's probably not going to play on my show. It was brain numbing work, and frankly made me view music as a chore for a few years afterwards... but you can listen to quite a few tracks this way and try and find the few gems amongst the chaff so to speak.
Actually you skip within the song…to do this well you listen to about 20-30 seconds of each song focused on intro, bridge, outro. Most songs aren’t air worthy so you’ll know in 5 seconds it’s a skip
It was a few hours a day
After months of this you get it down to a science
As other posters said, the CDs have stickers indicating the tracks they want you to play so sometimes I’d just play those if I was feeling lazy
But I also find it an idiotic attempt. It doesn't inform us about anything, and it makes any idiot with a mouse think they can create. Instead, they're listening to music that's worse than what YouTube has to offer.
Can you give some examples? In my experience, most music, even music we think of as banal or boring, had a pretty significant amount of effort, and creativity, put into its creation.
What's worrying is that even e.g. amateur film makers can now access something like Suno, and won't be offering their amateur musician friends an opportunity to learn. And once the incentive to learn and improve has gone, people will stop and music will wither.
I mean that’s where AI shines for me, letting 1 person (or few people) accomplish what would have required teams of skilled people to do. It’s the same story with tech in general. I write code that runs on web/android/iOS that has processed over a half a million dollars through it as a single-person company. That was all pre-AI, my point being that technology and the tools it produces allow us to do amazing things. I see AI as only furthering that.
Now I can create custom sounds using AI, let AI help with marketing copy that I’m not good at, have it generate cartoon/simple pictures to explain a process for my software, etc. A world limited not by my skillset but by my imagination.
I know people hate electron apps but the alternative is often no app (not a native app), in the same vein tools like this will allow people who can write lyrics but not music to share their songs. It won’t be for everyone but there is value in it.
There's a thought and intentionality to what's being created — I don't want what comes out of a prompt.
I want to follow artists and their output, not an algorithm or AI that mimics ones and is trained by questionably feeding on their output.
No, none of this is appealing. I can't control what anyone else does, but for me? No, no thanks.
The vast majority of people will just use a different AI tool to generate the lyrics, use Suno to make the music, and then use midjourney to generate the album art. And then they'll throw a tasteless post up on Instagram about how they "made a song".
In my mind this includes SUVs, cheap plastic crap, anything and everything disposable, stupid decisions like open/glass doored fridges & freezers in supermarkets, etc. We're allowing the environment to be destroyed so that we can buy a cheap plastic whatever that we throw away after less than a year.
What have you used and recommend for this? Gpt4?
What would you think of the piece if the artist was a young girl? A grown Turkish man? Someone who is deaf? A neo-Nazi Holocaust denier? A rich businessman using AI and just looking to make more money? A poor teenager using a computer at their library?
I think most people’s like/dislike/perception of the song would be colored by who made it and how it was made.
You can’t always separate the art from the artist.
https://app.suno.ai/song/67dac18c-973c-44b1-be89-96260a0ada9...
Might be effective PR, but it’s bad advertising.
https://app.suno.ai/song/ebe24f8b-f50f-43f3-8062-2c081d7c9dc...
(this is the old v2 algorithm available to free users)
most genres are more formulaic than people want to believe.
We’re destroying incentives at a rate of knots, oh well…
Similar to how with stable diffusion I can ask for a creature that doesn't exist.
Most of human creativity is just evolution is existing stuff after all. Look at all fantasy creatures, they're all based in part off real animals.
You can literally invent 4 genres in parallel in 10 seconds.
The biggest crime is copyrighting the best intro chord sequences in the 1960's, and since then everyone has to make slight variations or they would be called out for copying. We can't make good music today due to copyrights...
They really messed up then, because that's literally all I can see using this tool for.
We're not trying replace artists, but we're going sell a product whose only conceivable function is to more cheaply do the things artists do, and if they get replaced, it's not our fault!
I just can't get enough of that grey, training-data flavoured content. In years to come many of the top creators of this type of content will be creating content with this. This will democratise creation of this type of content for amateur content creators who will be creating wholly unique pieces of this content in seconds. How many years until we see a larger, personalised piece of this content created with this content generator in the place where people consume larger pieces of this content? Maybe this is actually how our brains work? But what _is_ intelligence? Made you think. Sure, it sucks now, but zero to one to infinity is how things work, it's going to be amazing. The letters are starting to blur together. Please invest.
https://open.spotify.com/episode/2c1yL8hlttlkCs6nPysVi0?si=O...
I find it interesting that at https://suno.ai the prompt "make a song for your workout" IMO failed. It's a nice song but it's not a song I'd think of as "for your workout".
Not as amazing because it's only providing the singing but I've been pretty impressed with the Synthesizer V demos.
https://www.youtube.com/watch?v=ugK78gC12wc
A Warrior's Dream: https://www.dropbox.com/scl/fi/sxe3ig5zynrd3pcs3oy8a/a-warri...
Ollie, the Orange Tabby: https://www.dropbox.com/scl/fi/amfudnsccpvl15arjurz3/Ollie-t...
the first one I made by describing an experience I had in an online text-based game that is now offline, long live Clok. Cogg is a good successor though. Anyway, GPT-4-turbo did pretty good making them into lyrics, besides the "profoundly floored" part; that was rather cheesy. I still like it though.
The second was for my cat that I used to have. I then tried out an AI Remaster tool. I kinda liked it better unmastered now, but ah well. It still reminds me of him and makes me happy when I hear it.
These tools, I think, should really be used for simple things like this, to give voice to personal experiences like this, for people who clearly have no musical talent like me, or to help those who do to jumpstart the creative process if that's what they need. People selling these songs... Well, I can't even fault them since it's hard to make any kind of money these days.
I would be interested in a tool that was trained exclusively on higher quality material, like a Richard Rodgers or Cole Porter AI...
https://app.suno.ai/song/7decbf5b-bd1b-4e13-8ed5-aa8c7a38b1b...
Lyrics are half generated and half hand-written.
It's a shame that I didn't have any control over which vocalist sang which part, and just kept having to re-roll the dice! In general it offers very little control over the performance.
I do wonder how much of this stuff is going to flood Spotify and co. and make payouts to actual artists even smaller.
I think the biggest problem with AI generated stuff is that the I in AI is hype, these things can't create real novelty. They're a new, poorly understood form of compression. If creativity is left to these AIs then the music will sound the same in 20 years at best, or more likely, degrade/converge when they're all trained on each other's and their own material.
There's a lot you're glossing over here. That's the part that AI simply can't do.
> If you are a paying subscriber to Suno, then you own the songs you generate while subscribed to Pro or Premier.
So, as soon as I stop subscribing, I no longer “own” the songs? What if they go under? That’s nonsense.
EDIT: I was wrong, that isn’t how it works
https://app.suno.ai/song/5e787b34-0c3a-4790-b88a-bcd54f4f078...
My fucking god, I'm really disgusted by "journalists" who hype up everything so that they can get a few clicks. Journalism ought to be critical, objective and investigative.
My god, these people are so melodramatic. If I could roll my eyes any harder they'd fall out of my head.
I got a female singer songwriter playing a downtempo acoustic song singing words describing my prompt, ha
Still impressive, but not "chatgpt for music".