Am I better at music than AI?
Vložit
- čas přidán 12. 07. 2024
- Yep, I'm the best at this.
Some of my predictions were inspired by this fascinating talk by Maggie Appleton - • The Expanding Dark For...
Some of my courses are on sale this spring!
Getting Songs Done - benlevinmusicschool.com/p/get...
Making Music for Yourself - benlevinmusicschool.com/p/mak...
Guitrongus - benlevinmusicschool.com/p/gui... - Krátké a kreslené filmy
I am an authentically human persona. This comment is real and organic.
Great video Ben! I love you!
I love you! Hope you're having a nice morning!
then why don't u marry him if u love him so much
@@elliotlangford824 I did!! And I would again 10/10!
@@JusticeCow You two are beyond adorable. I've been following both of your musical output for years now....and just adore that you two are together. Couple goals indeed. I have someone in my life right now I feel _might_ be that person. It's scary...cuz it's SO intimate. That scares me. But it's also inspiring to see you two...seeing that love in action. idk....I'm rambling lol. You both are the coolest. That's all!
The problem with AI is that it's a black box. A program has rules and logic for making decisions which a programmer can review to determine how or why a particular result is produced, but an AI has a network of equations generating their results that draw on statistical inferences about the training data. It's not really possible to look inside and see whose art they're drawing from to generate something. This is also why bias is such a problem in AI and the only way to resolve it is to try and unbias the training data beforehand. But even if you remove the obvious dimensions, the AI may end up biased anyways by finding highly correlated attributes. Copyrighted works should flat-out be left out of the training data unless the company has specifically requested and received permission to use it. Fair use never considered the sheer scale and capabilities of AI and should not be used as cover for corporations to use it to train their models for commercial benefit.
Generating good outcome instead of bad is also dependent on bias, bias of AI is not a bug, it's a feature. If I want to fine-tune AI to generate outputs in a certain style I want, I need it to be extremely biased to that particular type of output, AI that produces totally non-biased output is producing white noise, not very useful AI.
This has to be some of the most insightful music related discussion right now with the most fun visuals to go along with them. Love these always.
I spent a couple years making music with OpenAI's Jukebox (a now-old research project), and what was actually fun with it was coming up with interesting prompts (audio snippets, not text) to start the music, and then finding outputs that were broken in interesting ways, to then assemble into a song. I think even with AI that generates raw audio of a full song, it's boring to just churn things out. Once this stuff starts to feel normal and the novelty wears off, I think a lot of incredibly creative work will come out of artists being artists and breaking AI algorithms in interesting ways to create new sounds.
I wish Jukebox was still working.
Just want to say here, since you mentioned it - Liminal Onion is one of the most transcendent pieces of art, and I think about it often. I don't know what you did with it, but it reached me on so many levels that I can barely begin to understand what it made me feel. It's an experience unlike any other. Art like that would be impossible for AI to construct, because it deals with such an abstract representation of complex human emotions. We can only hope the future is full of more unbelievably unique artists such as yourself, to counteract the stale mundanity that will be AI generated beige music
I love hearing from people who really get what I was going for with that piece! Thank you so much for giving it a real chance!
What you said around 6:10-6:20 is a part of transcendental philosophy, "Speak what you think today in hard words and tomorrow speak what tomorrow thinks in hard words again, though it contradict every thing you said today" - Ralph Waldo Emerson: Self Reliance
Still hoping to see a Ben Levin Magic: the Gathering video! Love the display over the piano. ;)
Wonderful thoughts Ben! Thank you. ❤
Really great to hear your perspective on these issues, its more thoughtful and informed than anything else I've seen lately.
Thanks Ben. A good perspective. And we‘re in need of those
A lot of my instagram music suggestions are surely AI creations: The chord progression doesn't make sense at first but repetition makes it clear...
They even ramp the tuning and tempo, like you mentionned @2:13
Hineni, June, and Darla: A combo that AI could NEVER generate.
Good thoughts! I like your analysis.
Man you're so insightful with AI learning, I've come across the same thing and have wondered how we're going to make music in the future to know that its real. Great Video.
Your animations are so good
You're da best Ben!
amazing insight!
Best video about AI I’ve watched and best video I watched generally
"I would want the AI to be able to generate sounds I could never, ever, make on my own"
but then you did make those sounds! haha
Just discovered your channel, you're really interesting! Definitely got a cool brain. Love it
Yeee let's go Ben. Congratulations on being a better musician than AI!
absolutely amazing take on AI btw! super resonated with everything you said. Also never thought how pretty much we are all heavily influenced by the algorithm in almost every creative choice we make today.
We want human perspective on things. Otherwise it just doesn't matter.
Ya! Nice!
Lmfao 8:30 is so funny. and yeah, I want to explore with sounds that I Digitally or Physically can't make.
suno makes country that sounds like country radio hits - lol
The Maggie Appleton crossover that I never thought would happen!
this is cool
nice slivers bro
Superbowl level commercial, Ben 🙇♂️
Can’t wait to see an Improv-AI-generated live set. Artists may become so skilled at prompts that they can generate entirely new music on the fly in a live setting.. even THEY don’t know what the set is each night!
I love how comically long your microphone has gotten.
Ohhh, the old layout of yt is back! First video i clicked on in about half an hour and my wishes have been granted! Was this all mixed up on your end as well or just for me? The thumbnails huge af and right under the video? And the comments on the right where the recommended videos used to be?
I know OT, but i am so glad this is back, looked just horrible.
Damn, i guess it is not. Videos i click on opening in a new tab open in the new, awful layout, yours, as i clicked on by accident on my subscriptions page opened up in the old, cool layout. Sorry for bothering you with this, i know this probably seems petty but i do care about this crap.
Ben is that a Leviathan from the Dark on yr piano? What are the significance of the 16 cards you've framed?
They are cards that I had when I was a little kid that I liked the most because they look cool!
@@BenLevin Man you were right!
0:38 DUDE THIS IS FIREE🔥🔥🔥🔥🔥
Music Memes is a good point, same for Microgenres/Anti-Genre.
Also, Live Music definitely matters,
I would love AI assisted sound design. I personally struggle sometimes with translating a sound in my head into something I can work with on a VST like Massive or Serum. I also love anti algorithm ben videos.
nap and try to forget 😭😭😭
While machine-learned music/beats seems stale and unoriginal (and usually objectively bad), I have been experimenting with machine-generated *sounds* to find some specific samples, but they usually have a very filtered and processed sound. And ironically, the further from 'western music' you dive, the less authentic they usually are (like asking for taiko drum shots and getting clearly synth-generated sounds).
A nice vst I recently found called Visco lets you morph between drum samples (or any sample but it's length limit keeps it to one shots) in really interesting and creative ways. It has a really lenient trial/demo with just a timed splash screen (pro tip: keep the window open to avoid it)
I played around with Suno the other day and thought it was pretty interesting. I think it produces ear-candy that sounds fine for people that want to consume music as a product, but is ultimately empty for people that want to consume music as an artistic expression of humanity. Very similar to my thoughts on image generators.
I'm having some trouble formulating my full thoughts into words, but basically, I think things are gonna be fine.
You're probably right, but I'm uncomfortable with how good it is. What will it be capable of 5-10 years from now.
Ben, you are an incredible individual with tons of talent, but those animated doll things are CREEPY
I can see what you mean, other people feel that way about the cuties too!
nice
what a time to be alive
Welcome to 2 minute papers
i think youtubers playing music and filming themselves whilst playing, could become a thing in the distant future, as a way to prove they aern't random AI. Maybe you should try that Ben.. Food for thought huh?
I think it's very important to always mark your art as being AI-generated or parts of it being made by an AI.
What prompt would you recommend as input for suno to make a ben levin song?
For the next video, see if you can get a bigger mic. 🤣And nice use of deep plosives for headphone users.😖
If this doesn’t go viral I’ll eat my hat
We need to do more wackjob crazy things like you said about Autotune or tempo. Take Zaireeka for example, or a song using all English words, or a song using every single chord. We need human courage and passion in music.
Making music is even more of a niche
AI will never be able to create the story behind the music, music is emotion, and that, AI won't ever generate
this will change music, in both bad and good ways. i personally think its ok to use ai to create singular sounds (like a single kickdrum sound)
Dude I'm so with you on the man-in-refrigerator-dropped-into-pool-of-unharmed-goats sound generator! Primary input format: Make it happen, Elon.
MrBenLevin
As an AI chatbot, I don't know what you're talking about.
Edit: I actually typed this.
This shit seriously scares me and depresses me, but i do think it will help me insofar as I've always been very original and I think the incestuous pablum of AI music will make originality more appealing, at least for a while, until AI tools get so great you can work with them with as much precision as a DAW - which i would be ok with in some instances. I do have a gajillion songs I want to finish and can hear in my head, thousands of phone dictations and half baked lyrics, but the idea of producing them fully really intimidates me and would take so much longer than I'm willing to put in. It makes me very selective about what songs I finish, so its appealing in that sense to expedite an authentic creative process.
Ben you are our lord and savior
lol af 😂IRL RN 👍
Disclaimer: no goats were harmed during the making of this video.
2:28 There's a streamer called quin69 who has AI songs incorporated into his stream (you can use channel points and bits to generate an AI song), and I've already witnessed this being done by an AI. Probably not on purpose, but there have been instances where it will shift from being in tune to being completely out of tune from word to word.
That's super interesting, I like that as a use for generating songs. Just like using emojis or something.
@@BenLevin Well they're mostly used for calling Quin fat and bald, which might not be as interesting (definitely hilarious), but the technology is fascinating.
A.I. works by averaging out a collection of data. It's been proven that data can be corrupted in the field of image generation, so we might have some control over what data ultimately ends up in the hands of A.I. researchers. The difficult thing is that A.I. oversaturation could be a very simple way to censor an entire genre of music, making people jump ship to something else where there isn't enough data available. Music that was made today will become very, very rare. The value of everything you could possibly do right now will appreciate significantly, so make a bunch of cool stuff. Having the date and time next to anything made today that has just enough randomness an A.I. couldn't have thought of it is going to make your current actions a rare collectable.
In a way, nothing has changed or will change. Art that was made a decade ago with just enough randomness that a corporate algorithm working with the prompt "Art" couldn't have made has had its value appreciated. This comment is going to be very dated, very fast, but I see a pattern here that seems like it's going to continue. For instance: There is a trend of people trying to replicate micro genres for content on CZcams as a way of saying "I want to show you how easy this is." The thing is about those videos is they end up with pretty good approximates sometimes, but they almost never come back to making that kind of music again. If they made it again and again and again, they simply wouldn't be who they are. A single approximate, then they get bored.
A lot of the time they don't have the equipment available for a full recreation, so they wing it with VST plugins. They listen to the songs to figure out their instruments sometimes, sure, but a lot of the time they're probably working off of something someone else wrote about it on some obscure forum online. To face the music: There are songs where it is fundamentally impossible to tell what equipment was used to record it unless the person who made it were to tell you. If someone were to have made an approximate without knowing the intricacies of the equipment used, it's going to be like a cheated speedrun of Mario 64 that comes out just before someone finds a tech to always know when someone spliced their clips.
To add to that simile, in Mario 64 there's been a discovery that Mario always will blink after a certain amount of frames. If Mario blinks just a frame early or too late at any point in a posted speedrun of the game then the run is known to be spliced. Someone could hypothetically go through the effort of editing their run so that the blinks sync up perfectly, but that kind of effort is fundamentally uncharacteristic of a cheater. If they'd do it then they'd probably only do it once, and then they'd go on to do something else with their time. In the case of music, some frequency analysis could be performed on the equipment to verify the human touch of a work, appreciating its value significantly.
Ironically we might see one day an algorithm that sorts out everything non-verified to be human from people's playlists online. I have a lot of public domain music out there, so it's probably going to be easy to train a model on me. I've done a lot of random stuff over the years so it'd be neat to find out what the average of all of it would sound like. It wasn't posted to the public domain with the intention of allowing people to train A.I. models on my work, more for people of the future who are better at music than me who can post their better versions of my songs. More like a license to dream kind of thing than anything else, considering nobody's listening while I'm here to know about it anyways. Anyways, chaos will ensue. We'll see live renditions of A.I. compositions.
That AI music is better then a lot of popular music these days.
'intellectual property' laws should just go. they should have probably 30 years ago already, but hey better late than never. otherwise, having thought about all of this, i came to the conclusions much similar to yours
I have a suno track that blew me away, creatively. In AI's current iteration, it will mostly be used by companies and pop music.
But this track, one particular design choice is so good, so novel. I don't know how it did it. I can't imagine the models it used.
Like all all things, human creations as well, the key is to brut force the process. By creating 100s of pieces, you can select the very best for presentation.
Most will be average, but in the hundreds there is one that is exceptional.
Also, I have be working on an human album. Most tracks when finish have a fatal flaw or are best for personal enjoyment.
It is hard not to post some of the tracks that are especially interesting, but you have to try and maintain general aesthetics.
It is wild how much good music will go unheard. But that is our blessing as creators. We get to explore everything while the audience is exposed to a much smaller experiential universe.
Humans will always be better than AI at making music because they're actually making it
People are alarmed at AI as it is (and to an extent they may be right to be) but the current AI tech is limited to only reacting to prompts in a very complex but essentially merely knee-jerk manner. The software is still reliant on both the conceptual understanding of creators of their training materials and the conceptual understanding of the prompt-giver. Without both of those sets of humans, AI as it is now can only spit out random directionless garbage.
But the whole landscape will change if we can create an AI that actually understands concepts. Right now we can train an AI to recognize what things are and are not (for example) a dog. The revolution happens when the AI is able to recognize *why* things are and are not a dog. At that point, the real replacing of human work and creation will explode, as the AI gets exponentially better, faster, *and* cheaper than a human doing the same thing. An AI with an understanding of concepts will be able to employ all the strategies you posited, plus develop others faster than people can, to make novel and meaningful works of art.
The big question for artists and other human workers now is whether we, in any sort of foreseeable future, can get AI over the hump from reaction to contemplation.
This made me kind of optimistic. I don't think there's anything inherently wrong about generative AI making art, so long as there's still space for human artists. The issue is that capitalism will privilege the AI art because it's cheaper and less messy, and that hurts artists and art generally. In a world where that's not the case, an AI that can make better music than me doesn't really phase me. A lot of people/things make better music than me!
From what I can tell, capitalism really doesn't move away from expensive and messy. Just look at film production. It's a bit overly simplistic to view a capitalist economy as if its being driven by some optimized algorithm that seeks max quantity for minimal cost as opposed to the myriad individual desires of those with economic sway. AI's role will not be easily summarized as good or bad.
Fascinating takes, Ben. I personally see AI pushes music forward instead of destroying it. Listening to AI created songs feels like visiting a WordPress website using the default template. Sure, it gets the information across but it's forgettable. The folks who use AI in the most creative ways will bubble to the top (optimism!).
It's certainly possible that good things will happen!
Nobody ever calls me Al 😉
I can't really see AI replacing musicians in the way I listen to it. For a start I rarely listen to anything made after 1975, or before 1965. But I'm just not interested in listening to AI music, because I don't care about execution, I only care about the ideas, made by humans.
what is this ap
Bach was a greater musician than I ever will be, but that's not a reason not to make music anymore. So AI is also a greater musician than I am, but I still want to express myself through music. - no matter how bad it sounds... czcams.com/video/KikXazdfgG8/video.html
holy shit plastic pills ?!?!
I don't even need to watch. AI could never create Our Place, or Freak Machine, or Sea Stories, Invisible Paradise, Pulse, People, etc.
Idk, i think wer far off from ai making beach life in death
7:22 what? Wait. Hold on... is this true? I haven't paid anything yet. Is there a lifetime key option, or subscription only? Geez. How much, again? Is there a demo, or a free/lite version? I have this feeling that's the one I signed up for. Why didn't anybody else tell me we have to pay money to live?
Oh, you said MAKE money... Sure no prob, I have a printer. Anyone have directions to mod it so it will copy USD? IDKY I need 2 do that in order to live... but I want to be a team player. Which denomination bill is easiest to make? I guess probably $1 bills, but I like the dream big.
In 10 years, AI will be so advanced that it will sound more organic and human than music does now. To show we're human, we'll make music as predictable and lifeless as possible.
Having a surplus of information will never by itself create a sense of soul. AI can predict the future and we still won't be as compelled by its artistic output as we are by humans. (Just my belief)
@@ScarlettTheViewer What if "soul" IS the information and we've just been passing it along the whole time?
@@binarymystic Maybe soul is a type of information, but I don't believe it's ever going to be present in AI. You don't quantify or reinforce soul because you can't precisely define what it means to play with feeling or get in the groove, so we can't teach it directly. We could feed it tons of examples, but then it's just hearing raw data with no capacity to determine context and value what each component is.
“What I really DON'T like is when people use other artists’ voices in their music without their permission...”
I really like David Sylvian’s singing voice. When I imagine lyrics over my music, I often imagine them in his voice. I feel his vocal timbre would be The Correct One to use, the Ideal, if I wanted my music to sound Correct to me... which I do, as much as I can manage. If someone put his voice up somewhere as a free model, I would gladly take advantage.
I don't want to bother him personally. I don't have money to hire him, if that would even be an option, and wouldn't make him work freely/cheaply, especially considering I'm me. I'm picky, slow, inefficient, would struggle to articulate what exactly I want from him, usually struggle to see projects through to the end, and would want to use him kind of a lot for silly little things.
I could theoretically gain money, and the right kinds of experience, and connections, and work on my productivity, and carve out a possible future where I could work with him, but I have projects I'd prefer to have his voice in *now,* and tomorrow I could read that he's been executed by firing squad for one too many parking violations.
I also happen to be severely physically disabled, and have practically never left the bed I'm writing this in for the past eleven years, save for being wheeled out of the house on a gurney for a couple ER visits and appointments with doctors over the years.
My body sucks. A lot of why it sucks is because I had a pituitary tumor. One effect it had was a delayed/interrupted puberty, carried out later by replacement hormones. As such, my voice did not develop cleanly. My range is too narrow to sing the melodies I've written lyrics for. It awkwardly cracks off in and out of falsetto. Trying to force it to work for me in precise and unnatural ways quickly becomes physically painful. With time and effort, it might be possible to improve it anyway, but...
I am in *Bad* physical condition. I don't feel like I can safely assume I have enough time on Earth remaining to master any given skill I still want to. Especially given that I often feel too tired and empty to do much of anything. And covid is still very spooky.
I feel like not having David Sylvian’s voice is just yet another case of being limited by the particular body I'm forced to inhabit. If I happened to luck out and possess a body with vocal cords close enough to his to effectively imitate him, I could “steal” his voice as much as I wanted, because it would happen to also be mine anyway.
If he never consents to letting his voice be turned into a model, but someone who happens to sound exactly like him *does*, and offers up their voice for free, is there a practical difference?
I just want *the right instrument.* It's like I want an accordion, and what I've got on hand is a tin whistle, and if I hunt around I might find a harmonica. Other voices aren't suitable replacements.
If someone scraped my voice from live streams, and somehow managed to wring a good singing model out of that, I wouldn't feel deserving of royalties. I'd want it noted, for the benefit of listeners who'd appreciate knowing what to search for if they liked the sound of it and wanted to hear other uses of it, or if they'd want to use it themselves. Even if I had trained my voice well over the years, and recorded hours of clean a capella that had been scraped, that would have just been Something I Did, with an ability I could and likely would employ near-passively. You sampling my vocal timbre would be taking even less creative input from me than in the case of an unauthorized sample of my music. I just would have happened to have held a piece of your puzzle, that I couldn't have anticipated you would need to use in any of the exact artistic ways that you would. I as an artist would only incidentally be a part of your art, having no direct involvement with your process. Much of the particularity of my timbre would still have been beyond my control, anyway.
AI tools have been valuable artistic assets to me. Besides my voice being the way it is, I have a terrible recording environment. Constant background noise from appliances, seven cats doing cat things, people needing to sleep. My mics aren't great either. *None of that matters so much now.* A low-quality, noisy, half-whispered, masculine vocal recording can be converted into a strong, clean, feminine one, in a range I can't normally come close to singing in, without sacrificing subtleties of articulation or relative pitch, for free, in seconds. This is wonderful. This is enabling. This allows art to be made.
I've probably generated well over 20,000 images with AI. Before AI image generation, I wouldn't have even considered myself all that *into* static (vs. animated) visual art. Being able to specify what I'd like to see, receive it almost instantly, and rapidly iterate upon it has gotten me interested to a degree I can't remember ever feeling. It has, in fact, inspired me to take my neglected drawing skills more seriously. Being able to see things I've wanted to see, for which next-to-no specific pre-existing reference images may have existed, has made them easier to conceptualize in my mind’s eye, and unlocked understandings of certain concepts, such that I feel more capable of actually being able to draw them. I now feel like doing my own album art, and doing it well, is something I can realistically strive for.
There are many conversations being had, and to have still, about the nature of art, and how it all factors into a world with generative AI; conversations about the “redundancy” of human labor, the meaning of creating art, the inherent value of an artist and their right to survive under capitalism, repurposement/juxtaposition/discovery of something as a form of “creation”, whether art “belongs” more to its creator or those who observe and engage with it, etc.
At the end of the day, the important thing is that people are enabled to create the beautiful art they feel the world needs, and the people who most need that art can access it.
Shhhhh the ai is listening and already stole all of Ben's ideas lol.
I want an ai I can make noises with my mouth into and it figures out what noises I'm trying to make.
I can say "ai, I want a bass synth that goes BEOWWEWW." and it does it. Even if it's not perfect but takes feedback like "that was close but a little more BUVMMMM."
That's a cool idea!
I think there's a VST that does just this
it‘s not fun to “make” music with just one click，and if you can’t even express something with music，what the purpose of making it？I‘m not getting the point of AI
Some people don't find it fun to make music at all. They don't get the point either. Nothing new under the sun.
Adam Savage (tested, MythBusters) recently discussed the excitement of AI but also said he wasn't worried about it as a creative because it doesnt (by design) have a point of view (...so far). To your point if the music has nothing to say is it even really music?
@@jamescole7197 I’m not sure，but theoretically speaking AI doesn‘t have their “sense” or mind I guess， so it’s just gluing elements that exsisted right？people sometimes enjoy dissonances in music and i don‘t think that machines running 0 and 1 could figure out why，the ability of appreciating “wired stuff“ should be a unique trait of human。and after all，it‘s just a tool making stuff that people want it to make，so refer to who is programing it，it can only produce products under specific cognitive‘s effect i guess。no worries
@@jamescole7197 and i
would like to call those AI generated
”music“ as industrial product，as for whether it’s real music or not，leave it to history
🙂
Whatever the human comes up with that AI does not know, it will know it afterwards, copy it and make it even better and hopefully in the near future we won't have any real human artists anymore with their overpriced prices and concert tickets. Bright future ahead as AI will dominate everything!
I feel like we put the dead on pedestals... if I were to pass right now, I'd absolutely love for my art to be incorporated into the machine.
i feel like this is ego, "this is MY art it dies with ME"
as if you didn't get all your knowledge from the models before you.
How can it be ego when these artists weren't alive at a time to even know AI would recreate them in death? I'd also argue that taking a dead person's voice and using it to create whatever you want is not the same as repurposing their art. Yes, you got that knowledge from "the models before you," but being an artist is curating what you put out into the world. I don't think trying to respect the will of the artist is necessarily just ego or putting them on a pedestal.
@@jkdeadite it’s ego from current artists that are applying that sentiment to the dead.
this is why copyrights don’t last forever, people write songs, songs become popular and sprout sub-genres that have to use models from preexisting artists.
At some point, your art is no longer yours, it’s public domain.
Would you have the same argument for sampling music in hip hop?
@@gsharpenmydjencil We're talking about using Tupac's voice. The dude died in 1996. I agree that at some point, it does belong to the people. I don't agree with using their voice right now to make new shit and use his name and voice. I don't think respecting a dead artist has to be all or nothing.
@@jkdeadite it's all just speculation on what a dead person would want, either side, that's where the ego of us comes in.
i'm on the side of freedom of creativity though.
when conventions are challenged that's where we will find beauty.
Your free to allow people to use technology to mimic your voice, but you can’t assume everyone else wants that. It’s just a matter of respect and consent
"making music the old fashioned way"
is this not the same mindset that would've looked down on electric guitars and synthesizers?
everyone's digging their heels in, the more you let go, the more freeing it will be and we can get new music legends.
If it were static, yeah. Artists would have the time to fully explore the medium and adapt, find the coolest ways to use it and its limitations. Some things would be gained, some things would be lost.
But in the real world, it seems like by the time artists get a grip of the AI possibilities, a new version will come out. Dizzying acceleration until we stop one way or another.
@@Fiddler1990 the same can be said for musical instruments.
I’m not really sure what the issue is though?
Music is a tool of expression, which we use other tools to craft that expression with the end goal of sharing and making an emotional impact.
Music/art is one place where the ends justify the means.
Who cares what tools were used to make an impact emotionally?
Over history, new tools for music have come out (AI, new instruments, new styles, etc)
and those who don’t ADAPT get left behind with the times.
Ai is different, you're generating a whole song out of nothing.
@@simonsanchezkumrich8489 “out of nothing”…
Why is AI different?
AI (in its current paradigm) is different because it’s based in stealing peoples work, recycling it and selling it as something new.
Udio is better than Suno
I hate how everyone is entertaining AI...couldn't think of a bigger waste of time...