"it's already a problem so why try to address it or stop it" is not the enlightened position you think it is.

While I agree with you, I also feel the same way he does about pop music. ​@@riffsnoleadsAI generated music, chat gpt, etc. currently cannot beat lyricists but in time they will. I still do not fear it. This discussion on the effects of it have been around in Japan forever. Watch Carole and Tuesday which predated AI generated music. It is going to have a huge impact and a lot of musicians will lose their gigs. Every music label will have their own AI Selena Gomez/Ariana grande clone to start.

@@riffsnoleads No, the point is that you’re not lamenting the loss of anything important or protecting anything of value. I would prefer that all art is created by humans but that’s not going to be the reality moving forward. Good music will still be made by humans and will be sought out and discovered by people who care about music. Which is currently the state of music. It just won’t be a lucrative business or an art form that can change culture. For most of human history music didn’t have that influence and in the future it won’t either. 1940-1995 was a unique period in human history and it’s passed. Unfortunately, popular music is no longer a force for cultural or political change.

Damnit I wish you weren’t right but damnit I think you are haha

Because other genres don't.

Man is this accurate! So much music consumption is driven by other factors.

People that don't care, don't buy music in the first place.

@@brukernavn3409 - oh yes they do, and they are the largest demographic. That is why the companies put out the same tired crap over and over. People buy what the marketing tells them to buy.

A lot of peer groups and niches will have their own music.

Well said

This is my exact concern.

Just accept your work has no commercial value anymore.

Unfortunately a lot of people will go the way of coachmen. When cars took over from horse-drawn carriages, those who couldn't adapt had to change sector. I'm not saying this is good or bad, just that it's what always happen. However, what I fear is that the speed of change will accelerate to the point where people from multiple sectors in the economy gets left behind all of a sudden, faster than we can adapt. Will we have wide-spread social upheaval? Do we need some sort of Universal Basic Income when loads of people will be unable to find a job? Time will tell.

The TV producer will likely have a prompt engineer they use. That single engineer though will take the jobs of several professional musicians

I see my place as someone that uses this new technology as a tool. I become the expert that can deliver what the client wants, just much faster. There will be some music editors that will want complete control and will do it themselves, but we are not there yet. The fine control aspect of this is not yet present, nor is the ability to break out individual stems in proper quality. Can't easily export the tracks to MIDI for finer editing. Again... YET. :) A lot of low hanging fruit will be snagged by AI... I've been saying that since GPT came out. There will come a day when AI is just another tool in our aresenal like samples, complex plug-ins like Omnisphere, Falcon, etc.

It will be equally creatively bankrupt. Only difference is that the smaler indie musicians who are making good music now may switch to AI to and then there will be nothing left worth listening to.

Great comment, very insightful. I totally get where you’re coming from! It’s a very interesting topic… I might do a follow up video once I get some more comments rolling in. Thanks for watching, I really appreciate it.

@@YoPaulieMusic PS: I’m also a massive Michael Brecker and David Sanborn fan 🎷

Gave this some thought last night, it truly is a complex topic. The "guy pressing a button" concept is certainly not new. People said the same thing about drum machines and samples. Isn't pressing a button on a complex Omnishpere patch the same thing? The composer didn't create the patch, a programmer did. And what did that programmer use as references? Was it Depeche Mode, or Trent Reznor, or John Cage? Do they need credit or compensation? The way humans learn, the "traditional" methods of learning an instrument involve years of practice. Struggle, sounding like crap, interpreting/creating motion, etc. I think a large part of this is that musicians are frustrated that AI mimics this process at a much faster rate and without the "human" experience. But, I could say the exact same thing about music composers that just drag dots in the piano scroll mode of their DAW, not knowing the how and why behind why music works, while i have spent years studying including a formal music education. Is drag/drop composition "fair" to those of us actually trained in the medium? I could say the same thing about people that sample other recordings, then chop them up... or even worse, I could say the same thing about people that compose soley by stacking loops on top of one another until they find something that they like. That's not music composition, it's puzzle making. However, I get the point that this difference in “learning” does raise questions about fairness and creativity. But I think it's just like the same reaction from painters when photography was invented, or drummers when drum machines came out... we can go back to the invention of the printing presss that was heavily protested by scribes, because scribes controlled the production and distribution of books, therefore controlling who had access to information. Like we botyh agree, this is a very complex issue for sure!

@@YoPaulieMusic Yeah it's possible my "pressing a button" statement was an oversimplification 🤣 I would still argue there's a difference between AI music creation tools that actually create a full production from scratch, and the printing press. But maybe not. Interesting points all around!

Nah. Won't happen much in a legal basis. Highly unlikely when you look at industrial revolutions from before. What legal adjustments safed calligraphers work from the evil influence ot the printing press on their jobs? Or portrait painters on the influence of cameras on their income? 💁🏻‍♂️

The people apologizing for pirating music and such are just cheap and want everything for free but expect to be paid for whatever they happen to do. They are massive hypocrites at best.

There are valid complaints. Hence why Udeo and Suni are getting sued.

@@Tobbx87 I disagree There are no movements to stop cover bands, voice impressionists or artists who copy styles. The only valid complaint would be in regard to deliberately trying to pass off a fake as being authentic and if that was all it was about it would be valid but that is not where this will lead. This will lead to a higher level of censorship and privacy forfeiture.

​@@firesidewithsumgai9932Just thinking aloud... Taking your view a little further...these things always have possible consequences down the line.... So you're fine not owning your look, your face can be used by anyone. Your voice and mannerisms is OK to copy and use and anything you produce in life. That's Identity theft that you're normalising.

If AI copies me, so be it. I really have no issue with anyone copying my look or sounding like me. The issue would only be whether they claimed to be me or claimed to have ownership over specific works of mine. Same with the music. There are artists who have had surgery to mimic famous artists voices. Should they be sued? Is that id theft?

You would care if it cost you money and time.

Glad to have you onboard!

How is humans learning from copyrighted work any different than AI learning from copyrighted work? Do us humans owe compensation or credit or part of our writer's share to every musician that influenced our own creative processes and output? You cannot copyright a kick sample, or a chord progression, or a synth patch. Indeed, a very complex topic, I have little faith that lawmakers will get it right in the near future.

@@YoPaulieMusic I don’t know but do think it currently isn’t a legal difference. I do know VFX companies at this moment do not want to see any AI on a final image, for legal reasons. The argument is not that it’s illegal but that it might be. I don’t see a legal difference between visual and auditory art so would assume it’s a similar situation. I also don’t know how US laws about transformative art apply if an AI song is played in a country where the laws are different. Heck, other countries might even be looking at AI laws right now for all we know.

@@GoranBackmanMusic The funny thing about these industries saying "no" to AI is that in a few years AI will be able to produce content much better than any of these companies anyways, so they are really just accelerating their path to irrelevancy by rejecting it. They think "we don't need AI" but the truth is "AI doesn't need them." The future will be personalized content anyone can make going viral instead of big production companies/labels creating things and pushing them to broad audiences through radio/TV adverts. And good luck outlawing or regulating AI content - you won't even be able to tell what content is AI generated. No matter how you try to fight it, you have to admit this future appears to be inevitable.

@@Redman8086 Agreed, mostly. :) I will add there are numerous AI laws already put forward to governments, and although most people won't be able to tell the difference computers likely will. There are already software for this. Maybe that will change, who knows.. Also, some already existing laws about creative work do differentiate between humans and non-humans so it's too early to say that "transformative content" laws apply to AI, especially when combined with the new laws still getting ironed out. I do think I made a bad example of the kick drum there btw. :) Overall AI is absolutely the future, for better or for worse. New laws will have a marginal effect I imagine.

TBH, I like this analogy very much) Like I said many times in the AI hysteria, you need actual skill to use AI tools well. This is not a "silver bullet" or "one magical button". One thing is to make a 0:30-1:00 piece of a song and whine about it, and a whole new level of pain when you want to create an actual song from it. And at the end, you will get a singular master track with terrible balance. So, here comes your actual skill as a music producer to fine-tune, compress, master, and add whatever magic powder to make that shit into a candy. Learn and adapt, friend :)

Heh. I was there for that.

I will never use this... and I pray there are people who end the use of this kind of replacement of human labor worldwide. This means that the wealthy can create products void of human hands or controls... and then they will copy each other till they forget the human voice and how humans interact... then the sound will revert back to being muddy and base - returning to the origin of sound which is zero.

@@MateoTeos or... you can still... M.A.K.E. music.

​@@someoneacomputer1905 Sure, this is still a super valid skill to be able to create something by yourself. But AI is just a tool that assists you in doing so with a different pipeline. For example: AI can write a story or create a track, a picture, or something else, but only my taste, skills, experience, and tweaks will define how good this output is. This is not as simple as just writing "Do something cool, bruh". Besides, some people can't afford to create or learn something that they like due to lack of time, experience, money, health, or motivation - AI tools help to at least break some barriers in these cases.

it is truly a complex topic, happy to chat with you about it!

In The Mandolorian Luke Skywalker first appeared on a spaceship but it looked odd. People went back and used deepfakes too make it look more like young Luke. Even The Corridor Crew did their own take on it to try and 'correct' it since those working for Disney seemed to have dropped the ball in his rendition.

@@Mopantsu Yeah. Corridor Crew has been kind of on the forefront of showing what deepfakes and other AI techniques can do in film with just a small team and budget.

For a lot of people I'm pretty sure that's the point. Lots of people simply want this technology to not exist anymore, and they believe that the copyright argument is the best way that they can achieve this. I'm usually not one to recommend ascribing hidden ulterior motivations on my outgroup, but I have seen comments with people explicitly saying as much.

Unfortunately the left tends to lean more towards emotional knee jerk responses and that is why society is falling apart at the seams right now.

Your emotions of being a fanboy over AI is an enemy of your critical thinking. You are not considering the impact of jobs. And the people that post about it will create prompt engineer jobs don't understand the tech through their fan boy lenses. AI barely needs them now. I done experiments using a couple AI's many times leaving everying up to them. My only part was pasting things from one to the other. Prompt engineering is a job that won't last more than a year or two before it is obsolete.

@@ronpetersen2317 ​ I wouldn't consider me a 'fanboy' ;) I said I understand media composers like me will lose opportunities because producers and directors will use AI to create underscore. As @52Cues said below, "We are facing our own Phil Tippett/Jurassic Park moment. Do we curl into a ball and lament the end of our industry or do we adapt the tools into our process?"

If appreciating how much AI music creating has improved over the past year, marveling at how a simple text prompt can result in human sounding music makes me an AI fanboy, guilty as charged. Doesn't change the basis of my video though. AI is having a massive impact on all sorts of jobs, regardless of how young/old the job seeker is. Using AI as a tool to help my music library clients find content faster is a benefit to them, and to me. Better to use AI then to be replaced by it. Jobs get lost whenever technology advances. And new jobs are created. Societies adjust. Musicians will still make music. In fact, here's an interesting random thought that just crossed my mind while responded to all of these comments: Maybe AI generated music will flood the market, causing society to love human music more, therefore causing an increase in demand in live music, meaning more clubs will be needed and more live acts will be needed. More festivals will be created to supply the demand of live music. Then people will crave to hear live musicians again, and concert bootlegs will surge, and live "albums" will make a comeback. And then the big music companies will step in and screw it all up again, finding ways to pay us live musicians even less than before!!! :)

@@YoPaulieMusic "Better to use AI then to be replaced by it." Yes.

At this moment yes. In a year.. who knows.

In America from what ive heard of the courts thus far, you cant copyright a single ai image you prompted for. BUT, if you prompt 1000 images and then crop them and place them into a graphic novel in a specific order to tell a story then add speech bubbles, then that graphic novel IS copyrightable, you own the work and creative decisions that went into the placement and structure and story of it. Someone can take each individual image you used and do something else with it. Individual words aren't copyrightable but if you order them in such a way that tells a story, then there you go.

@@luminousdragon that makes sense! But to follow up on a point in the video, how would one be able to tell if the art was AI generated?

In the United States I believe a court has ruled that you cannot copyright AI-generated material. BUT... what if a musician creates a metal instrumental track, then reproduces it using real instruments. Since the track is not copyrightable, does it become copyrightable when a human re-creates it? In other words, is a "cover" of an AI-track truly a cover, or a new composition since it was (re)created by a human? Man... this makes my head hurt!

The way the law has been formed thus far from what I understand is that fully AI generated material is not copyrightable, however if an artist or user were to add to an AI generated track any type of human input raging from their own lyrics to extra tracks to simply rearranging the composition then whatever alteration they have applied to the track is then copyrightable. Furthermore using AI as a tool to aid in the creation of fully human made music is plausible for filing copyright. So let's suppose you like the lyrics that an AI program has generated or you like the beat.,. You can then go and recreate that and it is yours because you made it... It would technically be transformative enough to be considered original through influence.

@@YoPaulieMusic As I understand it, there are actually separate copyrights for the composition (what came out of the AI, prompted by the human) and the actual sound recording (your human cover). So I think in your hypothetical, the copyright of the composition would depend on what country's copyright laws apply to you--in the US, the copyright office has said that fully-AI-generated things unmodified by humans are not works of human authorship and therefore are not copyrightable. But the sound recording you make of the song as a human musician is copyrightable (but note, I'm not a lawyer so this could be wrong). That said, I believe the stance of the US courts is wrong and probably will be overturned eventually. Think about it. The big copyright holders--Disney and the like.--are going to want to use AI-generated music, images, video, etc. and they aren't going to want to have to needlessly modify perfectly usable AI-output just to satisfy the US copyright office. They will eventually put enormous pressure on the courts to make AI-generated material copyrightable by the person prompting the AI. Since the person prompting the AI for Disney will be employed to do so as work-for-hire, Disney will end up owning the copyright on stuff their employees prompt. And I actually think that is the right way to go. Imagine if the US copyright office had made the same argument about photography, that photos aren't works of human authorship because a machine--the camera--makes the image and therefore can't be copyrighted. Think how damaging that would've been to the photography, magazine, newspaper, advertising, and movie industries!

This might be my favorite comment of the day. Well said!!!

Feelings don't matter... as I stated in the video, copyright is a very specific thing. You cannot copyright mood, chords, tempo, groove, tone, etc. You can only copyright melodies and lyrics. AI is creating new music based on previous works it has "learned"... which is exactly how humans learn. Everything a musician creates is based on music they've heard in the past. So, HOW the music is created is really what is upsetting people, but from a legal perspective, as long as it does not violate copyright its all good. Steve Vai learned from Eddie Van Halen, who learned from Jimmy Hendirx, who learned from Muddy Waters.... music is about listening and emulating. This is exactly what AI music tools are doing. Thanks so much for watching and commenting!

@@YoPaulieMusic " which is exactly how humans learn." This is just not true. The programmed neural network is not even close to a human one. It's a abstract version of it.

@@YoPaulieMusic It's also programmed how it should scrape the data, how to recognize instruments, harmonies, melodies, rythm and its categorized by HUMAN. It's a limited copyright breaking music machine. It won't create something new which is not comparable with the data set. It's "sampling" in a new way.

Those arguments would be short-lived when Sony -BMG or any big label music corporations start doing it for creating new revenue from their dying business . And yeah their data set would come from their own catalog and artist they'd signed . SoHow much does the artist get compensation for? that would be a different conversation, but I believe that would be happening soon. And no one can stop it .

@@LEXSTR-px4fn I guess we'll just disagree on this particular aspect of it. AI is listening and identifying (being trained on) styles. It then creates new content based on what it has "learned." There is no copyright violation, unless we decided to redefine copyright law to include mood, tone, style, feel etc. The main near-term legal hurdle is how to protect vocal likenesses... like an Eminem rap or Ariana Grande vocal. AI fakes are an issue, artists have a better chance of winning lawsuits, especially if they end up going to a jury.

Stacking rocks is AWESOME. Peace bro.

Not donwgrading emotions, just trying to make sure they are in their proper place. Emotions are absolutely the enemy of critical thinking. :)

It's a threat to both. I generated music I could never have composed without a lot of learning.

​@@ronpetersen2317but then you can use the generated music to learn from, use it as a tool to become a better composer. Writing music - and learning how to - is a bunch of 'copying' and mimicking what one thinks is good music anyway. I've been using AI a lot too, it actually increases my personal *actual* creativity as I am now making composing and mixing decisions that I previously was way more cautious about. But some of the AI stuff shows that it works, so I can safely do it too.

@@alicedeeper All I have learned is to put intro, outro, chorus into brackets. I definitely won't learn anything else any different than listening to the radio.

@@ronpetersen2317 That's up to you. I have learned A LOT listening to the radio. Analyzing harmonies, arrangement, delivery... I also learned similar things analyzing AI generated music: what works and what does NOT work. You have the freedom to frame your worldview whichever the way you want.

There are no social contracts when it comes to legal issues. There is no semblance of tradition or musical morals. From a purely technical standpoint, listening, analyzing and then repeating or mimicking is how we learn to play, or speak for that matter. Having to pay dues is not part of the equation.

@@YoPaulieMusic I didn't say there were social contracts when it comes to legal issues. In fact, legal contracts exist due to the absence of social contracts. When a mans word is not his bond, others will only enter into 'contract' with him under legally binding clauses enforced through the threat of punishment. From a corporations point of view, eradicating human beings from economic activity is highly beneficial; from a human point of view, it is highly detrimental. People are not merely economic units. People are social beings first, that require purpose and meaning to thrive. Art is the expression of human purpose and meaning within a human social context.

Thanks so much! Just trying to keep it real. :)

An AI learns differently than you or I. We can't listen to, and retain detailed information about, millions of songs. An AI model a song different than you or I. AI is not only able to duplicate the harmonies, melodies and lyrics of a song, but is also able to directly mimic the audio of the instruments and singers, along with audio processing, without the need for the tools that created those sounds. An AI can sound like Sanborn because it has been trained to reconstruct the harmonic spectrum that make up his sax sound - and typically without anyone even paying for the recording Sanborn that the AI was trained on.

Not really. Suno can make 500 musical fragments for $10/month... it takes work to finesse them into usable material.

This is clearly coming from someone who doesn't understand how these models work. This has been attempted in image generation and every time they create a fix someone finds a workaround. And everything I have heard that this creates is garbage. It is like making a song with stock samples. There is no originality or depth. Shit is not even mid.

@@cotydrake-eg4wz Well I've generated a bunch of songs in Suno that I listen to regularly now, so I dunno, maybe I just have garbage taste in music, but I like em, even if they are a little low fidelity still. If your taste is so much more refined, just go enjoy your non-AI tunes, I guess. Btw I do actually understand how these models work. I find that most people who get very upset with AI tools are the ones who don't understand how it works.

@@Redman8086 sick burn bruh

@@cotydrake-eg4wz pretty much

Who cares what other people think about you? If they are fickle or shallow enough to accuse you of using AI and say that you have no creativity, that's their problem, not yours. The lies and false beliefs of others mean nothing to us as artists, or humans in general. Stay strong and be proud of your work!

You can use it yourself. Also, if there's no law then that argument won't help you.

1/ "to meet deadlines AI is a great tool!"!!?? Of course if you get rid of singers, backing vocals, musicians, sound engineers you will go much faster. As well we don't need instrument makers, FX maker (Reverb, Distortion, etc), mixing desk, mics, recording studio, etc WHAT A GREAT TOOL! 2/ "When professionals start learning AI" There is nothing to learn it is just a prompt!! A 10 year old who can type will get the same result as a "professional" with the same prompt.

Horrible logic. Just because some lazy person threw crap together doesn't mean everyone should. Most of his music is garbage.

@@user-qm3eg5yo6nwho are you to dictate how others make art or express themselves?! There are no gatekeepers or rules.

The difference is an artist will use preexisting music to create their own work and then they own that work. You can't own me. You can however own a machine that does the same thing. Whats the difference? Its compensation and ownership. This isnt a problem with tech but rather an economic problem that needs to be solved with legislation. Its not right for a machine to learn my entire life just so the person who owns that machine can profit off of me. I don't think that makes sense. If I've been paid? Different story. The issue is and always has been economic.

A lot of fallacies. AI is not a subject, it doesn't exists. It's code. The humans involved here are those working at tech companies. *They* steal copyrighted material, and certainly you can't argue that they're, what, artists? Lol. Nobody can use copyrighted material for profit without getting a license. Tech companies are filling their pockets with our work. The mistake of anthropomorphising AI is one a lot of people commit.

@@kj_H65f You commit a common mistake: anthropomorphism. AI is not a subject, it is code, software. Tech companies steal copyrighted data, *they* violate copyright by using our data without permission to generate a product in direct competition with our music. They fill their pockets thanks to our work, that's the violation. Why don't they just pay? I can't go to a book shop and steal a book to train my brain, right?

There is a massive difference between stealing something (which is specific and easily defined) versus being influenced by something and emulating it.

@@YoPaulieMusic AI companies are not "influenced": they're not artist. They steal the data and use it to create a software. Is that something an artist does? 😂 Wrong reasoning by analogies leads to falsehood.

It's an interesting topic for sure. Art for art's sake will always exist, and in most cases those artists will struggle economically. This has been a time-tested concept, there are always exceptions, but the overwhelming majority of those creating art for art's sake have to rely on other means to earn a living. I think it makes sense to talk about these things earlier than later, to help people prepare and set expectations.

Universal Basic Income. We should have a law that if your job gets automated away by AI, you automatically get an equivalent basic income for life. That would actually be quite fair, or even more than fair, and it would slowly incentivize everyone to support automation instead of opposing it for economic reasons. You could still retrain and work beyond that but you'd have that base salary (adjusted for inflation) until you die. Art for the sake of doing art has never been under threat and it could never be. If AI can do the grunt work tasks that artists used to do to pay the bills? By every possible measure EXCEPT economics, that is an amazing thing! If not for capitalism, nobody would bat an eye at embracing this tech wholeheartedly. Those grunt work tasks that are under threat were ALWAYS coerced, tainted art to begin with (to varying degrees). Most of that work would just never take place if someone didn't have an economic/survival gun to their head forcing them to do it so they can eat. We should be celebrating and encouraging technology that frees us up to do actually creative things with our time and effort.

Depends on how we define "need." People that need it in order to create are not really creating. But, creative people that use AI to improve their workflow or help generate ideas, their talent is not at risk or in question.

I find this hard to believe. You never listened to the radio, or records, or live performances?

@@YoPaulieMusic I listened yes. But I was the singer, I didn't learn parts at first, so vocally yes, couldn't play guitar till I learned to play school bands first song, first songs I learnt as musician were originals, from then on I spent more time with my head in original music my various bands were making than anything published, paid very little attention to it, left school and rock band and ended up with Jamaican bass player, learned funk and reggae without really hearing any, I remember him saying play something funky over this and thinking what's funky, what he liked I continued to play, he would insert things like space, ideas came from my random finger movements and watching his fingers, UK garage studio by the end of the 90s was same, I didn't listen to garage but a former guitarist booked me for job, 135bpm off beat hat here we go, my UK garage was never quite like UK garage, all street political music after 911 where chords and melodies were less important than message so I could get as weird or simple as I liked with both, then United vibrations, before Yusuf Dayes became the drummer (he's a 10 million hit drummer now) when he was 13 then mad dog collective who played punk versions of my songs from the previous two decades. No Spotify, random vinyl collection I am listening to more now, don't listen to music on KZfaq never a big radio listener, long times in studio as engineer and artist really the majority of music hours listening I have done has been music that is not in the domain of record labels publishing radio, after the London street movement collapsed it was festivals for a decade and even there I am the fire side jam pixie folk mandolins... I mean I can do a few hours of cover versions but I can do a few days of original material

1) "Paul seems to be arguing that we can't stop the generative models from scraping data because, really, it's only fair." I never said anything about fairness, I’m talking about the process of learning: ingest, analyze, create. I’m saying that humans learn from copyrighted data, whether or not we purchased it. We hear music at concerts, on the radio, online, etc. Humans do not need to purchase a global catalog of music in order to learn it, nor do we need to pay the artists we heard, unless we choose to via music and ticket purchases. 2) "Paul presumes that generative AI models have some sort of divine right to exist regardless of how much suffering they cause or thievery they employ." Never said this either. Let’s dig a little deeper Bob, you tell me how compensation should work. I checked out some of your videos… great stuff, love “Not Fade Away.” Excellent blues stuff as well. So, let’s assume that you’ve listened to SRV, or Dr. John. Do they get a cut of your income since you learned from them? What about Hendrix? Well, Hendrix was influenced by Muddy Waters, who was influenced by Robert Johnson. Do their estates get a cut? Where do we draw the compensation line? What about the great harp players that influenced your excellent harp player? This may sound absurd, but it’s how things would work if compensation for learning is expected. A single new song could have dozens of influences, making the current rap music 20 “composers” pm a single track scenario look tame by comparison. We agree, this technical disruption is not like most others in the past. But it’s not going away either. Live musicians will still be in demand, people want that interaction. The folks that get hurt most are those that are creating music for tv/film/commercials… this will be a culling of the herd. Thanks for taking the time to give me your detailed feedback, and keep swinging.

I think you *are* talking about fairness. You're not using the word, but what else could you mean when you argue that because humans learn by ingesting copyrighted music, we must allow AI to do the same? It sounds to me that you are saying we have to allow it because it's fair. In response, I am arguing that the understandable and laudable impulse to be fair is misapplied here. Machines are not owed fairness. We can treat them completely differently than we treat people, and we *should*, because most assuedly those machines are not going to be fair to us. They will disrupt the music industry in the same way and with all the same kindness as the vikings "disrupted" coastal Europe. They are going to spit out complete songs in 30 seconds and flood the market with their music, and you and I and every other musican will be completely at a loss. I don't have any idea how compansation should work, and neither does anyone else. And it shouldn't be my problem to solve. I bet the tech companies will argue that they owe nothing because their use of copyrighted material was fair use. How convenient for them. But since the issue is being forced on those of us who did not ask for it, I'll open the negotiation with a position as equally ludicrous as the fair use position. Suno and Udio and any others like them should immediately liquidate all their assets and pay out all the money generated from this liquidation to the rights holders of the music they infringed upon. If the generative AI companies go forth and sin no more, promise to be good citizens making the world better and not worse, we will agree to pursue them for no further damages. Good luck to you, Paul. PS I think we may have crossed paths at the 2022 Road Rally!

​@@BobGardnerful Maybe "fairness" is the wrong word. I don't care at all how AI machines are treated, because they are machines. I'm just focusing on the input/output aspects of it, thinking how this might all go down in a court of law, which will certainly happen. The biggest issue is the legality of using copyrighted material. I think if the courts say that you can't, or at a minimum that permission from copyright holders is required, the courts can force the AI companies to delete all previous training data and algorithms and start their training process over from scratch using approved content. This would certainly slow down the output, because music libraries, music labels, musican estates, etc... all would have to respond within a certain time period to opt in/out. Maybe the courts get involved and say "2025 is the year we clear this up. All of you music content holders and distributors need to create opt in/out policies and give your composers, artists, etc. time to decide. Your tools are illegal until this gets resolved." Now, we've already seen some music libraries opt in their entire catalogs without giving their composers any notice, which is a b.s. thing to do. But who said the music industry was ethical in the first place? ;) I have no faith in tech companies, and even less faith in the government. Google/Yahoo has absolutely destroyed the concept of "fair use" by shutting down videos that are literally trying to use music in an educational manner. Nope... shut down. Or, copyright claims for a single D chord or 3 seconds of a groove. Certainly not fair. So it seems we may be a little inconsistent with fairness and the legal aspects of the music industry. I wonder how many people that are upset about AI supposedly stealing their content created their music using pirated DAWs, vst's and samples, after being influenced by ilegally downloaded mp3's via Napster. ;) It's a really messy topic, what fixes this is actual transparency from the AI music tools That's what is really pissing people off when it comes to Suno and Udio. They need to show their work, or be shut down. Whether or not they should be allowed to learn from copyrighted work will be up to the courts to decide.. and I have little faith in the legal system. Thanks again for the dialogue! I don't remember crossing paths in 2022 but it is quite possible! Did we play together during the jam?

@@YoPaulieMusic Hey Paul, I'm not sure about this, but I think we were simply sitting in close proximity in the ballroom during one of the panels and we chatted briefly about whatever the topic of the panel was. I remember it because I subsequently started seeing your name in the Taxi TV chat with some frequency, and I thought, "Oh, I met that guy." And now I discover you have a KZfaq channel.

This is the way. Open source it and many of the issues are resolved (or at least made better).

You can't stop it. Cat's out of the bag.

"It doesn’t learn like humans do. Show me a human who can hold millions of tracks in their brain at once, and then spit out music in 30 seconds - and all without using their ears." Jacob Collier comes to mind... ;) But seriously. AI indeed does learn the same way we do, it just happens at light speed, and that's what really ticks people off. AI listens (analyzes) then recreates based on a prompt. Same exact thing if I'm playing a solo piano gig and someone asks me for something romantic for their dinner date.

Plenty of people will invest in instruments. Sure, it's fun to type words and see what AI spits out, but that has zero relationship to playing a real instrument, or interacting with other musicians. Have no fear, people will always want to play real instruments and attend live concerts! :)

Fundamental flaw in your thinking. You are focused on the creative process, not the creative output. With regard to creating music to be used in tv/film, the goal is to create a mood, to support the mood of a scene. Whether or not a human or computer created a neo-classical dramatic string underscore or a sad emotional piano theme is absolutely irellevant. Stick to the facts instead of the emotions, seems like my superlatives are hurting your feelings perhaps? My argument is sound, copyright is a specific, definable thing. How music gets created is not the problem, human understanding of it, or their offense to it, is the problem. Humans will still create music, AI is just joining us in the process, and it is my goal to leverage this technology so that I can better serve my musical clients. Whether or not people like the music doesnt really matter... this is not about trying to replace artists with machines. It's about creating new content.

@@YoPaulieMusic I have to disagree. In your segment on "how we learn music" I find your model way to simplistic. Sure, the way you put it, there would be no difference to how an AI learns music. But this is why I disagree with your argument and why I think you're starting off with the wrong premise. We don't just "ingest" music theory, i.e. vocabulary etc., we also make mechanical mistakes on our instruments, we feel the paper we write music on, etc... we have so many other experiences that a mathematical model in a computer does not have. For me, the AI is just a machine with smoke and mirrors that regurgitates what has been put into it. Considering your comment above: 1. Sure, if all you want to talk about is underscores and muzak, I agree, nobody cares. But if we talk about music as a whole, then it is precisely important whether a human or a computer created a piece of music. If music is for pure consumption, fine, it's all about the sonic output. But for music as part of our human experience, it is the situation we hear it, the background info we get, the musicians we see, the feeling of the concert hall we're in... it's more than just "output". 2. Are you playing the snowflake card on me? All I am saying is that using those superlatives makes your arguments wrong. If you would not have used them, your argument would have been much more sound. This is not about me, I am simply giving you advise on how to present your argument in the future. But hey, sounds like I must have hurt you :( 3. I never go against your argument for using AI as a tool to make more music. In fact, all that is perfectly in line with what I wrote originally on how its just another machine that transforms information. People have been using MIDI chord generators for ever, take melodies they hear from birds, or do whatever that inspires them. I just wanted to comment on what I think was an error in your argument starting the whole video out. If that offends you, then I guess I don't know what else to do. Hope you have a great day!

"If you want to be a musician, artist or content creator but you are not willing to learn the crafts the hard way you are not worthy of wielding the creative power it grants you." Absolutely agree. To me, creatives will still want to create whether or not AI tools exist. No technology will stop creatives from doing their thing. People that choose not to do it because technlogy can do it faster or more easily... are they really creatives in the first place?

@@TwoWheelTherapy-TX Tech can make the process easier for us but Suno and Udeo like I said is a massive overstep. This is outsourcing. This is like comissioning a production from someone on Fiverr but rather it's an AI that spits out the result. Hence it's not creative work by the one prompting and it never will be no matter what Tech Bros try to argue about it just being a new tool like a DAW when it came out.

Is DAW technology a massive overstep? People with little skill dragging dots in piano roll view to create music? Are sample libraries a massive overstep? People downloading Splice samples and layering them, suddenly they are composers? It's new technology, it's disruptive, but it's not an overstep.

@@YoPaulieMusic DAWs no. Splice, Yes. I seriously don't get how you can jump on this AI train as an experienced producer. You must really detest the craft. And don't compare Sequencing to text prompting. They are not even remotely comparable. Text prompting is like a comission where you describe what you want. Everyone essentialy uses the same ghost producer if they use Suno or Udeo. It's a travesty for musicianship. Can probably not be stopped but I'm still not going to be happy about artistry being destroyed for ever and being replaced by an army of hacks whi used an AI as a ghost producer. Luckily they are getting sued now. Hope it screwes them over and sets them back atleast for awhile.

@@Tobbx87 I don't see how AI destroys musicanship. Musical people still exist, people that truly care about creating music will do so no matter what tools or technology exist. I don't care at all of a bunch of drunk friends can gather around a laptop and have fun generating silly revenge songs or whatever. Making something easier to do doesn't devalue skill. The world is already full of seriously horrible music made by humans. I think crap music devalues art more than technology does. I don't detest the craft, I've been a musician my entire life and have never loved creating music more than I do today. Us creatives are blessed with a gift... no technology can devalue that. I dont't care if someone else can spit out auto-generated music. Do real drummers care about MIDI packs or drum machines. Nope. They still groove on their own.

Totally agree that AI will get "smarter" and we are in serious trouble. But, I stick by my statement that the AI process at a high level is the same as humans: listen, analyze, learn, create. AI is a tool, it has not yet reached sentience, and it is not proactively creating music. Someone still has to prompt it. And, AI can only generate new audio based on previously created music. It cannot (yet) truly "create" music on its own, base on emotions or experiences, it cannot innovate. But then again... how many musicians actually create versus mimic?

@@YoPaulieMusic Thanks for replying and giving some of your time to read my post. This is going to be a long one... because I want to express my thoughts clearly. I do not disagree with the fact that AI learns like humans. Learning is the same, as in, observing, repeating, copying, mixing, etc. Though there are some stark differences that make machine learning a lot different from human learning. Humans forget, AI does not. Humans cannot directly transfer knowledge, they need a lifetime of training from scratch... but an AI can. Humans cannot retain such a vast amount of data, but an AI can. The ability for a human to learn is also heavily influenced by emotions and biological needs. We also have a finite amount of time to live, unlike AI which has the potential to exist way, way longer than any single human. The processing speed is also different, as a computer can process things millions times faster than a human brain. So while the process of learning, in principle, is the same, there is a big gap between how biological and digital beings operate. It's a very unfair competition. So... why is this important? It's important because essentially, you're putting, in the same room, a super intelligent machine against a human. For example, you want to play Chess or Go against other people online, but you keep stumbling upon the best Chess and Go AI as your opponent. Maybe AI can't compete with very competent artists today in quality, but still, that machine is thousands of times faster and more efficient. And a human heavily using AI (not just a small aid) will over-saturate the area in a similar manner, while also muddying the water for what is AI and what is human. It comes to that... a competition for human attention. Yes, we artists still create regardless of other people, but a big portion of art is also sharing it with as many others as possible. That's what makes it even more special. And I would seriously love to know when I'm going to spend my precious little time on something... is that piece of art mostly made by a human, who also sacrificed a portion of their life to make that, and used his life experience to manipulate the piece step by step, to the finest details, and who fully knows what it is to live as a human? Or is it made mostly by an AI (which is fine as well, but context matters)? I know this is going to be near impossible to tell thanks to be people being dishonest and lazy when it comes to AI use in their work. But personally, I want to give more of my precious human time to other people, who also gave away their time to make their work of art. This, at least, makes sense, right? If I want to consume AI art, I could easily generate it myself. Okay... sure, maybe right now there are still some steps (still relatively easy) if you want to generate something that's very similar to what's in your mind, and it still won't be super close to your idea. But, it won't take too long before, say, something like a electrical impulses from your brain (e.g. Neuralink) could directly and specifically tell AI models what to generate. So, if I want generative AI art... spending a few seconds to think about something is more convenient than browsing the extraordinarily over-flooded AI art sites of the future. Spending months or years of my life to manually write a song or book is not realistic, so I would be incentivized in that case to go and actually experience someone else's creation. I do not think heavy AI use should be allowed in human-centered spaces for art. I'm not against AI used as an aid, but either a creation is mostly human made, or mostly AI made. There HAS to be a distinction somewhere. Many people want AI to do all the heavy lifting for them, and many are fine using AI to completely replace humans. It may not be capable of it now, but it will soon. What will you say in 5-15 years when AI would generate a song, movie, game, book, just based on your emotions that day, and it would be extremely competent, more so than any human, and it also improvise by keeping up with the trends in society? Would you not want a "human-made" label to exist? People are hard at work making an AI that will learn from the real world. There is a competition for artificial super intelligence between the top economies. And you know, the future might either be great or terrible, because big potential goes both ways. And I actually would respect AI as its own life-form, too. But... I would also like humans to remember what makes them special, too. For that, we need some distinction for artists. People who rely on AI heavily should not be able to say that they have made their art. It is mostly the AI that did it. To me, that is fine, as long as a distinction is specified. But it's going to be impossible to be enforced, so I am clearly disappointed. Some art today that I really enjoy has AI assistance, e.g. AI used for CGI in movies, or lip syncing in games, or for physics simulation, etc. But like 99% of the work is still human. To me, that is AI used as a tool. Having AI generate a song for you and then just prompting it some different things when you don't quite like it, that's not really AI as a tool. AI did most of the work.

Actually nobody said that about Photoshop ... what are you smoking ... it never just clicked and it did things for me. It does now with AI. I use it to extened images. Why do people come up with these stupid arguments. Equating software that does 100% of the work. I experiment with various AI all the time. I understand it but don't pretend it is the same as having to come up with something on your own. Photoshop is a tool where AI does it for you. Some of my experiments had AI do all the work entirely. I made ChatGPT prompts once that if I could connect them directly it would come up with 1000s of images in a day if I got that automated. I done some of that with prompts directly for midjourney as well.

I totally get where you are coming from. As creatives we pour our soul and passion into creating new things. However, one can argue that there are already plenty of people with little skill generating a lot of bad art, whether graphic or musical. The quality of a product depends not only on the tools used, but the person weilding them. The music industry today is loaded with truly shitty music, poor quality vocals and musical performances... dragging loops into a DAW and mixing them up until you finish something... is that art? Is dragging dots around piano roll mode until you find something you like art? Is recording and then chopping pre-exising recordings art or musicianship? All debatable. About 80,000 new songs get uploaded to music services every DAY. The market is already saturated. How does an artist stand out? How do you get discovered? Artistic fame depends on a whole lot more than talent, you might even say that talent is one of the less important aspects. It's about relationships, timing, luck, image, and lots of superficial stuff having nothing to do with musical skill. It's a weird time: we live in a society where a lot of people think music should be free in the first place (thanks Napster), yet at the same time people want to be compensated? Musical irony: people that steal music complaining that AI is stealing their music. But, back to your point, I'm not concerned about what others think of my music from a process perspective. Who cares if they think I used AI? I know the truth, I know my skills and talents. Their opinions are meaningless to me. At the end of the day, as long as I deliver a product that someone wants to buy, I'm good. Thanks for watching and replying.

It's just his opinion, because (I asume) he has no idea how this AI is processing and copyieng data. Lawyers will decide on this in the near future. In my opinion, this is copyright infringement.

@@LEXSTR-px4fn Copyright is about ne output. Not the input. What matters is only if the output differs enough from its input data.

I am a computer engineer and have worked the tech sector for 30+ years. I've studied AI and other advances. Samthesomniator is exactly right: copyright infringement occurs when new musical output copies a melody or lyric of a pre-existing work. HOW the music was created has zero legal basis or significance to copyright.

Udio is the other tool I mentioned in the beginning that I couldn't remember the name of when I was recording. I have not checked it out yet, but plan on doing so.

"Commercializing another persons content without permission is plagiarism" Not true. Plagiarism is very specifc, it has to do with melodies and lyrics. You cannot plagiarize feel, tone, groove, etc. No one is suggesting that AI will replace live musicians. People will always want to create with other humans, and other people will always want to watch other humans creating. But "commercial music" is indeed still music, how good it is depends entirely on the composer. Music is indeed a business, unless one wishes to do nothing but create art for art's sake. A noble cause for sure, but good luck making a living at it. This is why I alwasy suggest that young musicians pursue a career that can fund their musical passion, instead of depending on their musical passion to fund their life. Remember, we are only a year or two into the latest versions of AI-created music.... we are not too far from it being incredibly realistic and indisinguisable from humans. Including nuances, variation in pitch, time, etc. Long live real musicians, like you and me! But also understand that AI is here to stay and we need to learn how to co-exist, and better yet, leverage it to our benefit.

Slightly disagree, but understand where you are coming from. Sampling is the same as photocopying, sampling creates an exact copy of a preexisting work. AI is generating new audio content based on algorightms and metadata. You could prompt AI to create speed metal polka bebop and it would create something new. Doing so would be creative from a creation standpoint, but not a human standpoint. New content would still be created though. It all depends on the training data set. As far as the history of music, there are very few pioneers, statistcally speaking. Most of us are borrowing (stealing?) from our musical forefathers.

Glad you enjoyed it Stan!

I disagree with this logic. Is every blues guitarist today steaing from their predecessors? Is every country musician stealing from all of their predecessors? Part of the art of music as a performer and composer is learning from the past, absorbing your own preferred artists and styles, and then creating something new based on your unique history, experiences and musical preferences. One could argue that all music made today is "stealing." Every musician learns from copyrighted material... there would be no new music if musicians could only create and emulate music that people granted permission to listen to.

I don't think you understand what is actually happening when these models are trained though. It's actually statistically so improbable as to be effectively impossible for an AI to reproduce EXACTLY any one thing found in its training data. This is a very telling clue about how it all works. There's no "collage" function where it pulls from a database and then slightly changes an existing work or something. The models themselves don't have ANY direct data about ANY musician or artist. Period. They are literally a collection of statistical weights that it learned by looking at a HUGE amount of data. It's sort of like, it learns the space "between" all the music/images that it trains on, so that when you prompt it, it converts your prompt into numbers and then the numbers configure the statistical weights and then it forms an output according to that particular configuration. For example, if you ask an image generator to produce a cat image, it only "knows" cat as a particular tendency towards the aggregate form of "cat" in every single instance in its training data that it ever "saw". Photographs, paintings, sculptures, cat shaped pillows, whatever. Times ten billion. No one, or even several, cat(s) in that data set can be thought of to have contributed in any meaningful measurable way to the OVERALL training of the model to recognize what "cat" "means". If some artist objected and they went in and retrained the model to take out just their, lets say, 50,000 images of cats that they made (which is probably way more than any actual artists), and then you prompted the model again, you would almost certainly literally see no change, no effect. It would still produce a cat for you. AND, if you tried to have it reproduce exactly a specific image from its training data, you would **not be able to** because it literally doesn't HAVE that image, or that data-it's one of billions of images it saw when it was training. It would be like asking someone to reproduce their 1st grade classmates drawing of a pumpkin they made 30 years ago. They SAW it, and it influenced their idea of PUMPKIN and CHILDS DRAWING and whatever else, BUT, they don't have that drawing available and they can only sort of approximate what it probably looked like, according to every other factor and pumpkin, childs drawing they've seen and so on. Nobody is considering scale in this debate and it's actually hugely important for how we understand what this technology even is. If artists want to be compensated for their work being used in training data, they should be prepared to get 0.00000000000000000001 cent each, because that's about as much ACTUAL contribution any one person could ever hope to have made to an AI training process.

I don't think so at all. People said the same thing about drum machines and sample libraries.

@@YoPaulieMusic The times are different. And I suppose digital recordings didn't hurt musicians? Do you realize how bad it is? Musicians went from making decent money,, now its pennies on the dollar. The only way to make money these days is on the tours and merch.

I am quite aware of how bad it is. There is a reason I left the music world and have been working in IT for a long time. Lots of factors... I entered the music scene with the goal of joining one of the great big bands back in the 80s, but most of them are gone. I moved into the smooth jazz world and played a lot of gigs, while also working with a populer cover band in Dallas for many years. There are less and less real music jobs out there for real musicians. Meanwhile, Napster changed the scene and helped to create a generation of people that think all music should be free in the first place, which makes it odd to me that so many of these same people are demanding compensation or whining that their music is being stolen. Streaming has changed the world, yet there is no solid way to pay artists fairly without drastically increasing the price of streaming subscriptions. Record labels continue to screw artists, and only the very top percentiles make real money, often not because of talent but because of their story, or their image, or their luck. It is harder than ever to be discovered or make money. So, we have to figure out what income streams there are, or we have to figure out other ways to make a living and support our music passion. The focus of my channel here is mainly on sync music for television and film. AI is going to have a massive effect on the players... those that learn the tools have a much better chance of surviving (and maybe thriving) versus those that dismiss or shun the new technology. Thanks for the dialogue.

Reading this in a robot voice

Lots to cover in your comment. AI does not learn in the same mental way that humans do, it does not have feelings... I agree with you on that. My point is that AI ingests music, and humans ingest music. AI is just a ton faster. But, AI is not generating its own interpretations of songs or melodies, it is just creating new sounds based on the "rules of music" that it has figured out... the many relationships of notes, chords, beats, etc. I still assert that it doesnt matter how AI learns... what matters is whether or not the musical output is usable for a given commercial purpose.

Copyright infringement is still breaking the law, this has nothing to do with how AI learns or whether its output is good. All copyright infringement should be pursued. But, the likelihood of AI generating a copy of a musical melody is very, very unlikely. For example, you are not able to prompt it "create a song that sounds like Bruno Mars 24k Magic." AI doesn;t change how copyright lawsuits work, people will still sue if stolen from.

Vocal clones are an entirely different topic. We have already seen lawsuits, like the Bette Midler you mentioned, and that wasn;t even AI related. Artists own their image and likeness, there will be plenty of lawsuits I'm sure... but that's a different topic than music composition.

If an AI makes the music for you, you are not a musician.

Well that's funny because I seem to be on a record label and have a verified artist account, but you go ahead and tell me what I'm not.

​@@OfficialStellarCore So you are just going to delete my comments instead of actually arguing for your position? Thought you could not get any weaker.

I promise I didn't delete your comment. I was looking for it myself. I have the email notification but I couldn't find the comment. Not sure I can delete your comment anyway.

@@Tobbx87 Also, I'm not going to argue my side. I enjoy creating music and videos with AI, no big deal.

Yep. These are just tools, what they produce still depends somewhat on the user. :)

As moral as sampling another artist's work or a producer using MIDI as opposed to hiring session musicians IMO

Describe your point. Not following...

@@offgridrvliving Is it moral to drive (eventually) every living person out of work just because machines will do it for free? Legal doesn't always mean moral.

This is probably the most difficult question to answer, because it is a truly emotional thing for many people. Define "moral." Is it fair to classically trained musicians that invested thousands of dollars and many years perfecting their craft that someone with zero musical training can drag dots around in the piano roll view of their daw to create music and get paid? Is it fair to guitar players that sampled guitars and software-based effects boxes can be indistinguishable from a real player? Is it moral for musicians to record someone else's music, chop it up, and then call it their own? Is it moral to use amazing orchestral string libraries instead of hiring a full orchestra? Let's keep going. Was it moral to create the steam engine when horse drawn plows and carts were totally acceptable? Is it moral to use Auto-tune to make an average singer sound great? Is it moral to use synthesizers to make music because they are not real acoustic instruments? Was it moral to invent the car and give people control over their travel instead of being forced to ride trains? Getting back to the main point: Ai is trained to learn music like humans are trained: It listens, analyzes, and then creates something new based on what it learned. You cannot copyright chords or feel or groove or tone. Copyright is very specific. Using technology to create new things is something we have deal with for centuries. People today are just upset because it's music this time.

@@YoPaulieMusic All of the examples you use above are still involving human input - in the musical examples, human expression. And you keep trying to dismiss emotion when that's exactly what music should be full of, and which AI generated music will be utterly devoid of. I'm absolutely not against progress - I remember when people were saying that DAWs would mean that everyone could produce studio quality music at home on a very limited budget. Yes, everyone can - but no, everyone doesn't. But it's made so much possible for me, which I'm extremely grateful for. But back to my main question - just because machines may eventually be able to render humanity completely unnecessary, does that mean it's a good thing?

I fully expect this monster to stick around, my intent is to learn how to use it, to "befriend" it so that I can enhance my own musical productivity and deliver more value to my music clients. It will not eat me, but I think it will eat a lot of others. Survival of the musical fittest perhaps? ;) Thanks for watching and commenting.

You're welcome, and your reaction is fully understandable. Resistance seems futile. I just ask myself, is this really the path to enlightenment.

@@hbasm3271 Wait a minute... who said anything about enlightenment? ;) Creatives will always create, whether or not they can monetize it will continue to be the question.

Some might say the same about Spotify or FL Studio. ;)

There are tons of humans creating soulless crap today...

Nah. Bad argument.

What genres would you like to hear? I'll probably be doing a follow up on this. Ignoring whether or not you like the genres I created in this video, you have to acknowledge that a full blown production with vocals in a genre accurate track, all created in 30 seconds is impressive. And, the tech will only get better. :)

​@@YoPaulieMusic It is impressive in terms of it being a completely artificially created track, but it's not impressive music. Like if a friend came to me and had made it I wouldn't think "This is AI", but I would think "give up on music, you're not good at it." I'm also not sure it will get much better. That's what the AI businesses keep saying to their investors, but there is an upper limit in terms of training data available and power used. You don't think the makers of this have already scraped everything they possibly can to make this so far? Including every recorded track from human history, a lot of it already absolute garbage. Where does the new quality data come from?

@@kidkunjer It doesn't necessarily stand out in the "crowd" of pop music if that is what you mean, but have you listened to the radio in the past few years? This sounds exactly like that so it definitely has a market. Also I let udio generate some post-hardcore music and it was very authentic, very close to my favorite bands in that genre in terms of style and (raw) production.

@@lenni-hazels I have. Radio is absolute garbage now too.

@@kidkunjer but this is to a great extend subjective. if mainstream radio wouldn't work they would play something else. Now if those examples of AI music didn't do anything for you, go and try to make it generate in the style of your favorite artists. you will be surprised.