top of page

Adam Neely on the Hidden Cost of AI Music

  • Writer: Evan Nickels
    Evan Nickels
  • 4 hours ago
  • 21 min read

What happens to music when everyone is only listening to what they made themselves? 

In this episode, Jade Prieboy sits down with Adam Neely, composer, bassist, and YouTuber educator with over 1.8 million subscribers. They go deeper than the usual AI debate, exploring what we actually lose when music stops being something we share. 

Adam draws a clear line between stem separation tools he genuinely uses and commercial generative AI platforms like Suno and Udio, explaining why lumping them together under the AI label distorts how people understand the technology. He also shares why he is cautiously optimistic about attribution models and how AI-generated lyrics reveal the limits of what machines can feel. 

The conversation turns philosophical when Adam introduces the idea of solipsistic listening, the tendency to only love music you personally generated while tuning everything else out. He then offers practical advice for musicians trying to build community and makes the case for why constructive critique from a real listen might be the most valuable thing a musician can receive 

If AI is changing what music is for, Adam Neely is one of the people asking whether we actually want to go there

 

 

The news





Listen wherever you pod your casts:



Looking for Rock Paper Scanner, the newsletter of music tech news curated by the Rock Paper Scissors PR team? Subscribe here to get it in your inbox every Friday!



Episode Transcript

Machine transcribed


[00:00:00] Jade: Welcome to the Music Tectonics podcast, where we explore the intersection of music and technology. I'm joined here with Adam Neely. Um, he's here in Bloomington, Indiana for the algorithms event. he's joined us today to talk about kind of the intersection of music and, ai. Adam is an artist and composer, a touring bassist, and a content creator on YouTube who's amassed over 1.8 million followers.

He usually shares video, essays, lessons, and vlogs on the horizons. In music and music theory. Adam, thanks for joining us today.

[00:00:29] Adam: Uh, thank you so much for, having me. I'm excited to be here.

[00:00:32] Jade: Absolutely. So what's your take on the future of music? in a best case scenario,

[00:00:37] Adam: uh, a best case scenario. Everybody gets paid and everybody.

Listens to music that's made by humans, doing cool and interesting things with technology. I have no idea what that looks like. Um, I think quite frankly, the future of music is very up in the air, because of generative ai, because of companies like Suno and Udio, these companies, which, you can do text to prompt, you can prompt, these, songs, which sound pretty good, honestly.

 And I think because of that, uh, it's upended a lot of the ways that people interact with music. It's upended a lot of the ways that I think people will consume musical experiences To quote, Mikey Schulman, the CEO of Suno. and in the best case scenario, I think, musicians will be able to get paid for their work and,

People will be able to listen to and experience music in ways that honor that and honor the cultural tradition of making music and sharing that with, uh, with people around the world.

[00:01:36] Jade: Awesome. Have you heard about the companies kind of working on the ai, attribution engines and, and stuff like that?

[00:01:43] Adam: Yes. So I think that is actually. Um, that is actually a great way that this can happen. you know, like one of the problems is of course, like Suno and Udio, train their models. I don't even think allegedly, I think they've straight up admitted this on, all of the audio that's on the internet. And because of that, you know, musicians are not getting paid for that.

Like, I. Have music, I'm pretty sure that was scraped for those models. and you know, I haven't received a dime for that training data. there are, at least from what I understand, I recently learned this at the conference. there are means of, Trying to parse what, training data was used in particular outputs.

And those are attribution models that a lot of people are working on right now. I don't know any of the details, behind it, but these attribution models will hopefully get artists paid and get, compensated for their, their work and for intellectual property. So I think that is a, good way forward from where we are right now.

 and, you know, going towards a world where artists are, paid for their work and people can still generate music that are derivative, works from artists recordings. That's one way forward. I think it's a good way, but I don't know where it goes actually.

 but I, I would encourage everybody working on attribution models to, uh, to continue their work. 'cause I, I think I'm excited, about that.

[00:02:59] Jade: Me too. So there are, there are AI music companies that are fighting the good fight.

[00:03:03] Adam: Yeah, for sure.

[00:03:04] Jade: Do you use any, um, AI tools in your workflow currently?

[00:03:09] Adam: Yes, believe it or not, I use AI tools or at least, that's how they're marketed.

I'm not sure if I would call them AI tools. I use STEM separation. STEM separation is a big, part of my workflow on YouTube, believe it or not, like stem separation is just the ability to take the sound, like isolate a sound within a recording. And that's super, super useful for me as a music YouTuber because I can take a classic recording from like The Beatles or Metallica or just any classic song and isolate like just the bassline to.

That song and then show it and like talk about it and analyze it and say like, ah, here you can listen to this note and the way that this person frets this, this riff. And that's very useful for education and it's also useful for the production side of things because if I record in a noisy environment, I can isolate the sound of my voice and clean it up.

And that's amazing and that's super, super useful to me. That's a fundamentally different technology and system than. the generative AI systems that would, you know, you type in something or you prompt something and then out comes a produced track. To me that is only AI in the sense of both use machine learning, but they are not part of the same kind of general aesthetic of ai, which is, uh, one of like text to prompt or one of generation, like, you know.

Chat, GPT, the chat, GPT of music, the mid journey of music, et cetera. I don't use any tools like that. stem separation is something that like. I, I've been a huge fan of for a long time. Um, there's this guy by the name of Simon Fransman, who is a, wonderful jazz memer, from Sweden, I believe. And, uh, he,

Six years ago, he, like before Moises or perform at Laal dead ai, he had this meme where he takes, Pavaratti singing Nema, like the great Italian opera from Turandot. and on the big climactic note of Vincero, um, he ha he takes the sound of Pavarotti's singing. isolates it from the orchestra stem, separates it from the orchestra, and then auto tunes and melodyne's it into the, the lick.

And so six years ago, that was mind blowing. That was so awesome. Like, oh my God, I can't believe you can do that. And, uh, I remember being so inspired by the technology and the stupid little meme. and so because of that, I, you know, have, been a huge proponent of tools that can let you do silly, stupid little things like that, that you wouldn't be able to do.

Prior to those AI tools being introduced, I love that so much because it's just such a fun, it's a fun approach to music that, I don't see any necessarily ethical qualms about provided that, um, the models that these STEM separators are trained on are trained ethically. I do have a problem with, generative models, though.

 mainly because of the data provenance of their models, but also because, the way that they interact with the musical world in general is I think, not great in general. And because of that, I don't use Suno or Udio or any of those platforms in my workflow.

[00:06:18] Jade: I use, um, I use some AI tools that kind of what the valuable part of AI for me is this iterative nature.

[00:06:25] Adam: Mm-hmm.

[00:06:26] Jade: And I, but granted, I, the most of the AI tools I use are kind of in the corporate setting.

[00:06:30] Adam: Yeah.

[00:06:30] Jade: Which, you know, rewards, productivity, rewards, speed, all these things that are kind of like these corporate values.

[00:06:37] Adam: Mm-hmm.

[00:06:37] Jade: Um, do you see that these corporate values are kind of weaseling their way into.

This creative space around music. And I guess I asked that to kind of prompt the question, like, was music ever meant to be commercial? Um, and does, that take away from the part of music that is inherently kind of good for the soul?

[00:06:54] Adam: Yeah, I mean, so I didn't go into music so that I could make music.

Fast. I went into music because I wanted to, shred, uh, I guess maybe that is making music fast, but like, but in a different way. You know, I, I went into music 'cause it was, it seemed like a really fun thing to do and it was an important, you know, an important part of my life. Being able to make music and share it.

 I don't think anybody goes into music because they want to be productive. it's not a thing. wanna be productive because of the realities of the industry and because of the realities of society. You, if you're productive, you can, uh, make more music. You can sell more music, you can, accomplish more tasks and hopefully make more money so that you can survive, which is important for everybody needs to survive and pay bills.

And so because of that, productivity is a very useful thing in the music industry. But maybe. It's not a good thing for music in general. if we think back to, recorded music itself when we, you know, we have, Edison, Wright. Edison, yeah. He invented the phonograph. Uh, when we, we started recording.

Music. We turned music from this thing that, uh, was a performance from something that like a bunch of people got together and performed to something that could be sold and distributed, which is great because more and more people could listen to music and that was awesome. But it also turned music into this thing, a product that was subject to the same kinds of market forces than any other product would be.

 subject to like the sale of furniture, the sale of, Whatever, you know, mugs or whatever. And because of that, it's now in the same kind of, intellectual space as literally any other, practice that's in the corporate world. So it is subject to the pressures of productivity, and I'm not sure if that's a good thing.

 I don't think that's the reason why anybody wants to do music. Um, and I think giving in more and more to that is ultimately not a good thing.

[00:08:44] Jade: Mm. Mm-hmm. What do you think are some, you know, industry-wide misconceptions around AI music, and. Machine learning and all this, this has this whole topic that's been really, that's really taken over the music industry in the last two years.

But you know, you've been, you know, it's been going on longer than that.

[00:09:02] Adam: Yeah. I mean, I think the big misconception that I know, I know this is not, like not everybody has this misconception, but I think for the general public for sure, is separating the ideas of commercial generative ai, which is the term I constantly use to refer to things like suno and Udio, and then.

Any other thing which has the AI buzzword, which might use machine learning or artificial intelligence or whatever, but accomplishes a very specific task. And I, I use stem separation all the time. I think they're fundamentally different beasts. I think they represent very fundamentally different things.

And if people, you know, there's quite a lot of backlash against commercial generative ai. I'm a skeptic of commercial generative ai. I think if, other companies want to distance themselves from the backlash, they start a different brand, they start some different way of describing their technology or their aesthetic.

Um, I know AI is the great buzzword, but maybe just say machine learning or come up with some other different way of describing, uh, the technology that you're using. because I love things like tone Timbre, transfer. Again, if the models are trained ethically, it's great. It's awesome. And even, you know, generative models can be done, very well and very, ethically.

And I think there are ways of doing it, but I think, just as a fundamental beast, as a, part of the tech industry. As a part of this massive amount of money that's flowing into these companies, I think it, it's important to isolate them away from other aspects of the technology and the industry that, I don't have as much of a problem with because they're quite frankly just not the same thing.

Mm.

[00:10:34] Jade: Mm-hmm. so in regards to commercial generative ai, do you think you can tell by ear if something is AI generated?

[00:10:42] Adam: Interesting. Yeah, I think I can. for me, uh, the lyrics usually are the big tell and, you know, you the company's marketing and the way that a lot of people, uh, say is like, you know, the lyrics are the human elements.

And then, you know, the model will generate music around the lyrics. And to me that's really the indicator, you know, professional lyricists. Have a craft that they have honed for many, many years. And I can really tell, because usually the scansion, the rhythm of the words, doesn't quite feel right. Doesn't feel right to listen to, to sing along to.

'cause quite frankly, these aren't being designed to sing along to you're, you're just listening to them. But, you know, I have enormous respect for lyricists and how they're able to craft the rhythm of the word in a given language like English towards. The melody and the arc of the melody. You know, when we speak, we speak with a kind of rhythm and we speak with a kind of cadence and melody into itself with our speaking words.

And that you really have to be sensitive to if you're writing lyrics and matching the rhythm and the arc of the melody. And so what I found is I think the lyrics are the big tell. also the. The vocal quality sometimes has a really weird auto tune effect, but it's not even Auto-Tune, it's its own artificial thing.

 uh, it really is most evident for me in genres which are not popular genres like jazz. My, my thing, it's so obvious when it's artificially generated jazz, to me anyway, or classical music, for example, just it's, it's nowhere near being able to do it. But on, Popular music and pop songs. It often is, honestly, it, it sounds kind of like, other popular music, other human made popular music.

It gets like 95% there most of the time. so I understand why people might be fooled by it, quite frankly. Do, do you have, uh, do you have any tells for you? Like, in terms of how you can, uh, tell?

[00:12:44] Jade: I think so.

[00:12:45] Adam: Okay.

[00:12:45] Jade: And I think for me it's about like the top lines and the melodies.

[00:12:50] Adam: Okay.

[00:12:50] Jade: I think that personally, like music comes from the soul and it's, it's something that feels really good to do. Like when I'm playing drums and I'm gonna improvise a drum fill

[00:12:59] Adam: mm-hmm.

[00:13:00] Jade: There's a certain element of like. This is coming from a body. This is, this is how this should sound, this what, this is what feels good.

If I'm gonna sing, on a beat or anything like that, the melodies I'm coming up with are, they're gonna feel, they're gonna come from an inspired place, like from the soul. And they're gonna, they're gonna feel good to sing.

[00:13:16] Adam: Mm-hmm.

[00:13:17] Jade: And I think that it's really hard to measure, like this feel thing that we're talking about, but when I, when I hear certain songs, you know, commercial pop songs that I might have a feeling are AI generated.

I think it comes down to the feel, and it comes down to, if I was gonna try to sing this, does it feel natural? Does it feel good to do the, that the, the range of notes or that run?

[00:13:38] Adam: yeah, and, and I, I agree for sure. Like when I've, when I've learned suno basslines or learned things like that, sometimes there's stuff in them that aren't quite.

 they don't lay on the hands in the right way, or, you know, sometimes they do. Um, because, you know, they're training on recorded basslines. That sounded and felt awesome. Um. But it always is this weird regression to the mean, where it's like approximating what it would be like to play or hear something.

And I think it's always important, for musicians to write music that's not only fun to listen to, but fun to play and sing. And that's, uh, it's a hard thing to understand unless you actually do it with an instrument in your hand. And, Yeah, I think that's a great point though, in terms of stuff that is a, tell in terms of, can you understand or can you tell if it's AI generated?

Yeah.

[00:14:28] Jade: Have you ever tried to jam along with an AI platform or seen videos of people doing that?

[00:14:33] Adam: Yeah, I mean, I certainly did. Or like, I, I always will try out the platforms. I will always try out. Um, it's not particularly inspiring. I'd rather play play along to great famous recordings. Like the thing I always do this, my bass warmup when I'm just like warming up my, my hands and stuff is to play.

 Stephen Winter has a great tune called Do I Do, and it's a 10 minute long jam. It just goes on forever. And so. Just play that. And I love that song and I love playing along to that song. And I've been doing that for like five or six years just to warm up, hands up. I just play along to Stevie Wonders Do I do.

And I cannot imagine going to Suno to try and generate something that's as good or feels as good to jam along to. It might come. It might come along. Uh, I have not had that experience.

[00:15:22] Jade: So let's talk about this, this concept that you've been kind of poking at on on some of your YouTube videos. It's this.

Idea that AI music and this interacting with these platforms kind of leads to this like narcissistic behavior. And I've, found this in myself too, from watching your videos. I'm like, oh, oh, I'm doing that. And it's, for me, it's, I will take some lyrics off my phone, plug it into suno, write a prompt around some kind of pop punk thing or whatever.

And I'm really happy with the, with the result. I like my result and I'll listen to it, but I don't like other people's results. Um, and it also, there's, you know, we're also not, like, I could have easily just invited my friend over and we could have jammed and it could have been just as good or better. Um, so yeah.

Curious as to, you know, what you think about that, that kind of narcissistic behavior that this is, that this is leading to potentially,

[00:16:11] Adam: yeah, I, I mean I've heard it also called solipsistic listening. That's another fun term where you just, you become so in love with the thing that you made, that you don't listen to anything else.

 you know, if you've ever gone on ar Suno, this is, uh, something that I've, I've seen is like people saying like, Hey, you know, I, I love listening to my own music, but I don't like listening to other AI music. 'cause it, you know, my own music is amazing, but other AI music is cringe and terrible. And, you know, like for my own self, for like non-AI stuff, I love listening.

Well, maybe not listening to my music. Because sometimes, the recording, like, I might not get the right recording, but I love my own compositions. I'll say. I, I love what I, it is that I write. And, what's interesting is that doesn't make it so that I, when I listen to other people's music, I, I don't like it.

Which is, it's a weird thing. Yeah. I, I love many other people's kinds of music and I love listening to all kinds of music and not, you know. It's a weird effect, right? Where it's for the AI world, it's you only listen and you only love the stuff that you make, and you don't really care about what other people make.

And, I don't know what's the mechanism behind that. I think maybe, you know, it rewards like you constantly tweaking something to be exactly the right thing for you to listen to, but at the same time, that's what. Uh, regular music making is too, you're constantly tweaking and trying to find the best version of that idea.

 the lead investor of Suno from Menlo Ventures, I believe CC Gong in a tweet said that, you know, this is, what she's doing. Maybe it was just to like, Promote Suno or whatever. I'm sure it was that, but it's like she was like, I have moved all my listening habits from, Spotify to Suno as a means of encouraging that people to do this, of like creating their own music that they themselves listen to over and over again and.

On the surface it might not seem that terrible, but I think the being cut off from other people's music, from non-AI music is potentially a really dangerous thing in the long term. And what I mean by that is like we have a shared culture that comes from music. Like we know all the same songs when people,

 when it's somebody's birthday, you know, to sing the birthday song, right? It's like that's just hard coded. Like you have this set of songs, this repertoire that, you know, I used to be a wedding musician. I played, you know, these long four hour weddings, and people would get on the dance floor and start screaming.

ABBA starts screaming Bon Jovi start scream. Uh, like all these classic songs and there's something really beautiful about a bunch of people getting together and having that shared body of knowledge, that shared, framework for which to. Connect with one another. And if you're only listening to the music that you create on a suno or like a platform like that, you don't have that sign of shared cultural knowledge.

The way I describe it is like, if you've ever been in a conversation about a movie you haven't seen, or a TV show you haven't seen with a bunch of people excitedly sharing details and fan theories, it's a bit alienating. You don't like being in that social circle because you wanna, you wanna contribute, but they're talking about something that you have.

No idea what they're talking about. And when you personalize culture effectively, everything becomes the equivalent of that conversation where two people are talking about Breaking Bad and you haven't seen the show. or I don't know, what's it, what's the show? Uh, severance. I haven't seen severance yet.

So, you know, like that is that like, and uh, and it, it is just like, oh, I probably should see severance 'cause like these people are having these great conversations. But like, So to me that's, that's kind of like the direction of this, unfortunately. And I would, if you use Suno, I would encourage you to not just.

Listen to your own stuff, maybe also listen to some other stuff. That's my, the main takeaway here.

[00:19:58] Jade: Yeah, totally. I mean, I, I agree with you that this, community as this, this removing yourself from the community aspect, is just, it's such a slippery slope. I mean, think about going to a, basketball game or a soccer game where you're all singing the fight song.

[00:20:11] Adam: Yeah,

yeah.

[00:20:12] Jade: And you might even become really good friends with like the random person you're sitting next to with. Cheering together. They might even be on a different team. Who cares?

[00:20:19] Adam: Yeah.

[00:20:19] Jade: Like, it's that community experience that I think is what, uh, is really valuable to us as, as humans and human creators.

 do you think that there's a ton of intersection around like, music and sports and this kind of like, collaborative, you know. A big event. Like a shared experience?

[00:20:35] Adam: Yeah. No, I, I mean, a, a huge amount, like the spectacle of it all. Uh, a bunch of people getting together to witness greatness in real time.

Like of course, you know, uh, there's what is a great play, but a great guitar solo, you know, like there's the sa it inspires the same kind of reaction in us, I think. I, you know, I, I say this all the time. the things that an athlete. Does, um, in terms of discipline, in terms of training their body, their mental focus are exactly the same thing that a, you know, a musician does when they're learning an instrument.

Um, there's a great jazz educator named Hal Galper who says that we are athletes of the fine muscles, so we are athletes of the fingers rather than the large muscles. And I always think about that, because it, really speaks to something, True about, I think both music and sports in that, you know, you're trying to, through discipline and focus and craft, you're trying to, create something and better yourself and then also create this wonderful community experience.

And, it's both like very focused on the self and also the community. And I think that's, uh, that is like what music is and that, so I take a lot of inspiration from sports and athletics 'cause, It seems like it's the same thing in different clothes.

[00:21:46] Jade: It's the same, there's the same elements of magic there, I think.

For sure.

[00:21:49] Adam: Oh yeah, yeah.

[00:21:50] Jade: in terms of like this community building aspect that we're talking about, what are some, what are some tips you would share with someone that's really doesn't have anyone to collaborate with and is really looking to build that, musical community?

[00:22:05] Adam: Yeah, well, like thinking about a, a sports game, one of them, like, I'm not actually a huge sports fan.

I think, that's one of the ironies about being excited about athletics is like, I don't actually know that much about sports. but yeah, going to a game, you know, you want to, your chances are they're gonna be fellow fans at a game, so go to a concert. Chances are they're gonna be fellow fans at a concert and fans of, the same kind of music that inspires you.

 even if it's like, you know. Find a producer, find a dj, find a, you know, beat maker, find somebody who is putting on some kind of live event, and go there and talk to the people there, who very likely are gonna be into the same kind of music you are because you chose, you made that decision to get out out of your house and go to this place, which might be kind of uncomfortable, might kinda dark, you might not, you know, these places like are usually a bar, so you might not drink alcohol.

So it's like not great. It's not a great welcoming environment for you usually, unfortunately. But the thing that is there is community and the thing that is there is, like-minded individuals and you start to really, really realize how important physical spaces are in the creation of community. and concerts are usually a great place for that because you are always going to be in the place where, there's going to be people that you vibe with.

 I think about. Um, in New York City, it doesn't exist anymore, unfortunately, but there's a venue called Rockwood that's, it was like three or four stages all in the same building, and there was always different musical events happening all the time. Like on the weekends, they would be, have music from like three.

PM to 1:00 AM in a different band every hour. It was just crazy. And the thing I loved the most about Rockwood, was just the fact that I, anytime I would go there, I would always see somebody that I, I knew and it was exciting. It's like, oh, where are you playing, man? Yeah, I'm playing Rockwood two like in an an hour.

It's like, oh, I'll see you at that set. And it was something very. Very special to be able to be in a physical space with people that I knew, that I would see and hang out with and, build and grow a community that way. Now, I know a lot of people don't have access to, uh, physical spaces for whatever reason, and so I.

You know, for me, uh, internet forums are the, were the ultimate, and I think to a certain extent are the ultimate way now of finding, uh, like-minded people and sharing tips. Uh, I was active on mx tabs.net, a k Sputnik music. I don't know when it changed, but like. talk base.com. I know seven string.org is where a lot of the, uh, metal musicians that I look up to got their start.

They're like bedroom producers in the two thousands. Like, uh, Misha Mansour and TOS Abbasi were just posting on forums, like sharing their, like, uh, their riffs and stuff. And, you know, there's Discord servers now there's Reddits. find some kind of community and share the stuff. it's hard. It takes a lot of concerted effort, but I promise you it's worth it.

And it is ultimately what it's all about. even if you don't, you don't, you don't wanna share the music making, at least share the music with other people and talk about your music. And the first time somebody listens to your music and then shares and importantly critiques your music, that's a special thing because critique is.

Constructive critique mind you. But critique means that somebody really listens to what you're doing and really cared about what you're doing, enough to tell you enough to get inside your head a little bit. and so that I think is such an important thing. Uh, I think about my, uh, mentors, my elders and my community that have critiqued what it is that I've done, and how much I grew because of that.

And I could only have that because of community. But yeah, it's, it's hard, it's hard to, it's hard to make friends as an adult. Um, you know, so, uh, that would be my advice though. Yeah.

[00:25:48] Jade: Do you think anyone, uh, have, have you heard any stories about people. collabing, like from like YouTube comments, like on your videos or anything like that.

Has anyone like emailed you and said, Hey, we met in the comments on one of your videos and now we're working together?

[00:25:59] Adam: Not the comments. That's a great, great question. Not the comments I have. I have, uh, met people who have met at Sun Gazer shows, so my band shows who have gotten together and played, I mean.

Uh, there's definitely like people who are, uh, part of the general Adam Neely, in the general like music theory community on YouTube that have gotten together and started making music. it's very cool. There's like a very, very cool music theory, nerdy community that is surrounding, uh, my channel and a couple others, that, uh, they, they do good, they do awesome, awesome stuff.

 comment sections maybe. It's a, it's a bit of a, a wild west out there, but you know.

[00:26:36] Jade: Totally.

[00:26:37] Adam: Yeah.

[00:26:38] Jade: Awesome. Well, Adam, really appreciate your time today. Yeah. Really appreciate your insights.

[00:26:41] Adam: Thank

you

so much.

[00:26:42] Jade: Thank you so much.

[00:26:42] Adam: Yeah.

[00:26:42] Jade: yeah, look forward to, uh, hearing what comes next for you.

[00:26:45] Adam: Awesome.

Thanks for having me.





Let us know what you think! Find us on LinkedIn, and Instagram, or connect with podcast host Dmitri Vietze on LinkedIn.


The Music Tectonics podcast goes beneath the surface of the music industry to explore how technology is changing the way business gets done. Weekly episodes include interviews with music tech movers & shakers, deep dives into seismic shifts, and more.



bottom of page