Machine Music, the Velvet Sundown, and Whale Pop Stars with Drew Thurlow

Today, I chat with Drew Thurlow, a strategist focused on how AI might be democratizing the music industry, rather than destroying it. We talk about how AI is changing everything from the creative process to copyright law, whether or not we should be worried about AI-generated music like the Velvet Sundown, and how AI is being used to surface the most popular songs made by whales in the sea. We also talk about this former Sony, Warner, and Pandora exec's upcoming new book "Machine Music: How AI Is Transforming Music's Next Act,'" set for release in 2026. Drew will definitely have you thinking new thoughts today!

The News:

Top US Copyright official Shira Perlmutter reinstated, as Appeals Court blocks firing by Trump

Reddit, Yahoo, Medium and more are adopting a new licensing standard to get compensated for AI scraping

Spotify peeved after 10,000 users sold data to build AI tools

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!

Join the Music Tectonics team and top music innovators by the beach for the best music tech event of the year:

6th Annual Music Tectonics Conference October 22-24, 2024 Santa Monica, California

Episode Transcript

Machine transcribed

[00:00:00] Dmitri: 

I have with me here, Drew Thurlow. He's a former professional musician, turned music executive and strategist. He served as senior Vice President of A&R at Sony Music. He's a current strategic advisor to the music tech investment fund, Mindset Music, and he's held leadership roles at both Pandora and Warner Music.

A graduate of Brown University with a Master's in technology and leadership (I'm jealous, that sounds cool). His writing has appeared in Billboard and he remains a sought after voice on the intersection of music, tech and innovation. And his first book, Machine Music: How AI Is Transforming Music's Next Act will be published by Rutledge in 2026.

So I'm excited to have you here. Drew. 

[00:00:40] Drew: 

I am excited to be here. Long time fan and friend of all things Rock Paper Scissors and music tech. 

[00:00:47] Dmitri: 

Awesome. Thank you so much. Yeah, no, it's exciting to have you, and since you're working on this book, I'm glad we're gonna talk to you about a bunch of AI stuff and I think we could start pretty broad and start to narrow in.

I understand you also teach AI mastermind classes as well like on a consulting basis with universities, that kind of thing? 

[00:01:04] Drew: 

Yeah. Yeah. It's funny, I started my own company a couple years ago when I left Sony and I wasn't really sure what I was gonna be specifically doing. It coincided with this AI revolution and people just started asking me about it and then I wrote a series for Billboard about it and I got this book deal and I wrote this book and so a lot of people have been asking me, so I do have this masterclass that I've been doing for university programs, but also for private companies and individuals.

Just a state of AI. What do you need to know? How can you integrate some of this tech into your workflow and do so ethically and where are we gonna be going from here? This is a lot to keep up on. 

[00:01:36] Dmitri: 

Yeah, cool. Well, sometimes you may walk into a corporate environment or something where one person is all hot on AI, but they want their executive team or something to kind of get caught up on something when they ask like, “Why is all this AI happening now?” How do you respond? 

[00:01:50] Drew: 

Yeah, it did. Isn't it funny how it just exploded in our lives. And the demarcation line really is in October of 21 when Chat GPT launched and became this ever immediate presence and. I immediately saw that it was going to be a very big deal in music and have a lot of impacts on the music space.

So it really is a confluence of all these factors. I mean, AI as a term was coined in the late fifties. Machine learning has been around since it's least as long. There has been other research and development in AI over the last several decades, but just like a lot of emerging tech, it kind of started in research labs and it started in academia and then it would have fits and starts into the public. 

And then it was really a confluence of Nvidia making these really powerful chips. It's a confluence of machine learning algorithms, being able to run in the way that they need to to process this crazy amount of data. And then we also, in the music and tech spaces, we've had a lot of layoffs, a lot of people out of work, a lot with skills and talent.

So now we have this incredibly vibrant startup community and specifically in music tech as well. I'm tracking more than 750 music tech startups and that's seven hundred and fifty, and we just have a lot of people who have got skills outta work, so we just kind of relate itself into this like, really great ecosystem we have now.

[00:02:58] Dmitri: 

Wait, so are you saying like the growth of the music tech ecosystem in terms of startups you think is a direct result of people who got laid off and then started their own thing? 

[00:03:06] Drew: 

Yes, I do. 

[00:03:08] Dmitri: 

Okay. 

[00:03:08] Drew:

I do. I mean, there's a lot of reasons why we have so many startups now. If you talk to founders from 20 years ago, the kind of challenges they had are not what we have today.

“I have this great idea, but I don't know anyone who can build it. I don't know any engineers or coders.” And that's, you know, in the era of vibe coding, everyone can get dev work done, and then, “Oh, I have this great idea and my cousin can build it, but I can't afford to host it.” And these days cloud storage is essentially free or close to it.

So the democratization of the way people record and write music with DAWs- we're kind of seeing that with startups. It's easy to get a startup. You can get an idea up in an hour and vibe code it on Base 44 or something and then host it on any one of the sites and so you have a lot of people with a lot of startups and you know, 750 is a lot and they're clearly not all credible and the vast majority of them anyway aren't gonna go anywhere. 

But it's interesting to see the patterns of what founders are trying to solve and where their capital is flowing and the plurality of them are using ai. So it really has a huge mark on the scene. 

[00:04:04] Dmitri: 

Yeah. In a way I, what I'm hearing you say is like this technology and, elements or versions of this technology has been around for 50, 60, 70 years.

In some ways obviously developing. It's kinda like Moore's Law used to have something to do with the, the cost of of memory or the speed 

[00:04:18] Drew: 

Every 18 months that the cost of storage would double. Or the amount of storage you could get for the same cost would double. 

[00:04:25] Dmitri: 

Yeah, exactly. and now it's having this exponential impact, I think, with all the AI.

But in some ways what I hear you saying is that the interfaces for AI have reached the masses. Like you used to have to be, a PhD level researcher to even do anything with artificial intelligence, but now there's these layers of interface on top, mostly in the form of chatbots for starters.

[00:04:46] Dmitri: 

That has just made it accessible to everyone else. 

[00:04:49] Drew: 

Yeah, it's true. And we're in the early days of AI, at least as a consumer tool, and you know now, we're seeing these agentic platforms start to, and like I'm still, I come across people who still have not yet been able to use or play with, or don't know what an agentic AI platform is, but a chat bot is something that you can interface with and it gives you information. An agentic AI system is one that is an agent and it can actually make decisions for you and do things for you. So in a music context, you can run.

You are recording through Izotope’s Ozone, which is an AI powered mastering program. It's great. It's been around for a long time. I have at least eight years, but now you have these age agentic platforms, which are your kind of co-pilot and your writing and recording process, and then there's all these different kinds of implications that it's being used on the industry and business side too.

It's where this is at the very first time that we've actually had access to AI and consumer level, but it's just gonna get bigger and more impactful in what we're doing, 

[00:05:39] Dmitri: 

You know, let's dive into that. You brought up the recording process, the composing and recording process.

What other ways do you see AI transforming those creative elements of the music business? 

[00:05:49] Drew: 

Yeah, in some ways it's a little bit of a continuation of what we're already seeing. I was a session musician 20 years ago, and I came up in an era where, in order to write music, you had to kind of get in a garage with other musicians and compose that old fashioned way.

And then if you wanted to make a commercial grade recording, you had to go to a studio, which was expensive for someone who was making all their money mowing lawns in Rhode Island. So it's been really interesting to see how the digital audio workstation phenomenon has really lowered the barrier for commercial grade recording.

And now everyone who buys a MacBook or an iPad gets GarageBand for free. And they can make music. Well, that's an example of digital technology lowering the barrier of writing and composing and recording. And now AI is just furthering that same phenomenon. So there's all different kinds of generative features, and I don't mean just mean Suno and Udio, but other kinds of generative features.

Logic Pro, the most popular digital audio workstation that is used by a really professional grade recordings. Has generative bass, keyboard, and drum tracks right in their DAW now as a plugin. So if you're a composer writing and recording, you can use AI to help give you some ideas and it can end up in the final mix.

And then of course, people are using large language models to write lyrics, and there are some actual music centric ones that are designed for like music lyrics and poetry, and then the writing and then the mixing and mastering process. You know, there's AI powered. Mixing where you can get a reference mix with some machine learning technology. 

It's not great, you know. If you're trying to make a great recording, you still need a mixing engineer to clean it up, but it can get you 80% of the way there. And then same thing with mastering and all this stuff can be done right in one platform. Where even a few years ago, you would have to write in one area and then you'd have to record it in a DAW.

Then you have to export that track to a mixing engineer, have them mix it, and then they would send it to a mastering engineer, and then that mastering engineer would send it to you or your label, your distributor. But all that value chain now can just be done like in one platform, and AI really makes it a lot cheaper and easier.

[00:07:39] Dmitri: 

You know, I think some of the concerns in the music industry around what you're talking about, the creative process and how AI taking over and you said, you know, we don't need to talk about Suno and Udio to talk about how it's impacting composing. We had a responsible AI summit a couple months ago and a few artists showed up and one of them said, “I'm actually a pretty successful jazz saxophone player and I have millions and millions of streams.  I'm making a living at this, but I see what's gonna happen is all this AI created music is gonna flood the market even more, and it's just gonna make it harder for musicians like me who play an instrument to make a living.” 

So I guess, the examples you gave are that type of musician, right? The saxophone player who's probably using a DAW at some point to record. 

[00:08:20] Drew: 

Yeah. 

[00:08:21] Dmitri: 

And using other technology, possibly effects, possibly, even just reverb, you know, or whatever. And depending on what other genres he would get into. but I think. The other side is, are we creating a new generation of creative people?

Maybe you don't call them musicians or artists because they're using chatbots to create things in generative AI projects, rather than learning how to technically play an instrument or learning any music theory or harmony or songwriting or whatever. 

I guess I wonder what's your take on sort of like, will there be a generative AI class of artists that is separate and how will that impact the music industry? 

[00:08:59] Drew: 

Yeah, I mean, I have a lot of thoughts on that. First of all, the person who is a jazz musician making a living, making jazz music, he's worried about AI music flooding the market, Like I get the concern. I really do.

Being a musician is hard. and especially being an independent musician is really hard. It's hard to scale and there is a lot of competition, but like, can we just. Nothing that drives me more insane than the Velvet Sundown phenomenon of a couple months ago. 

[00:09:21] Dmitri: 

Yeah. 

[00:09:22] Drew: 

It just drives me crazy. I think the music industry is looking for this narrative that generative AI is taking mindshare away from real artists.

And if you look at the data, with more than two years of generative AI now flooding the market and the ecosystem, there's more than 10 million generative AI songs now on DSPs. It's not most of them, but it's a measurable percentage of the 250 million songs out there.

Soon to be at least, somewhere between five and 10% now, generative AI that will go up. all of that music. According to research done by Spotify and Geezer, it's like a third of a percent of the royalty pool. Like, fans are not listening to this stuff. They don't like it. The people's relationships to the art that they love and the music they love is so much deeper than just the audio they hear.

And I feel like this Velvet Sundown, I mean, yeah, they have a million streams now. Everyone's curious. Everyone's going to check this out. Of the 1.7 trillion on-demand streams that are going to happen in the world in 2025, it is like a rounding error of that generative AI. Now, that could change, but I say to the artists and I gave a talk at Berkeley a couple months ago on this very topic, and a couple of the musicians were raising the same concern.

Generative AI is not coming for real fandom. I think people have a hard time envisioning this technology and how it might expand the market. I think we are used to streaming and artists being a certain way, and then now this new music is coming along and like it's gonna replace it. But no, I mean it's, we're creating a whole new thing and I think the democratization of creativity is really good.

And the other point to answer your question is that you've heard this argument throughout history. 

[00:10:46] Dmitri: 

Yeah. Every few years 

[00:10:47] Drew: 

When the polyphonic synthesizer- polyphonic synthesis, a synthesizer that can play more than one note at once. When it was used, when it first became part of popular music in the late 1970s, the American Federation of Musicians literally tried to ban the synthesizer and they claimed it was a robot and it wasn't real artistry and it was gonna put musicians out of work and that was not ancient history. That was like 40 years ago. 

[00:11:07] Dmitri: 

Yeah. 

[00:11:08] Drew: 

And now then Peter Gabriel used it and everyone's like, “Oh, Peter Gabriel's using it, then it's totally fine.” And now you can't hear a pop song these days that doesn't have a polyphonic synth in it. 

So I mean, doing research from my book, I can give you so many examples. But I won't go into all of them- Plug for my book. Read the book if you wanna learn more

[00:11:23] Dmitri: 

2026. 

[00:11:25] Drew: 

Yeah. But there's this, the idea that technology is going to kill human artistic endeavors is just an argument that's been made over and over and over again through.

[00:11:35] Dmitri: 

I wanna ask you about that. Can you see a picture where the emerging technology that democratizes the creation of music leads to a larger music market, leads to more people creating more people engaging in a way that actually builds the total addressable market of music? 

[00:11:49] Drew: 

Yes, absolutely. And I think we've had an interesting inflection point because Up until now, when recorded music started in the 1930s, up until now, we've had the same dynamic where the thing that is valuable is the song. So as someone who ran an A&R team at Sony and then, did A&R at Warners, the writing and the recording process was a sunk cost.

You tried to get to the song as quickly as possible and as cheaply as possible because then you had the song and you would monetize the song. How we monetized music has changed over the decades, but it's the same paradigm. Well, now we have this thing where millions of people are using generative AI and other emerging tech to make music, but the songs aren't valuable.

Generative AI isn't eligible for publishing copyrights. First of all, that's something. Copyright offices around the world are pretty aligned on Number two, no one wants that dumb song I made for my five-year-old's birthday party. Like it's just not gonna have value. What does have value? is the process, and that's something that's kind of been a new phenomenon in music.

So I think there's a huge opportunity for the music industry to monetize that process. And MIDia has done some great research on this and they call them super fans, but people who really love music can engage with audio in a really interesting, cool way. And it's not replacing your favorite artist that you listen to on Spotify.

It's something else, and it's a different thing, and I'm really excited about the tech. Not only is the tech cool, but I'm excited about the opportunity for general people to be creative and for there to be this new market. 

[00:13:10] Dmitri: 

I wanna talk to you about music fans and listeners, and the entire music business, but first, we have to talk about copyright and AI a little bit here.

 I don't know how to dive into it, but that's the other side of this is like, is copyright at stake, when large tech platforms are claiming fair use of scraping all the world's music and then creating, sounds and music that could be competing or at the very least is violating copyright, depending on how the courts rule.

[00:13:37] Drew: 

Yeah. Right. 

[00:13:38] Dmitri: 

What are your thoughts on how this is, emerging right now and where things will go with AI and copyright? 

[00:13:42] Drew: 

Well, I'm definitely excited to have an outcome because right now they think there's this, it's what if you don't know anything about AI or music, but you work in music? You probably know, have heard the rumblings about copyright and lawsuits around training models on copyrighted IP. 

I'm excited to have an outcome. Once we have some business certainty around this, I think there's gonna be a lot more capital flowing to the space, and I'm excited for that to be a reality, but we do need to answer the question like, are you allowed to train an AI model on someone's protected works?

And if you are, what happens to the output? And what if that makes money? And if you're not, what happens to that when you make money? And I think I have a guess on where it's gonna go based, not only on just following this pretty closely, but what some of the things have been said, and then some of the early court decisions that we've had, mostly around literature and books and not really any around music.

But there's also like, I would just caution the industry that even if it is fair use to train a model on music. Even if an AI, generative AI platform can take the entire history of recorded music and train it on, and then it's fair use. Like I just it doesn't necessarily mean what you think it's gonna mean.

History is littered with unintended outcomes of things. It doesn't necessarily mean it's gonna be terrible for artists. I don't think we would choose that outcome, but there's a lot of uncertain things and there's also a lot of reasons why a generative AI company would wanna license the content, even if it is fair use.

You talk to the guys at these companies, like it's annoying to scrape the internet. the data's messy. It's missing metadata, like get a feed. And I think that's where it's headed. And I think what we've seen in these literature cases is that it's fair use to train the model. But how you get the information matters.

And so I think it's gonna be very convoluted, the answer. I don't think it's gonna be as simple as like black or white. It's fine or it's not. But I do think that we will at least have an outcome by the end of 2025 into 2026. And from there we can just move forward and like really kind of open up some capital and new markets around this.

[00:15:31] Dmitri: 

We will be good to find out. When we started this conversation you mentioned with the launch of Chat GPT, a lot of people are engaging with this, so there's just more awareness regardless of whether it's the music industry. But you also said that you sometimes advise folks about how they could be using AI.

I mean, we talked about the composing and recording process, but how's AI transforming the rest of the music business? 

[00:15:53] Drew: 

Yeah, it's, I mean, it's everywhere, right? So, let's go through the value chain. You're a label or you're an independent artist and you have a song ready to release.

Well, there are now at least two that I know of, Legal AI chatbots trained on music industry law. Now, I would not recommend anyone outsources their entire legal advice to a chatbot yet, but they can help, right. And not only are independent labels using them, but so are independent artists. And then you have the annoying part of marketing, which when I was running an A&R team at Sony, I spent most of my time sadly yelling at artists to be on social media more.

And that's not what anyone signs up for when they do A&R, but the reality is that's how you meet your audience. As an aside, I love how, we don't really respect the creation process enough and we make artists be their own digital marketing people and get online, but the answer to that isn't, let's just pause that and like let artists be artists.

The answer to that is like, let's use AI to make even more content and more slop, but for better or worse. That's how artists get discovered. So there's all different kinds of generative AI that can make your Spotify canvas a lyric video, write SEO copy, write ad copy, and create content for your social media.

There's a lot of great social for the music space, social media, calendar, platform generation, scheduling. That can help manage that unwieldy, time consuming process. And then there are a few companies that I really like that help you with streaming optimization. And there's a lot that's being done with catalog and deep catalog.

Everything from the avatars, the AB Abbas avatar and the ABBA voyage, to revitalizing IP to biopics. And a lot of interesting stuff being done with, deep catalog and AI. 

[00:17:28] Dmitri: 

I just had a weird thought that I've never thought before, and I don't know why I didn't think about it until now, but when you were talking about all the marketing AI slop that can be created as well as the useful stuff, I was just remembering screensavers.

Remember screensavers when everyone had to customize their computers? 

[00:17:46] Drew: 

Yeah. Wait, if I remember correctly, you needed a screensaver, right? Because like you didn't want the pixels on your. Monitor a freeze. It burnout 

[00:17:53] Dmitri: 

at first in privacy when you walk away from your desk.

[00:17:55] Drew: 

Oh yeah. 

[00:17:56] Dmitri: 

When people went to offices 

[00:17:58] Drew: 

Some artists out there should like take some of the popular like window screensavers and use it in their creative campaign 

[00:18:04] Dmitri: 

or vice versa. Maybe artists should be selling screensavers. 

[00:18:10] Drew: 

They're easy to make now. 

[00:18:11] Dmitri: 

Yeah. Yeah. I would assume they'd be very easy. I don't know how to tell your computer what to do with it, but, okay. we've gotta take a quick break, but I do want to come back to this issue about how AI's impacting fans.

We got into it a little bit, so when we come back, I'll ask you about that. We'll be right back.

Okay. Drew, this has been a blast. I kind of expected, since you said you teach masterclasses, I was sort of like picturing myself as a student in a class. But this has been a fabulous conversation, real dynamic. You've got me thinking about new ideas and so forth. I wanna dig into how AI's impacting music fans and listeners.

What are your thoughts there? 

[00:18:46] Drew: 

Yeah, that's really kind of interesting. I think of these three pillars, how it's the creation and writing and recording process for artists, how it affects industry, the business and rights holders, and the third pillar is how is it infecting fan listening and engagement?

And I think that this is probably the most important one because this is really what's gonna determine. The other two. This is gonna really inform where the money is and

[00:19:06] Dmitri: 

What do people want? 

[00:19:07] Drew: 

Yeah. What are fan behaviors telling us? What is it? What do they want us to build?

What do they want artists to be like? And I think we take it for granted that streaming is the way that we're all going to consume music, at least most people for the, yeah. And like that's, I wouldn't make that assumption necessarily. 

[00:19:22] Dmitri: 

Yeah. 

[00:19:23] Drew: 

There's a lot of interesting ways that this tech is being applied and I'll bring up MIDiA, again, I can't talk enough about how great MIDiA is.

I love MIDiA, their research is amazing and, they define a super fan as the top 20% of all music listeners, and there's a lot of data on how much these listeners like love just engaging with audio and that this tech now is available to put this in their hands. Like, you love Beyonce. Well now you can pull out.

The stems and just listen to each individual track and listen just to her voice or, you're curious, you wanna hear what Beyonce sounds like as reggae, press your button and make it reggae or you know, fan fiction. Like you could do mashups and remixes and, you know, there's a lot to work out.

So that's the fan behavior. Obviously Beyonce on some level should have some say in how her music and IP is used, and the licensing and rights holders need to decide how. Much interaction fans are gonna have with it. But I think we've learned through this UGC era and the social media era that relinquishing control over your music once you release it, can be very powerful.

And that's anthemic to me, you know. My first A&R job and marketing job at Warner is, I was working at Nonesuch and I had David Byrne and the Black Keys and Wilco and Emmylou Harris and a lot of West African music and composers 

[00:20:29] Dmitri: 

I just remember that's when we met Drew. 

[00:20:30] Drew: 

Yeah. That was your, yeah, and like, I was taught like David Byrne worked really hard to record this song and we're gonna release it the way he wants it and not gonna let anyone fuck with it. But that's, no, not the era that Gen Z and millennials really expect, they really wanna engage with this music.

And if you relinquish control over it, some of the most viral moments have happened when someone sped up your song on TikTok and has created a whole new audience for you. So I'm really excited and the rights holders are kind of open to this now. They're open to letting fans, so I'm excited to see what gets built in that space and is it like fan fiction?

Like what is it? And then a lesson that we've learned from AI is that sometimes tech takes a little while to mature. And I would say that's the same thing with the AR VR mixed reality future, which I'm a big, big believer in. And I think we're gonna be living in a world where a lot of us are spending our entertainment in virtual or augmented reality or mixed reality. 

And music is going to play a primary role. And those environments kind of think about it as merging traditional music, listening with gaming, especially when a lot of our time is freed up in the car and these. AR VR headsets get a lot more interactive and I just think there's gonna be a lot of interesting things that we're gonna do with not only the consumer creator, where people can make music in cool ways but they can engage with their favorite artists in interesting ways. 

And then even like on the live side, there's a lot of interesting things happening with, live, but both, everything from up, from the stadium level down to the club level in terms of how much easier it's getting for fans to buy tickets and see what they want and how much better the data is getting, with the touring artists at the club level.

And, it feels like this is all happening very quickly and it's because it kind of is. 

[00:22:04] Dmitri: 

I like the optimism of all this. Drew, it's making me excited 

[00:22:08] Drew: 

My thesis on AI is very optimistic and I wouldn't have said that before I did my book research, but I just kept coming across these examples where like my open my book with the story of when the camera was invented.

It was a common belief that it would kill painting and it like actually invented impressionism. You can draw a straight line between portrait painters being freed up to. Explore other kinds and there's just so many examples of that. So I'm just a big believer in this actually opening doors.

[00:22:33] Dmitri:

 Yeah. It's interesting if you put the emphasis on creativity, there's like a pendulum or spectrum between creativity and commerce, and sometimes maybe commerce swings the other way, but creativity increases. But then it swings back the other way because now that these creative categories have been created, there starts to be for fans to engage or audiences or patrons or whatever that becomes, monetizable.

Let's do a quick speed round. I have two questions for you that I thought we could just jump into real quick. I've got more questions than that as well, but you talked a little bit about working in A&R at Warner wearing your former record level hat and you, you worked at Sony as well.

How do you think labels are responding to AI and how do you think they should be responding? 

[00:23:10] Drew: 

Yeah, I gotta say I'm pretty optimistic there too. I think that no one in the industry wants to relive the file sharing days. And that was an example. I mean, could the industry have bundled that any worse?

And so I don't see that with AI. I think the people that are running point at the three majors and some stuff that, Merlin's doing and even some of the bigger indie labels, I think people are very like, have their ear to the ground. They understand this tech. I think they have to be careful, but they're being careful.

I think they're making smart investments when they need to. Of course, especially if you're a major label, you have to worry about your clients, which are your artists. Sony, I just started being an executive at Sony recently, like I know they would love to do a lot more, but you can't piss off Adele and I don't think Adele is super necessarily on board with a lot of this AI tech.

So you gotta move slowly. But they've made some interesting investments. They've made some interesting partnerships. And I think the industry is managing this much better. And that's not to say that they haven't laid off a lot of people the last couple years. They have. It's been a tough, tough economic climate.

But I think that the rights holders and the ecosystem and the artists community is gonna be a lot stronger because I do think this is not being managed the way file sharing was managed. I think the industry started to get their relationship with technology a lot, lot more by embracing streaming.

And I think I just see more learnings and they're gonna manage this pretty well. 

[00:24:19] Dmitri: 

Oh, great. Okay. Same question for streaming services. How are they responding to AI? How do you think they should be since you worked at Pandora? 

[00:24:26] Drew: 

Yeah. Yeah. I mean, I think that there's less that they really need to to worry about because they don't own IP.

And I know that the DSPs would love to offer some of these tools to consumers. They want to be able to offer these kind of mix up mashup, STEM separation, but that requires licensing. I think that they can build the product pretty easily. They can build the tech pretty easily, but they need the licensing and that's not gonna happen until the rights holders are ready and that's gonna take a while.

Spotify is releasing, we’ve been told, a what they call crossfade. The thing is called Spotify Mix. It's basically just a crossfade feature where, like a DJ feature, where you can play one song on Spotify and then mix it directly into another song so it doesn't have any space in between.

And there might be some like light beat matching there, basically like a DJ feature. That's like the first step Spotify's making in having the user actually interact with audio a little bit more than just passive listening. But I think in the near future, we're gonna have to start to see some of these tools built right into the DSP client where you're gonna be able to do a lot more than just listening.

[00:25:25] Dmitri: 

Okay, let's go back to the big context again. It appears that AI is gonna be as transformative as, say, the internet or mobile computing. I'm curious, do you agree? Why or why not? 

[00:25:36] Drew: 

Yeah, I would say even bigger, right? Like it's really the internet. I mean, I've heard the analogy around electricity. I mean, that seems pretty apt to be honest. It's 

[00:25:44] Dmitri: 

like the electricity of your brain. 

[00:25:47] Drew: 

Well, it's called a neural network for a reason. 

[00:25:49] Dmitri: 

Yeah. 

[00:25:49] Drew: 

I mean, the way that AI works, it's really, this entire technology is built off Alan Turing's initial research was, is that, you could classically compute something, but do it in a way that mimics the way neurons work in a brain.

So it's, it really is like a neural network for a reason. So I, yeah, I mean, it's gonna be that disruptive. It really is. And I think I completely agree. I think the mobile transformation is only looking bigger and bigger as we get further away from the proliferation of smartphones. I think it took a while for the industry to be like, oh, that was actually a pretty big deal.

[00:26:20] Dmitri: 

Yeah. 

[00:26:20] Drew: 

That gave us DSPs and I don't know that we would have DSPs if the mobile revolution haven't happened, but I think AI is like a, another magnitude bigger than that. 

[00:26:29] Dmitri: 

Yeah. Okay, now to come back to earth, what are some practical AI use cases that will become used by most people in the coming few years since you wrote this book you're looking at?

The music, but you're looking at the context beyond that too. I'll take either, whether it's music specific or just like what can we expect for all of us? Um, yeah, 

[00:26:45] Drew: 

I mean, I think the age of Agentic AI is really about to be a lot bigger part of our lives. We really are in early days. And the idea that you go to a chatbot, like Chat GPT or Anthropic or Claude and you say, give me a recipe for dinner, like, and it gives you a recipe for dinner, I think that's gonna be seen as like.

Very rudimentary very soon, because pretty soon you'll be like, give me a recipe for dinner, and it'll give you a recipe, and then it'll go to your shopping app and it'll fill out the cart for you, and then it'll order it, and then it'll set schedule delivery. Before you know it, you'll have your, and then when robotics get into the picture, then it'll cook it for you.

But that's another podcast conversation. but in a music context, there's still a barrier to using DAWs. There's still a barrier to using plugins and outboard gear, and that's something I struggle with as a musician. I'm not particularly technically minded.

I had trouble using gear that was made by engineers, but I can go to a do and I can say, Hey I really want my vocals to sound dreamy. And the copilot, the agentic copilot in my do will know that. I mean, I want short slapback delay and some reverb side chain compression. It'll go to my DAW and it'll do it.

And I don't need to learn how to use a plugin. I don't need to troubleshoot any complicated tech. I don't need to upload or download any updates to my software. And I think we're just gonna see more and more of those like hyper intelligent copilots. And that's same thing with the music space. I mean the, digital marketing is about to be radically reformed with a lot of automated processes for digital marketing. 

[00:28:04] Dmitri: 

Yeah. It's like your agentic marketer tells you everything you need for distribution and creates an editorial calendar for six months and produces all the content, the photos, the videos, the posts and your age agentic editor checks to make sure it meets all your criteria.

[00:28:20] Drew: 

Yeah. And the EU AI Act, which is the world's most comprehensive AI legislation, does a pretty good job of spelling out what they call human in the loop, where the human should stay centered in this process. You should not automate away all of this, but you can use these tools to replace some of the things that are tedious or kind of mind numbing or not particularly creative.

[00:28:42] Dmitri: 

Yeah, I'm curious. Like right now it feels like you need to keep the human component, like anything you've done on a chatbot or, any AI platform is super, super shitty and boring if you haven't looked at it and decided and tweaked it and really gotten it to what. A human would wanna see, I'm not just talking six fingers here or something like that on top.

Yeah. You can see it on LinkedIn. You're just 

[00:29:03] Drew: 

I know. Can we just, can we just not with the AI copy? It's like so clearly obvious to me when people are just using AI copy, there's a cadence to the way large language models work. Like the either or, and the, it's just like, can we just 

[00:29:17] Dmitri: 

In some ways that's a good thing because then you know when humans are involved.

Yeah. You know what I mean? Like in a, in a way it's like you can feel the soul in writing. and I think it's gonna be the same in music too. Yes. And you, at the very beginning of the podcast, you said something about it's not just the audio that people wanna hear. There's more, more to it as well.

But like you said, there are lots of use cases. I definitely would love it if I could get my meals suggested, purchased, and even made for me. That'd be awesome. I'd love that. 

[00:29:45] Drew: 

I know. 

[00:29:46] Dmitri: 

Okay. This has been an amazing overview and like I said, I did not expect to go down so many rabbit holes

It’s been really creative and really fun to have this conversation. Let's wrap up by asking you to go bananas. Drew, what is the craziest future of society and music look like with AI? When you think of your wildest imagination of where this is gonna go, paint us a picture. 

[00:30:10] Drew: 

Okay. I have this pretty banana theory that actually seems kind of logical to me.

But so one of the things that machine learning does really well with is taking large, massive data sets and making sense of it, and one of the earliest uses was, and I'm gonna get this research wrong, but someone in academia has recorded like every whale sound ever recorded on the planet.

And they've crunched this data and they're like, oh, whales are actually talking to each other in a language that's a lot more like a human language than we realize. And they've noticed that some of the whales, when they sing, some of the other whales all shut up. I have a theory that not only are we gonna be able to communicate with.

Using this really powerful, this might include we might need to get to like quantum computing, which is a little, much more powerful computing. But I have a theory that we're going to be able to communicate with animals because we've able to just process this data and that like we're gonna discover like animals, they're like animal pop stars.

 I bet there's whales like that. Other whales are like, oh yeah, that guy's songs are awesome. Or that girl's songs are awesome. And there's like, oh, shut up when you hear it because it's beautiful. I honestly think that, I think that we are gonna like, discover that they're like animal pop stars and that we're gonna be able to, that honestly doesn't, maybe I spend too much time thinking about the future, but that doesn't seem too bananas to me.

[00:31:20] Dmitri: 

No, actually, the way you say it, it doesn't sound bananas. I'm like, that's not as bananas as I expected. But then when I think of it, I'm like, wait, are we actually talking about this? When I was a kid, I used to listen to Paul Winter who was a, soprano saxophone player in New York. But he did recordings with whales.

[00:31:35] Drew: 

Yeah, 

[00:31:38] Dmitri: 

but maybe, maybe he knew something we didn't. 

[00:31:40] Drew: 

Yeah, like there's probably a bird out there who's like the best bird as far as other birds are concerned. The best bird singer. And like, they get a lot of attention when they open their beak or whatever. So, and I think we now have the computing power to be able to really make sense of all that.

[00:31:54] Dmitri: 

Amazing. Drew, what an amazing vision for how we can crunch all this data and be able to communicate across species and make music with whales. There's gonna be a new supergroup of whales and birds together. 

[00:32:06] Drew: 

I just lost like every potential masterclass client with that one.

[00:32:09] Dmitri: 

I don't think so.

I think what you're saying is like, there's so much data that we'll be able to parse. I mean,  animal language is something that we haven't, like, we've got the affection with pets that is a way of communication, et cetera, or, some of the primal stuff. Like there's a lion or a moose about to charge me.

You know, that's communication in some way. 

[00:32:29] Drew: 

Yeah. Right. 

[00:32:30] Dmitri: 

At the deepest level. And, it's a cool vision. All right, Drew, this has been a blast. I really appreciate you taking the time. I'm excited to see your book when it comes out. I definitely have to read it now. I think there's probably some examples of transformations and disruptions from the past that I would love to know about, as well as kind of a vision for where things are going.

Drew Thurlow, it's been a blast. 

[00:32:49] Drew: 

Thank you so much, and I'll see everyone at Music Tectonics in Santa Monica in early November.

Let us know what you think! Tweet @MusicTectonics, find us on LinkedIn, Facebook and Instagram, or connect with podcast host Dmitri Vietze on LinkedIn, Twitter, and Facebook.

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.