This is a live recording from our online Seismic Event of May 8th. Join us as we navigate the brave new world of generative AI music with some of the industry's sharpest minds. Amadea Choplin, COO of PEX, Charles Alexander, of Vinil and Remington Scott, from Hyperreal shed light on the pressing need for tracing, control, and monetization in an era where AI can replicate musical artistry with uncanny precision.
Next Seismic Activity event!
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
0:00:10 - Dmitri
All right, welcome back to Music Tectonics, where we go beneath the surface of music and tech. I'm your host, Dmitri Vietze. I'm also the founder and CEO of Rock Paper Scissors, the PR and marketing firm that specializes in music, tech and innovation. Get ready for seismic activity. So, if you're listening to the Music Tectonics podcast, you are time traveling. What you were about to hear was a live event you could have attended. Today's episode came from our online seismic activity event series. So if you're listening in the podcast land, check out musictectonics.com and you can be a part of our audience in a future edition as well. So in just a few minutes, I'll be bringing on some very special guests to dive in our conversation. But before we get started, I have a few announcements.
The music tectonics conference will be back on the beach in Santa Monica, California, October 22nd to 24th 2024. So please mark your calendars. A limited number of super early bird tickets are on sale now, so head over to musictectonics.com slash conference to get yours. Podcast listeners, make sure you're getting the musictectonics.com newsletter for updates like this about our conference, as well as this seismic activity you're listening to. And if you want to talk about making a splash at the conference, get in touch with Shaley Inkenbrook on our team. In other news, our next seismic activity online event will take place June 12th at 10 am Pacific, 1 pm Eastern.
And this one is on industry interface creator tools and the recording industry, that confluence we love to talk about it, music Tectonics. We'll have some demos from cool innovative companies before diving into a panel around the convergence of creator tools and the recording industry, so we'll actually have musical instruments that you can check out. So here we go into the full show. This seismic activity licensed to generate. Is this music real? So I want to bring on with me our presenters, our panelists, amadea, charles and Remington, if we can have them here on the mix with us. Welcome, guys.
Hey how's everybody doing?
0:02:27 - Amadea
Thanks for having us.
0:02:29 - Dmitri
So I want to ask what's going on with AI that we need to be talking about name, image and likeness, as well as attribution, all over again? Amadea, I'll introduce you. Amadea Choplin is the Chief Operating Officer at PEX, a global leader in digital rights technology. Pex is focused on delivering industry-leading identification technology to match audio, melody and video content across the web for rights holders, distributors and brands globally. On a mission to enable the fair and transparent use of copyright online, Amadea provides an expert understanding of the ever-evolving copyright and media landscapes. So, Amadea, I'm curious why do we need to be talking about these things today?
0:03:11 - Amadea
Yeah, I mean, I think we're talking about it because generative AI music advanced really quickly and is really good now.
So I think that's the first reason why we're talking about it. And the fact that it's really good and can generate music and content at scale means that there's a renewed fear of this loss of control over your image, likeness, name, et cetera. And so, just like with the birth of user generated platforms, where we had this fear over a loss of control over who's using our music and what are they doing with it and how are they sharing it, where is it once? I put it out there, now we have the same fear of loss of control when it comes to generative AI music and the use of our name, image and likeness. And so I think that's the main topic why attribution becomes essential again of how do we trace that music, how do we regain control over it, potentially, how do we monetize it, et cetera. And those are all the questions that are in discussion today and that are really interesting and complex, and I think that's why we're talking about it.
0:04:30 - Dmitri
Charles, Alexander, Billy Joel told you to move to Nashville. So, you did. Data is your love language. After working as a bioinformaticist in the biomedical research on projects related to the human genome project, you brought those skills to the music industry. You're a pioneer in the streaming, marketing, playlisting and strategy space and the CEO and co-founder of Vinyl V-I-N-I-L. Why don't you add a little bit about why you think we need to be talking about this right now?
0:05:02 - Charles
Yeah, and, by the way, my co-founders, sada Garba and Jeremy Brook, are in the call and I think the reason we're talking about it is because the tech is so good, right. So just in the last year, from you know, heart of my Sleeve to now, I think our minds were blown when Heart of my Sleeve came on, but then now we've got Udio and that is a whole next level kind of thing, and the guy who was, you know, deep faked on Heart of my Sleeve is deep faking other people to do diss tracks, right. So, and I think it's so good, and currently, you know, present company accepted, there isn't really an efficient attribution mechanism and consent and control is an issue and that's just the generative AI piece, that's the NIL piece. So we just need to establish mechanisms that are visible for folks to basically regain, you know, to give consent, regain, control and then, if it's appropriate, to be compensated for those assets, whatever they might be AI generated or not AI generated.
0:06:20 - Dmitri
So we also have with us Remington Scott, who's the founder and CEO of HyperReal, a global leader in digital identity, innovation, performance and engagement. His work has powered Academy Award-winning VFX productions of the highest-grossing films and the biggest-selling video games. He's currently focused on identity monetization and protection across all AI ecosystems. Welcome, remington. What are the challenges for rights holders and musical creators in this emerging era?
0:06:49 - Remington
The challenges are the technology is shifting sand. I mean, we try to get a grasp on what is going to be the next technology that moves this equation forward. That moves this equation forward and then it shifts and it moves faster than you think. So we have to think about how to be ahead of the game. You know and predict what this looks like, as it's happening and as it's moving very, you know, very quickly. So these are very real challenges and the challenges spread now like a spider web. It goes from the talent to their IP, to their management agencies and elements and the technologies that underlie it in a way that you can have consent. It's going to become devastating.
0:07:55 - Dmitri
Amadei, I'm curious from your perspective also about the challenges PEX, having been in this for a while trying to help figure out where music is going, how it's being used, where it's being uploaded. You've been watching this in a pre-AI kind of setting, but I'm curious now that you're watching this emerge what you think the biggest challenges are as well.
0:08:15 - Amadea
Yeah, I think a lot of the same intensified right. So I think one of the challenges around AI is its scale. So it's so much quicker and so much easier to generate this AI content and you can do it at a scale that's unprecedented. In some ways, that was some of the issues with user-generated content too, ray.
We track a number of UGC platforms. We have some 25 billion pieces of content and work in our database that we've tracked over the years in order to attribute that content back to its rights holders. With AI, that scale multiplies and infringement becomes faster, easier, much bigger. So I think that that's one of the key issues. And then, as we've developed technologies also to be able to identify works that have been, for example, heavily modified, sped up, pitched up, cover versions that might not be named as such, except to be able to attribute that back to the original rights holders, there now needs to be new technology that enable the identification of voice, image, et cetera, and so these are new technologies that need to be developed and lots of challenges ahead to be able to link back the AI generated work to its rightful owner content to video that they may not have had the rights to do, or the platform that it's on doesn't have a license or a way to either attribute or pay for the use of that.
0:10:12 - Dmitri
And so first you have this new upload ability that's creating this kind of crisis that that pecs kind of came in to sort of try to figure out how to help the musical rights holders and other rights holders figure out like where is their stuff, is it, you know, should they be collecting money on it? Should they be credited for it, or whatever? And then you start to get like the remix vibes, where it gets a little bit harder, or the cover song vibes, which gets a little bit harder. But then you have these tools that allow more and more people to do it. And also, I would assume, like the other challenge of that on the day is like trackability. You know, like if somebody just uploads an existing song, we had this kind of world of um fingerprinting of audio, where you could see where stuff is and be able to track it.
What's that?
0:10:55 - Amadea
music recognition technology in general.
0:10:58 - Dmitri
Yeah, right, um, but now you might not recognize it, but it's still derivative, right.
0:11:13 - Amadea
Exactly. I mean, I think you know it's right now some of the things that are recognized as AI outputs, but then and then trace back to original rights holders. Some of it is because, in a way, it's bad. Generative AI, right, it'll still include a sample of the work that it was trained on, or something like that, in a world where generative AI is doing what it's built to do that might not be the case, and so it becomes much harder.
That being said, you know you might be writing a song that you think is original and creative, and it turns out it sounds just like a Beatles song. Right, and it might be unintentional, but you're still infringing on that work. You know AI, in a way, can do the same, and so, in a way, there's a lot of the same problems of identification than before.
0:12:09 - Dmitri
Yeah, Charles, you are a musical artist yourself and have managed artists and been involved with promotion and playlisting and stuff like that. From that perspective, what are the biggest challenges you see for rights holders and musical musical creators with all this emerging?
0:12:25 - Charles
I think the biggest thing is trust, right? So whether I create an asset by going to the studio and playing and writing a song and all that stuff, or whether I choose to use an AI tool there was this amazing thing at South by a platform called Amy whether I created, create it in an organic fashion and that feels so old school, to say, the last year old school or whether I use an AI tool, whatever that might be, I want people to be able to trust that I put it up and if I so, the binary is okay, I put this out, you know, and I can choose to communicate with my fan base. All right, I used AI tools and I'll be upfront about. At least I will be upfront about it if I choose to do it that way. Or I can say, okay, this isn't me, don't touch it with a 10-foot pole, to be able to visually or audibly confirm to whomever is listening, watching what are consuming that yeah, I authorized it, I verified it, and so that should be the way it works.
And if you see something else out in the wild and I didn't authorize that, then don't trust that. It came from me. And that applies across the board, right, it's not just music, it's any kind of form of communication. It's an election year, so there's a lot of that kind of stuff running around, but I think the biggest thing is trust and then to be able to. You know, I want to be able to control and consent over whatever I put out and then, if it's appropriate, I want to be compensated for that stuff.
0:14:01 - Dmitri
I think we've done a good job of not making this a promotional panel, and I do want to get into what each of your companies do, because I think the reason you're here is because you have expertise. You're building out things that are super relevant. So I do want to get into that. But before we go there, let's keep it a little bit broad. How should we be protecting musical creators and rights holders?
Remington, we've talked a little bit about, kind of like why we're talking about this, what the big challenges are. Now let's go into what are some possible solutions. So let's go big picture and then we'll get into your individual companies. Also, before you get started, amadea, charles, all three of you feel free to. I don't want this to be like a round robin over and over and over again. Feel free to jump in. I'd love for you to ask each other questions and to make this a conversation. But let feel free to jump in. I'd love for you to ask each other questions and to make this a conversation, but let's let's start with you, remington, then we'll go there.
0:14:49 - Remington
How should we be protecting folks? Yeah, this is, this is what we're very concerned about the protection of your digital DNA and and that DNA is the fundamental parts. If you, if you atomize, you break down what it is that makes us human, what it is that we're doing in a performance space. That DNA is your voice, your emotion, it's your logic and it's how you look on its most fundamental levels. And these pieces, you have to now find a way to be able to protect it in AI ecosystems, because each of those components, there are companies building valuable valuations on their companies taking those components and sampling that DNA for use across broad markets, for use across broad markets.
So we have companies using voices and companies using images and all these different parts of logic, all these parts of this of the generative AI landscape. So it's about protecting your DNA, being able to make sure that you can authorize and authenticate how your DNA is being used and to make sure that you're tracking all of these uses. We know AI generally, machine learning can be detrimental to an industry. It could affect us negatively, but you can't negate all the positive outcomes of what AI is going to deliver for creators, and so we have to empower the creators to be able to use the tools with their DNA, and I think that's going to be part of what we're going to see, you know.
0:16:43 - Charles
I think Scott's right on there. You guys probably heard this last week about Randy Travis's new release. You know where that came from and you know. So if you don't know who Randy Travis is, steve Earle, like in the 80s and the 90s, called this thing called the country music credibility scare, you know. So there's a whole bunch of artists who are really substantial who came out of that movement, and randy travis was one of them, and so it about 10 years ago, in 2013, he had a stroke and now he can't sing, you know. And so there was this huge feature on CBS Sunday Morning this Sunday about how his producer of his entire career, kyle Lenning, warner Music all got together and decided to create a new single, them training the voice model on 42 tracks from his previous releases, and then Kyle lovingly and painstakingly working with the voice model and the surrogate voice and the the stuff that all the stuff was trained on and putting out this beautiful, beautiful, beautiful song. And so to Scott's point.
You know, I just think there's an example of what AI can do in the positive light, and this is all approved, by the way, by Randy and his estate and his wife and all that stuff. And so, whether someone chooses to use the tools or not, it should be completely in their purview and we need to have tools that allow that. And if it is okay, I'm going to go ahead and use AI to do this, or even if I choose not to use AI and create something, we just need to be able to enable, control and consent and so, and agency or what it is that you're doing if you're a creative artist. So I think, for me, and then just you know, to y'all's point tracking and just making sure people trust what they see, you know, I think that's where I think our focus is.
0:19:03 - Amadea
Yeah, I agree, I think it's all about control. But I think that control comes from two sides. One of them is a legal, regulatory one. I think contractually, name and engine likeness has been a little bit of an afterthought until now. It was maybe thought about for marketing purposes and things like that, but not from the central perspective that it becomes with AI. So from the legal perspective, the contracts kind of need to be adapted and clarified and tweaked towards enabling control over generative AI. And then I think and there might be a regulatory portion there too over what is, how do we standardize this? How do we ensure that some level of fairness across the board? And then there's a technological part to the solution. That's around tools and tracking. So giving tools for consent and use and then giving tools to and enforcing tools to track when infringement is in place.
0:20:29 - Dmitri
I think about the two sides of this too. There's the moment of creation, where there's this issue around training data, where there's opportunities to protect musical creators and rights holders, and then you have on the output side, once something's created, what do you then do to attribute people after the fact? So there's the training side, and then there's the kind of the derivative side, to who and how you get to monetize things as well. Those, those two sides, to me seem like an interesting point of sort of like the ethical dilemma and how to how to protect creators and rights holders.
0:21:09 - Amadea
Yeah. So I mean, maybe that leads into what we're working on, yeah perfect.
And I'm happy to kick that off, but that's how we come in at Pax. So I think there is a multi-pronged approach here. First, it's on the data training side. I think there's some good actors in the AI space that want to certify their training data and show that copyrighted works have not been included in their training data, and so we can run their training data sets through our search tool in order to find whether there's been any copyrighted works used in there. So that's one part of the equation is certifying those training sets.
0:21:55 - Dmitri
Can I ask about that, armadale Like? So they may have copyrighted works but they may have permission to use it too. Are they able to use that tool to sort of identify both copyrighted works and specific works that have been opted into their training data of?
0:22:09 - Amadea
course yes, yeah, okay, got it, cool okay um, and then, on the other end, if, um, if you're a rights holder and you're trying to track whether the name, image and likeness of one of your artists has been used One thing we've worked on and maybe to clarify. First, I think there's a lot of companies out there who are doing is this AI or is it not? We're a little less concerned about that. We think AI is getting better by the minute, and I think that the two main technologies that are out there to do that right now are either watermarking but the latest research shows that if you intend to remove a watermark, a watermark is always potentially removable. So if there's an intended infringement there, that's not necessarily the best solution and then the other one is artifact detection and so. But that needs to be trained on the output of the generative AI. So it's looking for things that are, you know, an artifact that's in the output. But so, by definition, you're always one step behind the generative AI there, because you have to, you have to train your tracking mechanism on that output and on the flaws of it. So that's always a cat and mouse game.
So we're more concerned about hey, what are the rights that are important here? What are we trying to protect. And so, for trying to protect someone's voice, for example, we've built a voice identification technology that's able to say okay, this is Dimitri's voice that's being used in this work. Some of it might be his own, some of it might be AI generated, because, you know, can't get enough of Dimitri's voice voice, um. And then uh, we uh. And then we run that through our catalog with known copyrighted works of rights holders and we can see, okay, is this part of dimitri's known work, um, and if it's not, then it's um, it's going to be an ai generated, versionated version, right? So that's what we're focused on tracking is the use of someone's voice right now and being able to detect whether that's been AI-generated or not.
0:24:37 - Dmitri
Super cool to see those pivots by PIX to adapt to what's going on, Charles. What about vinyl? How does vinyl fit into this landscape?
0:24:44 - Charles
Yeah, yeah, charles. What about vinyl? How does vinyl fit into this landscape? Yeah, yeah, and to Amalia's point, my colleague Jeremy likes to say it's whack-a-mole if you're trying to keep up with what's coming.
So vinyl is spelled V-I-N-I-L, which stands for voice, identity, name, image and likeness, and our particular approach is we work with the authorized either the artists themselves or an authorized representative to basically make sure that whoever gets an account with us is the actual source of where all these creative assets is going to come.
So that's step one for us.
We need to be sure that the source is someone we can trust, verify and say this is the official creator and then, once we've done that, what ends up happening is they can upload the asset to a media library and then, just like most people in our space, we are fingerprinting and then enabling tracking that asset across wherever you might see it.
So, whereas PEX incredible technology is looking into everything that's out there in the space which again is amazing we are taking the approach that if we ingest and upload, whatever the content is, if we've seen it and someone sees a particular asset, that particular asset out in the wild, they can then verify and make sure that that was actually what we call vinylized. And so you know, we get a little purple check mark on the thing and so you can you know, you can know that the creator or whoever the author is, actually authorized that. And then if you don't see a purple checkmark, then we can't vouch for that veracity of whether that ad is something that's authorized. And it also besides audio and video and whatever. We're also in the name image and likeness space. So in sports or in any other field like digital advertising, if you've got an asset that you want to use digital ads against, we would be able to sort of verify and authenticate all that content as well.
0:26:53 - Dmitri
Super interesting Two totally different approaches to sort of dealing with some of these issues, which is super interesting. Remington, what about HyperReal? How do you fit into this landscape as a company?
0:27:04 - Remington
Well, the music landscape is something that is relatively new to us.
The last couple of years, we came out of the VFX industry and feature film and we've been building these digital doubles of celebrities that you see in movies for the last 25 years or so.
So our perspective is like we're the guys that make the digital copies of the people and then we bring them to life, but all of it is under contract and it's under the direction of the studios and everybody involved's authenticated work.
Extension of this. So that way, talent can have these digital identities and the digital dna surrounding their identities and be able to use those identities to monetize new opportunities and to do that we had, you know, we're building the ecosystem for tracking, authentication, blockchain, encryption, um and and and then training on itself. So, for example, if we have a number of performances uh, with talent in the music space, like um, they're not just their voice but their emotion and the way they perform and we train that into ai, we could then generate multiple new ways that they perform to new music uh and new opportunities. So that's where we're coming from and we're really looking, our perspective, really looking about how to monetize new opportunities for artists and creators to be using AI and then ensuring that the work that they do has provenance, authenticity, there's consent on the new work and they're monetized appropriately.
0:28:53 - Dmitri
Again, really interesting. Each approach is so very different. And, Remington, obviously yours is not solely music, although Peck's also isn't solely music, and actually I think Charles is also. It's interesting. I mean, we're seeing a lot of this, where these applications apply beyond just music. Go ahead, charles.
0:29:07 - Charles
Well, I'm just gonna say you know this goes far beyond music, right. I mean, usually music to some extent becomes the canary in the coal mine, right? So whenever there's something new, we usually feel the pain first. But then, you know, we work in music because all of us live here in nash, nashville and music's close to our heart. But, like I said, it's going to get wild and crazy this election cycle and you know that's a huge problem or a potential problem. And then, yeah, it's going to in gaming and in, like I said before, digital advertising. There are so many other verticals for this this would apply. And so, if you just stay focused on the music piece, it's kind of a very narrow lens and short-sighted, because what's going to happen in music is going to be outside of music, even for music creatives, right? So, and you want to be able to verify and authenticate across all platforms and across all verticals, or at least most verticals. And Remington I just love the tech you guys are working on man it's pretty cool that Madison Beer stuff. That's crazy.
0:30:24 - Remington
Thanks, right back at you. Thank you very much. It's been great to be on this panel. You guys are amazing.
0:30:28 - Dmitri
We still have 10 minutes here and then we'll open it up to Q&A, so we've got some time here and, Remington, I'll ask you there's a lot that we raised here to be concerned about. Is it all bad? Are there opportunities there for this generative AI era, for creative content, music and so forth?
0:30:53 - Remington
It depends on yes and no. I it's a hard, that's a great, that's a great question. I knew you were going to ask it. I, I, I'm, I'm an optimist, and, and, and, and. The Pandora's box has been open and the tools are there, and the creators are using it and I'm using it. You know, and, and.
So I think that there's a future that we cannot disassociate from, in which we're all plugged into some kind of AI that helps us in many ways perform better than we can perform, do more than we can do, scale larger than we can possibly scale on our own or within our groups. And these are going to be the exciting parts of talent moving forward with a new technology for the first time in humanity, and how that can now be able to multiply, you know, the creative process without even worrying about technology, letting it do the things behind the curtains that it does, and talent is now able to just almost think and believe, and then it happens. In a way. The challenge and I think we talked about this a lot is that we have to just make sure that it's ethical, it's on consent, it's tracking you guys off, the authenticating it. It has to be able to be done in a way that we're not going to be. You know, everything we do someone else profits from, and that's the key, I think.
0:32:34 - Dmitri
Remington, I think I want to hear a little bit more about what this looks like in your vision with HyperReal, for an artist, an actor, a celebrity, a boxer, an athlete. Just paint a little bit more about what you actually see happening down the road here.
0:32:54 - Remington
I think we exist in the physical plane the way we always have. There'll be business and real world, irl things that we do and that we'll always do and hopefully those things don't change. We still go to events and we still participate with our friends, live, and it's all these things that are done and I don't think, hopefully that doesn't change. But I think that there's like an off-ramp from this highway and this off-ramp now might be like a little side road, you know, a little service road, but that's going to grow and that's where talent says here's a digital version of myself or here's a digital version of everything I have. That is my DNA. It goes off here as a sidecar and it grows with them. And as we are spending more time in virtual ecosystems and engaging with virtual identities and those identities' performances, we will find that these kind of like side roads become super highways and I think there's a world where they live together, a future where they're where they live together how about the rest of our panel?
0:34:11 - Dmitri
what? What do you think the music industry will look like as a result of AI? And, and say five years?
0:34:16 - Charles
well, you, know, at the beginning you said how I'm a big Billy Joel fan and it's true. He did tell me to move to Nashville and so I'm a massive Billy Joel fan. And this turns his life back on song that became, which you know was written by somebody else, but he sang it. But if you haven't checked out the official music video, you need to check out the official music video and I hope someone puts the link to that YouTube video in the chat because it shows Billy through his entire evolution and it's all. You know. The used AI to create different eras of his artist career.
And if you weren't a Billy Joel fan before because you never you know if you're young and if the only thing if you're a young person or Gen Z, the only thing you ever heard from Billy is Vienna, from 13 going on 30. And then you became a fan and you don't know anything else about it and you can see that here's a. It feels weird to say he's a legacy artist, because he's still a thriving, vibrant artist, but this is the kind of thing that's going to create fandom in new areas and in sort of new demographics, right? So I mean that one day, taylor Swift will be 80 years old and not touring, maybe, and there is the potential for AI to recreate what Taylor has done and create new assets. That's going to, you know, maybe introduce the grandchildren of some of the kids who are now super fans, right, and I think overall for art, that's a beautiful, beautiful thing, so that we don't have to have archaeologists excavating.
Oh, this is Taylor Swift's first album, you know, it's just, it's available and I think there's so much opportunity there from Billy, you know, and Randy, and then NIL. Revenue is life changing for a lot of these kids who will never play professionally at the NCAA level. They can still participate in revenues and have meaningful and profitable lives. You know, stuff like that. And yeah, I just think that there's so much opportunity in the music industry and beyond.
0:36:43 - Amadea
I think you've got a panel full of optimists, dimitri, which is very like you, but I think there's so many opportunities. I think AI tools in general really lower the threshold of entry to becoming a creator, a composer, and so, in a way, we're in front of a huge boom of creativity and creation, and I think that's really exciting, that you maybe no longer need to have done 12 years of musical theory to be able to compose something compelling with AI tools, and so I think that's really exciting, and so I think that's really exciting. I also think there's about to be some really cool, very personalized and generative AI outputs, right, and so I think stretching your imagination and creating, you know, generative playlists, I think is going to be something that's really personalized and really exciting in the next few years or months. Really of hey, I'm in this mood and I want this and that, and then getting a whole new playlist of songs that matches the exact mood and all the else that you're looking for, I think is is pretty cool.
0:38:20 - Charles
Yeah, I heard about this very thing music, tectonics, right. So Mark Mulligan and Tatiana Cersano were talking about lean-through experiences where fans get to and, at the end of the day, isn't that what it should all be about? You know, and again, with consent, control and being able to trust that what you're doing is actually authorized by the original artist.
0:38:41 - Amadea
But, yeah, I think there's so much opportunity we go back to the business side of things for a minute then. I think you know, at the beginning of user generated platforms, some rights holders thought that dancing babies might be an issue. I think we fully leaned into loving dancing babies and learning how to monetize that for the industry and the artists. I think the same will be true of generative AI.
0:39:09 - Dmitri
I am kind of surprised that we have such a positive sort of outlook as a group here, just because the topic was really about the threats of AI and how to know what's really real, and also who wrote this, who performed this, who recorded this, et cetera. So I am actually a little bit more surprised, Amadea, than you made it seem, Not because not because not a reflection on any of you guys, but it is interesting to me that three folks, leaders in this space of trying to figure out, like, how to adapt to and protect rights holders and help with monetization, do have a positive outlook and Remington your career as a creator. It makes it even more so in some ways. In a way, it's almost like you've thought 10, 20 years ahead. You've created these virtual avatars in movie, basically, like you know, done the motion capture and recreated this identity and then, instead of just like saying, okay, that's cool, I can do this cool creative stuff, I'll keep that to myself and keep pushing my career, You're saying like wait, can't this be accessible to everybody?
0:40:17 - Remington
Yeah, look I, this is perfect. I love, I love the high note here. I love all this. You know, the excitement we have for the future and the hopefulness that this, this group, has. Um, I'd hate to be the guy that talks about the snake eating itself and we end on that.
0:40:34 - Dmitri
We won't end on it. Bring it up, let's go.
0:40:36 - Remington
We got time good okay because I've always been looking at this from day one that these are tool sets for creators to be able to create better, you know, know and express themselves in new ways. You know, when we, when we were making Gollum, it was like here's an opportunity for an actor to be able to be outside of typecast. He could play a character you cannot play in reality. He was able to do something and and pull emotion into something that was, you know, beyond reality. And and now we're in that world. We're actually in that world where every single one of us can be inside of a new reality, expanding our capabilities and doing this, but at the expense of what? Like? Whose work? Whose backs are we all sitting on typing and prompting that? You have no idea.
That's been trained into some ecosystem, that and they're not opening up these doors. You know we've seen the interview with the CTO over there at OpenAI, who's effectively said she has no idea what's inside her training data, these. You know this is a dangerous world that we're in, right, so and and and. What is the real, the real possibility, is that just creativity becomes like this kind of revolving cycle of the same thing again and again, feeding on itself, and then you know who knows what 50 years from now looks like. I'm not going to worry about that, but but you know the future could be bleak.
0:42:11 - Dmitri
Yeah, yeah, well, but you know, like the other thing is we might be responding different than before. The technology is accelerating faster and so in some ways there's a danger of it getting too far ahead for humans to sort of like protect against or or integrate in a way that makes sense from a rights holder, ethical business sense or whatever. But I do wonder, like the Napster moment may have given the music industry a bit of a heads up about you know, what does it mean to just say no to things versus to say, well, the technology is here. One of you said that you know Pandora's box is open or the cat's out of the bag, or whatever metaphor we want to use there. It's not going away. And so the real question is like it's, it's feels almost like a judo move. You can't just stop it. It's not about building a wall. It's more about how do you integrate this new soci, societal, sociological shift that you know, even if you thought ethically, you were like not into user generated content or social video or algorithms predicting what your move is and then selling something to you, or you know so much tracking happening that companies, technology companies know more about you than you know about you and then can change your political persuasion or your purchasing patterns or whatever. We're still doing it, you know, like even if we protest it, it's still emerging.
And then you're like, hey, I got these socks from an Instagram ad. How did they know I had crazy pants and then I needed crazy socks, I you know. And then you're like, well, shit, I do have these socks now. But I guess, I guess I'm just saying to me it seems like we have learned a little something from the, from the Napster moment, and it seems like you guys are moving pretty fast actually compared to that moment to figure out what, what can this possibly be? Again, I don't mean to be rubbing, again, I don't mean to be Remington, I don't mean to be overly optimistic either. I do think it's just a little bit different than than time before, and the question is whether it's you know, it's like the matrix question really right. It's like are we, are we going to live trapped in a virtual world, or are we going to be able to be what Charles said at the beginning, kind of agents of what's happening?
0:44:30 - Charles
well, we can go dark too, if you want. You know, let's look at all those scenarios, which is, in part, why we developed our platforms like, oh, this will get really bad really fast if we don't, we're not careful. So I mean, yeah, I I share some of the same things, you know. I think that the recording industry learned about well, when new tech comes along, you shouldn't sue kids, you shouldn't sue grandmas. You know that's probably a bad strategy, but at the same time, you know, there's sort of these sort of larger existential questions about, okay, who gets to control what? Larger existential questions about okay, who gets to control what?
You know, at the end of the day, my personal belief and I think as a company, we believe that whatever gets created should be under the direct control of whomever created it. Right, and at the end of the day, it's their decision on whatever they want to do. What we can do is facilitate and amplify whatever that creative process is and enable it in such a way that, you know, one of the things we've heard about the generative AI stuff, especially with stuff like audio and whatever that production music libraries are going to pretty much go away. That sync is going to pretty much go away, and I find it ironic that an industry that fought so hard in collective bargaining to keep AI from intruding into their creative space would then maybe consider use creative assets that sort of, you know undermine another vertical right.
0:46:12 - Dmitri
So we'll see what happens. You're specifically Charles. You're specifically saying like the world of acting and filmmaking has taken a stance around AI, but then you sense that that same industry is looking at sort of using AI from another creator base music. Music to kind of replace the the uh. Sort of at least that's what I keep hearing.
0:46:35 - Charles
So it's, it's a separate segment, so the actors and stuff were fighting for that right. But you know, I can see directors and whomever going and going. Oh, I'll just use a text prompt to create this thing and then it'll be good enough to use in film or tv, you know. And if that's the case, well, all right, but people need to know that, because I think people would would respond differently based on what they think the source of their creativity is.
0:47:04 - Amadea
You know, yeah, yeah I think that threat is real if, if the industry doesn't move quickly to license and to make it easy to license and embrace the scale, the opportunity. So I think that's one of the hurdles of, you know, of sync licenses, is they're not easy to get, and so if it's not easy and then people work around it, and so I do think, and the bleak part of this is that the you know, the training has already been done on copyrighted works, and so it's it's, it's less about trying to stop that from happening and more embracing the fact that, well, it's already happened. So now how do we seize the giant opportunity that it can still be within this reality? So that's a little bit less optimistic maybe, for you.
0:48:04 - Dmitri
It's interesting no-transcript companies are doing the same thing they did with user generated content, which is scraping and kind of asking for forgiveness later or building a model later after the fact, which is troublesome from a kind of a legal perspective. But you know, in the case of, say, youtube, it seems to have worked out for them. At least they didn't get stewed into oblivion.
0:49:01 - Amadea
I think it's worked out for all of them and the big picture is right Grow first and negotiate licenses later and so I think that's the same play that the large AI companies are taking.
0:49:13 - Dmitri
I would say just the only reason I point out is it's easy to say everything has been trained on unethical data but to not realize there's this. Maybe it's a cottage industry at this point, but there's another movement as well that's trying to follow kind of ethical things as well, and you know, it could be that they this time around they end up on top in a different way than previously, just because there's enough of it happening as well.
0:49:41 - Amadea
Yeah, and if the music industry, I think, wants to support that, then there's a way to make it. You know, to figure out a formula to license music to those ethical AI companies, Because if it's easy people would rather pay upfront for the compliance safety. If it's impossible and hard, then they're going to do what those big tech companies are doing.
0:50:08 - Dmitri
All right. Unfortunately, we are running out of time here. This has been super awesome. Amadea, Charles, Remington, thank you so much for joining the panel and being part of this conversation. I think we brought up a lot of really interesting issues and it was really fun, Amadea, to hear kind of like the positive perspective that you guys are working on solving some things that are really challenging but also, you know, having having a good dose of of optimism. And, remington, you know some some beware of the snake eating its own tail, um as well. But thanks all for joining. Um. I've got a couple of quick announcements, but thanks to Amadea, charles and Remington Um, and thanks everybody for joining as well.
This has been great, great, active chat. Love to see that. Don't forget, my company, rock Paper Scissors, is around if you have any marketing or PR needs in the music innovation space. We love this world that we're talking about here. Also, remember our next seismic activity will be taking place June 12th. It's called Industry Interface, creator Tools and the Recording Industry. You can go to musictectonics.com and look for our online events and don't forget to mark your calendars for the Music Tectonics Conference happening October 22nd to 24th 2024 in Santa Monica, california. A limited number of super early bird tickets are on sale now through May 21st. If you like this conversation, you'll love doing it in person, at the beach, in the sun and by a carousel on a pier. Get in touch with shaley at rockpapersoosers.biz if you want to make a meaningful impact at this year's event. Thanks so much for listening.
0:51:43 - Charles
Thank you Dimitri, Thank you everybody, it was really fun to be on this panel.
0:51:50 - Dmitri
Thanks for listening to music tectonics. If you like what you hear, please subscribe on your favorite podcast app. We have new episodes for you every week. Did you know? We do free monthly online events that you, our lovely podcast listeners, can join? Find out more at music tectonicscom and, while you're there, look for the latest about our annual conference and sign up for our newsletter to get updates. Everything we do explores the seismic shifts that shake up music and technology the way the Earth's tectonic plates cause quakes and make mountains. Connect with Music Tectonics on Twitter, instagram and LinkedIn. That's my favorite platform. Connect with me. Dmitri Vietze, if you can spell it, we'll be back again next week, if not sooner.
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.
Comments