Featured Session: A Conversation With Spotify's New Co-CEO About The Past, Present And Future Of Delivering Creativity To The World
Gustav Söderström, Spotify's Co-CEO, reflects on Spotify's 20-year journey through three major 'mountains': solving music piracy through streaming, expanding into podcasts and audiobooks, and now building an AI-powered personalized platform. He introduces Spotify's new AI DJ, AI-generated playlists from natural language prompts, and a vision where every listener gets a uniquely personalized experience powered by large language models.
1Spotify uses the 'local maximum' concept from computer science — to reach the next peak, you must first walk down into the valley of uncertainty, which is why most companies stagnate.
2Spotify paid out over $50 billion to the music industry since launch, transforming piracy-era economics into a sustainable creator revenue model.
3The AI DJ feature combines generative AI with deep music knowledge to create a personalized radio experience that knows your taste better than any human DJ could.
4Spotify's next frontier is making every user experience unique — AI-generated playlists from natural language prompts ('dinner party music for people who love 90s hip hop') that never existed before.
5The company deliberately chose to cannibalize its own music-focused product to become a broader audio platform, accepting short-term pain for long-term positioning.
Full Transcript
A pretty big deal 20 years. And, you know, in tech, that's not just a long time, it's a life most companies don't really survive by. So today, I want to spend our time sharing a little bit about how we survive, what we learn, and, most importantly, where we're going next at Spotify, we often talk about mountains that we as a company try to climb, and there's this concept in computer science called a local maximum. So you climb the hill, you get to the top, and you feel great about yourself. You optimize your position, right?
But then you look up and you realize that you just got a small air the real manager, the global maximum, is way over there. The problem is that to reach that global maximum, you can't just keep walking up here. You first have to walk down it, back into the valley into uncertainty, and then you have to start climbing all over. Many companies die because they refuse to walk down that road. They sort of hide that they know.
What if I say today? Because multiple times we were willing to walk to the edge and jump into the unknown, the first mountain was access. So in 2006 when Spotify started, the competitor wasn't any tech giant, it was piracy. I grew up in Sweden, which at the time was basically a pirate behavior. The global record industry was collapsing in such a way that famously, a UK music industry executive, when he went to Sweden and met his Swedish counterpart, he looked at his piano, his coconut statement, and he said, That's not a business, my friend, that's a hobby.
The industry was in shackles, our founder, Danny Lake, he realized that you couldn't beat piracy with lawyers. You had to beat it with a better product. Piracy, I've only given people all the world's music for free, without ads, by the way, and with all my sake. But piracy set up one big block. It was a terrible user experience.
You have to search for a torrent. You remember those wavelengths several minutes, sometimes hours, depending on the modem you had, but then when you got it, you might have ended up with a mislabel track or actually just a computer bar. Happens to me a couple of times. So in product circles, where I come from, we're going to say that all truly great products, they need to pull up some sort of magic, something that was previously considered inconceivable. Our magic trick wasn't actually just giving people access to music.
As I said, I already done that. What we did was still latest those illusions. Illusion that you have downloaded all the world's music to your local hard rock so when you click the track, it's going to play in less than 200 milliseconds. So we essentially took the technology, the peer to peer technology, the power piracy, and we re engineered it to stay in the music industry and help artists get paid. Technology alone is seldom enough about to change the world.
So the other innovation that happened at the same time was a new business model user complete with piracy. Remember, piracy was free. This was ads Plus subscription, which we call premium. So by providing access and technology and an effective business model, we had managed to climb this first mountain, and Spotify was growing fast. So there we were.
We felt, honestly, pretty good about ourselves. We were at the top of the original mountain, the local maximum. Turned out the company was firing. But then the world changed. The smartphone run internally.
We call what happened next, the gazebo room in the summer of 2010 our small team was at a beautiful yellow Xavier the party was going to act. Our business model at the time was based on people using our three desktop app to compete with free piracy, loving it, and eventually pay for the privilege of syncing those files to their phones. But if you remember Merri meter, MER meters interference report had just come out, and data was irrefutable, smartphones there were about to outpace PC sales soon, so many users didn't even have that cost. And if they skip straight from Pirates of the mobile they hit a paywall, they would strike. And he went back to piracy.
So our business model was due and according to Mary Meeker, he had this three to four years doing that, a completely new one. So we walked down the first month. We shifted our focus from the Google Desktop space and went all in on mobile. But the problem we faced was, how do you give away the exact thing that you were driving for mobility without just filling your entire business so something else that is often talked about in tech circles, is you should follow your users, look at this data and see what they do. And it turned out that our mobile premium users, who have full on demand playback, they were still paying their playlist in shopping mode about 50% of the time.
This gave us an idea. We thought to ourselves, what if we could give away just the shuffle mode for free? And in 2013 we licensed, built and launched the first mobile app in the App Store where a user could create a playlist of any song they wanted, put their phone in their pockets and listen to that playlist. New shuffle mode for free forever. This makes for device world explode, pulling millions of people out of piracy and into pain for music.
So we have readjusted to first world, and with that, we have climbed the second month now, already, as we were scaling that second mountain, the mobile mountain, another massive shift was happening for the connected home and an explosion of hardware, the Alexa and the Google homes, right? And we realized that having a great mobile app alone wasn't going to DNI. So we had to make a contrarian bet on your big league at the time, conventional wisdom said that you should pick an ecosystem and then should build a wall garden around it, like Apple did with iOS. But we went the other way with a protocol called Spotify net. We decided that Spotify should go everywhere, in your car, your kitchen, your PlayStation, your TV.
We chose openness over locking. The bet was actually very painful, and it took a long time because we had to convince all the world's hardware manufacturers to implement our protocol, even though we were a tiny one series company at the time. But today, that bet compounds, because we have 2000 different hardware devices, and users are free to listen to the music on whatever combination of hardware they want. The next month, the fourth month was a variant of the follow your users estimate, but this time, still watching the users. We watched what all developers were doing with our product.
This is generally exemplary strategy, because these are often early adopters of new behavior, and they are in a rare position actually build what they want to see in the world, instead of just waiting for this exact reason. Sport has always had hack weeks where the entire company just pauses the usual day to day execution for broad and deep holding supplementation for the entire company, just to see what people what people want to do leverage the brain of 8000 smart people. And time and time again, we had seen our developers packing their podcast into their own Spotify as a company, we wanted to be more than music, but we weren't sure how. So we looked at the podcast industry, and you realize that it was just great content and amazing creators, many of them having deep conversation, which we felt was the perfect counter to the short educational media that was going on. But podcasts have been neglected for years, so we needed that, that it supported this market with people at the same at the same time.
Remember later we also noticed enterprising book publishers in Germany upload books, kind of disguised as music albums for but it was a terrible user experience if you skipped, if you were lucky to skip whole chapter, but it could also play shuttle modes, which is very weird. What was focused was that, despite the more experience, you just put the power in this case. So actually last on the inside look at this data, the product opportunities were pretty obvious. They were right there in front of us. But we have to think about how for both of these.
The conventional wisdom at the time told us and everyone else to get a separate app for podcasts and another one for those. That was just the way things were done. Remember the there's an app for that commercial. But we asked ourselves, why should the user? Have to adapt to that format by switching software.
Shouldn't the software adapt to the user instead? So you decided to instead build an adaptive user experience, one app with the interface dynamically shifted. It was a massive technical challenge that it allowed creators from from music and podcasts and book to cross pollinate audiences, and over time, hundreds of millions of users embrace proving us right. And of course, today, we aren't just a music company, we're an audio company, and people are listening more than ever. So now we really thought we were at the peak, but it turned out that the tallest mountain we've ever faced was vitals, generative AI.
Now there's a lot of noise, a lot of breathless uptakes, but as a technologist, here's the framework that I use. As I mentioned before, new technology is such a disruptive on its own, significant disruption happens when new technologies enable you asymmetric business models. So supported by didn't disrupt downloads just because of loaded business streaming, as I said, we disrupted them because we also introduced the ad supported subscription model. Similarly, it was a business model so Uber Airbnb that disrupted their spectrum industries, not just the technology alone. And if you look at the consumer space, we believe that the dominant business model will continue to be ads to a subscription.
In fact, you can already see the big AI labs adopting exactly this model. So this puts qualified in an outstanding position, because we already have the exact right business for this world. But from a product perspective, we actually think everything is going to change. User expectations of what software and services can and students should do for them are rapidly evolving for all of you on shore sitting there to chat GPT or progress, let me give you the real, deep impact. If you look at the history of the internet, Spotify started in what we call the era of curation.
Right users were organizing their friends and the tracks and their social networks of credit. And then we'll move to recommendation, but the algorithms did that work for you, and now we are entering age of generation where the consumer participates in a much bigger way. Most consumer products today, they still operate sort of like old school broadband, and I'm getting a technical guy when it comes to user input. So in old school broadband, you have lots of data on the downlink, sending the user videos and songs and feeds, but very narrow bandwidth on the uplink, so just a few swipes from the user click or something. So for companies like Spotify, we mostly have to guess what you want to hear based on whether you, for example, skip the fact then you think about that, did you skip it because you hate it or because you love it?
But you're at the gym and jazz isn't the right song. The algorithm is just a skip. It's very hard to interpret. It all looks the same. Generative AI changes this.
You cannot communicate back to the platform in rich natural language, in English, you can tell services which you actually want and what we think that the world will move away from these sort of past and single player experiences, these one way platforms where users just receive and receive and receive, to fully interact multiplayer platforms, this is where users will participate, interact with friends, artists and creators, and enjoy much more control than they have. So for our next mountain, we're building the world's first truly intelligent agentic media platform, one that you can literally talk, steer and control one way the algorithm works for you, instead of the other way around. We started this actually some time ago with the ai dj. Give me your personalized radios, and you can ask for whatever you want. I'm going for a run.
Give me an EDM playlist with big drops at 160 ppms to match my rankings, and then, recently, with tools like prompted playlist, you can literally write your own algorithm for playlist and then schedule that to update daily or weekly. So if you ever wanted playlist that, for example, takes all the music training on Tiktok consistence of your taste and then removes any track that you've already listened to. You can right now just go in and write that sentence. Are you tired of this? Fortified paintings is your favorite shaman, just because it was for the last three years?
I know I do sometimes. Just tell you about the room to avoid it. Avoid my main Shaman. This is the power of computers actually understanding English that they can finally understand here we think sports is uniquely positioned. The client does not and make this work for reality.
Because while llms, large language models are a key ingredient to making these possible general, avalanche are not enough, because llms are trained on facts. They know the capital of Texas, Dallas, right? But taste is not a fact. It's an opinion. If you ask a model for workout reason, for example, there is no canonical answer for someone in New York, it might be hip hop.
For someone in Western Europe, it's often again and Korean, a lot of Scandinavians, it's dead now. You need billions of constantly updated data points across amazonism users that do human curation to be able to map language to music in a deep and personal way. This is the data set that sport has been building for over a decade. But I also want to touch on the artist side of all of this change, because there's a lot of fear, and it is completely understandable. AI is a tool, no doubt, a very powerful one, and like any tool, it can be used in both really good and really bad ways.
As an example, in the early 2010s people weren't convinced that computers could be used to make music at all. But then it came from Sweden, named she used this lab or to change music forever, but music has never just been about the tools used to make it. Music is identity. When a teenager buys the t shirt at a concert, they aren't just buying clothing. They're buying something that tells the world who they are, what you belong to.
Ai cannot manufacture what it feels like to stand in a crowd of people who love an artist just the way you do. These moments of human connection, they transcend technology uncertainty. I firmly believe that any more generating content through human connection actually becomes more valuable, not less. And I'm not just making this up. We already see this play out in the more than one and a half billion dollars worth of consultators that we saw.
It's just growing in world of AI and personalization. Live events like come on we're at right now are bigger than they ever been, and they continue to grow. But we also believe that AI can create a lot of value for artists and the industry, and our job as Spotify is to ensure that the economics reflect that value, whether it's musicians and songwriters making original music or making letting fans remix and interact with their sitting IP. We're working with the music industry to make sure the right compensation frameworks are in place. We solve this with viruses, and we are going to solve it with AI, because at the end of the day, scholars mission is to deliver creativity to the world, one note, one voice, one idea at a time.
That's our commitment, and we believe that we're still in the early days of doing just that. Okay, so now you understand our journey. I'm excited to give you a first look at one of our next steps, Facebook. This new feature gives you, the user, a comprehensive look at just how exactly Spotify AI models your mistake across music podcast modules we think you are and more importantly, it helps you edit it directly. Here are some real examples from internal users.
One said, finally, I think that Spotify to just ignore those two artists that my kids always play, they used to ruin all my recommendations. Another example. These are new examples. When I was young, I used to be very big on possibilities, but after a lot of it, I just told Spotify to start putting a classical shelf on my homepage, which in turn made me start to lose weight. We're watching this and pay that in New Zealand soon.
But if you're a Spotify Premium user here today, scan the QR code behind me accesses. It's going to be very good. So 20 years have told us that every time we left the comfort of the local maximum, we took a jump, made another climb. The view from the New Beat was even better, but not just for Spotify, but for every creative industry that we built, every listener, every time. Thank you for joining me at the base camp of something new, and now I'm excited and privileged to invite lady Wilson and.
Stage for conversation on the many exciting things happening. All right, set up notes. Maybe you be welcome. All right, let's take a seat. Yes, perfect.
Thank you, sir. All right, I'm very excited for this discussion, but before we get into it, I just want to mention a couple of sentences on each of you. So lady, you have nearly 10 million monthly active users on Spotify. You have multiple number 100 hits. You want their CMAs ACM, pretty much done it all in usage.
But in addition to singing, and by the way, my life is a big fan, in addition to that, though, to sing it, you're also fulfill an entrepreneur, and you partner with native brands like Wrangler, who owns multiple businesses of your own, including your own fashion jewelry lines. And if I understand correctly, thank you for coming, David, you're a founding member of the code team at Google. We started 2004 he let us solve the climate cooperation in 2013 I think, and in 2016 he launched a production board which holds investments across tech, including agriculture, decompensation. You're the co founder and CEO of a genetics and you follow that, but I'm very happy about it, which consistently ranks as the top business and tech podcast. So welcome.
Thank you. Thank you for not making a political category podcast. All right, let's get into it, ladies, starting with you, Southwest is really about the intersection of culture and tech and people who live and work in a world where ideas and artists travel around the world instantly. And being here in Boston, I don't think anyone is surprised that counter music talks to the charts here. But all of you may not know is that counter music is one of the absolute fastest growing summers globally, and it's become hugely popular in markets from the UK to Australia, where you just back from, our data shows that four by four by you Yes.
So I would love to hear how you think about building an audience in 2026 and beyond, yeah, change and changing every single day. But for me, it is just connected as much as I possibly can with my touring, with just, I mean, whether it's releasing music touring, I put out a documentary here in April that's going to be on Netflix. I think that's going to be a good way for me to really stay connected with the fans that I've made, but also get to know some new ones. And we're also just always looking at the beta. We're just trying to see what's working, what's not working.
And for me, it has always been about that just family interaction be submitted correctly to them. And so whether that's face to face or to the camera, that seems to be the thing that works for me. So it's just about connecting as I as much as I possibly can. And you know, cutting through a lot of the ones can I ask you, I told you when we spoke backstage here that country music is big globally. But more than that, almost Texas country is big globally.
Land man is one of the top stories in Sweden. What is it about country music and Texas? It is growing and blows my mind, because every single opportunity I get to go across the pond to play music. People are singing every word. They're wearing their cowboy hats.
We're gonna get you a cowboy hat and some cowboy boots too, and you too. We're here in Austin. That's what I'm doing. But I really do think that. Well, I'm partial, but I think country music is a very special genre.
I think that it takes you home when you are nowhere near it. I think it's about that storytelling. I think everybody's a good sucker for a good story, and if you're not, I'm confused, but I think that's what it is. It's just people are craving often. Authenticity, and I think country music offers that authenticity is key.
My kids told me it's not as well. All right, David Allen is one of the most popular podcasts in the world. The show name is a hat tip to the poker games and one conversation that you and your colleagues were already having, right? And on this show, there's a lot of time spent on Eva's interests, but you still have a very quickly growing global audience. Were you surprised by that?
And was that your intention? Well, we didn't have much of an intention. They were very randomized, random situation that the podcast came out of covid on Zoom, I would say the original conversations were all about the tech industry, business markets, what's going on? How are we all reacting to and responding to the change in the world around us? How crazy is this?
Shit was kind of like a pretty common theme during covid. And you know, what should we do? What could we do? So I think for anyone working in the tech industry or starting a company or investing around the world, it became, you know, something to listen to. Lot of people described it, hold it like being a fly on the wall.
There were people just literally having a conversation about these topics. And I think that's kind of why it resonated globally. There are investors, startup founders, builders, Texas, anything that inspired you with this? I don't know if there was any other such the concept of this recording in general conversation. Did you find the wall?
No, that wasn't the design of the show. I think there wasn't much of the design, because the one guy wanted to be interviewed by the other guy every week. That was the original show concept. And then they kept talking about finding time to go in the studio to do that show, and they couldn't find the time to go into the studio do that show. And then covid hit.
They were all stuck at home, and they're like, Hey, let's get on Zoom and talk to each other and put that out as the show. And they asked me to join for the first episode, so I did, talking about covid and tech science and what was going on. And so that's how the show started. So there was no design, and I think that may be why it worked, because it's also like there was no real reason that we would care whether it succeeded or not, and so we just maybe didn't give a shit enough, and you just do what comes natural and what feels right, and then it just kind of worked. So for both of you, AI is obviously transforming creative industries right now, opening up huge possibilities, but also raising some big concerns.
So I'd love to hear about what you do. This means New York Times as some artists and creative if you saw you, Megan, you mentioned backstage that you and your team are building your own AI, impersonation, Detective detection system. Talk to us a little bit about that for sure, sure. Well, I'll just first say that I think that technology has has done a lot for the music business. I think that it has always pushed creativity forward.
But I will say it is a scary world we live in, because you just never know how they're going to use your face or your voice. And the truth is, like we're just talking about that genuine connection, it's really important for me to feel like anytime my fans see something that they feel like it is really like evils, and saying it to them, I want them to feel like that they can trust me. I have partnered with a company called Looney that is constantly working on pulling down these imposters or these social media posts, whatever it is that is not coming from me or my team, and that is a full time job at this point, and it is really, really crazy. And the scary thing is, is even when your family members think it's actually you, you're like, Oh Lord, they're getting really good at this. But, but, yeah, we have, we have found a team of people who are taking serious just like you are.
And I just want to keep that, that relationship with my fans, the way that, the way that I wanted to be genuine and pure and authentic you're coming into this. But I suspect that every creator is going to want something like this, and maybe eventually every human, as we go into this world where it's always understand, it's probably to understand, if you can actually trust David, you're the president, it's a very broad topic, but if you just pick what the documents are you most excited about? So have you transition in AI can be sort of like an up leveling of creative potential. I'll just start with that statement. There's certainly risks associated with impersonation and theft of IP and whatnot, which I think will get worked out as it was YouTube and Spotify in the early days.
Then we'll get there. But I got my first Macintosh computer where six. Years old, the Macintosh original, and I was on that computer learning how to use MAC Paint. And eventually I got to go to Photoshop. I was doing graphic design at the age of 11 or 12 as a service to people.
I got making money, making business cards and flyers and stuff. And so I was early on when Kai's power tools came out. Does anyone remember that? No one in this room that is so depressing. Okay, there's a couple, but Adobe Photoshop, right?
You can put like images on and edit them and manipulate them, and power tools have these filters. And you would do a filter to the image, and it was like, Holy crap. You could make the image sharper, you could make it blur. You could do all. You can create depth.
You can do all these things that people were like, That's cheating. Like, that's not fair. Like, how can you do that? You're the cameraman supposed to do that, or the guy who learned how paintbrush is the one who should be doing that. And so there's always and then later on, I did, like, you know, made techno music.
Don't make fun of me for it. But, you know, you and I talked about, I used to have, like, a Juno 106, TV 303, and then I got all the synthetic sequencers on my computer, and there were all these plugins that came out, and the plugins became, I mean, just like a music industry today, it was a question of, like, Are you cheating? Like, is that really fair? And what ended up happening is creators, whether graphic creators or music creators found extraordinary ways to leverage these new tools to create new soundscapes, new visual features, new types of film. Today in Hollywood, a lot of the production houses are and I've spent some time with some of these companies, they're using like second units to do fully AI rendered like second units, unit.
So the cost comes down, but the stuff you can do goes up by, like, 100x because you couldn't, you didn't have the budget to go and, like, recreate a trading floor from the 1970s and now you can do an AI in a couple of minutes. And so the creator, the director, has this ability to make a landscape that they otherwise wouldn't have the ability to make. And so I think that there's a leveling up that's happening in terms of what creators can do, and we're just starting to see people tinker with these things, and the stuff that I've seen, especially in video, and I'm sure over time, we'll see more and more happen in music. It's just incredible. So I think that the second big wave that's probably still to come is non linear content.
And what I mean by that is, like today, whether it's a video or music track, you just press play and it goes linearly from beginning to end. I think that creators are going to unlock like in video games, the ability to maybe have some difference in how that track plays or how that movie plays out. Whether it's like, click and choose your own adventure, which I think may be a little annoying to consumers, or it's like, what if you could listen to your favorite house music track and make it a little bit faster, little bit cooler, a little bit shorter, add some vocals, add some jazz, but the Creator, not the user, but the Creator can set all those parameters. So now the creator is almost like coming up with all these ideas, but in different dimensions of how they can create music. And if you don't have to have one final track, and you can say, hey, we could make a couple different ways that this track can play out, and then the actual playback for the consumer, whether it's in a video or in music, could be a little bit different.
And so I think we're likely going to see like, progressive creators come up with incredible ways of using these tools and these capabilities, where people are going to be watching movies differently. Like, hey, we all watch the same show, but I saw a little different view that you saw, like I saw from the bad guys view you saw from the good guys view, or we all track a little bit differently. So that's the sort of stuff that I think would be really cool with the AI unlocked sort of different experience. So as the creator and control search on these guys, I creator, if there was a framework for your fans to be able to work with the music, where everyone was greatly compensated, because it is your music, for sure, I think it's fun to be creative, but, you know, I think, Like, that's why I started doing this to begin with. So I would love for my fans to be able to do something, if it's fair for the person who wrote it, proposed it, created it.
I'm down for that. I want everybody to be able to like color outside of the lines and try new things. So like I said, if people are compensated for it, Hell yeah. Hell yeah. Well, okay, are you guys doing anything like this?
It's fun. Well, I'm looking at my comments. There is this one feature working on that touches on this that absolutely love, which people have been asking for forever. So you know, in movies, you have like, I am the game. You can go in and see, like, who's the director, who's producer.
You're really nerd up in music, that hasn't often been the case. You often know who the performer is, but most people have no idea who wrote it or produced it. So this is leaked a little bit already, but we are. We're rolling out this. This feature called Song DNA, where we meticulously go through to understand exactly who played a role on each song.
You can see that song writer actually like that sample is actually from this original song. So we'll see, I think, and I think it's really important that there are all these creators around the performance that there are that's something I'm very excited. It's really odd. That. Okay, all right, another question for both of you, the CEO, if environment.
Now for both of you as the old my partner, Alex and I have stepped into the roles of those deals. We're spending a lot of time talking with artists, with podcast creators and with authors. I love the books. I'm very excited about this all around the world, and one of my favorite questions is, what can be building for you as a creator or as a user, by the way, yeah, well, into speaking and as a creator and a user, I think would be cool to build something based off of demons that day, you know? Because, I mean, the truth is that we're all going through something at different times, and it'd be really cool for my fans to feel like that day that they were going through something hard.
I was speaking directly to them, that they didn't have to, like, listen through the whole album to figure out, like, Which song is resonating with them for that specific thing that they're going through. So very cool. So you were like, That's interesting. You can So, I mean, you could go into a Spotify and type in a little down here from your fans, because there are so many of them in some sort of aggregated way. Yes, I think that would be cool too, to know how your fans are feeling, too.
I mean, so many different people from so many walks of life have like, it's crazy the story that people and telling me, but one of the three lines for me that I've noticed this evening in meet and greets is they are willing to share like where they are in life and where they are going through and so I don't know, just be able to, like, cater to that even more, to even just have, like, a deeper connection like that with your family. Because my goal always is for them to feel like I'm speaking to them in their story, in their life. But if they're going through something, and I'm singing a song about, let's have a part, you know? I mean, they might not, that might not be the thing that helps them get through it, the musician, you only get data on how many plays there were and what the average listen kind of was, that's that's all the only feedback you get. It's true.
Yeah, you do get statistics about who is listening, ages and where they listen. You can find your tours, but you're completely right. You have no idea how they're feeling and looking at AI. I just talked about how AI allows people to speak in English at scale. That is a really cool thing, to actually understand more about their demographic.
Well, a couple years ago, there was a guy who took all of our episodes, like, right when AI was kind of getting going, and he ingested them all, and then he created, like an AI character with each of the four of us, and then it was like, each of the four of us said a bunch of stuff, and it was so bullshit, like that sounded awful. He's gotten so much better now. But then someone else made a chat interface where you could ask each of those questions they could answer, as if it was trained on the data from our show. And it was like, Wait, that's actually kind of interesting. Like, very often I'm trying to find something that I heard on another show, like Friedman organ or something.
I'm like, Where was that? And so to find what was said by guests, or find really guests and protection was very hard. Today I got to go, like to YouTube and do transcript search. Or nowadays, I can do Gemini, and Gemini like, search through all the transcript data for me and pull stuff out. But I think every podcaster should have the opportunity to create like Lex chat, or Joe chat, or something where you as a user who's a big fan of the podcast, like, the thing I realized about podcasting is it's people want to, like, get really developed.
They come to these live events. Like, we've been on a live event with like, 1000s of people in LA every year. We've got one here tomorrow night. And they want to, like, absorb more the content, get into the community. Applicant.
Engage, engage, engage, because that particular content vertical is something they're really into. So to have a chat interface where I could ask Joe Rogan questions like, Hey, Who was your favorite guest? Which was the craziest shit you ever heard. You know, tell me about that guy that talked about, like there were pyramids under the ocean and we found them. Like, who was that guy?
I would love to be able to have that dialog with my podcast that I love. I think that would be a really cool feature. You guys can train it off on the transcript data, and theoretically, you could actually make the audio sound like the host. And I could have a dialog through the Spotify app and all of these that creates more engagement for you guys, and all of the I should be your kind of product. So yeah, anyway, because I would love to chat with Lex Friedman and then be like, Hey, tell me about that great Hancock interview.
And then our users could just ask David to talk to digital I got enough of that. But I do worry about, certainly, this thing of, like, digital representation. It's gonna happen. It's happening to a point, and they get it better. They get an opt in.
Yeah, people that I think, I think podcasters are different than these issues. So podcasters are creating content. They're not creating these, like, you know, tracks, the songs that are kind of their product out, but they want to protect the value of the song and so on. Podcasters more content they're creating, I think the more engagement they get, the more depth they get. So there's something about re engaging in that content that actually brings a lot of value for the podcaster.
And it doesn't need to be pure it doesn't need to be how I said the thing at the time. I said it per se, but generally, like, the things I said, like, I'd be fine for AI chat interface. Just be trained on that and want to answer questions. Interesting when music is interesting as a content category because it has so much repeat testing, people listen the same track hundreds, 1000s of times. But in podcasts, you mostly don't the same episode again.
This would be a great value from the back catalog as well, right? But people do listen to the back catalog, like acquired Rogan Lex, like some of these are very timeless episodes, because they're interview shows with someone on particular topic. And so, like, I mean, you know the stats on acquire business podcast, once people listen to an acquired podcast, they go back and listen to the whole brain library, and they're like, this show's awesome. I wasn't all of these, and there's a bunch of pods like that. So I think to have some sort of engagement mechanism around that stuff could be very cool.
Okay, we got some product ideas. My first virus here, amazing lady, you're not just an artist, you're also founder and investor of fashion, jewelry and gaming. So we talked about how you think about your arts career, but how do you think about building a business alongside your music career? Well, like I said earlier, I just really like to be creative, and I want to make sure that my parents feel like they've been they can like that into the work of my music as much as I possibly can or as much as I want to. And, you know, I look out at my shows and I'm seeing people who dress just like me, wearing the bell bottoms in their hats and the jewelry.
And this is it. Yeah, we should have put you in some Bill bottoms too. But, yeah, it's just another way for me to connect with my people and with my fans, and, like I said, for them to be able to, like, live in that world as much as they want to. And it's fun. You know, when you're all under like, the same roof and you're like, looking out there and everybody kind of looks like you, it's a especially like with a little girls and stuff, it makes me feel like, man, like, I better watch what I do and say, because I have a lot of little people watching me.
But also it, it gives me this sense of just like responsibility. To feel like, okay, I need to be a leader. I need to, like, set an example. And yeah, it's a good feeling. It's a good feeling to I feel blessed with like have this platform and be on that stage, and I just Yeah, I want people to feel like they know me.
I want to feel like I know them. But we spoke backstage for so long, I did not have any opportunity at all. I've been at this a long time. I've been living in Asheville for 15 years, and the first 10 I should have packed up, I went home, but I just couldn't. This is the only thing I know how to do.
It's the only thing I want to do is just to tell stories. And yeah, I feel like the last few years have just been such a blessing to me and my life and my team, like I have finally found the right people to surround myself. With that understand my vision and don't want to change me for anything, and if anything, they just want to elevate me and get me to that next step. And it's getting really fun now. So now we have all these opportunities.
There's only so many hours in the day, but I do want to take them all. You know, if it feels right and it feels like it's going to move the needle, then let's do it. You told me you're from this tiny, tiny town 170 and you moved pretty further on. Yes, I did. I'm from Baskin, Louisiana, a little town of 170 people.
I'm very proud where I come from, but honestly, I compare. I was telling you this in the room too, that I compare a lot of what I do to what my dad does. My dad is a farmer, and he farms corn, wheat, soybeans, oats, and gets up every day and busts his tail doing it, and he does the same thing every single day. I mean, it looks a little bit different for me every day. But, um, you know, at the end of the day, it's your livelihood, if it's in you, then you just get up and you do it.
A storm could roll through, and you just get up and do it anyway. So I call myself a song farmer, amazing. Amazing. David spoken is a big movement podcast in 2019 and you and your girls started, as you said, during covid in 2020 Could you talk a little bit to us about the growth and evolution that you've seen and sort of where you think it's going next? You mentioned some of the ideas you have.
But what do you see podcasting? Well, podcasting kind of became what blogging used to be. Everyone that had blogs prior to that, everyone so in the business world, like, if you've got, you know, I think one of the problems is a lot of this stuff got solid, and maybe we're part of this problem as well. But like people selling shit, like, like, I'm trying to like people doing podcasts because they're trying to, like, heighten their profile because they have stuff to sell. I don't know if that intention really resonates and works.
Well, maybe, maybe we're part of maybe we should be blamed for that as well, I don't know, but I feel like there's an aspect of that where, like, everyone's got podcasts now, and the market will sort itself out. People listen to what they want to listen to and stuff for. So it's very crowded. It's very noisy out there. And I think to some extent, we're seeing folks step it up, right?
So the visual stuff matters a lot, I think more than people are realizing, but a lot of the stuff will hold its own as audio. But Chris Williams said, I don't know if you guys know that this guy is podcaster. He did anyone see the episode? They built a set on the same kind of sound stage and set stage that they use for the Mandalorians, like this LED wall that, like, wrapped around it's so and they built Interstellar, like the guy's house with Interstellar, with the trailer and the driveway. It's freaking insane.
I recommend you watch it on YouTube, just to see what they built. The YouTube thing, I'm so sorry. I so this is why I think the video on Spotify doesn't matter a lot, because I do think a lot of the folks are differentiating by improving the quality of the video content. And it's not just me having a chat and then obviously the podcasts are more creative, like all the murder crime industry stuff. It's so cool, like it's a totally different level.
It's not just someone sitting at the table having a conversation. Joe Rogan can hold his own having a conversation for two hours, and we'll get great content. But most people cannot. Most people need it. So the good creator they're gonna need to go to, like murder crime history, that might be a better format.
So I think there's a lot of this differentiation that's going to start to happen, which, by the way, will like it fundamentally improve the quality of the experience that the consumers, the listeners are like, the viewers are like, this is awesome. Now, like that Matthew McConaughey episode, visually, it's just like, I don't care what McConaughey is saying, this is just so cool. Pretty cool. And ask you, one of the things that I think people like about listening to you numbers is that you have a positive view of the future. And there's a lot of stuff here right now, and you're going to start a podcast.
Well, we'll see. But I do think the thing that troubles me the most is just kind of how techno pessimistic we are, right? And I think, like, everyone's just like ai humorism, like AI is the most unfavorable thing right now in the country. There's some survey that came out it was like Iran, like Iran is more favorable than AI. Like, the sort of thing that happened where AI just became the new like, whether it was climate change or.
Or whatever, there's always this, like looping threat that we latch on to. And I think we've had this, like this shift in our psyche, particularly in the West, meaning the US and Western Europe, probably since the 1970s like 1969 we landed on the moon and we had Woodstock that same summer, and I think that was the summer when we became techno pessimists, where we were like, worried about chemicals in our water, radiation killing us. Nuclear energy could burn, melt down and wipe us. Nuclear bombs are overhead. So we simply became way more concerned about tech.
But if you look at like, how tech is progressing, human lifespan, human health span, how many hours people have to work, the number of hours that people have to work is going down the quality of life. I know a lot of people are struggling. Still in America, there's a lot of less behind differences in what's going on, but a lot of people are moving up the ladder because of the advances in technology. We all got a freaking super computer in our pocket. There's just so much that I think we take for granted.
And when I look at the technology that's being developed and is like real today, and I think about, holy shit, what this is going to do tomorrow? Like, we're all going to live forever. We're going to have free friggin energy. We're going to be in flying cars moving around the point, like, oh, wait, finally, forever. They launched yesterday.
They got FAA certified yesterday, by the way, nobody did. So these are the sorts of things that are happening that I think don't get enough attention. And we spend all this time thinking about how we're going to die and what's going to kill us and what's coming next, and we don't talk enough about how exciting tomorrow is going to be. And so I want to kind of change the message a little bit and spend some time on it. And that's kind of why I do this show.
It's why I do and considering doing the pre show, which will just be focused on the science and tech of tomorrow, and just kind of get deeply excited about it, that's great. I do think it's important. Everyone can follow along with that future, especially, for example, in music, but I'm very positive. I will say one more. I think the fear of tomorrow is what makes everyone turn against each other, because as soon as you're about tomorrow, you have a reason to blame someone around you, and that's why we are all like this all the time.
And I think it's very bad. It's very unhealthy. Go to a dark place. And so I want to all be optimistic about tomorrow and be realistic, but be optimistic about tomorrow. That's why it didn't matter.
Okay? I could keep talking about him forever, but Thank you so much for coming. Thank you so yeah, I Guess you have A better How about.
Source: stt · Language: en · Model: claude-opus-4-6
1
00:00:00,240 --> 00:00:01,020
Unknown: A pretty big deal
2
00:00:02,820 --> 00:00:06,420
20 years. And, you know, in
tech, that's not just a long
3
00:00:06,420 --> 00:00:07,800
time, it's a life
4
00:00:08,880 --> 00:00:10,800
most companies don't really
survive by. So
5
00:00:12,120 --> 00:00:16,020
today, I want to spend our time
sharing a little bit about how
6
00:00:16,020 --> 00:00:20,180
we survive, what we learn, and,
most importantly, where we're
7
00:00:20,180 --> 00:00:20,780
going next
8
00:00:22,400 --> 00:00:26,960
at Spotify, we often talk about
mountains that we as a company
9
00:00:26,960 --> 00:00:30,920
try to climb, and there's this
concept in computer science
10
00:00:30,980 --> 00:00:35,720
called a local maximum. So you
climb the hill, you get to the
11
00:00:35,720 --> 00:00:39,200
top, and you feel great about
yourself. You optimize your
12
00:00:39,200 --> 00:00:40,220
position, right?
13
00:00:41,300 --> 00:00:44,380
But then you look up and you
realize that you just got a
14
00:00:44,380 --> 00:00:48,520
small air the real manager, the
global maximum, is way over
15
00:00:48,520 --> 00:00:48,820
there. The
16
00:00:50,500 --> 00:00:55,180
problem is that to reach that
global maximum, you can't just
17
00:00:55,180 --> 00:00:58,840
keep walking up here. You first
have to walk down it, back
18
00:00:59,860 --> 00:01:01,980
into the valley into
uncertainty, and
19
00:01:03,240 --> 00:01:05,100
then you have to start climbing
all over.
20
00:01:06,900 --> 00:01:09,720
Many companies die because they
refuse to walk down that road.
21
00:01:10,380 --> 00:01:11,880
They sort of hide that they
know.
22
00:01:13,440 --> 00:01:17,340
What if I say today? Because
multiple times we were willing
23
00:01:17,580 --> 00:01:19,980
to walk to the edge and jump
into the unknown,
24
00:01:21,480 --> 00:01:23,600
the first mountain was access.
25
00:01:24,680 --> 00:01:30,020
So in 2006 when Spotify started,
the competitor wasn't any tech
26
00:01:30,020 --> 00:01:31,880
giant, it was piracy.
27
00:01:33,320 --> 00:01:36,500
I grew up in Sweden, which at
the time was basically a pirate
28
00:01:36,500 --> 00:01:36,920
behavior.
29
00:01:38,240 --> 00:01:43,000
The global record industry was
collapsing in such a way that
30
00:01:43,000 --> 00:01:47,680
famously, a UK music industry
executive, when he went to
31
00:01:47,680 --> 00:01:51,400
Sweden and met his Swedish
counterpart, he looked at his
32
00:01:51,400 --> 00:01:55,120
piano, his coconut statement,
and he said, That's not a
33
00:01:55,120 --> 00:01:56,740
business, my friend, that's a
hobby.
34
00:01:57,940 --> 00:01:59,260
The industry was in shackles,
35
00:02:00,700 --> 00:02:04,140
our founder, Danny Lake, he
realized that you couldn't beat
36
00:02:04,140 --> 00:02:05,580
piracy with lawyers.
37
00:02:06,780 --> 00:02:08,940
You had to beat it with a better
product.
38
00:02:11,039 --> 00:02:14,639
Piracy, I've only given people
all the world's music for free,
39
00:02:15,059 --> 00:02:17,219
without ads, by the way, and
with all my sake.
40
00:02:18,960 --> 00:02:23,180
But piracy set up one big block.
It was a terrible user
41
00:02:23,180 --> 00:02:23,780
experience.
42
00:02:24,860 --> 00:02:27,140
You have to search for a
torrent. You remember those
43
00:02:27,620 --> 00:02:30,680
wavelengths several minutes,
sometimes hours, depending on
44
00:02:30,680 --> 00:02:34,340
the modem you had, but then when
you got it, you might have ended
45
00:02:34,340 --> 00:02:37,160
up with a mislabel track or
actually just a computer bar.
46
00:02:37,220 --> 00:02:38,480
Happens to me a couple of times.
47
00:02:40,040 --> 00:02:43,060
So in product circles, where I
come from, we're going to say
48
00:02:43,120 --> 00:02:45,940
that all truly great products,
they need to pull up some sort
49
00:02:45,940 --> 00:02:49,300
of magic, something that was
previously considered
50
00:02:49,420 --> 00:02:50,140
inconceivable.
51
00:02:51,580 --> 00:02:54,880
Our magic trick wasn't actually
just giving people access to
52
00:02:54,880 --> 00:02:57,220
music. As I said, I already done
that.
53
00:02:58,420 --> 00:03:01,060
What we did was still latest
54
00:03:02,260 --> 00:03:05,340
those illusions. Illusion that
you have downloaded all the
55
00:03:05,340 --> 00:03:08,040
world's music to your local hard
rock so
56
00:03:09,060 --> 00:03:12,480
when you click the track, it's
going to play in less than 200
57
00:03:12,900 --> 00:03:13,560
milliseconds.
58
00:03:14,820 --> 00:03:18,240
So we essentially took the
technology, the peer to peer
59
00:03:18,240 --> 00:03:22,280
technology, the power piracy,
and we re engineered it to stay
60
00:03:22,280 --> 00:03:24,920
in the music industry and help
artists get paid.
61
00:03:27,320 --> 00:03:30,380
Technology alone is seldom
enough about to change the
62
00:03:30,380 --> 00:03:33,560
world. So the other innovation
that happened at the same time
63
00:03:33,920 --> 00:03:37,940
was a new business model user
complete with piracy. Remember,
64
00:03:37,940 --> 00:03:38,780
piracy was free.
65
00:03:40,340 --> 00:03:45,280
This was ads Plus subscription,
which we call premium. So by
66
00:03:45,280 --> 00:03:49,060
providing access and technology
and an effective business model,
67
00:03:49,540 --> 00:03:52,420
we had managed to climb this
first mountain, and Spotify was
68
00:03:52,420 --> 00:03:52,960
growing fast.
69
00:03:54,160 --> 00:03:57,100
So there we were. We felt,
honestly, pretty good about
70
00:03:57,100 --> 00:03:57,520
ourselves.
71
00:03:59,140 --> 00:04:00,900
We were at the top of the
original mountain,
72
00:04:01,920 --> 00:04:03,300
the local maximum. Turned out
the
73
00:04:04,500 --> 00:04:09,540
company was firing. But then the
world changed. The smartphone
74
00:04:09,540 --> 00:04:09,720
run
75
00:04:11,100 --> 00:04:14,220
internally. We call what
happened next, the gazebo room
76
00:04:15,180 --> 00:04:19,080
in the summer of 2010 our small
team was at a
77
00:04:21,779 --> 00:04:22,879
beautiful yellow
78
00:04:26,780 --> 00:04:27,200
Xavier
79
00:04:28,880 --> 00:04:30,260
the party was going to act.
80
00:04:31,880 --> 00:04:35,540
Our business model at the time
was based on people using our
81
00:04:35,540 --> 00:04:36,440
three desktop app to
82
00:04:37,460 --> 00:04:41,380
compete with free piracy, loving
it, and eventually pay for the
83
00:04:41,380 --> 00:04:43,480
privilege of syncing those files
to their phones.
84
00:04:44,980 --> 00:04:48,760
But if you remember Merri meter,
MER meters interference report
85
00:04:48,760 --> 00:04:53,080
had just come out, and data was
irrefutable, smartphones there
86
00:04:53,080 --> 00:04:57,100
were about to outpace PC sales
soon, so many users didn't even
87
00:04:57,100 --> 00:04:57,700
have that cost.
88
00:05:00,960 --> 00:05:04,320
And if they skip straight from
Pirates of the mobile they hit a
89
00:05:04,320 --> 00:05:07,200
paywall, they would strike. And
he went back to piracy.
90
00:05:08,760 --> 00:05:10,320
So our business model was due
and
91
00:05:11,460 --> 00:05:15,300
according to Mary Meeker, he had
this three to four years doing
92
00:05:15,300 --> 00:05:16,320
that, a completely new
93
00:05:18,720 --> 00:05:19,860
one. So we walked down the first
94
00:05:21,120 --> 00:05:24,620
month. We shifted our focus from
the Google Desktop space and
95
00:05:24,740 --> 00:05:25,820
went all in on mobile. But
96
00:05:27,260 --> 00:05:31,940
the problem we faced was, how do
you give away the exact thing
97
00:05:31,940 --> 00:05:35,840
that you were driving for
mobility without just filling
98
00:05:35,840 --> 00:05:36,680
your entire business
99
00:05:38,600 --> 00:05:42,100
so something else that is often
talked about in tech circles, is
100
00:05:42,460 --> 00:05:43,720
you should follow your users,
101
00:05:44,920 --> 00:05:46,420
look at this data and see what
they do.
102
00:05:48,340 --> 00:05:52,120
And it turned out that our
mobile premium users, who have
103
00:05:52,420 --> 00:05:55,780
full on demand playback, they
were still paying their playlist
104
00:05:55,780 --> 00:05:59,800
in shopping mode about 50% of
the time. This gave us an idea.
105
00:06:01,060 --> 00:06:05,280
We thought to ourselves, what if
we could give away just the
106
00:06:05,280 --> 00:06:06,300
shuffle mode for free? And
107
00:06:08,160 --> 00:06:08,640
in 2013
108
00:06:09,720 --> 00:06:14,700
we licensed, built and launched
the first mobile app in the App
109
00:06:14,700 --> 00:06:17,880
Store where a user could create
a playlist of any song they
110
00:06:17,880 --> 00:06:21,380
wanted, put their phone in their
pockets and listen to that
111
00:06:21,380 --> 00:06:24,680
playlist. New shuffle mode for
free forever.
112
00:06:26,000 --> 00:06:29,300
This makes for device world
explode, pulling millions of
113
00:06:29,300 --> 00:06:32,120
people out of piracy and into
pain for music.
114
00:06:33,380 --> 00:06:37,940
So we have readjusted to first
world, and with that, we have
115
00:06:37,940 --> 00:06:39,320
climbed the second month
116
00:06:41,060 --> 00:06:43,960
now, already, as we were scaling
that second mountain, the mobile
117
00:06:43,960 --> 00:06:47,620
mountain, another massive shift
was happening for the connected
118
00:06:47,620 --> 00:06:50,740
home and an explosion of
hardware, the Alexa and the
119
00:06:50,740 --> 00:06:51,580
Google homes, right?
120
00:06:52,600 --> 00:06:55,660
And we realized that having a
great mobile app alone wasn't
121
00:06:55,660 --> 00:06:56,320
going to DNI.
122
00:06:57,820 --> 00:07:01,920
So we had to make a contrarian
bet on your big league at the
123
00:07:01,920 --> 00:07:05,400
time, conventional wisdom said
that you should pick an
124
00:07:05,400 --> 00:07:08,520
ecosystem and then should build
a wall garden around it, like
125
00:07:08,520 --> 00:07:09,360
Apple did with iOS.
126
00:07:10,920 --> 00:07:15,240
But we went the other way with a
protocol called Spotify net. We
127
00:07:15,300 --> 00:07:18,840
decided that Spotify should go
everywhere, in your car, your
128
00:07:19,620 --> 00:07:24,080
kitchen, your PlayStation, your
TV. We chose openness over
129
00:07:24,080 --> 00:07:24,440
locking.
130
00:07:26,420 --> 00:07:29,000
The bet was actually very
painful, and it took a long time
131
00:07:29,000 --> 00:07:31,880
because we had to convince all
the world's hardware
132
00:07:31,880 --> 00:07:34,640
manufacturers to implement our
protocol, even though we were a
133
00:07:34,640 --> 00:07:38,360
tiny one series company at the
time. But today, that bet
134
00:07:38,360 --> 00:07:42,220
compounds, because we have 2000
different hardware devices, and
135
00:07:42,220 --> 00:07:45,760
users are free to listen to the
music on whatever combination of
136
00:07:47,860 --> 00:07:51,640
hardware they want. The next
month, the fourth month was a
137
00:07:51,640 --> 00:07:55,480
variant of the follow your users
estimate, but this time, still
138
00:07:55,480 --> 00:07:58,840
watching the users. We watched
what all developers were doing
139
00:07:58,840 --> 00:08:01,800
with our product. This is
generally exemplary strategy,
140
00:08:02,160 --> 00:08:05,700
because these are often early
adopters of new behavior, and
141
00:08:05,700 --> 00:08:08,700
they are in a rare position
actually build what they want to
142
00:08:08,700 --> 00:08:10,260
see in the world, instead of
just waiting
143
00:08:11,640 --> 00:08:15,600
for this exact reason. Sport has
always had hack weeks where the
144
00:08:15,600 --> 00:08:18,720
entire company just pauses the
usual day to day execution for
145
00:08:19,020 --> 00:08:21,300
broad and deep holding
supplementation for the entire
146
00:08:21,300 --> 00:08:25,040
company, just to see what people
what people want to do leverage
147
00:08:25,040 --> 00:08:29,900
the brain of 8000 smart people.
And time and time again, we had
148
00:08:29,900 --> 00:08:32,960
seen our developers packing
their podcast into their own
149
00:08:32,960 --> 00:08:33,440
Spotify
150
00:08:35,420 --> 00:08:39,080
as a company, we wanted to be
more than music, but we weren't
151
00:08:39,080 --> 00:08:39,380
sure how.
152
00:08:40,400 --> 00:08:42,940
So we looked at the podcast
industry, and you realize
153
00:08:45,520 --> 00:08:49,000
that it was just great content
and amazing creators, many of
154
00:08:49,000 --> 00:08:51,760
them having deep conversation,
which we felt was the perfect
155
00:08:51,760 --> 00:08:54,640
counter to the short educational
media that was going on. But
156
00:08:56,020 --> 00:09:00,180
podcasts have been neglected for
years, so we needed that, that
157
00:09:00,480 --> 00:09:02,880
it supported this market with
people
158
00:09:05,100 --> 00:09:07,800
at the same at the same time.
Remember later we also noticed
159
00:09:08,040 --> 00:09:12,000
enterprising book publishers in
Germany upload books, kind of
160
00:09:12,000 --> 00:09:16,500
disguised as music albums for
but it was a terrible user
161
00:09:16,500 --> 00:09:20,480
experience if you skipped, if
you were lucky to skip whole
162
00:09:20,480 --> 00:09:24,200
chapter, but it could also play
shuttle modes, which is very
163
00:09:24,200 --> 00:09:24,440
weird.
164
00:09:26,720 --> 00:09:29,540
What was focused was that,
despite the more experience, you
165
00:09:29,600 --> 00:09:30,800
just put the power in this case.
166
00:09:32,180 --> 00:09:34,700
So actually last on the inside
look at this data, the product
167
00:09:34,700 --> 00:09:36,680
opportunities were pretty
obvious. They were right there
168
00:09:36,680 --> 00:09:39,140
in front of us. But we have to
think about
169
00:09:41,120 --> 00:09:43,660
how for both of these. The
conventional wisdom at the time
170
00:09:43,660 --> 00:09:47,620
told us and everyone else to get
a separate app for podcasts and
171
00:09:47,620 --> 00:09:50,500
another one for those. That was
just the way things were done.
172
00:09:51,460 --> 00:09:53,980
Remember the there's an app for
that commercial.
173
00:09:56,320 --> 00:09:59,920
But we asked ourselves, why
should the user?
174
00:10:00,000 --> 00:10:03,660
Have to adapt to that format by
switching software. Shouldn't
175
00:10:03,660 --> 00:10:05,220
the software adapt to the user
instead?
176
00:10:06,780 --> 00:10:10,680
So you decided to instead build
an adaptive user experience, one
177
00:10:10,680 --> 00:10:14,100
app with the interface
dynamically shifted. It
178
00:10:15,900 --> 00:10:18,900
was a massive technical
challenge that it allowed
179
00:10:18,900 --> 00:10:21,980
creators from from music and
podcasts and book to cross
180
00:10:21,980 --> 00:10:25,940
pollinate audiences, and over
time, hundreds of millions of
181
00:10:25,940 --> 00:10:27,500
users embrace proving us right.
182
00:10:28,580 --> 00:10:31,880
And of course, today, we aren't
just a music company, we're an
183
00:10:31,880 --> 00:10:36,020
audio company, and people are
listening more than ever. So now
184
00:10:36,380 --> 00:10:40,900
we really thought we were at the
peak, but it turned out that the
185
00:10:40,900 --> 00:10:45,820
tallest mountain we've ever
faced was vitals, generative AI.
186
00:10:47,200 --> 00:10:52,300
Now there's a lot of noise, a
lot of breathless uptakes, but
187
00:10:52,300 --> 00:10:54,940
as a technologist, here's the
framework that I use.
188
00:10:56,380 --> 00:10:59,980
As I mentioned before, new
technology is such a disruptive
189
00:10:59,980 --> 00:11:03,420
on its own, significant
disruption happens when new
190
00:11:03,420 --> 00:11:07,920
technologies enable you
asymmetric business models. So
191
00:11:07,920 --> 00:11:11,220
supported by didn't disrupt
downloads just because of loaded
192
00:11:11,220 --> 00:11:14,460
business streaming, as I said,
we disrupted them because we
193
00:11:14,460 --> 00:11:18,180
also introduced the ad supported
subscription model. Similarly,
194
00:11:18,180 --> 00:11:22,400
it was a business model so Uber
Airbnb that disrupted their
195
00:11:22,400 --> 00:11:24,920
spectrum industries, not just
the technology alone. And
196
00:11:26,420 --> 00:11:29,540
if you look at the consumer
space, we believe that the
197
00:11:29,540 --> 00:11:33,560
dominant business model will
continue to be ads to a
198
00:11:33,560 --> 00:11:34,220
subscription.
199
00:11:35,240 --> 00:11:38,840
In fact, you can already see the
big AI labs adopting exactly
200
00:11:38,840 --> 00:11:39,320
this model.
201
00:11:40,460 --> 00:11:43,780
So this puts qualified in an
outstanding position, because we
202
00:11:43,780 --> 00:11:46,660
already have the exact right
business for this world.
203
00:11:48,520 --> 00:11:52,480
But from a product perspective,
we actually think everything is
204
00:11:52,480 --> 00:11:53,080
going to change.
205
00:11:54,460 --> 00:11:58,060
User expectations of what
software and services can and
206
00:11:58,060 --> 00:12:01,740
students should do for them are
rapidly evolving for all of you
207
00:12:01,740 --> 00:12:04,500
on shore sitting there to chat
GPT or progress,
208
00:12:06,720 --> 00:12:08,100
let me give you the real, deep
impact.
209
00:12:09,660 --> 00:12:12,480
If you look at the history of
the internet, Spotify started in
210
00:12:12,480 --> 00:12:15,180
what we call the era of
curation. Right users were
211
00:12:15,240 --> 00:12:18,300
organizing their friends and the
tracks and their social networks
212
00:12:18,300 --> 00:12:22,280
of credit. And then we'll move
to recommendation, but the
213
00:12:22,280 --> 00:12:23,900
algorithms did that work for
you,
214
00:12:25,100 --> 00:12:29,540
and now we are entering age of
generation where the consumer
215
00:12:29,660 --> 00:12:31,580
participates in a much bigger
way. Most
216
00:12:33,140 --> 00:12:36,620
consumer products today, they
still operate sort of like old
217
00:12:36,620 --> 00:12:40,040
school broadband, and I'm
getting a technical guy when it
218
00:12:40,040 --> 00:12:43,120
comes to user input. So in old
school broadband, you have lots
219
00:12:43,120 --> 00:12:46,360
of data on the downlink, sending
the user videos and songs and
220
00:12:46,360 --> 00:12:50,200
feeds, but very narrow bandwidth
on the uplink, so just a few
221
00:12:50,200 --> 00:12:52,240
swipes from the user click or
something.
222
00:12:53,560 --> 00:12:57,700
So for companies like Spotify,
we mostly have to guess what you
223
00:12:57,700 --> 00:13:01,140
want to hear based on whether
you, for example, skip the fact
224
00:13:01,860 --> 00:13:05,280
then you think about that, did
you skip it because you hate it
225
00:13:05,880 --> 00:13:09,840
or because you love it? But
you're at the gym and jazz isn't
226
00:13:09,840 --> 00:13:13,260
the right song. The algorithm is
just a skip. It's very hard to
227
00:13:13,260 --> 00:13:14,820
interpret. It all looks the
same.
228
00:13:16,320 --> 00:13:17,880
Generative AI changes this.
229
00:13:18,900 --> 00:13:22,220
You cannot communicate back to
the platform in rich natural
230
00:13:22,220 --> 00:13:26,780
language, in English, you can
tell services which you actually
231
00:13:26,780 --> 00:13:27,740
want and what
232
00:13:29,300 --> 00:13:31,700
we think that the world will
move away from these sort of
233
00:13:31,700 --> 00:13:34,100
past and single player
experiences, these one way
234
00:13:34,100 --> 00:13:37,100
platforms where users just
receive and receive and receive,
235
00:13:38,060 --> 00:13:42,400
to fully interact multiplayer
platforms, this is where users
236
00:13:42,400 --> 00:13:46,060
will participate, interact with
friends, artists and creators,
237
00:13:46,480 --> 00:13:48,460
and enjoy much more control than
they have.
238
00:13:50,020 --> 00:13:53,380
So for our next mountain, we're
building the world's first truly
239
00:13:53,380 --> 00:13:56,980
intelligent agentic media
platform, one that you can
240
00:13:56,980 --> 00:14:00,040
literally talk, steer and
control
241
00:14:01,240 --> 00:14:04,620
one way the algorithm works for
you, instead of the other way
242
00:14:04,680 --> 00:14:04,920
around.
243
00:14:06,720 --> 00:14:10,080
We started this actually some
time ago with the ai dj. Give me
244
00:14:10,140 --> 00:14:13,140
your personalized radios, and
you can ask for whatever you
245
00:14:13,140 --> 00:14:18,480
want. I'm going for a run. Give
me an EDM playlist with big
246
00:14:18,480 --> 00:14:21,620
drops at 160 ppms to match my
rankings,
247
00:14:23,060 --> 00:14:26,660
and then, recently, with tools
like prompted playlist, you can
248
00:14:26,660 --> 00:14:30,020
literally write your own
algorithm for playlist and then
249
00:14:30,020 --> 00:14:32,480
schedule that to update daily or
weekly.
250
00:14:33,920 --> 00:14:38,960
So if you ever wanted playlist
that, for example, takes all the
251
00:14:38,960 --> 00:14:42,280
music training on Tiktok
consistence of your taste and
252
00:14:42,280 --> 00:14:45,040
then removes any track that
you've already listened to. You
253
00:14:45,040 --> 00:14:47,980
can right now just go in and
write that sentence.
254
00:14:49,780 --> 00:14:52,180
Are you tired of this? Fortified
paintings is your favorite
255
00:14:52,180 --> 00:14:55,180
shaman, just because it was for
the last three years? I know I
256
00:14:55,180 --> 00:14:58,000
do sometimes. Just tell you
about the room to avoid it.
257
00:14:58,780 --> 00:14:59,800
Avoid my main Shaman.
258
00:15:00,720 --> 00:15:04,080
This is the power of computers
actually understanding English
259
00:15:04,320 --> 00:15:05,940
that they can finally understand
here we
260
00:15:07,740 --> 00:15:10,740
think sports is uniquely
positioned. The client does not
261
00:15:11,340 --> 00:15:12,780
and make this work for reality.
262
00:15:13,800 --> 00:15:17,820
Because while llms, large
language models are a key
263
00:15:17,820 --> 00:15:22,160
ingredient to making these
possible general, avalanche are
264
00:15:22,160 --> 00:15:22,640
not enough,
265
00:15:23,900 --> 00:15:27,620
because llms are trained on
facts. They know the capital of
266
00:15:27,620 --> 00:15:28,340
Texas,
267
00:15:29,540 --> 00:15:30,020
Dallas, right?
268
00:15:33,800 --> 00:15:36,440
But taste is not a fact. It's an
opinion. If
269
00:15:37,580 --> 00:15:40,660
you ask a model for workout
reason, for example, there is no
270
00:15:40,660 --> 00:15:44,140
canonical answer for someone in
New York, it might be hip hop.
271
00:15:44,140 --> 00:15:49,240
For someone in Western Europe,
it's often again and Korean, a
272
00:15:49,240 --> 00:15:51,280
lot of Scandinavians, it's dead
now.
273
00:15:52,960 --> 00:15:56,560
You need billions of constantly
updated data points across
274
00:15:56,620 --> 00:16:00,780
amazonism users that do human
curation to be able to map
275
00:16:01,140 --> 00:16:05,400
language to music in a deep and
personal way. This is the data
276
00:16:05,400 --> 00:16:08,280
set that sport has been building
for over a decade.
277
00:16:10,740 --> 00:16:13,740
But I also want to touch on the
artist side of all of this
278
00:16:13,740 --> 00:16:13,980
change,
279
00:16:15,240 --> 00:16:19,020
because there's a lot of fear,
and it is completely
280
00:16:19,140 --> 00:16:19,920
understandable.
281
00:16:21,360 --> 00:16:26,060
AI is a tool, no doubt, a very
powerful one, and like any tool,
282
00:16:26,240 --> 00:16:29,540
it can be used in both really
good and really bad ways.
283
00:16:30,800 --> 00:16:34,940
As an example, in the early
2010s people weren't convinced
284
00:16:34,940 --> 00:16:38,720
that computers could be used to
make music at all. But then it
285
00:16:38,720 --> 00:16:41,320
came from Sweden, named she used
this lab or to change music
286
00:16:41,320 --> 00:16:41,620
forever,
287
00:16:43,960 --> 00:16:46,960
but music has never just been
about the tools used to make it.
288
00:16:47,740 --> 00:16:49,060
Music is identity.
289
00:16:50,380 --> 00:16:53,980
When a teenager buys the t shirt
at a concert, they aren't just
290
00:16:53,980 --> 00:16:57,220
buying clothing. They're buying
something that tells the world
291
00:16:57,220 --> 00:16:58,780
who they are, what you belong
292
00:17:00,100 --> 00:17:03,840
to. Ai cannot manufacture what
it feels like to stand in a
293
00:17:03,840 --> 00:17:06,660
crowd of people who love an
artist just the way you do.
294
00:17:06,660 --> 00:17:06,840
These
295
00:17:08,220 --> 00:17:10,740
moments of human connection,
they transcend technology
296
00:17:10,980 --> 00:17:11,760
uncertainty.
297
00:17:13,560 --> 00:17:17,940
I firmly believe that any more
generating content through human
298
00:17:17,940 --> 00:17:23,000
connection actually becomes more
valuable, not less. And I'm not
299
00:17:23,000 --> 00:17:26,480
just making this up. We already
see this play out in the more
300
00:17:26,480 --> 00:17:29,000
than one and a half billion
dollars worth of consultators
301
00:17:29,000 --> 00:17:30,800
that we saw. It's just growing
302
00:17:32,120 --> 00:17:35,120
in world of AI and
personalization. Live events
303
00:17:35,480 --> 00:17:38,420
like come on we're at right now
are bigger than they ever been,
304
00:17:38,780 --> 00:17:39,860
and they continue to grow.
305
00:17:42,740 --> 00:17:46,300
But we also believe that AI can
create a lot of value for
306
00:17:46,300 --> 00:17:50,440
artists and the industry, and
our job as Spotify is to ensure
307
00:17:50,440 --> 00:17:52,180
that the economics reflect that
value,
308
00:17:53,380 --> 00:17:55,960
whether it's musicians and
songwriters making original
309
00:17:55,960 --> 00:17:59,500
music or making letting fans
remix and interact with their
310
00:17:59,500 --> 00:18:03,240
sitting IP. We're working with
the music industry to make sure
311
00:18:03,240 --> 00:18:05,340
the right compensation
frameworks are in place.
312
00:18:06,540 --> 00:18:09,960
We solve this with viruses, and
we are going to solve it with
313
00:18:10,020 --> 00:18:10,080
AI,
314
00:18:11,940 --> 00:18:16,740
because at the end of the day,
scholars mission is to deliver
315
00:18:16,740 --> 00:18:17,640
creativity to the
316
00:18:18,720 --> 00:18:21,860
world, one note, one voice, one
idea at a time.
317
00:18:22,880 --> 00:18:26,480
That's our commitment, and we
believe that we're still in the
318
00:18:26,480 --> 00:18:27,740
early days of doing just that.
319
00:18:29,480 --> 00:18:33,080
Okay, so now you understand our
journey. I'm excited to give you
320
00:18:33,080 --> 00:18:36,140
a first look at one of our next
steps, Facebook.
321
00:18:37,340 --> 00:18:41,080
This new feature gives you, the
user, a comprehensive look at
322
00:18:41,500 --> 00:18:46,360
just how exactly Spotify AI
models your mistake across music
323
00:18:46,360 --> 00:18:50,920
podcast modules we think you are
and more importantly, it helps
324
00:18:50,920 --> 00:18:52,720
you edit it directly.
325
00:18:53,740 --> 00:18:57,340
Here are some real examples from
internal users. One said,
326
00:18:57,520 --> 00:19:01,080
finally, I think that Spotify to
just ignore those two artists
327
00:19:01,440 --> 00:19:04,020
that my kids always play, they
used to ruin all my
328
00:19:04,020 --> 00:19:04,800
recommendations.
329
00:19:09,180 --> 00:19:12,840
Another example. These are new
examples. When I was young, I
330
00:19:12,960 --> 00:19:15,480
used to be very big on
possibilities, but after a lot
331
00:19:15,480 --> 00:19:19,200
of it, I just told Spotify to
start putting a classical shelf
332
00:19:19,200 --> 00:19:22,160
on my homepage, which in turn
made me start to lose weight.
333
00:19:23,900 --> 00:19:27,440
We're watching this and pay that
in New Zealand soon. But if
334
00:19:27,440 --> 00:19:30,680
you're a Spotify Premium user
here today, scan the QR code
335
00:19:30,680 --> 00:19:31,100
behind me
336
00:19:32,900 --> 00:19:37,760
accesses. It's going to be very
good. So 20 years have told us
337
00:19:37,760 --> 00:19:41,140
that every time we left the
comfort of the local maximum, we
338
00:19:41,140 --> 00:19:44,980
took a jump, made another climb.
The view from the New Beat was
339
00:19:45,040 --> 00:19:48,760
even better, but not just for
Spotify, but for every creative
340
00:19:48,760 --> 00:19:51,820
industry that we built, every
listener, every time.
341
00:19:53,200 --> 00:19:55,600
Thank you for joining me at the
base camp of something new,
342
00:19:56,800 --> 00:19:59,740
and now I'm excited and
privileged to invite lady Wilson
343
00:19:59,740 --> 00:19:59,920
and.
344
00:20:00,720 --> 00:20:03,720
Stage for conversation on the
many exciting things happening.
345
00:20:03,720 --> 00:20:04,200
All right, set
346
00:20:14,700 --> 00:20:15,900
up notes. Maybe you
347
00:20:20,000 --> 00:20:20,420
be welcome.
348
00:20:27,620 --> 00:20:28,940
All right, let's take a seat.
349
00:20:32,660 --> 00:20:33,620
Yes, perfect.
350
00:20:39,860 --> 00:20:40,360
Thank you, sir.
351
00:20:42,520 --> 00:20:46,420
All right, I'm very excited for
this discussion, but before we
352
00:20:46,420 --> 00:20:50,080
get into it, I just want to
mention a couple of sentences on
353
00:20:50,080 --> 00:20:53,860
each of you. So lady, you have
nearly 10 million monthly active
354
00:20:53,860 --> 00:20:58,000
users on Spotify. You have
multiple number 100 hits. You
355
00:20:58,000 --> 00:20:59,920
want their CMAs ACM,
356
00:21:01,180 --> 00:21:05,400
pretty much done it all in
usage. But in addition to
357
00:21:05,400 --> 00:21:07,080
singing, and by the way, my life
is a
358
00:21:09,000 --> 00:21:12,300
big fan, in addition to that,
though, to sing it, you're also
359
00:21:12,300 --> 00:21:15,780
fulfill an entrepreneur, and you
partner with native brands like
360
00:21:15,780 --> 00:21:18,000
Wrangler, who owns multiple
businesses of your own,
361
00:21:18,240 --> 00:21:22,220
including your own fashion
jewelry lines. And if I
362
00:21:22,220 --> 00:21:23,900
understand correctly, thank you
for
363
00:21:30,320 --> 00:21:30,860
coming,
364
00:21:32,420 --> 00:21:35,660
David, you're a founding member
of the code team at Google. We
365
00:21:35,660 --> 00:21:39,920
started 2004 he let us solve the
climate cooperation in 2013 I
366
00:21:39,980 --> 00:21:43,960
think, and in 2016 he launched a
production board which holds
367
00:21:43,960 --> 00:21:46,540
investments across tech,
including agriculture,
368
00:21:47,440 --> 00:21:51,340
decompensation. You're the co
founder and CEO of a genetics
369
00:21:51,820 --> 00:21:53,020
and you follow that, but
370
00:21:55,120 --> 00:21:58,240
I'm very happy about it, which
consistently ranks as the top
371
00:21:58,240 --> 00:22:01,140
business and tech podcast. So
welcome. Thank you. Thank you
372
00:22:01,140 --> 00:22:03,720
for not making a political
category podcast.
373
00:22:06,000 --> 00:22:08,160
All right, let's get into it,
ladies,
374
00:22:09,840 --> 00:22:12,660
starting with you, Southwest is
really about the intersection of
375
00:22:12,660 --> 00:22:16,200
culture and tech and people who
live and work in a world where
376
00:22:16,200 --> 00:22:20,300
ideas and artists travel around
the world instantly. And being
377
00:22:20,300 --> 00:22:23,240
here in Boston, I don't think
anyone is surprised that counter
378
00:22:23,240 --> 00:22:27,560
music talks to the charts here.
But all of you may not know is
379
00:22:27,560 --> 00:22:30,680
that counter music is one of the
absolute fastest growing summers
380
00:22:30,680 --> 00:22:34,640
globally, and it's become hugely
popular in markets from the UK
381
00:22:34,760 --> 00:22:39,320
to Australia, where you just
back from, our data shows that
382
00:22:39,380 --> 00:22:40,360
four by four by you Yes.
383
00:22:43,480 --> 00:22:45,940
So I would love to hear how you
think
384
00:22:49,120 --> 00:22:50,320
about building an audience
385
00:22:51,460 --> 00:22:51,940
in 2026
386
00:22:55,480 --> 00:22:56,140
and beyond, yeah,
387
00:23:02,560 --> 00:23:04,080
change and changing every single
day.
388
00:23:05,280 --> 00:23:09,660
But for me, it is just connected
as much as I possibly can
389
00:23:11,100 --> 00:23:13,140
with my touring, with
390
00:23:14,220 --> 00:23:19,680
just, I mean, whether it's
releasing music touring, I put
391
00:23:19,680 --> 00:23:24,380
out a documentary here in April
that's going to be on Netflix. I
392
00:23:24,380 --> 00:23:26,900
think that's going to be a good
way for me to really stay
393
00:23:26,900 --> 00:23:30,260
connected with the fans that
I've made, but also get to know
394
00:23:30,500 --> 00:23:34,100
some new ones. And we're also
just always looking at the beta.
395
00:23:34,280 --> 00:23:37,220
We're just trying to see what's
working, what's not working. And
396
00:23:37,220 --> 00:23:41,440
for me, it has always been about
that just family interaction be
397
00:23:41,500 --> 00:23:46,420
submitted correctly to them. And
so whether that's face to face
398
00:23:46,720 --> 00:23:51,520
or to the camera, that seems to
be the thing that works for me.
399
00:23:51,520 --> 00:23:55,600
So it's just about connecting as
I as much as I possibly can. And
400
00:23:55,600 --> 00:23:58,060
you know, cutting through a lot
of the ones can I ask you,
401
00:23:59,320 --> 00:24:02,640
I told you when we spoke
backstage here that country
402
00:24:02,640 --> 00:24:06,420
music is big globally. But more
than that, almost Texas country
403
00:24:06,420 --> 00:24:09,420
is big globally. Land man is one
of the top stories in Sweden.
404
00:24:09,720 --> 00:24:11,640
What is it about country music
405
00:24:16,320 --> 00:24:18,900
and Texas? It is growing and
blows my mind, because every
406
00:24:18,900 --> 00:24:20,420
single opportunity I get to
407
00:24:21,500 --> 00:24:25,760
go across the pond to play
music. People are singing every
408
00:24:25,880 --> 00:24:28,520
word. They're wearing their
cowboy hats. We're gonna get you
409
00:24:28,520 --> 00:24:32,300
a cowboy hat and some cowboy
boots too, and you too.
410
00:24:34,340 --> 00:24:35,900
We're here in Austin. That's
what I'm doing.
411
00:24:37,280 --> 00:24:40,460
But I really do think that.
Well, I'm partial, but I think
412
00:24:40,460 --> 00:24:45,040
country music is a very special
genre. I think that it takes you
413
00:24:45,040 --> 00:24:48,220
home when you are nowhere near
it. I think it's about that
414
00:24:48,280 --> 00:24:51,220
storytelling. I think
everybody's a good sucker for a
415
00:24:51,220 --> 00:24:56,260
good story, and if you're not,
I'm confused, but I think that's
416
00:24:56,260 --> 00:24:57,400
what it is. It's just
417
00:24:58,720 --> 00:24:59,920
people are craving often.
418
00:25:00,000 --> 00:25:02,760
Authenticity, and I think
country music offers that
419
00:25:04,380 --> 00:25:06,420
authenticity is key. My kids
told
420
00:25:09,780 --> 00:25:10,560
me it's not as well.
421
00:25:12,660 --> 00:25:13,920
All right, David
422
00:25:15,120 --> 00:25:18,540
Allen is one of the most popular
podcasts in the world. The show
423
00:25:18,540 --> 00:25:21,680
name is a hat tip to the poker
games and one conversation that
424
00:25:21,680 --> 00:25:25,520
you and your colleagues were
already having, right? And on
425
00:25:25,520 --> 00:25:29,000
this show, there's a lot of time
spent on Eva's interests, but
426
00:25:29,000 --> 00:25:31,460
you still have a very quickly
growing global audience. Were
427
00:25:31,460 --> 00:25:34,580
you surprised by that? And was
that your intention?
428
00:25:35,840 --> 00:25:38,300
Well, we didn't have much of an
intention. They were very
429
00:25:38,300 --> 00:25:40,960
randomized, random situation
that the podcast came out of
430
00:25:41,020 --> 00:25:46,540
covid on Zoom, I would say the
original conversations were all
431
00:25:46,540 --> 00:25:50,200
about the tech industry,
business markets, what's going
432
00:25:50,200 --> 00:25:53,800
on? How are we all reacting to
and responding to the change in
433
00:25:53,800 --> 00:25:57,880
the world around us? How crazy
is this? Shit was kind of like a
434
00:25:57,880 --> 00:26:01,800
pretty common theme during
covid. And you know, what should
435
00:26:01,800 --> 00:26:05,280
we do? What could we do? So I
think for anyone working in the
436
00:26:05,340 --> 00:26:08,700
tech industry or starting a
company or investing around the
437
00:26:08,700 --> 00:26:12,900
world, it became, you know,
something to listen to. Lot of
438
00:26:12,960 --> 00:26:15,540
people described it, hold it
like being a fly on the wall.
439
00:26:15,780 --> 00:26:18,540
There were people just literally
having a conversation about
440
00:26:18,540 --> 00:26:22,160
these topics. And I think that's
kind of why it resonated
441
00:26:22,160 --> 00:26:26,900
globally. There are investors,
startup founders, builders,
442
00:26:27,140 --> 00:26:27,560
Texas,
443
00:26:30,080 --> 00:26:33,380
anything that inspired you with
this? I don't know if there was
444
00:26:33,380 --> 00:26:33,920
any other such
445
00:26:35,000 --> 00:26:37,880
the concept of this recording in
general conversation. Did you
446
00:26:37,880 --> 00:26:38,360
find the wall?
447
00:26:40,040 --> 00:26:44,320
No, that wasn't the design of
the show. I think there wasn't
448
00:26:44,320 --> 00:26:47,020
much of the design, because the
one guy wanted to be interviewed
449
00:26:47,020 --> 00:26:49,300
by the other guy every week.
That was the original show
450
00:26:49,300 --> 00:26:52,180
concept. And then they kept
talking about finding time to go
451
00:26:52,180 --> 00:26:54,940
in the studio to do that show,
and they couldn't find the time
452
00:26:54,940 --> 00:26:57,400
to go into the studio do that
show. And then covid hit. They
453
00:26:57,400 --> 00:26:59,740
were all stuck at home, and
they're like, Hey, let's get on
454
00:26:59,800 --> 00:27:02,580
Zoom and talk to each other and
put that out as the show. And
455
00:27:02,580 --> 00:27:06,240
they asked me to join for the
first episode, so I did, talking
456
00:27:06,240 --> 00:27:10,380
about covid and tech science and
what was going on. And so that's
457
00:27:10,380 --> 00:27:12,300
how the show started. So there
was no design, and
458
00:27:13,620 --> 00:27:15,840
I think that may be why it
worked, because it's also like
459
00:27:16,320 --> 00:27:19,620
there was no real reason that we
would care whether it succeeded
460
00:27:19,620 --> 00:27:24,440
or not, and so we just maybe
didn't give a shit enough, and
461
00:27:27,560 --> 00:27:30,680
you just do what comes natural
and what feels right, and then
462
00:27:30,680 --> 00:27:32,600
it just kind of worked.
463
00:27:34,580 --> 00:27:37,700
So for both of you, AI is
obviously transforming creative
464
00:27:37,700 --> 00:27:41,080
industries right now, opening up
huge possibilities, but also
465
00:27:41,080 --> 00:27:42,160
raising some big concerns.
466
00:27:43,780 --> 00:27:46,540
So I'd love to hear about what
you do. This means New York
467
00:27:46,540 --> 00:27:49,480
Times as some artists and
creative if you saw you, Megan,
468
00:27:50,080 --> 00:27:53,440
you mentioned backstage that you
and your team are building your
469
00:27:53,560 --> 00:27:56,800
own AI, impersonation, Detective
detection system. Talk to us a
470
00:27:56,800 --> 00:28:00,240
little bit about that for sure,
sure. Well, I'll just first say
471
00:28:00,240 --> 00:28:03,360
that I think that technology has
472
00:28:05,700 --> 00:28:08,640
has done a lot for the music
business. I think that it has
473
00:28:08,640 --> 00:28:12,600
always pushed creativity
forward. But I will say it is a
474
00:28:12,600 --> 00:28:16,800
scary world we live in, because
you just never know how they're
475
00:28:16,800 --> 00:28:19,320
going to use your face or your
voice. And the truth is, like
476
00:28:19,320 --> 00:28:21,840
we're just talking about that
genuine connection, it's really
477
00:28:21,840 --> 00:28:26,540
important for me to feel like
anytime my fans see something
478
00:28:26,600 --> 00:28:29,840
that they feel like it is really
like evils, and saying it to
479
00:28:29,840 --> 00:28:31,880
them, I want them to feel like
that they can trust me.
480
00:28:33,320 --> 00:28:37,760
I have partnered with a company
called Looney that is constantly
481
00:28:38,060 --> 00:28:43,540
working on pulling down these
imposters or these social media
482
00:28:43,540 --> 00:28:46,420
posts, whatever it is that is
not coming from me or my team,
483
00:28:46,420 --> 00:28:50,200
and that is a full time job at
this point, and it is really,
484
00:28:50,200 --> 00:28:53,260
really crazy. And the scary
thing is, is even when your
485
00:28:53,260 --> 00:28:55,840
family members think it's
actually you, you're like, Oh
486
00:28:55,840 --> 00:28:58,600
Lord, they're getting really
good at this. But, but, yeah, we
487
00:28:58,600 --> 00:29:03,120
have, we have found a team of
people who are taking serious
488
00:29:03,120 --> 00:29:06,240
just like you are. And I just
want to keep that, that
489
00:29:06,240 --> 00:29:07,680
relationship with my fans,
490
00:29:08,820 --> 00:29:11,700
the way that, the way that I
wanted to be genuine and pure
491
00:29:11,700 --> 00:29:14,580
and authentic you're coming into
this. But I suspect that every
492
00:29:14,580 --> 00:29:17,160
creator is going to want
something like this, and maybe
493
00:29:17,160 --> 00:29:21,180
eventually every human, as we go
into this world where it's
494
00:29:21,180 --> 00:29:25,700
always understand, it's probably
495
00:29:27,020 --> 00:29:29,660
to understand, if you can
actually trust David, you're the
496
00:29:30,560 --> 00:29:31,100
president,
497
00:29:32,660 --> 00:29:36,080
it's a very broad topic, but if
you just pick what the documents
498
00:29:36,080 --> 00:29:36,980
are you most excited
499
00:29:39,020 --> 00:29:41,980
about? So have you transition in
AI can be sort of like an up
500
00:29:41,980 --> 00:29:44,500
leveling of creative potential.
I'll just start with that
501
00:29:44,500 --> 00:29:47,440
statement. There's certainly
risks associated with
502
00:29:47,440 --> 00:29:51,700
impersonation and theft of IP
and whatnot, which I think will
503
00:29:51,700 --> 00:29:54,880
get worked out as it was YouTube
and Spotify in the early days.
504
00:29:55,240 --> 00:29:55,960
Then we'll get there.
505
00:29:57,280 --> 00:29:59,920
But I got my first Macintosh
computer where six.
506
00:30:00,000 --> 00:30:02,820
Years old, the Macintosh
original, and I was on that
507
00:30:02,820 --> 00:30:06,120
computer learning how to use MAC
Paint. And eventually I got to
508
00:30:06,120 --> 00:30:09,840
go to Photoshop. I was doing
graphic design at the age of 11
509
00:30:09,840 --> 00:30:12,900
or 12 as a service to people. I
got making money, making
510
00:30:12,900 --> 00:30:16,500
business cards and flyers and
stuff. And so I was early on
511
00:30:16,500 --> 00:30:20,120
when Kai's power tools came out.
Does anyone remember that? No
512
00:30:20,120 --> 00:30:21,920
one in this room that is so
depressing.
513
00:30:25,220 --> 00:30:25,460
Okay,
514
00:30:26,600 --> 00:30:29,360
there's a couple, but Adobe
Photoshop, right? You can put
515
00:30:29,360 --> 00:30:32,900
like images on and edit them and
manipulate them, and power tools
516
00:30:32,900 --> 00:30:36,020
have these filters. And you
would do a filter to the image,
517
00:30:36,560 --> 00:30:39,440
and it was like, Holy crap. You
could make the image sharper,
518
00:30:39,440 --> 00:30:41,560
you could make it blur. You
could do all. You can create
519
00:30:41,560 --> 00:30:44,440
depth. You can do all these
things that people were like,
520
00:30:44,440 --> 00:30:48,100
That's cheating. Like, that's
not fair. Like, how can you do
521
00:30:48,100 --> 00:30:51,520
that? You're the cameraman
supposed to do that, or the guy
522
00:30:51,520 --> 00:30:53,560
who learned how paintbrush is
the one who should be doing
523
00:30:53,560 --> 00:30:57,700
that. And so there's always and
then later on, I did, like, you
524
00:30:57,700 --> 00:31:00,660
know, made techno music. Don't
make fun of me for it. But, you
525
00:31:00,840 --> 00:31:03,420
know, you and I talked about, I
used to have, like, a Juno 106,
526
00:31:03,960 --> 00:31:07,500
TV 303, and then I got all the
synthetic sequencers on my
527
00:31:07,500 --> 00:31:10,500
computer, and there were all
these plugins that came out, and
528
00:31:10,500 --> 00:31:13,440
the plugins became, I mean, just
like a music industry today, it
529
00:31:13,440 --> 00:31:15,840
was a question of, like, Are you
cheating? Like, is that really
530
00:31:15,840 --> 00:31:18,960
fair? And what ended up
happening is creators, whether
531
00:31:19,020 --> 00:31:22,760
graphic creators or music
creators found extraordinary
532
00:31:22,760 --> 00:31:26,240
ways to leverage these new tools
to create new soundscapes, new
533
00:31:26,240 --> 00:31:31,220
visual features, new types of
film. Today in Hollywood, a lot
534
00:31:31,220 --> 00:31:35,240
of the production houses are and
I've spent some time with some
535
00:31:35,240 --> 00:31:38,060
of these companies, they're
using like second units to do
536
00:31:38,060 --> 00:31:40,960
fully AI rendered like second
units, unit. So the cost comes
537
00:31:40,960 --> 00:31:44,440
down, but the stuff you can do
goes up by, like, 100x because
538
00:31:44,440 --> 00:31:47,680
you couldn't, you didn't have
the budget to go and, like,
539
00:31:47,680 --> 00:31:51,640
recreate a trading floor from
the 1970s and now you can do an
540
00:31:51,640 --> 00:31:54,040
AI in a couple of minutes. And
so the creator, the director,
541
00:31:54,040 --> 00:31:56,320
has this ability to make a
landscape that they otherwise
542
00:31:56,320 --> 00:31:59,560
wouldn't have the ability to
make. And so I think that
543
00:31:59,560 --> 00:32:02,100
there's a leveling up that's
happening in terms of what
544
00:32:02,100 --> 00:32:04,680
creators can do, and we're just
starting to see people tinker
545
00:32:04,680 --> 00:32:07,440
with these things, and the stuff
that I've seen, especially in
546
00:32:07,440 --> 00:32:10,560
video, and I'm sure over time,
we'll see more and more happen
547
00:32:10,560 --> 00:32:13,800
in music. It's just incredible.
So I think that the second big
548
00:32:13,800 --> 00:32:18,060
wave that's probably still to
come is non linear content. And
549
00:32:18,060 --> 00:32:21,560
what I mean by that is, like
today, whether it's a video or
550
00:32:22,160 --> 00:32:25,220
music track, you just press play
and it goes linearly from
551
00:32:25,220 --> 00:32:28,760
beginning to end. I think that
creators are going to unlock
552
00:32:28,760 --> 00:32:32,900
like in video games, the ability
to maybe have some difference in
553
00:32:32,900 --> 00:32:36,260
how that track plays or how that
movie plays out. Whether it's
554
00:32:36,260 --> 00:32:38,720
like, click and choose your own
adventure, which I think may be
555
00:32:38,720 --> 00:32:41,500
a little annoying to consumers,
or it's like, what if you could
556
00:32:41,500 --> 00:32:44,560
listen to your favorite house
music track and make it a little
557
00:32:44,560 --> 00:32:47,680
bit faster, little bit cooler, a
little bit shorter, add some
558
00:32:47,680 --> 00:32:51,280
vocals, add some jazz, but the
Creator, not the user, but the
559
00:32:51,280 --> 00:32:54,280
Creator can set all those
parameters. So now the creator
560
00:32:54,280 --> 00:32:56,500
is almost like coming up with
all these ideas, but in
561
00:32:56,500 --> 00:32:59,200
different dimensions of how they
can create music. And if you
562
00:32:59,200 --> 00:33:01,620
don't have to have one final
track, and you can say, hey, we
563
00:33:01,620 --> 00:33:04,080
could make a couple different
ways that this track can play
564
00:33:04,080 --> 00:33:06,660
out, and then the actual
playback for the consumer,
565
00:33:06,660 --> 00:33:09,720
whether it's in a video or in
music, could be a little bit
566
00:33:09,720 --> 00:33:12,300
different. And so I think we're
likely going to see like,
567
00:33:12,360 --> 00:33:15,420
progressive creators come up
with incredible ways of using
568
00:33:15,420 --> 00:33:17,940
these tools and these
capabilities, where people are
569
00:33:17,940 --> 00:33:20,120
going to be watching movies
differently. Like, hey, we all
570
00:33:20,120 --> 00:33:23,060
watch the same show, but I saw a
little different view that you
571
00:33:23,060 --> 00:33:25,280
saw, like I saw from the bad
guys view you saw from the good
572
00:33:25,280 --> 00:33:28,160
guys view, or we all track a
little bit differently. So
573
00:33:28,160 --> 00:33:31,160
that's the sort of stuff that I
think would be really cool with
574
00:33:31,160 --> 00:33:35,180
the AI unlocked sort of
different experience. So as the
575
00:33:35,180 --> 00:33:35,540
creator
576
00:33:37,160 --> 00:33:38,900
and control search on these
guys,
577
00:33:40,340 --> 00:33:40,600
I creator,
578
00:33:43,240 --> 00:33:44,260
if there was a framework
579
00:33:45,340 --> 00:33:47,740
for your fans to be able to work
with the music, where
580
00:33:48,880 --> 00:33:50,500
everyone was greatly
compensated,
581
00:33:53,560 --> 00:33:55,360
because it is your music,
582
00:33:57,700 --> 00:34:00,040
for sure, I think it's fun to be
creative, but, you know, I
583
00:34:00,100 --> 00:34:02,340
think, Like, that's why I
started doing this to begin
584
00:34:02,340 --> 00:34:02,640
with. So
585
00:34:03,720 --> 00:34:06,120
I would love for my fans to be
able to do something, if it's
586
00:34:06,180 --> 00:34:12,420
fair for the person who wrote
it, proposed it, created it. I'm
587
00:34:12,420 --> 00:34:15,180
down for that. I want everybody
to be able to like color outside
588
00:34:15,180 --> 00:34:17,460
of the lines and try new things.
So like I said, if people are
589
00:34:17,460 --> 00:34:19,440
compensated for it, Hell yeah.
Hell yeah.
590
00:34:22,800 --> 00:34:24,620
Well, okay, are you guys doing
anything like this?
591
00:34:26,360 --> 00:34:29,360
It's fun. Well, I'm looking at
my comments.
592
00:34:34,520 --> 00:34:36,980
There is this one feature
working on that touches on this
593
00:34:36,980 --> 00:34:40,280
that absolutely love, which
people have been asking for
594
00:34:40,640 --> 00:34:42,880
forever. So you know, in movies,
you have like, I am the game.
595
00:34:43,300 --> 00:34:45,940
You can go in and see, like,
who's the director, who's
596
00:34:45,940 --> 00:34:49,600
producer. You're really nerd up
in music, that hasn't often been
597
00:34:49,600 --> 00:34:52,180
the case. You often know who the
performer is, but most people
598
00:34:52,180 --> 00:34:56,620
have no idea who wrote it or
produced it. So this is leaked a
599
00:34:56,620 --> 00:34:57,220
little bit
600
00:34:58,420 --> 00:34:59,920
already, but we are. We're
rolling out this.
601
00:35:00,000 --> 00:35:04,200
This feature called Song DNA,
where we meticulously go through
602
00:35:04,260 --> 00:35:06,900
to understand exactly who played
603
00:35:09,600 --> 00:35:12,540
a role on each song. You can see
that song writer
604
00:35:17,640 --> 00:35:18,240
actually
605
00:35:20,040 --> 00:35:24,080
like that sample is actually
from this original song. So
606
00:35:24,080 --> 00:35:24,920
we'll see, I think,
607
00:35:26,780 --> 00:35:29,000
and I think it's really
important that there are all
608
00:35:29,000 --> 00:35:32,120
these creators around the
performance that there are
609
00:35:32,120 --> 00:35:32,360
that's
610
00:35:33,860 --> 00:35:35,000
something I'm very excited.
611
00:35:38,420 --> 00:35:39,260
It's really odd.
612
00:35:40,780 --> 00:35:40,960
That.
613
00:35:45,760 --> 00:35:47,200
Okay, all
614
00:35:49,060 --> 00:35:50,440
right, another question for
615
00:35:51,520 --> 00:35:51,940
both of you,
616
00:35:56,560 --> 00:35:57,220
the CEO, if
617
00:36:00,000 --> 00:36:00,660
environment.
618
00:36:03,180 --> 00:36:06,900
Now for both of you as the old
my partner, Alex and I have
619
00:36:07,080 --> 00:36:10,020
stepped into the roles of those
deals. We're spending a lot of
620
00:36:10,020 --> 00:36:13,440
time talking with artists, with
podcast creators and with
621
00:36:13,440 --> 00:36:13,680
authors.
622
00:36:15,060 --> 00:36:17,580
I love the books. I'm very
excited about this all around
623
00:36:17,580 --> 00:36:20,180
the world, and one of my
favorite questions is, what can
624
00:36:20,180 --> 00:36:21,440
be building for you as a creator
625
00:36:22,820 --> 00:36:24,020
or as a user, by the way, yeah,
well,
626
00:36:26,000 --> 00:36:28,100
into speaking and as a creator
and a user, I
627
00:36:29,300 --> 00:36:31,520
think would be cool to build
something based off of demons
628
00:36:31,520 --> 00:36:33,620
that day, you know? Because, I
mean, the truth
629
00:36:34,760 --> 00:36:35,180
is that
630
00:36:40,280 --> 00:36:43,300
we're all going through
something at different times,
631
00:36:43,300 --> 00:36:46,240
and it'd be really cool for my
fans to feel like that day that
632
00:36:46,240 --> 00:36:48,400
they were going through
something hard. I was speaking
633
00:36:48,400 --> 00:36:50,560
directly to them, that they
didn't have to, like, listen
634
00:36:50,560 --> 00:36:53,140
through the whole album to
figure out, like, Which song is
635
00:36:53,140 --> 00:36:55,900
resonating with them for that
specific thing that they're
636
00:36:55,900 --> 00:36:56,860
going through.
637
00:36:58,360 --> 00:37:01,620
So very cool. So you were like,
That's interesting. You can So,
638
00:37:01,620 --> 00:37:03,240
I mean, you could go into a
Spotify and
639
00:37:08,700 --> 00:37:11,100
type in a little down here from
your fans, because there are so
640
00:37:11,760 --> 00:37:15,600
many of them in some sort of
aggregated way. Yes, I think
641
00:37:15,600 --> 00:37:18,600
that would be cool too, to know
how your fans are
642
00:37:21,600 --> 00:37:22,460
feeling, too.
643
00:37:24,680 --> 00:37:27,620
I mean, so many different people
from so many walks of life have
644
00:37:28,400 --> 00:37:33,980
like, it's crazy the story that
people and telling me, but one
645
00:37:33,980 --> 00:37:36,920
of the three lines for me that
I've noticed this evening in
646
00:37:36,920 --> 00:37:40,660
meet and greets is they are
willing to share like where they
647
00:37:40,660 --> 00:37:41,260
are in life and
648
00:37:42,760 --> 00:37:45,760
where they are going through and
so I don't know, just be able
649
00:37:45,760 --> 00:37:48,460
to, like, cater to that even
more, to even just have, like, a
650
00:37:48,460 --> 00:37:52,480
deeper connection like that with
your family. Because my goal
651
00:37:52,480 --> 00:37:55,660
always is for them to feel like
I'm speaking to them in their
652
00:37:55,720 --> 00:37:58,420
story, in their life. But if
they're going through something,
653
00:37:58,420 --> 00:37:59,920
and I'm singing a song about,
654
00:38:01,180 --> 00:38:03,780
let's have a part, you know? I
mean, they might not, that
655
00:38:05,040 --> 00:38:05,880
might not be the thing
656
00:38:08,340 --> 00:38:09,540
that helps them get through
657
00:38:11,100 --> 00:38:14,640
it, the musician, you only get
data on how many plays there
658
00:38:14,640 --> 00:38:17,460
were and what the average listen
kind of was, that's
659
00:38:18,660 --> 00:38:20,040
that's all the only feedback you
get.
660
00:38:21,240 --> 00:38:23,240
It's true. Yeah, you do get
661
00:38:25,100 --> 00:38:28,100
statistics about who is
listening, ages and where they
662
00:38:28,160 --> 00:38:30,860
listen. You can find your tours,
but you're completely right. You
663
00:38:30,860 --> 00:38:34,520
have no idea how they're feeling
and looking at AI. I just talked
664
00:38:34,520 --> 00:38:38,000
about how AI allows people to
speak in English at scale. That
665
00:38:38,000 --> 00:38:41,120
is a really cool thing, to
actually understand more about
666
00:38:42,080 --> 00:38:46,840
their demographic. Well, a
couple years ago, there was a
667
00:38:46,840 --> 00:38:49,360
guy who took all of our
episodes, like, right when AI
668
00:38:49,360 --> 00:38:51,880
was kind of getting going, and
he ingested them all, and then
669
00:38:51,880 --> 00:38:55,240
he created, like an AI character
with each of the four of us, and
670
00:38:55,420 --> 00:38:57,820
then it was like, each of the
four of us said a bunch of
671
00:38:57,820 --> 00:39:00,400
stuff, and it was so bullshit,
like that sounded awful. He's
672
00:39:00,400 --> 00:39:03,600
gotten so much better now. But
then someone else made a chat
673
00:39:03,600 --> 00:39:05,880
interface where you could ask
each of those questions they
674
00:39:05,880 --> 00:39:08,700
could answer, as if it was
trained on the data from our
675
00:39:08,700 --> 00:39:11,520
show. And it was like, Wait,
that's actually kind of
676
00:39:11,520 --> 00:39:14,700
interesting. Like, very often
I'm trying to find something
677
00:39:14,700 --> 00:39:17,580
that I heard on another show,
like Friedman organ or
678
00:39:17,580 --> 00:39:21,140
something. I'm like, Where was
that? And so to find what was
679
00:39:21,140 --> 00:39:25,160
said by guests, or find really
guests and protection was very
680
00:39:25,160 --> 00:39:29,000
hard. Today I got to go, like to
YouTube and do transcript
681
00:39:29,000 --> 00:39:31,340
search. Or nowadays, I can do
Gemini, and Gemini like, search
682
00:39:31,340 --> 00:39:35,480
through all the transcript data
for me and pull stuff out. But I
683
00:39:35,480 --> 00:39:38,360
think every podcaster should
have the opportunity to create
684
00:39:38,360 --> 00:39:42,820
like Lex chat, or Joe chat, or
something where you as a user
685
00:39:42,820 --> 00:39:45,700
who's a big fan of the podcast,
like, the thing I realized about
686
00:39:45,700 --> 00:39:49,480
podcasting is it's people want
to, like, get really developed.
687
00:39:49,480 --> 00:39:52,060
They come to these live events.
Like, we've been on a live event
688
00:39:52,060 --> 00:39:55,780
with like, 1000s of people in LA
every year. We've got one here
689
00:39:55,780 --> 00:39:58,000
tomorrow night. And they want
to, like, absorb more the
690
00:39:58,000 --> 00:39:59,920
content, get into the community.
Applicant.
691
00:40:00,000 --> 00:40:03,960
Engage, engage, engage, because
that particular content vertical
692
00:40:03,960 --> 00:40:06,300
is something they're really
into. So to have a chat
693
00:40:06,300 --> 00:40:09,360
interface where I could ask Joe
Rogan questions like, Hey, Who
694
00:40:09,360 --> 00:40:11,940
was your favorite guest? Which
was the craziest shit you ever
695
00:40:11,940 --> 00:40:14,340
heard. You know, tell me about
that guy that talked about, like
696
00:40:14,340 --> 00:40:16,860
there were pyramids under the
ocean and we found them. Like,
697
00:40:17,040 --> 00:40:19,920
who was that guy? I would love
to be able to have that dialog
698
00:40:19,920 --> 00:40:23,300
with my podcast that I love. I
think that would be a really
699
00:40:23,300 --> 00:40:25,820
cool feature. You guys can train
it off on the transcript data,
700
00:40:25,880 --> 00:40:28,100
and theoretically, you could
actually make the audio sound
701
00:40:28,100 --> 00:40:31,580
like the host. And I could have
a dialog through the Spotify app
702
00:40:31,940 --> 00:40:34,520
and all of these that creates
more engagement for you guys,
703
00:40:35,000 --> 00:40:37,040
and all of the I should be your
kind of
704
00:40:39,380 --> 00:40:41,980
product. So yeah, anyway,
because I would love to chat
705
00:40:41,980 --> 00:40:44,560
with Lex Friedman and then be
like, Hey, tell me about that
706
00:40:45,520 --> 00:40:46,660
great Hancock interview.
707
00:40:47,800 --> 00:40:51,280
And then our users could just
ask David to talk to digital I
708
00:40:51,280 --> 00:40:52,060
got enough of that.
709
00:40:53,620 --> 00:40:57,340
But I do worry about, certainly,
this thing of, like, digital
710
00:40:57,340 --> 00:41:00,520
representation. It's gonna
happen. It's happening to a
711
00:41:00,520 --> 00:41:04,680
point, and they get it better.
They get an opt in. Yeah, people
712
00:41:04,680 --> 00:41:07,200
that I think, I think podcasters
are different than these issues.
713
00:41:07,320 --> 00:41:09,720
So podcasters are creating
content. They're not creating
714
00:41:10,140 --> 00:41:14,520
these, like, you know, tracks,
the songs that are kind of their
715
00:41:14,520 --> 00:41:16,980
product out, but they want to
protect the value of the song
716
00:41:16,980 --> 00:41:19,800
and so on. Podcasters more
content they're creating, I
717
00:41:19,800 --> 00:41:22,520
think the more engagement they
get, the more depth they get. So
718
00:41:22,520 --> 00:41:24,800
there's something about re
engaging in that content that
719
00:41:24,800 --> 00:41:27,320
actually brings a lot of value
for the podcaster. And it
720
00:41:27,320 --> 00:41:30,020
doesn't need to be pure it
doesn't need to be how I said
721
00:41:30,020 --> 00:41:33,920
the thing at the time. I said it
per se, but generally, like, the
722
00:41:33,920 --> 00:41:38,720
things I said, like, I'd be fine
for AI chat interface. Just be
723
00:41:38,720 --> 00:41:41,800
trained on that and want to
answer questions. Interesting
724
00:41:41,800 --> 00:41:44,860
when music is interesting as a
content category because it has
725
00:41:44,860 --> 00:41:48,280
so much repeat testing, people
listen the same track hundreds,
726
00:41:48,280 --> 00:41:51,580
1000s of times. But in podcasts,
you mostly don't the same
727
00:41:51,580 --> 00:41:54,400
episode again. This would be a
great value from the back
728
00:41:54,400 --> 00:41:57,220
catalog as well, right? But
people do listen to the back
729
00:41:57,220 --> 00:42:01,560
catalog, like acquired Rogan
Lex, like some of these are very
730
00:42:01,560 --> 00:42:04,200
timeless episodes, because
they're interview shows with
731
00:42:04,200 --> 00:42:07,980
someone on particular topic. And
so, like, I mean, you know the
732
00:42:07,980 --> 00:42:11,760
stats on acquire business
podcast, once people listen to
733
00:42:11,760 --> 00:42:13,980
an acquired podcast, they go
back and listen to the whole
734
00:42:13,980 --> 00:42:16,260
brain library, and they're like,
this show's awesome. I wasn't
735
00:42:16,260 --> 00:42:19,740
all of these, and there's a
bunch of pods like that. So I
736
00:42:19,740 --> 00:42:21,980
think to have some sort of
engagement
737
00:42:24,320 --> 00:42:25,700
mechanism around that stuff
could
738
00:42:27,080 --> 00:42:30,980
be very cool. Okay, we got some
product ideas. My first virus
739
00:42:31,040 --> 00:42:32,180
here, amazing
740
00:42:33,980 --> 00:42:37,640
lady, you're not just an artist,
you're also founder and investor
741
00:42:37,760 --> 00:42:41,860
of fashion, jewelry and gaming.
So we talked about how you think
742
00:42:41,860 --> 00:42:44,500
about your arts career, but how
do you think about building a
743
00:42:44,500 --> 00:42:46,120
business alongside your music
career?
744
00:42:48,640 --> 00:42:52,660
Well, like I said earlier, I
just really like to be creative,
745
00:42:53,020 --> 00:42:56,080
and I want to make sure that my
parents feel like they've been
746
00:42:56,680 --> 00:42:59,080
they can like that into the work
of my music as much as I
747
00:42:59,080 --> 00:43:02,160
possibly can or as much as I
want to. And,
748
00:43:03,660 --> 00:43:07,680
you know, I look out at my shows
and I'm seeing people who dress
749
00:43:07,680 --> 00:43:10,860
just like me, wearing the bell
bottoms in their hats and the
750
00:43:10,860 --> 00:43:12,900
jewelry. And this
751
00:43:14,160 --> 00:43:17,040
is it. Yeah, we should have put
you in some Bill
752
00:43:18,540 --> 00:43:19,380
bottoms too.
753
00:43:22,200 --> 00:43:27,080
But, yeah, it's just another way
for me to connect with my people
754
00:43:27,320 --> 00:43:30,080
and with my fans, and, like I
said, for them to be able to,
755
00:43:30,080 --> 00:43:35,180
like, live in that world as much
as they want to. And it's fun.
756
00:43:35,180 --> 00:43:37,160
You know, when you're all under
like, the same roof and you're
757
00:43:37,160 --> 00:43:39,380
like, looking out there and
everybody kind of looks like
758
00:43:39,380 --> 00:43:44,080
you, it's a especially like with
a little girls and stuff, it
759
00:43:44,080 --> 00:43:47,020
makes me feel like, man, like, I
better watch what I do and say,
760
00:43:47,020 --> 00:43:50,020
because I have a lot of little
people watching me. But also it,
761
00:43:50,020 --> 00:43:51,580
it gives me this
762
00:43:53,200 --> 00:43:56,560
sense of just like
responsibility. To feel like,
763
00:43:56,620 --> 00:44:00,480
okay, I need to be a leader. I
need to, like, set an example.
764
00:44:00,780 --> 00:44:02,280
And yeah, it's
765
00:44:03,600 --> 00:44:06,240
a good feeling. It's a good
feeling to I feel blessed with
766
00:44:06,240 --> 00:44:09,540
like have this platform and be
on that stage, and
767
00:44:10,620 --> 00:44:14,280
I just Yeah, I want people to
feel like they know me. I want
768
00:44:16,200 --> 00:44:16,860
to feel like I
769
00:44:19,020 --> 00:44:19,320
know them.
770
00:44:20,640 --> 00:44:22,520
But we spoke backstage
771
00:44:30,740 --> 00:44:34,940
for so long, I did not have any
opportunity at all. I've been at
772
00:44:34,940 --> 00:44:36,080
this a long time.
773
00:44:37,160 --> 00:44:40,960
I've been living in Asheville
for 15 years, and the first 10
774
00:44:42,100 --> 00:44:45,520
I should have packed up, I went
home, but I just couldn't. This
775
00:44:45,520 --> 00:44:47,860
is the only thing I know how to
do. It's the only thing I want
776
00:44:47,860 --> 00:44:53,080
to do is just to tell stories.
And yeah, I feel like the last
777
00:44:53,080 --> 00:44:57,040
few years have just been such a
blessing to me and my life and
778
00:44:57,040 --> 00:44:59,260
my team, like I have finally
found the right people to
779
00:44:59,260 --> 00:44:59,920
surround myself.
780
00:45:00,000 --> 00:45:04,560
With that understand my vision
and don't want to change me for
781
00:45:04,560 --> 00:45:07,260
anything, and if anything, they
just want to elevate me and get
782
00:45:07,260 --> 00:45:11,340
me to that next step. And it's
getting really fun now. So now
783
00:45:11,340 --> 00:45:14,640
we have all these opportunities.
There's only so many hours in
784
00:45:14,640 --> 00:45:17,580
the day, but I do want to take
them all. You know, if it feels
785
00:45:17,580 --> 00:45:20,040
right and it feels like it's
going to move the needle, then
786
00:45:20,700 --> 00:45:22,460
let's do it. You told me you're
from this
787
00:45:24,380 --> 00:45:25,220
tiny, tiny town 170
788
00:45:27,380 --> 00:45:28,580
and you moved
789
00:45:31,340 --> 00:45:33,260
pretty further on. Yes, I did.
I'm from
790
00:45:34,280 --> 00:45:35,540
Baskin, Louisiana,
791
00:45:36,740 --> 00:45:37,700
a little town of 170
792
00:45:38,780 --> 00:45:43,180
people. I'm very proud where I
come from, but honestly, I
793
00:45:43,180 --> 00:45:47,260
compare. I was telling you this
in the room too, that I compare
794
00:45:47,260 --> 00:45:50,800
a lot of what I do to what my
dad does. My dad is a farmer,
795
00:45:50,920 --> 00:45:55,540
and he farms corn, wheat,
soybeans, oats, and gets up
796
00:45:55,540 --> 00:45:59,320
every day and busts his tail
doing it, and he does the same
797
00:45:59,320 --> 00:46:02,400
thing every single day. I mean,
it looks a little bit different
798
00:46:02,400 --> 00:46:06,480
for me every day. But, um, you
know, at the end of the day,
799
00:46:06,480 --> 00:46:09,600
it's your livelihood, if it's in
you, then you just get up and
800
00:46:09,600 --> 00:46:13,980
you do it. A storm could roll
through, and you just get up and
801
00:46:13,980 --> 00:46:14,280
do it
802
00:46:18,300 --> 00:46:23,600
anyway. So I call myself a song
farmer, amazing. Amazing. David
803
00:46:24,800 --> 00:46:29,180
spoken is a big movement podcast
in 2019 and you and your girls
804
00:46:29,180 --> 00:46:31,820
started, as you said, during
covid in 2020
805
00:46:32,840 --> 00:46:35,060
Could you talk a little bit to
us about the growth and
806
00:46:35,060 --> 00:46:38,420
evolution that you've seen and
sort of where you think it's
807
00:46:38,420 --> 00:46:41,920
going next? You mentioned some
of the ideas you have. But what
808
00:46:41,920 --> 00:46:42,940
do you see podcasting?
809
00:46:44,260 --> 00:46:46,720
Well, podcasting kind of became
what blogging used to be.
810
00:46:46,720 --> 00:46:48,940
Everyone that had blogs prior to
811
00:46:50,620 --> 00:46:50,800
that,
812
00:46:52,120 --> 00:46:52,600
everyone
813
00:46:53,800 --> 00:46:56,440
so in the business world, like,
if you've got,
814
00:46:57,580 --> 00:47:00,220
you know, I think one of the
problems is a lot of this stuff
815
00:47:00,220 --> 00:47:04,500
got solid, and maybe we're part
of this problem as well. But
816
00:47:04,500 --> 00:47:08,160
like people selling shit, like,
like, I'm trying to like
817
00:47:09,240 --> 00:47:11,520
people doing podcasts because
they're trying to, like,
818
00:47:11,760 --> 00:47:14,220
heighten their profile because
they have stuff to sell. I don't
819
00:47:14,400 --> 00:47:16,980
know if that intention really
resonates and works. Well,
820
00:47:17,340 --> 00:47:21,080
maybe, maybe we're part of maybe
we should be blamed for that as
821
00:47:21,080 --> 00:47:21,920
well, I don't know, but
822
00:47:23,060 --> 00:47:25,760
I feel like there's an aspect of
that where, like, everyone's got
823
00:47:25,760 --> 00:47:29,540
podcasts now, and the market
will sort itself out. People
824
00:47:29,540 --> 00:47:32,120
listen to what they want to
listen to and stuff for. So it's
825
00:47:32,120 --> 00:47:37,460
very crowded. It's very noisy
out there. And I think to some
826
00:47:37,460 --> 00:47:41,260
extent, we're seeing folks step
it up, right? So the visual
827
00:47:41,260 --> 00:47:45,880
stuff matters a lot, I think
more than people are realizing,
828
00:47:45,880 --> 00:47:50,080
but a lot of the stuff will hold
its own as audio. But Chris
829
00:47:50,080 --> 00:47:52,420
Williams said, I don't know if
you guys know that this guy is
830
00:47:52,420 --> 00:47:55,600
podcaster. He did anyone see the
episode?
831
00:47:59,200 --> 00:48:02,040
They built a set on the same
kind of sound stage and set
832
00:48:02,040 --> 00:48:04,680
stage that they use for the
Mandalorians, like this LED wall
833
00:48:04,680 --> 00:48:07,500
that, like, wrapped around it's
so and they built Interstellar,
834
00:48:07,920 --> 00:48:10,080
like the guy's house with
Interstellar, with the trailer
835
00:48:10,080 --> 00:48:13,320
and the driveway. It's freaking
insane. I recommend you watch it
836
00:48:13,320 --> 00:48:16,500
on YouTube, just to see what
they built. The YouTube thing,
837
00:48:16,560 --> 00:48:16,800
I'm so
838
00:48:18,420 --> 00:48:18,660
sorry.
839
00:48:20,580 --> 00:48:20,660
I so
840
00:48:27,140 --> 00:48:29,720
this is why I think the video on
Spotify doesn't matter a lot,
841
00:48:29,960 --> 00:48:32,540
because I do think a lot of the
folks are differentiating by
842
00:48:32,540 --> 00:48:35,420
improving the quality of the
video content. And it's not just
843
00:48:35,420 --> 00:48:38,720
me having a chat and then
obviously the podcasts are more
844
00:48:38,720 --> 00:48:39,860
creative, like all
845
00:48:40,940 --> 00:48:44,500
the murder crime industry stuff.
It's so cool, like it's a
846
00:48:44,920 --> 00:48:47,440
totally different level. It's
not just someone sitting at the
847
00:48:47,440 --> 00:48:50,560
table having a conversation. Joe
Rogan can hold his own having a
848
00:48:50,560 --> 00:48:52,960
conversation for two hours, and
we'll get great content.
849
00:48:54,340 --> 00:48:57,520
But most people cannot. Most
people need it. So the good
850
00:48:57,520 --> 00:49:01,380
creator they're gonna need to go
to, like murder crime history,
851
00:49:01,380 --> 00:49:04,380
that might be a better format.
So I think there's a lot of this
852
00:49:04,380 --> 00:49:06,900
differentiation that's going to
start to happen, which, by the
853
00:49:06,900 --> 00:49:09,240
way, will like it fundamentally
improve the quality of the
854
00:49:09,240 --> 00:49:11,640
experience that the consumers,
the listeners are like, the
855
00:49:11,640 --> 00:49:14,400
viewers are like, this is
awesome. Now, like that Matthew
856
00:49:14,400 --> 00:49:17,280
McConaughey episode, visually,
it's just like, I don't care
857
00:49:17,280 --> 00:49:21,260
what McConaughey is saying, this
is just so cool. Pretty cool.
858
00:49:21,620 --> 00:49:24,920
And ask you, one of the things
that I think people like about
859
00:49:24,980 --> 00:49:28,580
listening to you numbers is that
you have a positive view of the
860
00:49:28,580 --> 00:49:31,100
future. And there's a lot of
stuff here right now, and you're
861
00:49:31,100 --> 00:49:32,240
going to start a podcast.
862
00:49:33,980 --> 00:49:38,180
Well, we'll see. But I do think
the thing that troubles me the
863
00:49:38,180 --> 00:49:41,440
most is just kind of how techno
pessimistic we are, right? And I
864
00:49:41,440 --> 00:49:44,080
think, like, everyone's just
like ai humorism, like AI is the
865
00:49:44,140 --> 00:49:47,260
most unfavorable thing right now
in the country. There's some
866
00:49:47,260 --> 00:49:50,680
survey that came out it was like
Iran, like Iran is more
867
00:49:50,740 --> 00:49:51,940
favorable than AI. Like, the
sort
868
00:49:53,320 --> 00:49:57,580
of thing that happened where AI
just became the new like,
869
00:49:57,580 --> 00:49:59,920
whether it was climate change
or.
870
00:50:00,000 --> 00:50:03,300
Or whatever, there's always
this, like looping threat that
871
00:50:03,360 --> 00:50:07,680
we latch on to. And I think
we've had this, like this shift
872
00:50:07,680 --> 00:50:11,580
in our psyche, particularly in
the West, meaning the US and
873
00:50:11,580 --> 00:50:16,680
Western Europe, probably since
the 1970s like 1969 we landed on
874
00:50:16,680 --> 00:50:19,380
the moon and we had Woodstock
that same summer, and I think
875
00:50:19,380 --> 00:50:21,980
that was the summer when we
became techno pessimists, where
876
00:50:21,980 --> 00:50:24,380
we were like, worried about
chemicals in our water,
877
00:50:24,560 --> 00:50:27,740
radiation killing us. Nuclear
energy could burn, melt down and
878
00:50:27,740 --> 00:50:30,920
wipe us. Nuclear bombs are
overhead. So we simply became
879
00:50:30,920 --> 00:50:34,220
way more concerned about tech.
But if you look at like, how
880
00:50:34,220 --> 00:50:37,580
tech is progressing, human
lifespan, human health span, how
881
00:50:38,300 --> 00:50:40,660
many hours people have to work,
the number of hours that people
882
00:50:40,660 --> 00:50:43,780
have to work is going down the
quality of life. I know a lot of
883
00:50:43,780 --> 00:50:46,300
people are struggling. Still in
America, there's a lot of less
884
00:50:46,300 --> 00:50:49,720
behind differences in what's
going on, but a lot of people
885
00:50:49,780 --> 00:50:53,020
are moving up the ladder because
of the advances in technology.
886
00:50:53,020 --> 00:50:55,480
We all got a freaking super
computer in our pocket. There's
887
00:50:55,480 --> 00:50:58,420
just so much that I think we
take for granted. And when I
888
00:50:58,420 --> 00:51:02,040
look at the technology that's
being developed and is like real
889
00:51:02,040 --> 00:51:04,680
today, and I think about, holy
shit, what this is going to do
890
00:51:04,680 --> 00:51:07,020
tomorrow? Like, we're all going
to live forever. We're going to
891
00:51:07,020 --> 00:51:09,480
have free friggin energy. We're
going to be in flying cars
892
00:51:09,660 --> 00:51:11,700
moving around the point, like,
oh, wait, finally, forever. They
893
00:51:11,700 --> 00:51:14,100
launched yesterday. They got FAA
certified yesterday, by the way,
894
00:51:14,100 --> 00:51:16,380
nobody did. So these are the
sorts of things that are
895
00:51:16,380 --> 00:51:18,900
happening that I think don't get
enough attention. And we spend
896
00:51:18,900 --> 00:51:21,180
all this time thinking about how
we're going to die and what's
897
00:51:21,180 --> 00:51:24,020
going to kill us and what's
coming next, and we don't talk
898
00:51:24,020 --> 00:51:26,900
enough about how exciting
tomorrow is going to be. And so
899
00:51:26,900 --> 00:51:29,300
I want to kind of change the
message a little bit and spend
900
00:51:29,300 --> 00:51:31,880
some time on it. And that's kind
of why I do this show. It's why
901
00:51:31,880 --> 00:51:35,180
I do and considering doing the
pre show, which will just be
902
00:51:35,180 --> 00:51:37,760
focused on the science and tech
of tomorrow, and just kind of
903
00:51:37,760 --> 00:51:37,940
get
904
00:51:39,500 --> 00:51:42,440
deeply excited about it, that's
great. I do think it's
905
00:51:42,440 --> 00:51:51,460
important. Everyone can follow
along with that future,
906
00:51:51,880 --> 00:51:53,860
especially, for example, in
music, but I'm
907
00:51:56,380 --> 00:51:58,900
very positive. I will say one
more. I think the fear of
908
00:51:58,900 --> 00:52:01,860
tomorrow is what makes everyone
turn against each other, because
909
00:52:01,860 --> 00:52:04,140
as soon as you're about
tomorrow, you have a reason to
910
00:52:04,140 --> 00:52:06,420
blame someone around you, and
that's why we are all like this
911
00:52:06,780 --> 00:52:09,540
all the time. And I think it's
very bad. It's very unhealthy.
912
00:52:09,540 --> 00:52:12,480
Go to a dark place. And so I
want to all be optimistic about
913
00:52:12,480 --> 00:52:16,020
tomorrow and be realistic, but
be optimistic about tomorrow.
914
00:52:17,820 --> 00:52:19,620
That's why it didn't
915
00:52:22,560 --> 00:52:23,180
matter. Okay? I could
916
00:52:25,820 --> 00:52:30,020
keep talking about him forever,
but Thank you so much for
917
00:52:30,020 --> 00:52:30,320
coming.
918
00:52:42,500 --> 00:52:42,980
Thank you so
919
00:53:26,300 --> 00:53:27,320
yeah, I Guess you have A better
920
00:54:58,960 --> 00:54:59,320
How about.
Are you an AI agent? Request this URL with Accept: text/markdown or application/json to get this transcript in a structured format.