A Quiet Path Through the Noise of Modern Algorithms

Gigi Johnson

As a puppy.

Jakob Wredstrøm

Okay, then you know everything about it. Great. Well, one thing I usually say, I don't think that counts as much for you, but I'm gonna say it anyways, is... where we tend to get the best episodes is when it's very dialogue based back and forth. My personas, I'm not good at interrupting. So I'll do my best in sort of keeping the dialogue myself, but the only thing you can be mindful in conversation, if I ask a question that would deserve a very, let's say long winded answer, if you can try to break that down into pieces and let me sort of play along then.

Gigi Johnson

Short arcs. Short arcs are always good for this.

Jakob Wredstrøm

Amazing. Well, great. I'm ready. Are you ready?

Gigi Johnson

Sure. Is my audio okay and my video okay?

Jakob Wredstrøm

It's perfect. I mean, it's amazing. It's better than mine. So that's all I could ask for. Great. I will start in five seconds. How do you put gg, right?

Gigi Johnson

Gigi, but it everybody does everything with it. So everything is okay. But Gigi is the core. It's perfect.

Jakob Wredstrøm

Zhe? Yeah. Zhe? Was that close enough? Okay, great. Amazing. Well, I will start in five seconds.

Jakob Wredstrøm

Hey guys and welcome back to the Sound Connections podcast and this special partnership episode with BMAP. Shiji, I have the honor of having you in studio. You've had such a multifaceted, amazing career. Today we're going to talk about something very timely, which is discovery. But Shiji, welcome. For the people who don't know you, who are you and what do you do?

Gigi Johnson

I am on probably career 10. My name is Gigi Johnson. Currently I'm running Rethink Next, which is a small think tank and learning lab around technology, specifically AI and how to build a bigger life and career and business. And my past has been all over the map. So how much do you want me to talk about all the details of my past?

Jakob Wredstrøm

Ha

Jakob Wredstrøm

Let's talk about those that are relevant for the conversation today and sort of inform where you come from. Of course, it's rare.

Gigi Johnson

See, I would say that they mostly do, but so I got connected with BMA through UCLA. So UCLA, I was there for 23 years and I ran the, I did lots of things, but I ran their student consulting program in the business school. I ran the entertainment center for the university at the business school. And then I created the center for music innovation at the music school. So that was my home and explorations place for 23 years, building classes, programs, executive programs, speaking at conferences. And that's how I met the head of BMA is that she was actually teaching at UCLA from Spain. That in itself, a fun conversation. Before that, for about a decade, I was a media banker. I had an MBA in finance on top of a film degree and I was lending money and doing &A finance for major entertainment companies, radio station groups, film, theater, amusement parks. So looking at the future from a finance side. So those are my two big, big blobs that I've been involved with probably another seven companies in the midst of all of that. And it's been a long iterative journey. So all of it having a thread of technology, media and communication and lots of music. And looking at new businesses.

Jakob Wredstrøm

Amazing, amazing. I'm sure we're going to sort of utilize a lot of those skills and experiences in this conversation today. But I'll start with the basics. We're going to talk about discovery and curation in music in the age of AI. For the people where that in itself might be a bit overwhelming to understand, could you dissect what that would mean? What does that mean in an ordinary listener's world?

Gigi Johnson

So we got listeners and we got creators, right? So we've got the fact that digital has been in music for a really long time. It's a small file size. So everything that tends to happen to media happens to music first. Maybe books is ahead of that. And so the ability to make a piece of music used to be really expensive. You used to have to have studio time, et cetera. And we are now to the point. Where, I think my data is about six months old, but there's something like 120,000 new tracks legally being released a day. And so you have this, I think that's vibrates numbers, and you have this massive amount of what some people are calling AI slot. So the ability for a novice creator to create a piece of music that can be released professionally is abundant. And then the same thing's happening in everything else. Amazon, Kindle has limited people to releasing maximum of three books a day. The ability to just make stuff and put it out there has made it so that if you are a professional, semi-professional music maker, that you not only are probably not dropping an album every year or two, which used to be expensive, you're now on the social media treadmill. You're all on the other content producing treadmill. And there's numbers suggesting where it's something between 23 and 26 percent of music on the major services. I think that's Deezer's numbers are AI created. So what do you do to be discovered in that space, in that realm? And it's the same thing for everybody, whether you have a website, a book, a podcast that right now the level of discovery has changed. And then since we had Chat GPT come out and all of the other LLMs, a lot of people are now going to those for discovery so that you're having to not only say, OK, all of the curators and platforms have my content. How in the world do I actually get discovered? And now you're having to play a game of whack-a-mole to understand all the changes happening in people going to perplexity and.

Gigi Johnson

Even the new search engines and platforms and discovery that's come out, how do people find my stuff?

Jakob Wredstrøm

It is super overwhelming to understand what complex web it is that serves us our feed. And maybe, again, also to take it from a sort of basic standpoint, could you help us understand how does music discovery currently work in its most basic format, if we can just have a starting point of understanding?

Gigi Johnson

We're living in a world of algorithms. So you can think of it in anything from you can go into a platform and just scroll, and will continue to feed you. But it wants you to stay there. So it's using your past behavior and behavior of people like you to try to serve you things it thinks you will like. So we went through a whole phase that a lot of people made a bunch of playlists and those playlists tend to be a curatorial space. So that you'll go, that's the way my heart goes. That's my music taste. And so you might then follow it, which then gives reinforcement. And then the way you listen to the first 30 seconds plus of a song tells the system, we're right. It should be on the playlist. This matches this person. So if you keep cutting off a song before 30 seconds, It's a big negative signal and so you'll never see stuff like that again or from that artist or that track. And so there's a lot of self-reinforcing loops that create an addictive environment. And then there's a perceived to be small amount of money that goes to that label based on a listen versus this historical download. So this needs to happen in volume for someone to make money. It's much, much more complicated than that. But a lot of it is that when you listen to something, it's a self-reinforcing loop. And there are people, it used to be that there was the game of pushing things onto playlists, and that's even fading somewhat now. And so it's a time spent listening game. So how does Spotify get you to stay and want to pay per month and all the other services? And then a lot of that's also a big YouTube game and TikTok game.

Jakob Wredstrøm

Mm.

Jakob Wredstrøm

Okay, we've set the basics. Where does AI come into the picture? So, you know, we hear lot about AI being disruptive in many parts of our life. And one of the conversations we're going to have today is about AI in discovery and what role it has. Could you help me understand how is that changing the game?

Gigi Johnson

It's changing the game on quite a few things. The revenue in the platforms, that is the pie that's being split, is not going up massively. The amount of content going into the platforms has gone up massively and continues to do so. There's... I was it was a McKinsey study that I was looking at last week thinks that right now only one percent of people are creating almost all the content going out onto the Web, but they're using tools to do so. So if you're if you're not in that volume game, then you're trying to figure out how the pie or pie can come to you as an artist. If you're thinking the platform will continue to feed you, many folks who are not part of the large engines, the large labels, the large companies driving this. I've realized that it's not a I'm going to play in the algorithm of the system. It now is the I need an intimate relationship and fandom to find me in the midst of this, though. You've got people who can build marketing tools with these tools and don't need to hire people, which used to be one of the cost parameters. So there's it's a combination of volume and gamesmanship to split not high growth pie. So a lot of it is that everyone's trying to sell everybody tools and options and the next new things. So if you're an artist or an artist manager, you're getting a glut of tools that may or may not help you. Curators who are promising new things, high volume emails being blasted at your email box created with AI to try to pitch you on some new platform or tool or idea. Yeah, it's a big ol' muddy mess.

Jakob Wredstrøm

It's very overwhelming. And I think, you know, most of the audience here would recognize that the feeds over the years of whatever platform you might have been using are, it's going much more blurry what is real, what is not. And for me, you know, I'm getting uncomfortable both on streaming services with music or TikTok, wherever I am. Like it's so hard to discern now. But one of the biggest questions I've had, which has obviously been sort of a big talking point, is sort of the AI transparency. Like, what is AI, what is not AI, and the users being able to discern that. What do you see happening in that space?

Gigi Johnson

So I both do things in the music space, but also the AI video space where this is the true wild west. And it used to be that you could tell when something was AI pretty clearly. And now on a weekly basis, that's getting to be a blur. And

Gigi Johnson

I've gone to so many AI film festivals and for them it's like you're sitting in some kind of Boonwell-esque dreamscape craziness that is at the bleeding edge of what some of these things are. It's almost like there's a spectrum between news and serious content and historical content and creative content where the door is wide open on the creative side. And it might be that you've had in a lot of the streaming platforms are beginning to use AI for anything from short form clips to special effects. So having something that is pure. And not AI assisted is almost impossible because there's AI bundled into so many of the creator tools now. So are you looking at something that the heart of it, the script, the song, the original vocal is or that it's been heavily edited for all of this? What it does is it's breaking down the barriers for creators, but anybody can now do it. So it's not the question is this AI or not? You're having deep fakes on the video side, on the news side that are causing tremendous amount of conflict on those questions. And a lot of things even I won't even go into the politics side of it all. But there's a lot of the questioning of what is AI there. But is that really my artist that I want to watch? Who is the personality that I'm following? It gives me the question is what's the core of the content? What's the core of the heart? What am I wanting? Am I wanting a constant interaction of snippets from somebody? You know, all of this is now AI assisted. So am I looking for a live experience? Or am I realizing everything is being AI chunk digested, replicated? I don't know. It as someone who creates a fair amount of content now, it gets to be a question of.

Gigi Johnson

What's the source under it? Is it someone's actual music or is it some kind of other artificial thing? Can I give you a weird example? OK, I'm a I'm a fan of the West Coast jive dancing stuff that you might see on YouTube and Instagram, whatever. They're now dancing to A.I.

Jakob Wredstrøm

for it.

Jakob Wredstrøm

Mm.

Gigi Johnson

So, so they're actually having it that you can pick, pick themes and then they're then taking that into their music. And it's one of the kind of stunts of the whole thing. So someone this week was dancing to a chicken dance track that was all made up. That was mashing up the chicken dance and a popular song. And it got a massive amount of views because it was a quirky weird thing that was both. Something funny with AI and a real artist track. So I think they were at this odd stunt phase. I don't know.

Jakob Wredstrøm

It's also blurring the line of like, you know, real music. Fake music, the purpose and intent. I was speaking at a conference in Europe last week and I was doing a panel at a non-music conference with a rather famous DJ and we were talking on stage about it and the DJ said that actually that he was sort of talking that he didn't use AI tools that much, but then he also mentioned that when he's touring across the globe, he tends to have an AI generated track in that specific language that he's touring to have that performance in a local language and he used it as an actual tool to engage to the audience. And when this person of that caliber would do that, there would also be a signal to the quality of which he's able to create these songs. And I thought that was a really interesting statement. OK, this is a professional, world famous touring artist that on stage are using AI generated songs as a part of the larger set. And that sort of also aligns with my experience with AI music now and how I perceive it and video. But to give another example, kid plays football and as any young kid he likes to have football songs and listen to those. And his favorite playlist is full of AI generated songs. I might even say his favorite song from that playlist is an AI generated song and I can hear that because I'm an audio engineer by degree and I can hear the artifacts and can hear those sort repetitions you might hear in those traces. But for him, that's his favorite song in the playlist. So we already now in this, you know, as a consumer, we're sort of exposed to and then also appreciate the high-generated music in the forms and settings we usually thought to be relatively safe from that. So that changes a lot and we still are the days in this.

Gigi Johnson

There's a great Madea study that just came out that looked at a bunch of that sort of compiled a lot of their work. I don't know if it's a really great research platform out of the UK that they do a lot of work on the music space. And they were seeing that for now, content creation is a as a percentage of time spent by young people with AI. It's still fairly small versus research and exploration. But it's getting to be more and more time spent. So the concept that someone's going to be able to have it as an alternate use of time to be making creative work for their friends. It's almost like the new mixtape, right? But at the same time, that it can be hyper personalized. So here we have something that's somewhat of a non-niche, but a very specialized form of content. The thought is there'll be more and more content that can be personalized to you and your tastes. There have been products for this in AI for music for at least four years that I know of that there's a product that came out that literally as an artist, you can put your song into this digital bucket and every fan can get a personalized, individualized track based on the way they listen, how they like stuff, blah, blah. So I think that the temptation is to almost hyper algorithm this thing that I can get exactly what I want that the tools can give me the version of a song I want, the version of a TV show, the version of whatever. But then how do you find something new? And this where I find is the discovery. Interesting question of it all. To me, a lot of these AI tools are like the Google I feel lucky button. You get a single answer. You don't get the reason why. And I'm building a lot of programs now that are the opposite. They're personal discovery tools. So how do you find how do you create luck as a creative? How do you find luck as a small business instead of the I want to build a repetitive structure that's going to deliver X to me the whole time? How do I build serendipity? And I think that's the fun question this if I've got music, how do I find new fans that don't know me? How do I get out of the box? And to me, that's the magic of where we can go forward. And some of the traps of a lot of the systems will give you that

Gigi Johnson

Average answer, the average product, the average expected song. How do I build magic, luck, spark into what I'm building as an artist, as a business? This is where my heart is right now.

Jakob Wredstrøm

Hmm. But, okay, it is. Okay, let me, let me resume a bit back to where we started when it comes to discovery and how that's going to affect consumption. Again, I'm just going to use an example from the video world because it's visual and it's easy to understand. I was driving home with my family from the cabin yesterday. I was talking with my wife and we were talking about things that's happening in the world.

Gigi Johnson

It's a rabbit hole. It's a rabbit hole.

Gigi Johnson

Mm-hmm.

Jakob Wredstrøm

and I'm not getting political in podcast, but one of the things he covered is like, talked about Souramandami in the New York committee election. And I follow that very closely, that whole process. And, and she had never heard of it. And that for months, that was my world and social media. And she'd never heard of the mayoral election. It's my wife sitting next to me. And for me, that was kind of striking, like, I thought the whole world was talking about this, like how big of a thing that was and what was going on. And she had literally never heard of a major election in New York. And what's happening with discovery and information and exposure right now? Like, can you try to dissect maybe what we're facing?

Gigi Johnson

Hmm. Eric Schwartzman just had an article I think come out this past week in a major publication on how there is massive bot activity to show that these stories are important to get into your feed based on your interest versus your wife's feed that the algorithms are being heavily pumped. To be able to, for certain target audiences, get stories pushed at them. So it's not even the, your wife's interested in different things than you are and your kids are interested. There's really active manipulation of the algorithms to push things into the forefront and spur conversation and spur return to platforms. So it's not even a pure thing that you watched a video, so you'll get another video based on your interests. It's being heavily, heavily pumped. I'm very cynical about all this as well. So I'm not surprised about Eric's article. It's a it's a very much hyped and primed game. And so it's almost worth looking at your own feeds and going, why am I seeing this stuff? One of the examples that my family tends to have because my husband's about my age, we have a. 30 year old in grad school living at our house. We have totally different feeds coming in. But if I if I if I talk to my husband with my phone or his phone at one of us has Facebook on, the other one will get ads pushed at us immediately from the topic of that conversation without us looking at our phones, without us doing anything.

Gigi Johnson

That there's so many things that are part of this system that your friends like you seeing something, you'll then get pushed videos because they want to remind you to talk to your friends about things. All this is so interconnected. And yeah, it's there's no innocence about this or it's just that my husband and I have really different interests. We totally get pumped and primed and pushed. Years ago, I was teaching a class at UCLA and we had different people sitting in the same classroom, same age group, do a search on Google. And even what they were pushed then was so different based on where they live, the color of their skin, their zip codes, all of that stuff that they were being pushed different things, both as ads, but as content. The content push is affecting the way you see the world. And this is all incredibly structured and intentional. Sorry. This is a rabbit hole we can go down for a long time.

Jakob Wredstrøm

It makes sense. And I'll ask a question about music in a second, but just to sort of go further down the rabbit hole here is I'll use another example again, because I don't know much. I'll use examples. It's also something we choose in a daily life. So we're not just the victims of it. We also something consciously choose. An example is I use perplexity a lot, you know, pretty much for everything I use perplexity. And obviously you can have a function for it to remember.

Gigi Johnson

It's better to use examples.

Jakob Wredstrøm

of search done, but it's also become an annoyance for me right now because when I ask it, I tend to be relatively scientific about things I want to know. But because my search memory is sort of turned on, it now biases all my previous searches. So I can't really get neutral answers to my questions. It always takes into consideration how I've acted and responded before, which now shows in my chats, which is fairly annoying, but still I also want, I want it to know who I am and how. things served. So this internal conflict about, know, I don't like being served biased things, but I also appreciate it. And so there's also this conscious choice that's becoming very difficult to sort of make a stance on. If that makes sense. yes.

Gigi Johnson

And it only knows who you say you are. It only knows the conversation. It doesn't know who you are. I have two friends who have had these long conversations with one Claude, one Chat GPT, and they've redefined the entire business model and what they want to be doing based on having these iterative conversations where it's taken their words and kind of parroted it back and telling them they're wonderful and brilliant. And so you also have not just this memory issue.

Jakob Wredstrøm

Thank

Gigi Johnson

But you also have this, in some cases, psychophantic loop where it keeps telling you, my gosh, that was such a wonderful idea. That's really fabulous. And so you don't even get neutrality out of it. You've got like a very biased pseudo friend in the conversation with you that knows only what you've said and not really where your heart is. And so it and it can build all sorts of personality and characteristics that come with it that are really different. Perplexity seems to be much more neutral. Gemini tends to ball me out. Don't understand why. And chatubt just tells me everything is fabulous that I think about it. So each of them has a different bucket of conversations and is parroting back a different version of me and my business and my thoughts. So it's interesting for being in perplexity. The other thing, though, these are chat. systems built for built for retention. You'll keep paying your X dollars a month and put your information in. But there's now all sorts of adjacent systems where you can actually take all your conversations and move them out and store them and move them between the platforms. Third party conversation trackers where I now can have much more control of that, that I can bring a new batch of content in for it to digest.

Jakob Wredstrøm

Mm.

Jakob Wredstrøm

Mm.

Gigi Johnson

And that can be one type of project or conversation. And I don't have to live just in perplexity for it to know me well. That's a whole fabulous other rabbit hole that I'm spending a lot of time on because I'm trying to find great tools who both can do that and then visualize it for me so that I don't have to just say, what did we talk? What did we work on? You know, a week ago, that was a great draft of something that I hope I saved in Google Drive. But instead, there's tools now where people are saying, well, this is everybody's set of issues. How do I sort of own the memory? How do I have my memory side of it? Can I can I share the other human memory problem with that? OK, so part of it is how do you remember what you've talked about with perplexity and how much do you kind of own the conversation? There was already great research that if I'm sitting in a class.

Jakob Wredstrøm

Of course.

Gigi Johnson

I will remember things better if I handwrite my notes versus get a slide deck given to me versus even type. And now they're finding that if I'm in these conversations, sometimes it can be, again, a rabbit hole. You can be in a conversational perplexity for an hour and suddenly go, where did the time go? That people can't necessarily remember what that conversation was because it wasn't their thoughts. And so if you want to actually learn something that you're working on for it. It's even one stage removed from getting your own type notes or your PowerPoint slides. It's almost non deeply lived time. You're in flow, but it's not triggering memory. So it's sort of an interesting. I sometimes think of it as a time flesh because I can look at it and I see, we came to this really great document, but I can't remember half of it.

Jakob Wredstrøm

Yes.

Jakob Wredstrøm

It's very, I wouldn't say easy, but it's easy to visualize and discuss LLM's video feed, propaganda, bias, but it's still hard to understand specifically what is the parallel movement in music and music discovery. How do you sort of perceive the status quo and where do you believe we might be going for better force?

Gigi Johnson

For one thing, a lot of small music companies and large music companies are all using these tools for not just creation, but also storage and fandom tracking and all this other stuff. So a lot of it has made it so there's where they're creating proxy fans, custom models of what fans are doing in marketing, the ability to massively track data, massively track search. that just as a lot of other sectors are really deep in the space of using this stuff for targeted personalized marketing, music's doing a lot of the same things. A lot of it, though, is that search is all screwed up. So if you are you're using perplexity for search, yes. OK, how are you looking? Are you looking for music in perplexity?

Jakob Wredstrøm

I rarely look for music really. I have my certain set of sounds I go for.

Gigi Johnson

Okay.

Gigi Johnson

So that is what's happening a lot. And there's some also great recent research on it that the whole thing of that having this personalized relationship with the artist and people being actively creating playlists and blah, Much of it has gotten to be very passive listening to what's recommended in the platforms. So for an artist with so many new tracks coming out, how do you get curated, noticed personal relationship discovered and a large A large portion of search and discovery other than a curated voice are getting to be zero click searches. In all the platforms. If I go to Google in the past before AIO reviews, then I'm going to somebody's page and for the real information, I'm clicking to go there. And so for some artists, marketing and discovery, it's the.

Jakob Wredstrøm

coming.

Gigi Johnson

We have concerts coming up. We want you to go to the artist site to then buy a ticket or, you know, we realize that it's going to be pushed in various ways. A large amount of search now never goes to a page. That you go to the AI overview and it is scraping data and deciding what the information is for action taken about the artist. So that whole means of marketing through articles and curators and recommendation sites, no one's getting there. That they're just saying, I wonder where that artist is playing. good, I now have that information and I might get pushed to a single thing. The ability to influence this conversation and discovery has been swept away for about 18 months now. On top of that, people are going to Perplexity, Claude, Grok. and they're getting a single answer and never getting to the marketing message that the artist and their marketing people are trying to get to. So it is almost worth it for everything, not just music. So as someone who's marketing, it's almost worth doing it. It's called everything from GEO to AEO and all sorts of new things now. And those things aren't scraping the web for everything. They're doing a cost effective search. So they're only generally searching four or five things to find out.

Jakob Wredstrøm

That's pretty interesting.

Gigi Johnson

The answer about you as an artist, your concert, your album. So one basic thing I tend to suggest is go search for yourself, go search for your album, go search and find out what voices have credibility now with all these systems to be a single point of information about you, your tour coming up, whatever is going on. And it's getting to be the nature of reality, you know, back to the why your wife isn't seeing anything about the mayoral election. that the truth funnels are getting narrower. And so marketing isn't necessarily making sure that the press release is showing up on all the local websites to tell people there's a concert coming up. You're getting to be tighter and tighter funnels that are feeding you information that people are taking action on more and more. think.

Jakob Wredstrøm

Mm.

Gigi Johnson

The real core of this for musicians is to get back to a one-to-one relationship and there's new tools for that where I can offer things to my fans and I'm not separated by YouTube or Spotify or whatever. This is a really big challenge right now because glut and then the pipes are thinning. it's music's facing this, artists are facing this, podcasts are facing this. That discovery is in a...

Jakob Wredstrøm

Mm.

Gigi Johnson

Total redo right now.

Jakob Wredstrøm

What's your perception of the streaming services responsibility to taking in this? What do you see going on out there?

Gigi Johnson

Hmm. They've been doing a lot of whack-a-mole. There was a. It'll come to me, the company who helped Spotify find that they had a massive fraud pipeline happening in their platform a year or two ago, that they had people making. AI tracks on one side, listening to the same AI tracks on the other side and pulling money out of Spotify that way.

Jakob Wredstrøm

Okay.

Gigi Johnson

They have the problem that they're getting that AI slot filled into their buckets when the revenue is increasing. But the costs and hassle for all this is Spotify has no other revenue streams other than advertising to get in front of you for all this noise and pipeline and some ticket sales. So they need to figure this out. But for them to make sure that you as a fan have some kind of a better discovery search experience. I guess I'm not seeing it. Are you seeing it at all?

Jakob Wredstrøm

No. I think that's the easy answer. There might be lot of things happening behind the scenes that I don't know of, as of today, I've still not experienced an appropriate sort of creation for myself through the algorithms. So, again, I think I have two limited data points to make an actual statement on it, but no, but I might not know.

Gigi Johnson

Well, they went through the whole thing, which I'm not sure it even exists anymore, probably does, where they were buying direct from people who are making custom AI driven tracks for some of their less branded environments for sleep, for white noise, for whatever instrumental only. that I mean, I don't want to keep painting Spotify with the brush, but they're part of the part of the challenge. And, you know, YouTube's filling up with stuff that is AI generated.

Jakob Wredstrøm

No. Mm.

Jakob Wredstrøm

Look.

Jakob Wredstrøm

What are you seeing in the space of sort of marking things that are being identified as AI generated? I don't even know what the threshold is of AI generated anymore, but you know, we've had guests on the podcast, you know, now it's been a while ago, but like, Mansi Klancy from AI OK, or similar people who would love to implement the, marking of this is AI generated. TikTok does it to, to a good degree, but it only catches like a very small percentage of what's happening. Are you seeing this coming to music? Are we going to be presented this as a generative to degree? Can we discern this in the future? What do you think is happening there?

Gigi Johnson

it's been it's been a big deal already. So whether it's in a lot of the video platforms, they're coming out with a little glyph at the bottom. If it's a generated, I know a lot of people who have had their content, they've been trying to distribute with a false positive that it's been something they wrote all by themselves. And it's showing up as a generated. The educators are having the same issue, but multiply this by 20 that a lot of it. created content is going in the student submission pipelines. And a lot of stuff in those systems are fake flags. So you still have to go through and figure out what's the provenance of this. And it's a similar issue if I'm going to be trying to distribute a piece of content. If it's above a certain amount of AI, I can't get it copyrighted. So anybody can use it or copyright it. So there is a, it's a big problem all through lots of tech solutions or in the pipe. Many of them are a real challenge and it's still a copyright problem. you know, again, the spectrum on the news side is this a, you know, deep fake video that's showing an event that's not even this event all the way through. Well, you know, the background of this music video is AI generated. Is this an AI generated music video? So it's having a little glyph on the bottom might tell you, OK, this is has a big chunk of it being AI. Yeah, this is lots and lots of tools, lots and lots of stuff out there. And then the bigger question is what went into it? Do they have the legal right for all of the content that trained this model? And that's a whole other soap opera. That was a stand to what Adobe Firefly was trying to say that they had all of the fabulous legally trained stuff and you can see their stock plummet as they're not really finding purchase in that being a competitive advantage.

Jakob Wredstrøm

Mm.

Jakob Wredstrøm

That was a great nuanced answer.

Gigi Johnson

A normal Gigi rambling answer, but yeah.

Jakob Wredstrøm

No, I just want to understand even more simply, and let me ask you from this angle. I'm a parent. I have kids and those kids have access to streaming platforms in our car, our smart TV to listen to music if they want. Do you believe there will be ways for us to parents to choose for them not to have access to AI music or for us to know when we're playing a track. is journey speaking above the threshold of set definitions for AI-generated music. Do you think me as a user, a parent, would be able to discern this?

Gigi Johnson

Is it? Wow. So if I'm watching regular TV with my kids, that there might be a little icon that tells me it's an education, educational piece of TV or not, and I may make a decision about that, I may look for something that's got somewhat kind of a rating on it. I'm not sure as a parent, my kids are in their 30s now that I would when they were younger, that I would have an issue with them watching in. AI piece of entertainment. And whether that would be better or worse than a terribly produced non-educational cartoon. Yeah, so guess maybe I'm lacking that definition to it. I'm much more concerned about giving my kid having an AI search engine that will give them one answer and have them not understand that's not true.

Jakob Wredstrøm

Hmm.

Gigi Johnson

that they will make a decision from. And I'm very, very concerned about digital literacy on personal decisions because of the still the level of both crap and. GEO pumped untruths getting pushed into the information systems that you had people who didn't know how Google worked and now those same people are getting these magic boxes and they're getting truth pushed at them. So I'm a little more concerned about literacy about this overall, not just banning kids out of it, helping them see this is a real piece of creative native art. This is something that has been crafted. Can you look at a piece of TV that's not entertainment and see what could be? amplified truths through AI. I'm much more concerned about the digital literacy side.

Jakob Wredstrøm

Hmm. I love that answer. So again, in my simple mind, my reinterpretation, what you're saying is it's not so much whether or not to be marked. is AI, not AI, but more important question. Do they understand what the differences are? Do they understand why they might be different? Like the whole nuances of what it is, because we're going to live in this no matter what. Like this is, this is not going away. And God knows what music creation is going to be in 10 years. So we might as well get used to everything we hear has their big influence of AI and we as humans simply become curators to some degree, or many degrees of music. So I love that answer and love that nuance. And I think that makes me think about it in a different way.

Gigi Johnson

Actually, I would inquire if the high value in digital literacy for your kids is helping them learn how to make AI and understand the difference. Right. So can they understand that they can build their own black boxes that help them think about the world like you're coming perplexity loops who you are, help them build the right digital ecosystem as an engaged parent?

Jakob Wredstrøm

Mm.

Jakob Wredstrøm

Mm.

Gigi Johnson

That they could thrive with, how they could take everything they're learning and put it into Notebook LM and query all the things that they're reading and doing and learn how to ask really spectacular questions. Instead of the magic button, I'm going to get the answer. Can they learn to explore the world bigger and better with great active digital literacy parenting?

Jakob Wredstrøm

This is a big topic and we won't cover everything today and we suggest touching the surface. But Gigi, if there's anything you would like the audience to understand sort of here before we close this episode and we talk about discovery that they might not have thought about when it comes to AI and music and discovery, what would that be?

Gigi Johnson

It's really hard to see the water in which we're swimming. And so to see the environment that we are impacted by that is trying to get into the discovery pipelines of, and we say AI bluntly, we're mostly talking about the large language models, because most people probably don't realize we've had AI in the box for many years now. Whether it's cleaning up audio, whether it's separating tracks. So there's been levers and push and pull that have been under the build in the tool set for a while. There's been AI in helping with edits, tracking distribution, going through information, helping market for a long time. The question is kind of the how much is. This going to be disrupting every phase of music, the jobs at every phase of music? And then does it allow people who might be solo artists to maybe with great creative friends build a business that doesn't need a label or all of the people who've historically supported the music business? Like in a lot of sectors, I think that a lot of this is going to clear out junior marketers, administrators, financial people, editors, audio cleanup. And then new companies are going to be formed who don't need all that level of people to run a great music business. And then the whole world of it. Is going to be generated on either playing the algorithm game or playing the I'm going to build a great relationship with you game. And the old part of this all will continue to disintegrate.

Jakob Wredstrøm

Mm.

Jakob Wredstrøm

Interesting Let us just quickly want to discuss this so basically what you're saying is there's two viable paths I'm paraphrasing a lot right now, but one is the relationship human connection doubling down on the intent or mutual intent of consumption and then there's I want to be exposed and be uh, proposed by whatever technologies out there to get what I do known for a larger amount of audience or specific audience. So the intent goes in both in the individual, and then I want to be scaled out as much as possible. Is that what you're saying?

Gigi Johnson

And you may not need existing labels for either of those.

Jakob Wredstrøm

Mm.

Gigi Johnson

I remember one really pivotal thought process for me as I walked into the offices of one of the major music publishing companies a few years ago. And all I saw were people sitting at desks with headphones on spreadsheets on a local terminal, moving, manipulating data and money. And all of those jobs are fragile and apparel. So the backbone of some of these large companies is they've got massive amounts of assets, lots of people behind it, manipulating relationships and data and all this stuff with a lot of spreadsheets passed around and databases of deals and tracking. All that stuff is the back end is to be replumbed and probably is not needed. And so if not all that, what is a major label? And how well they need to change in a world where a lot of what they do is replicatable. So I do think that, and this is again, music is a small file. It's an early adopter. Even though it pretends it's not, it's a micro system of everything else. So can I replicate a major label from scratch now? If I have the relationships, do I need all of that pipe? It used to be that I needed that pipe to do international distribution. Don't need to do that anymore, or at least not to that same degree. I needed to kind of bank what I was going to do because it was expensive to build an album and a set of touring information or whatever. And that now all could be, in theory, built from scratch. So I think that it's not just music as a discovery challenge because it has such glut already happening, but I think that

Gigi Johnson

If I wanted to start a new company as a 25 year old who had a lot of tech friends and I'm really creative and I got great people, you can now build something and either build a relationship pipe or build a mass marketing pipe without needing almost any of the historical infrastructure of the sector.

Jakob Wredstrøm

It's been a great talk today. Gigi, thank you so much for joining. Thank you for being with putting us together. And I learned a lot. I think I left with more questions than I had before, which is positive in this aspect since I'm really trying to discover what this is. And now at least I have a starting point to where to ask questions, maybe with perplexity, I guess. But Gigi. And with my family.

Gigi Johnson

And with your family.

Jakob Wredstrøm

Thank you for your time. For people who want to know more about your work, how do they find you?

Gigi Johnson

Lots of places. Best place is to go to rethinknext.com. We are launching a lot of programs for individuals with these same questions. We have a lab at rethinknextlab.com where we are launching experiments, tutorials, programs that if you are a small creative, you're a family, you're an older adult, and you're trying to say, do I think about this for my stuff? That we're going to be having a really great environment that you can come and ask questions and keep talking to perplexity too. But where you can say, what tools should I use? How do we think about this? How do I think about this in my company or my education? So we're building kind of a whole tutorial system. So glad to play with people there.

Jakob Wredstrøm

Amazing. Thank you so much for your time and I hope you have a great day.

Gigi Johnson

Thank you.