Technology has always played a key role in the world of music, with things like digital production software, loop pedals, and multi-track recording transforming the industry forever. Artificial intelligence (AI), however, is about to be the most disruptive technology in music.
In today’s episode, Ben Taylor sits down with Alex Mitchell, Founder, and CEO of Boomy, to discuss the future of AI in music.
Boomy is an industry leader, using artificial intelligence to create "instant music.” Alex shares how he has tried to solve the question of what good music is with data and why he believes that, ultimately, good music is whatever people want to listen to. The music industry is evolving as we move into a time of growth, as evidenced by the explosion of content on TikTok and the billions of streams that have resulted from it, and why he believes that the least efficient way to invest in musicians today would be to start a traditional record label.
Boomy presents the opportunity for a new consumption dynamic, using AI to look forward, not backward, and develop a native format for the next generation of listeners. On the topic of AI, Alex reflects on whether this emerging technology helps or hinders artists, saying that Boomy is not designed to replace musicians but rather provide the necessary tools to those who don’t have the same resources and access that many musicians do. In the end, great songs aren’t created by optimization algorithms, but by people with the tools to express themselves!
Taking a minute to look toward the future of technology in music, Alex speaks to the likelihood of using real-time EEG neurofeedback or brain-computer interfaces (BCI) to accelerate trends, saying that, while there are companies that are already working on this technology, we probably won’t see it realized in our lifetime. When asked about the future impact of AI on music, Alex explains how Boomy facilities personalization or what he calls ‘context-aware algorithmic music’, which he believes is currently hindered by the difficulty to monetize it. What is missing, in his opinion, is a way to incentivize the business models of those systems, which is one of Boomy’s core goals.
Reflecting on the two components of what musicians do, the first being skill and the second bringing taste, Alex explains how Boomy is building an interface where technology fills in the skill gap and allows everyone to become a composer in their own right. As the technology evolves, there will be greater opportunities to create and consume hyper-personalized content more rapidly and, as our appetite for short-form video content grows, Alex also thinks that the way the music industry operates will continue to evolve too.
Regardless of what we think is ‘good' music or ‘bad’ music, the only data we have to go on is consumption and Alex believes that it’s what we each bring to music that makes it good, not some set of universal criteria. As a musician himself, Alex isn’t on a mission to replace musicians; he is leveraging the latest technology to create a whole new generation of creators and explore a new consumption dynamic that blurs the line between performer and audience.
Key Points From This Episode:
Technology has always played a key role in the world of music, with things like digital production software, loop pedals, and multi-track recording transforming the industry forever. Artificial intelligence (AI), however, is about to be the most disruptive technology in music.
In today’s episode, Ben Taylor sits down with Alex Mitchell, Founder, and CEO of Boomy, to discuss the future of AI in music.
Boomy is an industry leader, using artificial intelligence to create "instant music.” Alex shares how he has tried to solve the question of what good music is with data and why he believes that, ultimately, good music is whatever people want to listen to. The music industry is evolving as we move into a time of growth, as evidenced by the explosion of content on TikTok and the billions of streams that have resulted from it, and why he believes that the least efficient way to invest in musicians today would be to start a traditional record label.
Boomy presents the opportunity for a new consumption dynamic, using AI to look forward, not backward, and develop a native format for the next generation of listeners. On the topic of AI, Alex reflects on whether this emerging technology helps or hinders artists, saying that Boomy is not designed to replace musicians but rather provide the necessary tools to those who don’t have the same resources and access that many musicians do. In the end, great songs aren’t created by optimization algorithms, but by people with the tools to express themselves!
Taking a minute to look toward the future of technology in music, Alex speaks to the likelihood of using real-time EEG neurofeedback or brain-computer interfaces (BCI) to accelerate trends, saying that, while there are companies that are already working on this technology, we probably won’t see it realized in our lifetime. When asked about the future impact of AI on music, Alex explains how Boomy facilities personalization or what he calls ‘context-aware algorithmic music’, which he believes is currently hindered by the difficulty to monetize it. What is missing, in his opinion, is a way to incentivize the business models of those systems, which is one of Boomy’s core goals.
Reflecting on the two components of what musicians do, the first being skill and the second bringing taste, Alex explains how Boomy is building an interface where technology fills in the skill gap and allows everyone to become a composer in their own right. As the technology evolves, there will be greater opportunities to create and consume hyper-personalized content more rapidly and, as our appetite for short-form video content grows, Alex also thinks that the way the music industry operates will continue to evolve too.
Regardless of what we think is ‘good' music or ‘bad’ music, the only data we have to go on is consumption and Alex believes that it’s what we each bring to music that makes it good, not some set of universal criteria. As a musician himself, Alex isn’t on a mission to replace musicians; he is leveraging the latest technology to create a whole new generation of creators and explore a new consumption dynamic that blurs the line between performer and audience.
Key Points From This Episode:
EPISODE 21
“AM: If I thought this was a threat to musicians, I wouldn't be doing it. That's why I jumped in. This has to be done ethically, it has to be done with good economics, right? You have to actually have a business model behind the thing to support those ethics, frankly. We're a couple of years into this experiment and there's so much that's gone right, and there's so much that we've learned, that I see those principles guiding us as we go forward. There will always be purists who are going to point fingers and say things, but I've been that purist my whole life. So, it doesn't faze me. I completely understand it.”
[INTRODUCTION]
[00:00:30] ANNOUNCER: Welcome to More Intelligent Tomorrow, a podcast about our emerging AI-driven world. Critical conversations about tomorrow's technology today.
On today's episode, host Ben Taylor is joined by Alex Mitchell, Founder and CEO of Boomy.
[INTERVIEW]
[00:00:53] BT: Alex, I'm thrilled to have you on the podcast today and I'm excited to understand your journey with your startup, and also the future of AI and music and the intersection there.
[00:01:05] AM: Awesome. I'm super excited to be here.
[00:01:07] BT: Alex, I think for every new guest, we always want to understand their life story. It's always fascinating to me, but it's interesting for junior talent, because many of them want to follow you. I think most people know that that's quite difficult, because left and right and two rights and then a left, careers can be kind of a random walk. So, I'd love to understand your backstory.
[00:01:27] AM: Sure, and I've done a lot of random stuff. But I think the thread has always been driven by passion for music, and driven by passion for technology, and those two things. The quick version is I was violin player, so I was an electric violin player and had some YouTube stuff happen when I was like in high school. The idea of putting music out people listening to it, and then like making money from that, as a teenager, was pretty incredible, got me into the Internet, got me into trying to understand, like how many people out there in the world that their worth, right?
That was, I think the main lesson. I would put up these videos playing violin and I would get back hundreds and thousands even of emails from people all over the world talking about the music. So, from there, I went to school for music, graduated, played in probably a dozen or so bands all around New York City and got to play internationally. I got to play at some really cool venues as part of that experience and did sort of the performing thing. As I was playing, and as I was learning about the music industry, mostly through my manager at the time, it was just obvious that there was this whole world, this whole music industry world, behind the performance, behind the music, and that stuff was just really, really interesting to me. This was also during a time where music was in decline. Music was like 10 straight years of year over year of revenue declines. I looked at that, and I said, “I'm going to start a business in this space that makes tons of sense. I can't wait to see how that goes.”
So, I started a company that was an A&R research company. We did basically like similar to political polling, but for whether or not songs were good or not, and that sort of launched me into music tech, that launched me into selling to the music industry and selling to musicians and artists. I sold that company in 2016. And then, during the period afterwards, and after I sort of finished up at the company that bought my last company, it was thinking a lot about AI thinking a lot about automation. I had met some of the early players and some of the early researchers in that space, because they were using our technology to kind of assess how good their music was. When looking for something else to do, it was just kind of a natural, obvious fit. It was okay, this is creative. This is potentially an absolutely massive business if you couldn't figure out certain technical pieces of it. And I called the smartest person I knew in music, Matthew, who's Boomy’s co-founder and we went after it.
That was 2019, and here we are seven million songs later; a huge number, hundreds and thousands of monthly users later, and we're learning so, so much about what the next chapter music looks like. I think there's a lot that we're seeing that people are not aware of.
[00:03:58] BT: So many interesting thoughts listening to you. One of the thoughts was, I thought in the past, how nice would it be if you could invest in an artist. So, when you hear the very first Post Malone song ever, or the first Britney Spears ever, if you'd been able to think, “Wait a second, this is different. I anticipate that this will be a hit.” And this goes into this bigger topic of what is good music? I feel like that is the billion or even the trillion-dollar question. If you can engineer good music, or if you can predict good music. And I'm curious what your thoughts are there. How do you define good music?
[00:04:32] AM: I mean, we could talk about what is good music for like the whole however long we're doing this, because I literally spent four or five years and a whole company's worth of effort trying to solve that question with data, right? Trying to say, “Well, here's an idea. Let's just go ask people. Is this a good song to you? Yes or no? Can we correlate that into hit prediction? Can we try to use data to figure out what the next hit is going to be?” There's a long line of companies. There are companies going after that question today with AI, and I think that my model, there's a lot of models for what good music means.
Ultimately, whether we like it or not, whether I like this or not as a musician, the answer is good music is whatever people listen to, full stop. Good music is what people want to, want to listen to. That’s the principle that in my career over and over again, I've seen to be true. I think historically, we've thought of music as being major and indie, right? And then maybe there's levels of stratification within major levels of stratification within indie. And then this longtail of just I'm picking up a guitar, and I'm a bedroom artist, and nobody ever pays attention to that stuff. When the music industry was small, and especially when it was contracting and not growing. It was pretty much just focused on – and was easy just to focus on – and the money to your point about investment, came from focusing on the absolute best artists in the world, as defined by music industry tastemakers, by historical hitmakers, by that sort of upper levels of the major record label and sort of major publisher major music, industry. It was basically those people deciding what was good and playing it for lots of people. And then those people thought it was good too.
The industry that we're really moving into now is one where it's a time of growth. It's a time of explosion of content and UGC. Look at the billions of streams and videos coming out of TikTok, as the obvious example of this. Now, you have niche genres. Now you have tiny pockets, right? For example, Boomy enabled Turkish EDM as a genre. We have tons of users in Turkey who are creating EDM albums, almost all for the first time. We've heard from these users, they told us, I have emails from users that are like, “Thank you for making this web based. Thank you making it free. I'm using this on my phone and I've never been able to make this music before. We don't have a store for this where I live and this is super fun and I want to keep doing that.”
Now, you have those niche genres, and I think that's where the growth is going to come from. I think that's where the investment is going to have the greatest upside, because we're shifting into a market where that line between creation and consumption is blurring, and I think it's going to get so blurry that this audience performer major, indeed, that paradigm that we're used to, it's going to become something new, and we spent a lot of time thinking about what that new thing is going to be. But what we're certain of, is that that's the bet you want to place. That is the bet with the outsize return. If you want to invest in musicians, the least efficient way to do it would be to start a traditional record label the same way everybody did in the ‘90s. It would be let's find three great artists that I think are great, or that my A&R team thinks are great, let's put half a million each behind them, and then hope, which incredibly is the business model of the vast majority of new record labels out there.
In my experience, I just saw so many of those labels. I got to know some of those artists, and saw their journeys and was like, “There's got to be a better way to do this. And I bet the way to do this is to focus on those artists who haven't broken yet.” Lil Nas X is a perfect example. Billy Eilish another great example. These were artists that didn't come through maybe the traditional channels that the hits of yesterday came from. So, maybe that's a lot of words to say. I think we're transitioning from a hit driven business, or a business that's primarily or solely hit driven to something new, and that new thing has a lot more room in it because we're back to growth in the music industry.
[00:08:34] BT: That's really interesting. The other interesting thing about you is you’re a bit of a hybrid, because if you have a technical background, I'm curious if you relate to this, because my background is technical and I think there were moments in my career, where I didn't have as much respect for musicians and artists that the smart people do this, and other people do that, which I disagree with now, because I see real genius. There's just as much genius in an artist or musician. And if anything, I would argue there's more. They have to be even more courageous because there's not a right answer. Even though you and I are talking about, there could be a right answer or there is a process that's developing. But if you're a musician or an artist, there's an infinite number of ways that you could have modified that or changed that. And as someone who studied math, or physics or AI, there's typically a right answer. It's more minor. So, I'm curious how that evolution has been for you. I guess it's more of an embarrassing thing for me to bring up that maybe I discounted these geniuses before. What are your thoughts with that statement?
[00:09:34] AM: Maybe I just have a very broad definition of artists. I guess you could say I have a technical background. I taught myself how to code when I was younger and I think there are a lot of musician hackers. That's true. That's something we know is true. It's a similar part of the brain in some ways and maybe that's why my perspective and this is something that I bring to the company. This is something I talk to other founders about all the time. And tell me if the following scenario, is familiar to. You've got a bunch of business people in an organization, and you've got a bunch of developers, right and engineers, and the timeline is a matter of debate. The timeline is why isn't this thing done yet. I really view programmers and especially the most talented programmers, they're artists, they're not mathematicians. What I say sometimes to managers is like, you're going to a bunch of painters and you're saying, “Why isn't the painting done? I expected this painting on this date.” That’s not really how painting works.
Especially with something like Boomy, which is creative, which reflects the creativity of us as a team making decisions about algorithms about interfaces. There is no right way to build a Boomy. Boomy has never existed before. You've never had vertically integrated and music creation and distribution. When you're going into these unknown areas, looking at this as well, these are just a bunch of mathematicians that need to solve an equation. The equation is x. This isn't really a solve for x kind of a business. This is something where we constantly have to listen to users, we constantly have to reprioritize what we're building, and I don't think there's anybody inside of Boomy who I wouldn't consider an artist in some way or another. Some explicitly, they're also musical. Some better just very talented programmers and maybe society doesn't look at that person and say that's an artist. But I failed to see the difference when you're building something creative.
[00:11:23] BT: I love that. I completely agree and I love that sentiment. Because to be an artist is to create, it's to be creative, and engineers and developers are definitely that. They're trying to think outside the box. There's not a one way to build something. So, I am kind of stepping back from that earlier statement that things are binary, I agree with you.
One question I had for you, which I haven't made this connection before, is, I'm a big fan of storytelling. If I bring you into a scene with the hero, the thing of the hero's journey, going through the chaos, and the hero does something incredible. If I bring you right into the moment with the hero, you don't appreciate it. It doesn't change you emotionally. You really have to go through the story to really appreciate the heroics. I guess the question I had for you is, is music the same way? Because when I think about a very powerful riff, or think about a song that's made you cry, if I brought you right there into that, into the five or 10-second portion, it doesn't do it for me. I need the whole thing.
I'd love to dive more into music theory. It's not about the five-second window. How do you think about creating a song or the diversity of notes? Or maybe bring our listeners up to speed on music theory? Because a lot of them are AI professionals and developers and data scientists. This is a new territory for them to understand this.
[00:12:40] AM: Well, sure. And I think that the answer is that we're trying to figure this out. We don't know. With my musician hat on and with my producer hat, the guy sitting there with a guitar has a better end hat on. I agree with you. There's this way that we think about creating music and what music is. There's this sense of I'm sitting there, I'm thinking about it, I'm bouncing off of other highly talented, creative musicians. If you make a parallel, or if we sort of make a parallel to the way AI systems work, we're trying to find our local minimum. We're trying to find our results of this set of variables, which is my creativity, your creativity, my ability guitar, all this stuff, to create what is probably influenced by what we already like, which becomes this like two-minute 30, or three-minute 30 work, if we’re playing rock. I used to play a lot of classical music and whatever they would call modern works, modern classical.
In those environments, it's a lot different. You're probably going to listen to what's happened before. You're going to try to not do Bach. We've done Bach. So, what is there left to do with a string quartet? And maybe you end up with a 15-minute atonal work that 98 percent of people would say sounds like crap, and then the 2 percent of music theory trained, highfalutin Juilliard attendees will say, “Oh, my gosh, what a stroke of genius.” So, if what we're trying to do is make good music. Again, I'm going to go back to what I said before, it's just music that people want to listen to. And I would have agreed with you of like, you can't throw somebody into it five seconds, you have to build a story. You have to express artistry. You have to get somebody involved with this art and with this music, and then TikTok happened. It's like, “Well, nope, wrong. Definitely not.” Because TikTok is full of 10 to 15 seconds. I mean, we have hundreds of millions of streams of Boomy music on TikTok. I'm not saying there wasn't artistry in it. I'm not saying our users aren't artists. They obviously are.
But the mode of consumption shifted so dramatically, that what people expected and what people wanted, as it turns out, throwing you right into the best part of the song, and TikTok actually allows us to do this. TikTok allows us to pick which part of the song is going to be defaulted to a video and TikTok and that's an interesting In function and we spend time thinking about, “Well, is it the build that you want to put in TikTok? Is it just go straight into the beat?” We're trying to shove these old ideas about what music is into radically new forms of consumption.
So, I'm not answering your question about music theory. But I think from a theory perspective, and from an optimization perspective, and what we work on every day, we have to start from this place where there's no perfect song. If you had the most magical AI machine in the universe, and you asked it, write me a perfect song, there's probably a 90 percent chance it's just going to write you a Beatle song, right? I mean, because what does that – look at history. It's going to write you, Beatles, and Michael Jackson over and over again, until you move on. So, I think the biggest difference between Boomy and maybe other projects in the space, other research in the space, certainly maybe a beige or difference between us and like open AI, is that we are using this technology and these systems to look forward not backward, meaning we're not trying to make a best guess, regurgitation of what Michael Jackson could sound like, if you continue to make music on a certain timeline. That's a very cool problem to solve with machine learning and AI. And there are definitely some teams that are looking at that problem.
But we're looking at this differently. We're saying there's an opportunity for a new consumption dynamic. Clearly, those new consumption dynamics are happening in market. We're taking these old formats. We're trying to shove them into these new formats. But what is the native format for this upcoming generation? Not even really, you or me? But this upcoming generation and the way that they're consuming music and the way they're creating music, how do we serve them because they're the ones who have shown up in force to create on Boomy, right? It's unsurprisingly much more of the younger generation than it is my generation or your generation.
That's our problem. Our problem is, there's no perfect song, there isn't a music theory answer to what a perfect song is. All there can be as a perfect song to you, and all there can be is, well, what works in the market. So, that's why our approaches put hundreds of thousands of songs, hundreds of thousands of releases, millions of songs, actually, relatively quickly into the market, and then look at the data. What are people listening to? Why are they listening to that music? Can we track that back all the way down to specific sounds, specific types of hits, specific structures? What do people edit? What do they change? What are the people doing right with these algorithms, because that is the equation that's going to lead us to consumption, which again, is kind of radically The only thing that we care about. Ultimately, what's good is what gets consumed. That's the principle that I set out on anyway.
[00:17:34] BT: The fun thought listening to you is we're novelty seeking creatures and genres change over time, and so there's always that need for fresh. There's always that need for new. It's not like we would just continue to invest in music we've heard. The thought I had listening to you is AI should be very good at helping us generate hybrid mixes of combining this genre with this genre. Normally, you'd need an expert musician to try to kind of get through it. But AI I could just give that to you immediately if you had a question.
What does hard rock jazz sound like, with kind of a New Orleans flair to it? If it can start giving that to you, it might allow us to – one thought is AI will accelerate our creative ideation or experimentation, because I think one of the concerns potentially for the audience's is this an area in AI that is unfortunate where it is, replacing the artists? That it is replacing the musician. Was this kind of a sacred side of humanity that we don't want to be replaced? What's your response to that? Does this help the artist or hurt the artist?
[00:18:40] AM: I agree, it's like the most interesting question. I'd be lying if I said it wasn't a big reason why I wanted to jump in. Because not to make any enemies here, but I saw a lot of technologists who were experimenting with AI and doing it, and when I sat and I listened to their business models, it was one to one about replacement. I won't name the company, but one of them, for example was, “Hey, check it out, we can just replace your composer for your film. We can just get rid of that person. And we can get rid of that whole orchestra full of people. And we can draw some lines, and we can do some things and press some buttons. Now, you have a perfectly matched score to your video.”
What was interesting, I think, in that experience was budget wise, it still kind of made sense just to still hire the musicians versus use maybe a more complicated or almost perfect solution in that. But the issue for me was the premise was, this is a thing that people are doing, that we can replace with technology and that is how we will make lots of money. I didn't like that. This probably isn't true. But at the time, if you looked at the major AI music companies and the ones that were really there, I just saw a lack of musicians and a lack of people who had maybe the cultural experiences of spending a lot of time with struggling artists and struggling labels and really kind of understanding the business. And that if you tried a strategy like this, or if you came at it from, there's this skill out there, and we can just replace it with technology, then you would have pitchforks outside of your office and that's just not viable, in my view.
So, that's kind of the first. I want to give context on why I have this perspective, before I dive into it. The other thing to think about is that I'm a violinist, right? I’m a violin player. Violin is hard. Violin is like, insane. If you actually think about what we're making, like four-year-olds do. I started violin at four and I had like an intense Russian teacher who was like, it was highly disciplined, it was highly structured, it was like, “You're going to Juilliard. You're going to be the greatest violinist in the world, or you're a giant failure.” Anybody who studied music or competed or have ever played these competitions. You're like nine, and you're competing against like eight other nine-year-olds to be like, who can play Bach the best? It's kind of bonkers.
I remember the first time I picked up a guitar, and I was like, I play a lot of guitar, too. I picked up a guitar, and I was like, “This is cheating. There are frets on a guitar. As long as your guitar is in tune, you could just put your finger in a one inch space and you could just have be in tune. You don't have to learn this for 10 years, starting from very young age to get the muscle memory.” It's like, “This is totally BS, this is totally cheating.” My point being you can draw purity line, anywhere you want. And me coming from a classical background is, believe me, they draw purity lines all over the place. They draw purity lines of like, anything past 1910, sometimes.
It can say, is it real? Or is it replacement? Or is it whatever kind of about anything? That's before you get into DJs, and of course, before you get into something like Boomy. So, that's a lot of words to say. I believe that this technology, the underlying set of technologies, that underpin what Boomy does, is not about replacing a skill. It's about giving a skill to people who wouldn't necessarily have the same kind of resources, the same kind of access that I did, that probably a lot of musicians did. From a musical perspective, it's about doing something new. I actually don't even like that we put genres on Boomy. We actually launched without that. People got really confused. People are like, “I want to make hip hop.” We're like, “But don't you want to make the music in your soul?” And they're like, “No, we want to make hip hop.” So, it's like, “Alright, fine. These six ones are going to get you like close to that.”
But the way Boomy works is to make mistakes, is to blend genres, like you were saying, is to use algorithms that can kind of get a little bit weird, a little bit strange, sometimes necessarily bad. But to get you to where you can take that, edit it, add your voice, imbue it with your humanity, then you have a great work of music. That's how you get to great songs. Not by an optimization algorithm, but by giving people tools to express themselves. So, that's where I just think my perspective is that no, we're not replacing anybody. We're creating a new generation of artists who are creating in a different way, and in the same way that I can look at guitar and say, “Well, that's BS. That's too easy, or that's whatever.” People are going to look at Boomy that way too, but there are hundreds of thousands of Boomy users who consider themselves artists, because they are. They're just creating with algorithms or with our tool or with our instrument in a different way than what they recreating with traditionally, because Boomy has never existed before. So, that's the interesting sort of space that we're occupying.
[00:23:40] BT: One thought I had was sometimes it's the very subtle imperfections that make it perfect. So, when you're listening to a musician, you can imagine AI playing the violin where it could actually play it digitally, at a level that or even maybe physically a robot can play some musical instrument with no mistakes. I think there's something beautiful and raw about especially for a live performance. These minor details I think, make the difference. Something you might appreciate, Alex, is I've grown up – this will sound too harsh, because I was going to say I grew up listening to garbage like Warped Tour and like all of these different bands.
[00:24:18] AM: It’s great stuff. What are you talking about?
[00:24:20] BT: So, I went to a jazz concert, a private jazz concert in New Orleans, and I felt like I've never heard music before. It was a spiritual experience. But also, just the way that it was unique. That experience can't be repeated the way they were playing the music and how they were doing it. And also, I think for people, especially on the West Coast, they're not used to brass, and just the power that brass brings in the song, it's hard not to be emotional, which is pretty interesting.
I did want to fast forward in the future. Maybe this is being too futuristic, but this podcast is called A More Intelligent Tomorrow. Do you think you see a scenario in the future where We are using real time feedback from EEG or directly from your brain to try to accelerate what is good music. Because rather than people rating, could you just monitor my brain and say – especially at Boomy, you try all these variations, and within a few days, you've really narrowed in on what I'm going to react to.
[00:25:20] AM: I don't think it would surprise you to learn there are already teams working on that. There are already – I can think of at least one and I can remember the name Nicent, I think, that we went through – ASCAP did a program last year, and one of the teams in the program for just future facing music stuff in general. And one of the teams there was a BCI brain interface music company.
So, we thought about that exact question. Can you replicate that experience? Can we give you that experience? I don't know. Probably not. In our lifetimes, I think it's the safest answer I can give, based on how little we know about, how the brain works. Based on the signal to noise ratio of the ways that we measure and the limitations of that measurement. I think we'll probably limit the applications for a while. But if you can get to the point where we really understand, and maybe what you're hitting on is we don't really understand why you standing there with brass in that environment gives you that – the reason you're using the word spiritual is because you don't know how to describe it, or like words like spiritual, I don't know, I don't get it. I completely understand that with music.
I've got 20 stories that are like – and I was in New Orleans two months ago. Yeah, it's like, get me to the music ASAP, because there's something amazing and special here that's happening here because of that tradition. The only way I can explain it, the only way I can rationalize it is that we've been evolving for a really long time. We've been a social species for a really long time and there are things about music, and there's things about certain styles of music, that just click for I what I think are evolutionary reasons, Brass is loud. Traditionally, it's been like the loudest noise is the thing that makes us all pay attention. I think there's really basic sort of human behavioral reasons. But then there's this whole other question mark category of I have no idea, why is jazz so great? Why is jazz in New Orleans so much greater than your average club in in some other place? That's crazy stuff. And will BCI be part of that, in helping us to understand that?
I don't see another way. Because until we have a full understanding of what it is that's making you feel that way, we can't even measure it. And if you can't measure it, you can’t analyze and we can’t analyze it, we can’t build towards it. So, I don't know. Hopefully, that's a good enough answer. But I have a similar question and we're trying to solve it with again, like one of the things that we know we can do is make you listen to a song that you made, because you like it, and you'd like it because he made it. I'm working on that, that we haven't really had a major exploration of that phenomenon in music yet. Because we've never had a time where it didn't require a huge amount of skill or training and some tradition to get a song that is mine, that is part of my identity, not just created, but out into the market and out into the world.
So, that's what I care about right now. It's the consumption dynamic around I want you to be able to say, “Wow, that brass was incredible. This is really awesome. I want to do that.” If I want to do that, you don't have that thought today, because you're like, I'm going to go learn how to play a brass instrument, and I'm going to find a band, I'm going to move to New Orleans for eight years, and then I'm going to get into this band. I'm pretty sure we can get you there with the technological principles that we're developing pretty soon. So, it becomes, well, I want to make Ben's version of what I just heard. I want to build on this tradition, or change that tradition.
What Boomy is doing right now, I think of it very much as a musical tradition machine. How do we create not just music and not just people creating that music, but actual traditions that can last for 50 years, 100 years, 1,000 years out of what people are maybe calling AI music today? But that will almost certainly have better terms for it as the stuff proliferates and as we start thinking about music a little bit differently.
[00:29:20] BT: I feel like there are 20 different threads to pull on when it comes to future technologies. So, maybe I'll throw out three really quickly and then you lean into the one that's the most interesting to you. So one is, when we assign music to think of video, a lot of times we're just selecting like, Oh, these are the five tracks. I'll pick one. This one sounds better, there will be a future where AI comprehends the scene where it's much more intentional about the music design. That's the first topic. And then the second topic is if I go on a two-hour run, I'm physically running. I need music that inspires me to help me run my fastest to hit my goals.
Think about sure music that's dynamically being written based on feedback for people that like to do sports and exercise. And honestly, I forgot the third. So, the first two.
[00:30:09] AM: Well, it's the same personalization, or context-driven music. And maybe I'm going to go on a slight tangent here and just say that, I think one of the things that's held back context aware, algorithmic music is the difficulty to monetize, right? Because some of the things you're describing, I've seen them, in some cases, like 5 to 10 years ago. For example, products that would have a dynamic, specifically for the running – I can’t remember what it was called, and I'm not sure if it's still around. But it was basically this app, and I saw a demo where as you ran faster, it would sort of increase the BPM of the track. As you were in your environment, it would change, right? It would basically change depending on what you're doing.
I think the clearest example of this is these meditation apps, like calm. Where you've got what is functionally algorithmic music, but it's changing, it's going faster or slower, or augmenting depending on what a person is doing and falling asleep is a big one. Meditation is a big style for us, actually. People like making their own reputation music. So, that context dependent algorithmic music that changes is very cool, is obviously useful, is not really able to be copyrighted, is not really able to be this is just the word exploited in the global rights system, right? The recorded music industry, the one that I'm so obsessed with, which by the way, had its biggest year ever in 2021, if you weren't paying attention, we’re back to the ‘90s, and then some in the recording music business, and with a huge amount of growth on the way in the next several years.
If I have a song or if you have a Boomy song, or if you have any song that you created, and you're playing that in a restaurant, you're playing that to a movie, to your point about, it's being written to the scene, you got to pay for that, and you got to pay the rights holder for that. And there's a whole massive, beautiful, sort of broken, but always self-healing and fixing system of performing rights organizations and record labels, and just a bunch of different players who are all about extracting that value and paying it to copyright holders, and that's just what exists. If you have a work, like a composition and recording that changes that's dynamic, similar to what a Boomy algorithm could do if we made it real time, or what some of these systems like calm have, it's not really copyrightable, because it's a shifting composition. It's not an instance of a recorded work. It's an instance of a composition.
In order to be eligible for copyright, there's a requirement basically, or maybe an unspoken requirement that it is a recorded work or a written down notes on paper. But notes on paper can change, and you get into sort of this weird gray area, when you have a dynamic composition, that's not going to be the same all the time. You can't really license that. You can’t monetize it in the same way and that just doesn't make any sense to me. So, I think the use cases you're describing are totally possible. The way you incentivize the business model of some of those systems, I think, is what's missing. Boomy maybe has two goals, right? One is purely creative goal. Let's draw creativity out of people.
But the other and why we've had this feature for so long is well, how do we monetize that? How do you build a solid business around that idea? Right now, we're creating recordings, right? People are creating recordings, and we're in that global copyright system, and we're getting you paid. We pay 80 percent of the royalties that we receive out of distribution fees to the users, and the users can actually earn revenue, on our platform from music that they're creating with our technology and that is great. I would love for another type of creator, somebody who knows how to make algorithms. Maybe somebody could create the perfect algorithm for you to run to, but how do we compensate that person? How do you pay for that? Would you pay 20 bucks a month for that? Would you pay 100 bucks a year for that? Or actually, what I suspect, would you pay nothing for it? But you're getting it as a benefit from your Netflix or from Apple or from some other existing provider, some features, some streaming service.
If that's the case, then how do we pay for that? How do you compensate algorithmic music creators? That's really dicey question. We don't have a good answer to it. So that's what I see holding it back. And I think, as we try to develop this stuff, we'll eventually get to, not how do you do it, because I think we know how to do it. It's how do you make money from it? And how do you incentivize making it really, really great? So, that's something we spend time thinking about.
[00:34:42] BT: So, it seems like it would definitely have to be a service. If you are having custom tracks for every listener. The third problem that came up that I forgot about was I saw a white paper where someone was showing that you can play sound tones during REM sleep and it will improve sleep, where you could imagine having AI go chase that, which maybe it's fascinating, but it might be less interesting because ultimately when it does optimize that, you and I might agree that this is not music, whatever this is. I'm glad it helps us sleep better. But I'm glad I'm asleep when I'm hearing it. It's not something I've listened to.
The one analogy that I think would be really exciting. So, I'll give you the art analogy and then we can move this to music. You've seen DALL.E 2, where they can create these amazing works of art with tech. Now, you can imagine a scenario where there could be a company in the future, and I think there will be, that you can have an 80-inch OLED picture frame in your kitchen and it will have dynamic, evolving art. But then me as the owner, I can speak to it. So, I can come down in the mornings, “Oh, the magician koala in the field of wheat, make it the ocean or add some sharks,” and I can just tell the painting and it will do it.
Now, to take that over the music. I imagine in a smart home system, it’s always playing music in the background. But I, as the owner, it's almost like I'd become a composer. I could say, “I want to hear more brass.” I could say, “Make it more like this, 20 percent like this and that.” Do you think that is also a future we could realize where every human becomes a unique composer in the moment when they need it?
[00:36:17] AM: I think that there are two things that musicians do, and the first thing we do is we create a song. The second thing we do is we listen back to what we just made, and decide if it's good or not. So, the first part is skill and the second part is taste. The first part is possible to be done algorithmically and that's just true now. Maybe musicians and purists like me debated this for many years. Can you actually create a skillful work from a “machine” that people wouldn't be able to tell the different word there? The stuff coming out only can pass any tuning test, in part because it's people, it's not just a machine. All of our best stuff is people creating with algorithms.
But the interface that you're describing, is really one where the automated system in your home system is doing part one for you. It's filling in the skill. That's what Boomy does. It fills in the skill. And then the second piece is the taste. What's your taste? Maybe you have horrible taste, as compared to your peers and everybody and you're going to create something that sounds terrible. Maybe you have amazing taste, but no skill, and you're going to be able to tweak it and figure it out until you've got something that you really like or something that fits you in the moment. That is exactly the interface that we're building, that people in the space are thinking about. Whether it's happening through voice, whether it's happening through buttons and sliders, whether it's happening through cool three-dimensional blobs that you're pulling in a Metaverse. There's a lot of different questions about what is that ideal interface for somebody who is potentially a non-musician.
When you say, is everybody going to be a composer? I believe that, aside from the teeny, teeny part of the population, that actually has an inability to hear me. It's an interesting group of people. There are people who don't really like music, and they don't really understand music. It's a tiny group of people, but like, 99.9 percent of people really like music of some kind. What is that? Why babies dance and sing? It's this innate thing that we have, an ability to understand it. So, so what I think is that, it's not replacing musician, because it's really just filling in the skill gap that you need to do part two, because you're doing part two anyway. You're just doing part two with playlists and your friends playing new stuff, and you hearing things out in the world and Shazam-ing it. That's part two. That's listening to music back that someone else made and saying, “Hey, I like this. I want to tie my identity to this.” As the skill gap for part one, for I need to make a song closes, that's where I just think that what is a composer? You've solved it.
Yes, I think that interface is coming. I think you'll see voice interfaces with that. I think you'll see Metaverse interfaces with that. When I look at DALL.E, and when I look at the arc, again, I think I spent a lot of time thinking about consumption. There are some my YouTuber friends are going crazy on this for last couple of weeks, and they're like, “Oh, look at what this did. I put this prompt and here's this content and then it's a tweet, and then you move on.” So, you have much faster creation, you have much faster consumption. Does that tweet matter in 30 years? Do you put it up at MoMA? Remains to be seen.
So, I think that very fast content cycle is what we're going to get used to. It's not going to be let me go listen to the song or look at this work of art over and over and over again, because it's a great piece or it's great master. I think it's going to be, “Hey, check out what I just did with this song. Now, I'm going to make a second one. Hey, check it out what I just did with this release. You know what, I'm going to make a whole release of just Lo Fi but it's going to be about my school, and how much I'm annoyed my teachers.” That's the kind of, I guess, hyper-personalization that I think is going to start defining how we look at a lot of different kinds of content. That was traditionally just wait – it just too way longer to do because you didn't have technology filling in that skill gap.
[00:40:04] BT: That's fascinating. I think the other thing I was thinking about is you have these layers of technical competency, so someone like yourself that you studied music, you could actually really dive into the weeds of being a composer, if you have the AI, if you're unable to AI, but someone like me who's maybe more naive, or very basic. I'm really limited to, “Please review my Shazam, or please monitor the Pandora,” and that's kind of my limit. I might say, “Introduce this instrument" or “Do that.” But besides that, I can't really dive into the weeds like you could. So, I think you could have different users. The other thought I had, you could also imagine a smart home, that is – it's always monitoring you. If you're dancing, it's pretty good. So, if you don't dance at all, then it's got work to do, so that that could be an interesting feedback.
[00:40:55] AM: The real question becomes, what happens when your home starts playing you the song that it knows, you would want to hear without you prompting it at all? Because that's the insane end point of all this stuff, that I don't know what that looks like, right? It becomes like, Ben, you seem stressed, I'm going to play the song that I just wrote for you that, based on all of the data that I know about you, is going to make you feel a certain way. And now you have an automatic way to manipulate people's emotions and feelings if you unlock that bit. If you unlock what it is about music that makes people feel certain ways, and then you have automation in that. So, what does that look like? I think that's again, far future. I think, maybe not in our lifetimes.
But the question of what is a musician in that world? Maybe there's no human that can match? But that's sci fi. That's. We can get close to that, I think. But I don't know that we're going to see that in our lifetimes. It's certainly the natural endpoint, I think of what like the neural links and some of the research that we do might lead to in like, 200 years or something. But I don't think we have to worry about that anytime soon.
[00:42:06] BT: There's a fun thought that if we actually reach that, then we're very close to the singularity. If AI can create the perfect song for the perfect moment, then it's very advanced. Maybe it'll just end up playing what is the – maybe the tribute Jack Black wrote, the world's perfect song, that'll – it'll all regress to that. I was also thinking about it, can we have some general consensus about what is bad music? Even the joke about work tour, I'm thinking about there are definitely some artists, Kid Rock, or I don't know, pick your least favorite band. There has to be – maybe Kid Rock is a bad example. I think like, St. Cloud, we all see or there has to be some music out there that we can definitely agree that this is bad music.
[00:42:52] AM: I'm going to share a quick story about early Boomy days. So, when TikTok started becoming bigger in the United States was around when Boomy was in its like, earliest testing phases, there were a small number of users. We were doing distribution. We were seeing numbers coming out of TikTok, and we were watching these videos and trying to figure it out. And there was just one video I will never forget of a bird, and this is the video I'm going to describe for you. It's about a six-second video of a bird pooping on a bed. That's what it is. It was racking up views for whatever reason, right? Some kid or something like how to pet bird, it pooped on the bed, there's your story beginning, middle, and end.
Our song, and at the time, our systems were let's just call it pretty rudimentary. You can accept or reject tracks on Boomy that are within the style that you've selected. Back then we had a 98 percent rejection rate. So, 98 percent of what's coming out of our systems was just like, totally garbled nonsense, and nobody liked it. Two percent people would save and like and release. I would say that my view on this particular song that was on this particular video is not what I would call a traditionally good song. It had some of those early mistake type features that made it really abrasive and weird, and somebody had to use the word poop in the title of the song. It had been recommended for this particular video. And this video went, I wouldn't say viral, but it was getting a lot of attention.
And I sent to the whole company, and I was like, “This is so important, guys. Everybody needs to watch this video of the bird pooping.” I swear to God, at an industry conference in front of every major record label and all these big wigs and all these like fancy music industry people, when I gave a presentation on what we're seeing is the future, I started on like, “Everybody we need to watch this video, this bird pooping.” There were some people in the room who got it, and some people don't. Now, with the benefit of two years, people still joke about this moment, where like, think about it. It's the music industry. Everybody's like friends with somebody famous and trying to make the best possible song and I'm up here saying here's a horrible song by traditional standards in this video. That's not a good – why would we want to watch this? This isn't going to go in Netflix. This is a six-second video of a bird pooping. There’s money here. There's value here. There's consumption here. This is what hyper personalized consumption looks like. This is what a place for a bad song is.
So, going back to my principle of like, what is good music? Good music is what people listen to. That means good music can be bad. Having good music can find these corns, and of course, since then, what have we seen? We've seen 10s of millions of videos get created with Boomy content, hundreds of millions of streams, right? Coming weekly, out of all kinds of videos for all kinds of different purposes. It would be like saying, “Hey, why would anybody want a 15-second video at all? Because clearly what we all like, is our long Netflix specials, right? And eight episode runs. That's the only thing anybody could possibly” – well, as it turns out, nope. There's this ravenous appetite for very short form video content, to such a degree that it's completely changed the way we think and the way the music industry operates in like two years, and that's just going to keep happening.
What makes a bad song is, in my view, maybe if you force me into a definition, it would be a song no one wants to listen to, for any reason at all. That's a bad song. I would exempt anybody who you would traditionally say is bad. Because when you're saying, “Oh, this is bad”, you're seeing it from a perspective of like popular culture from a pop culture perspective, right? What is good music and what is bad music? But pop culture and what that is, is shifting very, very rapidly, and your pop culture and my pop culture is not the same pop culture as this upcoming generation. They consume differently. They think about the stuff differently. They're like, who is the Beatles and why would I care? Like some of them, that's extreme. But you know what I'm saying? When we do customer research, and when we really understand this generation of creators that's coming out the people who are famous to them, like I've never heard of, and vice versa.
So, that's what I think is going to shift. You and I can decide, like, here's some artists that are definitely bad. But those artists are making millions and millions of dollars off of music because people want to listen to it. So, that's just my opinion and my principal. I'm not saying I'm right. But I do think that pop culture is shifting that “bad music” or bad music in the context of traditional pop culture, that has a place now that it didn't before. I can tell you, I don't think the music industry is at large has really shifted there. But that's what creates such a huge opportunity for those of us who are maybe willing to not stick up our noses and really look at like why would somebody watch this song with this video of this bird pooping? And how can we replicate that thing? That's how we've been able to grow so quickly is by bringing unpretentiousness to it, I guess.
[00:47:49] BT: The thing I like about this topic of music is you're constantly surprised. So, let's take a song like Rebecca Black, that ‘Friday’ song that went viral. Initially, when we hear it, we might say this is so bad, it's good. This is so bad, we think this will actually go viral. So, I think the thing I'm thinking about listening to you is there's different types of good, or you're saying it a good song is something that someone listens to. But there's different outcomes, because I might watch something on TikTok that is really bad. It's not a song I'm going to listen to for hours when I run, but it's a song I'll listen to in the short form. And it actually adds to the engagement adds to my attention because it's different. It's not – so I guess there's different outcomes. You have music that inspires you, music that makes you run music that mellows you out, but also music that might just pull you in with short form attention.
[00:48:40] AM: I mean, it kind of goes to my point before about how you can draw a period line anywhere. I remember, I was at a party, and I was in charge of the music, because I was the music guy there. I have a friend and she pretty much just listens to pop music, top 40. That's what she listened to. So, I put LCD Soundsystem on and like LCD Soundsystem is great music. I don't know, maybe we could debate that. But I love LCD Soundsystem. It was the intro to like a particularly odd track, and so she comes over to me and she's like, “Hey, can you put on some music?” It's like, “What is this? What is this crap?” Right?
So, that's my point about good versus us broken indie kids, there are certain artists that are that's obviously great and amazing. But then somebody else who has a totally different context is not going to think that and that's before we get into – and really what we're talking about is like domestic US driven pop culture. Let's start going to the rest of the world for a second. There's a whole part of the world that listens to a different scale. There's an Eastern scale. It's not the same scales that that we're used to with music and other cultures. I remember when I was studying Klezmer, or studying Indian Ragas, they play and they think about music in this completely different way, so much so that if you weren't raised in that tradition, you would hear it as being atonal and like definitely not good. And similarly, they would listen to our music as being atonal, because they were raised in a different tradition or raised in a different scale.
I just keep going back to this point that like, we could argue, and probably these arguments tend to be about, well within the context of US domestic pop culture over the last 30 years, who is good and who is bad. And I'm going to use that to like, define my identity here or I'm going to use that to say, I love LCD Soundsystem and you don't even think it sounds like music. Both of us are wrong. If we're like in Bhutan, or somewhere that they don't listen any of this stuff. So again, if you ask me, what are the universally good musicians, the data answer, if you're curious is the Beatles and Michael Jackson. They have like this insane global appeal.
But as somebody who obviously thinks a lot about music, my challenge to that idea, is that true because there's something innate about the music itself? Or is it because it happened at a time where mass media could simply play the best thing of that time for the maximum number of people on four channels, or on the first time they got radio, you turn on a radio and that's what's there? Is that why people like that? And is that why we tie songs that we hear now back to some of those historical influences? Or is it something innate about the music itself? We can't really know that. But separating out those questions kind of gets you to this point where I don't think it's going to be that way forever. I think it's been that way for a while. But that was limited by mediums reach, that was pre-internet. I guess I'm just saying that, can we identify objectively good? Can we identify objectively bad? All we can really go on is consumption, and we're always going to have different views of it, whatever we bring to it, right, I think is what's true.
[00:51:47] BT: Yeah, it's complicated. I was laughing after that New Orleans jazz concert I went to because the individual that invited us to go, he was asking, what type of music do you like? I was so overwhelmed with the experience that I thought, you're like this.
[00:52:04] AM: I like this music.
[00:52:06] BT: I think in the future, I was thinking one of the issues with some of these jazz concerts is sometimes you get lucky. Sometimes they're on, sometimes they're off. So, if I was inviting someone to something like this, this experience that I had, I don't give a crap about what music you'd like. You're going to go and experience this. Normally, that is up for debate and it's complicated, and if you're listening to my radio, or my track, or my Spotify versus yours, we have different taste, different experiences. But like you've hit on multiple times in the interview, you've got generational changes. Someone who's younger is going to have completely different tastes than we might have, where you and I both have a strong appreciation for a lot of these classic artists because they were part of our childhood.
Alex, I've had a blast talking to you. This is such a fascinating topic. And it's also one of those topics that I can't see this ever becoming boring for you. It's never – the search for the perfect song that can never be and how complicated this all is. But also, I think the thing that is very inspiring is the good that this can do. The more people that can be uplifted, or even the more people that lean into music that didn't have that ability, that is a very beautiful gift to give. So, that's really inspiring that you can align technology and your passion, but also align it with something – it's an inspiring mission.
[00:53:30] AM: And it's fun. It's fun. It’s fun to work here, also. For all the product talent and technical talent listening, we're a fun place to work because it's a fun set of problems. And you're right, it keeps us interesting and engaged probably for the rest of our lives. Because a Boomy is evolving every three months, as new people come in with new use cases and as we improve the underlying algorithms, as we improve the interface, and as we get closer to, I think that moment that you described, where you're just maybe speaking your house or crossing a couple of buttons and describing what you want, and we can get you there by filling in that skill gap. But education and use of Boomy in the classroom, I think has been one of the surprises and maybe delightful surprises that we saw, and that we're continuing to see in terms of a use case for Boomy that we never really thought about in the beginning.
We know that teachers are using Boomy in the classroom to teach not just how to create music or what music is, but what AI is. It's a really easy way for students to talk to teachers about this. It's an easy way for students to grasp, like, what is AI? What is automation? How does this work? Because they know what music is, and they know that they would want to make a song or some of them are already kind of musical. By showing it in a classroom setting, it gets music students over the hump of I suck. It gets them over the hump, because that's what makes most students quit. They try to play an instrument or they don't have enough time to practice with all their other stuff and they have no going on, and they quit music.
Same thing with programming with technical skills, it's like, this is just too hard, or this isn't really a concept I can grasp. We know that Boomy has really significant future in the classroom, just like it does in a bunch of other contexts that we've seen it be used. To the good, that you're describing, that's what we're here for. We're here because I was a musician, who saw a lot of technologist working on technology that would replace musicians by using tech that looked backwards and looked at how do we regurgitate the past. I said, “Let's do it differently. Let's do something that's going to make something new for a new generation of creators and pay for it by exploring this new consumption dynamic, which is the blurring line between the audience and the performer.” And your point about it being interesting, there's really, really fun stuff that falls out of all those things and we're doing it all at once and it's been pretty crazy. So, that's where we are and I think that's where we're going to continue to go.
[00:55:56] BT: Real quick, before we wrap up, what's the size of the company? How many employees? Where are you in this startup journey, just for people that are considering reaching out?
[00:56:06] AM: Yeah, sure. So, we're 20-ish now. We're actually hiring pretty rapidly looking to be 30 to 50 by the end of this year, even given the slowdown, and maybe the overall venture market. We're maybe unique in the startup sense that we really focused on revenue. We focused on revenue, and getting our users paid, so that they can build little businesses and reinvest and create more. We found a pretty good business here and we found a pretty good group of – actually, a really great group of people to be able to get this mission done. So, we are pretty efficient, right? We aim to be more of like a smaller, but very high output team. People are sometimes surprised. They think we have a lot more resources than we do.
But we've been building a “public” not an open source sense. But we've been building this thing with users, I guess, for two years and change. To see this much progress and this much interest in this relatively short period of time, has given us a lot of excitement about the next phase. So, you're going to see Boomy be a lot really loud, late this year, early next year, because when you go from half a million users to a million to 10 million to 20 million, that's when you get into the billion song per year future in a market that's only ever had 70 million songs on it up until now. So, that's what I'm spending my time stressing out about. What does that world look like? And that's imminent.
[00:57:27] BT: In that experience, you already have the experience that you'll be capturing in the future. That is a massive competitive moat, the insights you'll be capturing.
[00:57:35] AM: That's the idea.
[00:57:35] BT: Yeah. Well, congrats. Congrats on the success and I'm excited for the future of what this means for all of us and thank you so much for your time.
[00:57:43] AM: Appreciate it, Ben.
[END OF INTERVIEW]
[00:57:46] ANNOUNCER: Thank you for joining us on this More Intelligent Tomorrow journey. Discover more and join the conversation at moreintelligent.ai. The future is closer than we think.
[END]