Intro 

Arjun: That piece of music that you just heard was not played, sung or recorded by anyone. This was generated. This was generated by ai. Today we'll be learning and questioning whether the feeling music gives us can exist even without the hands and the minds that once made it. And for anyone who's ever created or listened to music, with your eyes closed, this may hit deep.

Okay.

I wanna begin today's episode by introducing music and what it means to an artist and what it means to a listener. Music means different things to different people. For an artist, it's an act of creation. For a listener, it could be a form of connection. Regardless of what kind of music you listen to, whether you play an instrument or how you understand it, how you perceive it.

Music has a place in everybody's life. It might live as a memory, some form of a sound that marks a moment back in time. For some people, it might live as an emotion that gets triggered when a song ends, or it might even live as a state of mind. Something that for me is true a place where you go where sometimes words fail, words probably are not enough.

We can begin this conversation by saying one simple thing, that music is an integral part of our lives. And if you're listening from a field like design, or you come from a culture where music runs through everyday life in one form of the other. You will probably know exactly what I mean, because music isn't just entertainment.

It's a framework. It shapes how we think. It shapes how we move. It shapes how we react. It can shape how we remember things as well. And today, we'll be speaking about a major shift that's happening in that framework. It's not entirely new. It's been building on for a while, but in 2025, it's finally visible.

And to understand what that means. I'd like to look at it from two sides. What music means to a creator who has experienced the process of being in that meditative state with the surge of the hormonal chemicals, the surge of dopamine, oxytocin, serotonin that come in within that being of the state of creation.

And for any regular listener who, who might feel the same, chemistry is reflected back through that sound growing up for me. the famous Greek pianist, Yani was a childhood hero, He still is, and I remember particularly in one interview he said. If I did my job right, the listener will feel exactly what I felt while I was writing a piece of music.

And that state, that shared state where the creation and the experience are meeting, that's exactly what is being changing right now. And we are speaking today specifically about AI generated music. How for the first time in 2025, things have changed so dramatically that the word generating music is now going to be a colloquial term, and we are gonna see the multi-side reactions that come as part of it.

How can we look at it? What's a way to embrace this or reject it? And I didn't wanna share my thoughts in a reactive state for the last nine months, I've had the opportunity to sit with it. I worked on over 600 tracks that are AI generated with different models. I've learned the technology behind it.

 I've tried to understand the process of it, and today is a conversation between me and Dinesh exploring a new normal of a world where the sweat and the blood of creating art forms may be missing, and whether this is good or bad, whether. There's a reaction that we can try to embrace a point of view that we can probably share.

 maybe the real question isn't anymore whether AI can make music. It's probably whether we can listen with the same innocence. And for people who in the creative field, we've all chased that moment of flow in code and design and sound.

So the question is, is AI mimicking that state of flow? And how good has it gotten till now? If you're a creator of some sort, whether you're a designer or a musician, or a performing artist, this is an episode that you may relate to and we'd love for you to be with us till the end. More importantly, we'd love to hear your questions, your thoughts, your feedback, and whether this relates to an experience we all can.

Probably share. This is the inflection point series. Let us begin.

Context setting - Audio and AI

Arjun: The context is this

ever since. The recent prominence of ai. In the last three years, particularly, there has been LLMs that have come in large language models and a space around generative AI has emerged, which has not only, worked on, images and video content, but it has now worked on audio content as well.

What it means is you can generate or create a prompt to generate audio artifacts,

and these audio artifacts can be text,it can be a speech, it can be a song, it can be a piece of music. specifically in the last 18 months, we have seemed to have turned a corner where. It's, stunningly developed to a point whereit's incredibly hard to differentiate between AI generated artifacts and human produced high, incredible quality artifacts.

 in simple terms, the world of AI is now able to generate music, sound, speech, voice in different, different ways with a large amount of accuracy and incredible artistic nature.

When I say incredible, this means. it gathers a reaction of, wow, this is great music, 

Dinesh: You some examples in mind? 

Arjun: there's this famous YouTuber who exposed this band called the Velvet Sundown. 

Velvet Sundown was a fictitious band created by ai, and they became extremely popular, getting incredible number of views, like in a very short period of time.

Only later to be, revealed and exposed by the,audience as ai. 

Dinesh: years back, even Indian artists started using AI to, simulate voices of artists who were already, passed away. Like we had SP Balasubramaniam voice in one song. We had Vijaykant's voice in another movie, right? So we have Vivek acting as a character in a movie Indian, right?

Arjun: The idea that you can now, create or synthesize data, to create or, fictional art forms, which are not there. And the quality of the output is becoming very artistically, real and in its review sounding. Very good.

Okay. This has been sample tested. When I say sample tested, many people have started listening to these artists just like they would other artists as well. you now have a lot of people listening to music, feeling, moved by it and not knowing whether it came from an artist or not.

How is the world is reacting to this?

Arjun: Right. This is the background to it. Now, when people asked certain questions around it, this is clearly divided the world into many different opinion schools, 

 one of the reactions that we're getting is that, hey, AI is becoming so powerful that we can't really differentiate.

So how can you trust,music right now? The other thing is it's so easy to make money. if big people are getting money from music and if it's so easy to create music like this, now can we also create music and make money? 

Dinesh: Hmm, 

Arjun: Right. So suddenly that access to creating good quality music is now becoming easier.

hence

there are people who are saying that their jobs will be out because of this. This is the context. So first of all, questions, clarity on the context. Anything that you wanna ask.

Dinesh: this context is fairly, synonymous with 

Arjun: other situations where AI is in the writing space or AI is in the image creation space, or AI is in the video space. So is AI in the audio space, is it anything different or is there something about it which is different from the other areas?

Dinesh: why the audio space is, more, interesting to you. 

Arjun: the image, generative AI is fundamental to several disciplines, not just media or entertainment it's important for science itself, the amount it's important for engineering. Also, the amount of, image processing and image based AI techniques you can improve on will. yield, technological advancements in several disciplines.

This is only for images. The moving pictures and audio has other derivatives. For example, the world of cinema and world of audio has many people linked to the same artifact creation. So one song may require 17 people to create. One movie may require hundreds of people to create one song, one movie score may require a lot of things to do it, it touches several roles and job descriptions we knew earlier to exist.

It just not the artifact creation alone, right? It kind of reduces several disciplines, pipeline or workflows together into this artifact creation. Finally. But this is one fundamental difference, that not the job of editor or a graphic designer or one artist doing it. it's almost like a, the artifact that required culmination of several art forms together.

Dinesh: Right. 

Arjun: that's one difference I see. The second difference is there is a huge emotive angle to it, or an incredible emotive quotient towards it, both in the moving pictures as well as in audio, right? and hence more importantly, and this is something we will speak in the next segment as well.

the businesses surrounding making money from these two, the business models that exist today in the world, that try to make money from movies and try to make money from art is very different from how people make money from pictures.

AI's incredible accuracySpacer

Dinesh: you had spoken about two aspects. One, the creation, the AI creation, the AI audio creation has this incredible accuracy and delivery, artistic delivery. So what do you mean by that? 

Arjun: Good question. So we'll take two samples. An artist incredible talent and genius of a mind, creating a piece of music and a AI tool, creating a piece of music right, given to a same sample set of audience. 

Dinesh: Hmm. 

Arjun: For simplicity of this conversation. Same genre, same kind of music, same all things remaining. Same. This artist delivered an audio format of the event, the final product.

This, AI model delivered the same thing. When I say accuracy, I am looking at several indicators of what a good audio product will look like, 

 well mastered, well produced, good mixed sound, pleasing to hear. I'm not talking about the composition or the artistry behind it yet.

I'm only speaking about the output format of it. AI models have, and the techniques and the algorithms around it have identified what pleases the human ears. And there are companies who are working on it to make it even better. we've got all the check boxes, right? Right. The output format, the arc of emotions, 

 the signature styles of the different kinds of music are already captured. they have now gotten very close to what a human would deliver even with incredible number of years of experience, talent.

Training knowledge has delivered. This is what I mean by accuracy.

Artistic delivery has lot more depth to it, that you convey an idea, conveying an emotion, conveying a certain situation. For example, if I, I could generate a song, the, the term now generate a song is, is now normal. 

Dinesh: Hmm. 

Arjun: I could generate a song which is about pain and suffering, and I could say the words, I broke my heart, for example, and the way the I broke my heart would sound by the AI vocalist would convey pain.

 So it has got that nuance to it now.

Dinesh: I heard AI created music, about a year back, I guess, right? I think Google just launched something on that. I don't know where created music is in, I actually tried to do something, and then hear it out. at least at that point of time, it was not very impressive. it, it didn't have the, 

Aesthetic or it didn't have that artistic resolution to it Some sounds are there. sounds like music, but it's okay

 but now you are saying that you can, it's like, it is now like a well produced is that where we are right now? 

Arjun: Exactly.

Dinesh: Oh, okay. 

Arjun: I want you to process this, okay. And then ask me questions, and then I want to lead into something that tells us how real this is.

If I want to compose like Hans Zimmer

Dinesh: So I'm, I'm just, brainstorming over here. Okay.

I don't have the skills for creating music. I can't play an instrument, but I know the philosophy and thinking behind Hans Zimmer's music creation process. Say for example, dune, it's like, it has to be out-worldly and so, it's, it's technologically advanced, so I'm gonna derive music from, metal sounds and, sounds inspired from garage, things like that.

And then I'm gonna try to create a music that is.

that's how Hans Zimmer would approach each movie and then overlay the emotion, of the movie on top of this fundamental music. Be. If I'm able to kind of understand, well, are you suggesting that I will be able to sort of create about 50% of what Hans Zimmer could deliverfor a movie? 

Arjun: Here's my shortest answer to that, you'd be close to 95% of what Hans Zimmer could do.

Dinesh: Oh man, This is not you having a hyperbolic arc to whatever it is This is how it is right now.

Arjun: This is me working with AI for over 600 songs

Dinesh: Hmm. 

Arjun: and telling you this,

Dinesh: What do you mean by that? 

Arjun: I worked with ai, worked with AI Tools services before even launching QuestX. have a collection of 400 tracks,

 and since then, 200 more.

Which are completely AI generated.

Dinesh: So you're saying you created those tracks and you have them 

Arjun: Yes. 

Dinesh: Oh, you're coming from that, background. I mean, let me just process this. 

Arjun: okay, I gave it a little gravity, right? 600 tracks since on la

Dinesh: Now, I'll tell you this, this is where we are at right now.

I have spent,

Arjun: hours every day for the last 90 days, on this area. 

But the amount of time it takes to create. Two tracks, three tracks is one minute.

Dinesh: But still, you have put in a lot of effort. 

Arjun: Yeah. To reach where I am right now to form this, my set of opinions right now come from this place.

Dinesh: yeah, so good that you brought it up. the creation time, though it is less, but still the f did put that effort to, kind of make it work for you, right? 

Arjun: Very, very important question. This particular thing that you just said is probably the most important question, but when you phrase it differently.

Dinesh: hmm, 

Arjun: It took me 30 seconds to create one song 

Dinesh: hmm 

Arjun: of incredible good quality that I would not have achieved myself if I was a solo musician doing something else. 

Dinesh: Right. 

Arjun: If I was aiming for a certain output as a musician without using ai, is the gratification coming in from the effort or the outcome?

 there is this desire to sweat it out to create that outcome. You know, as an artist, there is that. That's where there's no sweat. I'm chilling and doing right now. It's like absolute chill.

So that's statement number one. two, for you as a listener, 

Dinesh: Mm-hmm. 

Arjun: and this is something I've been disagreeing for a long time and almost the reason why Innocence Theory started long back as a music thing was when you listen to a piece of music, without that innocent mind of looking or listening to that piece of music, we attribute a large part of that Art to the Creator.

We immediately associate the people behind it to create it how it was created. It's Hans Zimmer music, it's created with this particular mindset.almost like the way it was created is part of the art artifact itself. Which is true, which is genuinely true for many, many, many authentic things, right?

the outcome is not packaged as a single mp3 or a WAV file, or a Spotify listen, or a YouTube video. There is so much that happens in the effort involved in it, that that comes part of the artifact that is gone. There's zero involvement that right now zero.

Dinesh: But how can you say zero.

Like why kind of contest that 

Arjun: I'll play you a piece of music right now, 

Dinesh: You tell me if it's not good. 

Arjun: the reason why I'm saying that is you'll reach a state where, Hey, this is good music. 

Dinesh: hmm, 

Arjun: Okay? That part I can guarantee.

The moment it was, there is a immediate reaction when, you know it is AI generated or generated without much effort. that diminish the outcome? one question that we have to deal overall around these artifacts.

Dinesh: So, this is a very deep question, we will probably come back at it with layers.

Comparison with 3d animation

Dinesh: Yeah, I have a parallel analogy to this 

animation, right? there are different types of animations. Like classical animation is where use sketch every frame. 

Arjun: hmm.

Dinesh: and then came this, 3D animations, which, 

Arjun: Very good. 

Dinesh: pic 

Arjun: I like where you're going with this. Yeah. 

Dinesh: Yeah.

With pixer and things like that. and then we got drawn to that. Right now, classical animation, does not exist in mainstream so much.

So at least I think this is my, sort of my assessment or hypothesis 

Yeah.

and if you look at classical, animation, it is very effortful. And I personally like watching those movies, right? they come with a certain, uh, philosophy.

Like things will be stretchy, the eyes will bulge, they'll read this rubbery feeling and it defies gravity to some extent. you know, all of that.

let's say you are watching a cartoon, which is, frame by frame sketched, 

and you like, it likes it. And the way he describes it is that the eyes are bulgy. It has got a little stretch to it. you'll see a certain, aberrations, which kind of signature style to that form of, animation.

Arjun: Now your statement was, I like frame by frame. What do you call that? 

Dinesh: Classical animation. 

Arjun: Today - Create an animation style that creates stretchy, bulgy eyed animation. In the form of this is a text prompt, 

Dinesh: right? 

Arjun: right? You can create the outcome that Dinesh wanted in an animation much faster than the classical animation people did it.

Dinesh: Right. 

Arjun: And when Dinesh sees it again, because it is highly accurate, the Dinesh might not know the difference between how it was made. 

Dinesh: Yeah. 

Arjun: Right. 

Dinesh: Yeah, that's that, I just want to go with another example on this. I are, they may stop motion animation on, you might have heard 

Right. 

Arjun: capture frames and then you 

Dinesh: it together 

Yeah and that's a very laborious work. 

Arjun: Absolutely.

Dinesh: and you can, even make something look like a stop motion animation, I guess. with technology, you can make it look like stop motion animation. but that's not the important point. Stop motion animations. they have a certain character to it again, right? They look real, they are animated. But if you freeze a frame, what you see is realistic.

It's not sketched, or it's not generated. it's not a computer rendered image. It'll be like, okay, you have kept a clock and then you have kept a thing, and then the clock is cotton mouth and it seems to talk. and then it has a little, jag, um, what's the word? Um.

Choppy feel to it. Those are all the, features of a stop motion film. And you do fall in love with the stop motion film and that whole format. And then once you watch the movie, you go and see how the movie was done. And then you see these beautiful sets, which these guys make. And then there is all these tiny, tiny face expressions.

They would have a, like, if you have a character, then you will have a box of faces for that character. Eyes perusu aagara maari, Sogama irukkara maari, Mooku idhu agaara maari, no, all of that. And they keep SW swapping the thing and then they do this, it's a very miniature, very dextrous job. so to answer the question, which you asked, like, is it for the effort or is it for the outcome? I have a feeling. it's not just one camp, always. There are times when in an artistic work you appreciate, the effort more than the outcome, and then there are times when you, maybe you don't care so much about the effort that went to it, but you just, you like what you're seeing or hearing.

Hmm. 

 Say something controversial so that I can take a side on it. 

Arjun: The creator does not matter.

Creator does not matter Spacer

Arjun: The creator does not matter.

Dinesh: Okay. I say it'll matter,

feel the creator does not matter. when you look at creations for what it is, you only observe, and suddenly you wear this hat of, Hey, I'm so inspired by it. I want to know more about it. Or, a part of my appeal comes from knowing how it works, right? Some things are magical because you don't know how it works.

Arjun: The moment you know the magic, it does not, it's not appealing anymore, 

Dinesh: Hmm, hmm, 

Arjun: I mean, when you break down the magic, then you appreciate for the skill. You don't appreciate for the awe of the art,

Dinesh: Yeah. So there's that. Yeah. So magic I think will be a little, it not 

on 

Arjun: the spectrum 

Dinesh: But even if you take, say for example, this is one famous, silence track. You, you're aware of that. 

There's this guy who comes, famous artist. It's there in, 

music and all of 

Arjun: that 

Dinesh: The track is nothing in the track. 

Arjun: Hmm Okay. It's just silent. So this guy comes, opens the, his, piano that whatever, shutter and things like that sits in the table for like three minutes then he goes off.

Dinesh: 

Arjun: Mm

Dinesh: his art is that silence. 

Arjun: mm. 

Dinesh: So. is this silence different from any other silence? 

in that, then in that case it doesn't, make sense. but we, as people, I think that the human nature wants to relate to art, just as an artifact. We somehow, we want to learn more about it. Like if you are connected, that's why we have any form of, say movies or say Game of Thrones or anything.

We have this fan following. We know who's actor, what happened, what's the story? Trivia. There's so much of background. Uh, say Harry Potter, like fan theories. See, they are the, you know, there's so much that gets, created. 

I think it all stems from the idea that, okay, if I can relate to this work of art, then I would like to with it deeply. 

Arjun: Engage with it deeply. 

Dinesh: Huh. 

Arjun: that's a very good word to say. 

Dinesh: Yeah. So that engagement, I think so far it has been to figure out the process, maybe know more artist and things 

Arjun: that 

Dinesh: But now I think it can be a little different. 

Arjun: wonderfully put, I think that is a core premise of this episode also, that the stages of engagement of the art will change. You look at it with certain, certain neutralize You, you listen to it, you hear it for the first time.

A sense of appeal happens, a sense of awe, wonder angst the emotional arc happens, and then you reach a certain point. there has been a time, at least for two years. In the last 15 years, we have listened to only one song for the entire year. I would not listen to anything else. 

Dinesh: mm-hmm 

Arjun: Yeah, I listened to only one song day in, day out. And it did not saturate on me. 

 the newness in the track came to me in the hundred and seventh Listen, and the 256th listen. Right. they used to be that cycle to it. For me at least. This is, I'm normally talking from my personal experience, right?

 I can serve certain confidence there, there could be people like, like me who are doing this right.

I think we need to baseline this thing. How we engage with art stages has changed. that's the conclusion of this entry of AI for us.

Intermission 

Dinesh: When I first began studying how these AI systems work, there are two words that keep returning to me, approximation and repeatability.

Arjun: We have said this before, AI is not intelligence as we understand it to be. Its intelligence approximated. It looks close enough and behaves close enough and produces results close enough to what we think a skilled human might do, think or act.

And close enough is where you start reacting emotionally. You nod, you feel, and you believe, and that's what is happening in music as well. The models that are used to create music, they don't understand the art. They approximate it. They have studied millions of patterns and learn how to recreate the, the idea of how someone would feel.

This is somewhat like watching Ben Kingsley play Gandhi a performance. That feels true even though it's, it's an interpretation and it's not life itself.

And then there's, there's the other word, repeatability. If, if you can look at approximation. As about getting close, repeatability is about consistency, and that's where AI behaves like lightning. They say lightning never strikes the same place the same way twice. You ask the same melody to be generated the next time it'll give you something else.

Well, art as a funny way of looking at imperfections or approximations or lack of consistency, even in that inability to repeat itself, even in that inability to be consistent. There's a sense of beauty that all of music that is generated is kind of in the moment

 being able to come back to the same idea and refine it and, and, and make a perfect version of it. It, it kind of gives you in finite variation, not about the same song, but of the same idea in many different ways.

And somewhere in that difference, in, in that almost familiar place, it opens up a new question that we are all trying to answer, which is, if. Art, if music can be endlessly approximated, what does originality even mean anymore? 

AI is not repeatable

Arjun: the reason I landed up with 600 songs is that I was not able to get to every aspect of the song that I wanted.

Dinesh: what do you mean by that? You mean you were not able to resolve it to the finest detail the way you wanted.

Arjun: Yes.

Dinesh: That's because when you went to change the second time, something else changes. So Is that what you're saying? 

Arjun: Yes. Not only changed, it changed the entire meaning of the first piece itself. 

Dinesh: Hmm 

Arjun: Right? From a, from my musical, goal point of view. Right. But it would tease you just enough to show that, hey, this could have some more potential than what you had thought of.

Dinesh: Mm. 

Arjun: I was in this rabbit hole of things. It would tease me just enough to know, oh, if this is possible, that could be possible. But then when I go after that, oh, but this meant off. It was already good here. And this is true for code generation. This is true for image generation. This is true for video generation.

This is true for music generation. AI per se, itself is not repeatable.

Dinesh: what's the reason behind it?

Arjun: 

it's a huge thing. when I say huge thing, some part of it is designed by intent.they have very short memory of what you spoke to them. You don't have the context of it. The historical context and the flavorsare, are less. Okay. It knows what the next thing to do incredibly well, but it does not know the history.

Dinesh: And for every current state, there are many possibilities of the next state.

Hmm. 

Arjun: Right. And because it chose the next one, state it need not choose the same next state every single time that arc will keep changing, which is why it is like all random, variable based, very modeled approximation of things.

Is everything an approximation?

Arjun: Any form of music generated by AI is an approximation by design.

It's not the art, it's the approximation of that art.

unless you say that approximation is the totality of everything. 

Dinesh: Hmm. 

Arjun: And this is coming from many, many angles I'm a keyboardist as a musician, 

Dinesh: Right 

Arjun: My, tool for the trade is a keyboard. A keyboard by design has several tones and things that it's not supposed to be having.

It's by design and approximation of the rest of the world, the music world, 

Dinesh: Correct. 

Arjun: right? And that tool is given to me and said, hey artist, can you make something out of this? 

Dinesh: hmm. 

Arjun: Okay, world begins there. 

Dinesh: your entry point is there 

Arjun: Entry point itself is there, I don't know to play the piano. I don't know. When I say, I don't know, I don't know.

To play 

Dinesh: in The purest way, 

Arjun: The purest way, right? I don't know. Know to play the saxophone, the bras, the bass. I have been in places where people have hated me because I have a, I'm a keyboardist and I didn't want me to join the bands, I felt this okay, and my tool for the trade is an approximation.

Today, AI by definition, is based on models and model by definition is an approximation of a real world phenomenon, right? So while we may have gotten accurate, we have got approximately accurate in all of this. While it might make you feel the wonder of an artist created music and what AI created music is, that means.

Your trigger to say Wow has been quite approximate. It's not pristine.

Arjun: We attributed celebrities. 

Dinesh: hmm 

Arjun: award-winning, Grammy winning creations of music or other forms of art in their pinnacle, award-winning categories for the artistry or something else. And the scale has changed now.

Dinesh: hmm 

Arjun: You have auto tune that came in the industry many years back, So you always had tools that would do more than what a human can do, given to certain people who have the access and the talent to use those tools. Now, suddenly there's a tool which is given to everybody, right?

Business model is the problem

Arjun: The music that you, Dinesh that you have claimed that you don't know to make music in this episode, you will create better music than me if I played the my keyboard. Okay? Now is my battle with you on basis of the tool It's when you decide that you are a better musician because you have this tool, and then come in the way of my livelihood as a musician, then we have a different debate.

It's the business model debate that we have. It's not the tool debate at all, I'm saying

the way I earn money. If it is threatened by the way, you are now having access to ai, then it's the money problem. It's not the AI problem,

right? So I think we should flavor that problem in a separate debate because we have out ways to make money out of music, movies and all those in a certain way, 

There are distribution agencies, there are labels, there are record companies, there are people who make contractors, artists, they sell the music, they make the money in the business model designs, all of these are already extremely challenged right now.

They to change, and I'm off the premise that they have been wrongly footed for years now.

Probably with AI for the first time, we may have a chance to correct it.

Dinesh: Oh, okay. You're coming from that side. 

Arjun: Yes, we may have a chance to correct the way business models have been designed for this.

Dinesh: So at the foundation level, if you've, look at the business model, right? I, I, I don't know how music industry works, but if you have to like, break it down to 

Arjun: I make music, I release it I get a distributor, either a label or some form of a service like cd, baby tune, code distro, all of this. I give my music almost no review happens on my music. 

Okay? hmm. 

Almost no review happens on my music. And if at all there is a review, it is extremely, biased, highly biased reviews that happen.

And I have case studies to, to talk about that maybe later. But once the play hits the distribution platforms, my music is now available on Spotify, Amazon, and all the big players who are all digital platforms who have recently come up throwing the CD business cassette business all out of auto shape, right?

Dinesh: Hmm. 

Arjun: And I say recently, in the last 15, 20 years, right now, the money I make comes from plays that happen online. The number of streams, any playlist that I've adopted, my song, and any covers and royalties other people have taken.

And what it means is that I need to get money for the music I make by attributing my name to the music, which means it's a copyright thing. Right now, that world has also changed. The fundamentals of copyrights have changed is my view on this. And when I say my view, I'm coming at it from not so technical backing, but at the same time spend my time doing this study in my own way.

I am of the belief that the foundations of copyright were always wrongly footed. Foundations of making business out of art forms are always wrongly footed.

And probably this is for the first time, we are put in a very highly conflicting situation, which will throw a lot of people out of lives, 

which is almost a corrective mechanism. And that's the view I, I'm likely to take. 

People making, trying to make a livelihood or doing this one thing that they have learned all their life, right? If somebody spent years and years learning only Excel, and that is now automated, what do they look like?

Their life? Does that mean that we don't automate that? No. We gotta find some other mechanisms there. So that is a very hard problem, but it's a solvable problem, and it's a self-correcting problem. Over a period of time, it'll change. thing is what you said last, right? How can you create AI art?

The word "AI Art" is a very important aspect in the modern, language itself, right? it'll soon try to only prove what I'm trying to emphasize that the creator does not matter. Will you now say ChatGPT art or Gemini art or copilot art, or this art or that art?

So far in the very early stages of ai, we are now blanketing AI as one ai, then it's a battle of services in the ka next next Generation loan, it'll be, oh, that service is better than this service andha maari aidum So that's one aspect.

Third thing is now the art has become very personal and it is how it is supposed to be.

Dinesh: Right. 

Arjun: with the 600 tracks that I have. I cannot tell you how therapeutic it was for me. I have reached places with this music where I would not reach myself,

right? the patterns, the music, the things that it, it was able to do with things that I wanted to do is incredible. Right now, it has opened it. It has basically broke my inflammation in my body. It's like that, right? It, it broke new angles. It gave me new perspectives. I made music for me for the first time, right? I used to play music. When I played music, it had very physical limitations of what I could play right now. I was able to idea, I was able to travel faster. I was able to travel faster from one idea to another. I made tracks. These are, I will not release these tracks.

my artistic integrity of all of this. there's an ethical angle that I've super imposed about all of this over and above, right? This is going to be personal. This is going to be extremely personal. my work with AI as an art, as a tool will be like a tool, how I use it as a tool, and that tool comes and shapes a certain fine grain aspect of what I was not able to deliver as an artist because I I don't wanna be evil to my trade.

Also, like when I say evil, if I just did something else and tried to fool everybody about, that's not something I, I'd like to do right now. But there are people doing that. There are incredible amount of people doing that. And the latest news that I have, is that several established artists, several established big, high revenue, high stream artists, can be super easily thrown off all distribution platforms by what are known as AI playlist bots.

So what you do is basically, let's say you are AR Rahman, and now your music is of a certain caliber, certain kind, and you're already of a certain stature. And now I could take your music, put it into shitty, playlists across the world and report them, and it would report the artist. And soon AR Rahman one can be wiped off from all distribution platforms because

Dinesh: When, when you say shitty playlist and report them, meaning what, what does that mean? 

Arjun: I could create a fictitious playlist, AI generated playlist of all bad music. And report the, each of the artists as spam creators, bad music. Like there, there are ways to flag artists, flag content, 

Dinesh: Ah, 

Arjun: when you attribute all of those, you are essentially now attributing it back to the artist.

And today the business model is if something is flagged, it doesn't take much for these distributors to flag your entire artist profile. 

There's no conversation involved in all of this, 

Dinesh: Hmm. 

Arjun: Right? 

Dinesh: Which is why, thus the sanctity of all distribution providers right now have changed. There are new platforms that have now added the AI generated music as part of their content policy. We will accept AI generated content if it has this conditions.

Arjun: There is no regulation around all of this, which means we are having the luxury of a conversation.

There are so many people who do not know that this is what is shaped their way of approaching art has been very traditional and they'll continue to react to these forms of art in a very different way. It's like having a AI girlfriend, 

Dinesh: Hmm. 

Arjun: You don't want to be in the place where the girlfriend breaks your heart.

Yeah. What I like about this, what we spoke is that, somewhere it feels that we are going to be empowered. So we have some, leverage in the game.

The leverage comes after the hard acceptance that, we've got it wrong. We, have to let go of something to get the new leverage. The old leverage is not there at all, 

Summary

Arjun: Alright, for someone who Who listened to it partly, or maybe they were driving, they didn't like fully pay attention to it, and then they just want to get a crux of, okay, what is it that this episode was about? So how would you close it? 

If you are a musician listening to it and you didn't listen to it, please park the car on the side of the road, go back to the beginning and start from the beginning, is where I would say. For everybody else who are enjoying music. when you hear a piece of art or when you see a piece of art,

appreciate the art form for the artistry. It is intended to be less analytical because everything that you have analyzed so far, the parameters of that is changing. We will get back to you shortly once that parameters have been defined.

So basically chill and enjoy right now so that let the world settle and your law about how to look at art. If you are a parent whose children are now studying animation or any other fields which are going through all of this. Seek a career counselor once, because we may not be the experts, but life in all of these artistic areas have changed.

If you are a live performer, if you are a per person who's intending to perform, probably there is a, longer shelf life for you, so keep at it.

Outro

Arjun: So, where does this leave us? Right. AI can now compose, perform, produce, music that even sounds real. How do we as listeners and creators stay connected to the art itself?

For a long time, I also used to feel, hey, is is this slipping away from us? When I first heard about AI generated music and before I had the chance to fully absorb all of this. It felt too easy. It felt too quick, and it was discrediting the traditional sense of interacting, feeling, and making music.

Now, I've completely shifted my perspectives after having spent time with over 600 tracks. And I say this with a certain set of observations. Number one, they say that music or art generally evokes a reaction, a thought, an action. And if AI generated music is able to do that, then good for the listener, 

Two, if. AI generated music helps me as a creator reach a place which I wouldn't have reached as a creator without the tools or the techniques that are available to me or the limitations of my skillset, then great for the creator. three. AI is now really exposing the flawed business models of arts, and specifically business of music.

There's probably too much attributed to the creator and not the music itself. Maybe this is the point where we begin to start to listen and engage with music differently. Not to find who made it, but notice how it makes us feel and to stop hunting for authorship and kind of return to feeling the experience.

It's not about streamings, it's not about the play counts. Somewhere. The metrics of the business is probably not relevant and we might need to take a re-look at all of this. And it's not like we have a solution yet, but it's important to question this. Maybe This is where our work really begins.

It's about learning to engage with art, not as something we own or defend, but as something we share and grow through. it's about to collaborate with the tool and not compete with it. Because even if the method changes, the reason we make music hasn't changed, the reason we listen to music hasn't changed.

 We still create music in ways to express, we still sing, to express and recognize each other. And somewhere there in, in that space where we are able to express and create that, that part is still human.

It's still us. And the spectrum just got wider because of AI generated music. the change is here 

 As always, this is just this inflection point of music and many things have already happened since we first drafted this episode, and we will be covering more of them in the upcoming parts of this series. If you like this episode or if you'd like to talk to us, please do reach us. At listen@innocencetheory.com.

Share your thoughts, comments, feedback. We'd love to hear them all. We'll feature all your thoughts on the episodes. Thank you for listening. See you on the next one.