Michael Walker: If you’re listening to this then you likely already know that being an independent musician is a lonely road. And maybe your friends and your family don’t fully understand why you do what you do, or why you invest so much time, energy, and money towards achieving your music goals. And especially early on, it can be hard to find people who really understand what you’re trying to accomplish and how to make it happen. So, that’s where Modern Musician comes in!

My name’s Michael Walker and I can understand and relate to that feeling. I’ve been there myself, and so has our team of independent artists. The truth is that basically everything good in my life has been the result of music. It’s the reason I met my wife, my 3 kids, it’s how I met my best friends. And now with Modern Musician, we have seen so many talented artists who started out with a dream, with a passion, without really a fanbase or a business. And you’ll take that and turn it into a sustainable full-time career and be able to impact hundreds, maybe even thousands or millions of fans with your music. We’ve had thousands of messages from artists who told us we’ve helped change their lives forever. It just gets even more exciting and fulfilling when you’re surrounded by a community of other people who get it, and who have shared their knowledge and success with each other openly. So, if you are feeling called into making your music a full-time career and to be able to reach more people with your music, then I want to invite you to join our community so that we can help support your growth and we can help lift you up as you pursue your musical dreams. You’ll be able to interact in a community with other high-level artists, coaches, and industry professionals, as well as be able to participate in our daily live podcast, meet these amazing guests, and get access to completely free training. If you’d like to join our family of artists who truly care about your success, then click on the link in the show notes and sign-up now!

Edward Balassanian: Let's say you're a vocalist or let's say you're an instrumentalist and you really don't want to go on to hire a studio, hire a producer, get a bunch of accompanying material created. Now you can just have the generative music service. Fill in the blanks around your beautiful vocal or your amazing chord progression. That's essentially what you do when you hire a producer anyhow, right? It's kind of a black box in a room. They make a bunch of sounds for you around your melody and they give it back to you. Think of that as the generative music service now. So that technology is working today and we're going to be releasing that very soon to the world so that producers can use this as a service to augment their workflow.

Michael Walker: It's easy to get lost in today's music industry with constantly changing technology and where anyone with a computer can release their own music. I'm going to share with you why this is the best time to be an independent musician and it's only getting better. If you have high-quality music, but you just don't know the best way to promote yourself so that you can reach the right people and generate a sustainable income with your music, we're going to show you the best strategies that we're using right now to reach millions of new listeners every month without spending 10 hours a day on social media. We're creating a revolution in today's music industry and this is your invitation to join me. I'm your host, Michael Walker.

All right. So I'm excited to be here today with Edward Balassanian. So, Edward is a tech entrepreneur. He has over 25 years of experience. Currently he's the CEO of a company called Aimi and Aimi is a groundbreaking music platform that's leveraging technology to be able to create generative music. He has worked with startups like BCom Corporation, B Labs, Vital Abuse and Strings, Inc. and is the inventor of over 80 patents related to different technologies. And specifically today, I'm really excited to connect with him and have a conversation about the future of music as it relates to generative platforms and AI, and to hear a little bit more about Aimi and for all of us who are here right now, who are musicians, there's different ways of looking at what's happening right now with AI in particular with generative music. And I think for a lot of people, there's a sense of fear or a sense of, I don't know, resistance or sort of this denial, like: Oh, it's not like that impressive or, Oh, like it's, it's pretty bad cause we're sort of afraid of a future that might involve more… I think the fear is that music we might lose our creativity. We might lose it because of “the machines”. And so excited to have a conversation with you today, just to hear from your perspective, how can we align with this technology and use it in a way that actually amplifies our creativity and helps us all connect more. So Edward, thank you for taking time to be here today.

Edward Balassanian: Thanks, Michael. Pleasure to be here.

Michael Walker: Awesome. So to kick things off, I would love to hear just a little bit about your story in terms of your background of how you started Aimi.

Edward Balassanian: Aimi technically the idea is pretty old. I had the idea back in 2015 and it was just a moment watching a really gifted live DJ, basically producing live. And just watching this individual craft this musical journey in real time while taking into account not only the story that he was creating, but the audience and the reception to the music. And this went on for four or five hours and it was really mind blowing to watch. And part of the sense of fascination I had with this was how much of this journey that he created was this combination of engineering and music.  And that's something that really struck me that there's so much engineering in music making, especially when you talk about this new digital frontier for music making, and that was really the genesis for Aimi was helping creative individuals do more creatively without being bogged down with the engineering side of music making. 

Michael Walker: Interesting. So when you describe like engineering versus the music, can you describe a little bit more about what you mean specifically by like engineering as it relates to the music standpoint?

Edward Balassanian: Sure. But if you look at your traditional DAW, there's a lot in the interaction between you and the DAW that has less to do with music and a lot to do with the interface of the DAW and putting together sound and engineering sound and creating arrangements. A lot of that is formulaic. I know a lot of it is also creative, but there's a significant part of it that is formulaic. And if you remove some of the burden from the user/the creator having to do things routinely, it frees them up to do more creative stuff. So they don't have to get bogged down in the routine. And that's really the goal of technology, right? The goal of technology is to remove the burden of routine from humans because we thrive in our creative spirit and our mechanical robotic side that we can train ourselves to do that isn't what inspires us. And certainly isn't what separates us from machines. 

Michael Walker: Super interesting. Yeah. That's a great point that DAWs and a lot of technology in general is like a tool that we can use to sort of express what we're trying to express creatively, but it does seem like for a lot of us that's the thing that gets in the way. We have this vision, we have this creativity and it's like: ah, I don't know how to use the technology or how to express myself through it. And so being able to more directly interface between the artists and their music and their ideal outcome without necessarily needing to handle all the engineering stuff in the middle ground is really interesting. So tell me a little bit more about Aimi and how does that tool work exactly? Like what kind of generative music is it making, and how does an artist interface with that tool?

Edward Balassanian: Well, let's back up a bit and look at where we started. Cause when we first started, the name Aimi is the AI Music Initiative. So the genesis of the company really was around this idea of bringing artificial intelligence to bear for music making. Keep in mind, generative music has been around for a while going back to Brian Keno and even in your DAW, you've got tools to create drums or other instruments that are generative. So this concept isn't new. It's really the degree to which music making is going to be transformed over the next decade to where a lot of the tasks that you do as a musician will be automated. And again, just going back to my point earlier, the more you remove this need for the human being to be the automaton or the automated piece of this, the more you can free them up to be creative. So the way Aimi works, when we first started, we did look at training monolithic neural nets on finished music and you can get a neural net to make music. There's no question about it. You can see a bunch of examples of this out there right now, but it's really fraught with a lot of problems that approaches. First of all, there's a data set problem. The good stuff is all copyrighted. And while it's a little ambiguous right now whether you can use copyrighted material or not, it's kind of unambiguous ethically, whether it's okay to take somebody's creative work that they've spent their human capital or their creative capital making, and then to usurp that and essentially turn it into this amalgam of people's creativity. In many ways, a neural net is going to give you the most common denominator, right? So it's going to essentially dumb down the creative elements and give you the most common denominator. So we didn't like that model. So we moved away from that pretty quickly. And instead we approached music making the way producers do. When a producer makes music, it's much more than the finished sound: It's the way the music is made and that has a big impact on that finished sound; the way you arrange samples, the effects you apply on those samples in the arrangement, the way you master the entire combination, all of that feeds into the music making. An analogy I often use is: a chef and a bowl of soup. If you walk up to a bowl of soup and taste it, it's hard to recreate that recipe. The recipe is so much more than the finished taste. It's the ingredients, the quality of those ingredients, what happened to those ingredients along the way. So Aimi makes music the way producers do and to facilitate that, we created a programming language called Aimi's Script. And this is essentially a programming language for generative music. Our vision has always been to release Aimi's Script to the world so that anybody can build generative music apps. And we're on our way to being able to do that right now.

Michael Walker: Cool!
Edward Balassanian: Essentially what it does is it gives you the ability to express music as code. Because it is code, you can be as declarative or as imperative as you wanna be. Meaning you can get as specific as you wanna be about how Aimi should mix, master, and arrange and produce music. Or you can be less specific and let Aimi kind of just do its thing. Either way, one of the benefits that you get from this programming language is the ability to have steerability in the music making, which is another issue that we had with monolithic neural nets. As a creator, as a musician, I mean, it might be interesting to prompt a neural net and get a bunch of music out, but good luck trying to change the fourth bar, or trying to make the chord progression a little bit different than what it was. It gets very, very cumbersome to steer these neural nets because it's difficult to reason about their output. In fact, if you give the same prompt repeatedly, you're very likely to get very different outputs. You can see this when we go to ChatGPT. So all of that kind of pointed us towards a programming language instead. And what we do is we use machine learning to generate the plan, if you will, the script. So we're using machine learning to generate Aimi's script which in turn, generates the music and a human being can also use Aimi's script as well. So for those who don't want to write code, we have ways to avoid the code writing. For those who do want to write the code, you can jump in there and use Aimi's script to create a full featured generative music hack. 

Michael Walker: Wow. That's so cool. So yeah, you're developing it kind of code first. And so that means that you can develop apps around it, but also you can create interfaces to be able to interface with it on a more human level. I'm curious, I mean, just a few days ago from the time of us recording this right now, there is a new model that OpenAI just announced called Sora. And for anyone that hasn't seen this yet, this is a total mind blowing release because now they have generative video that is very hard to distinguish between what we'd normally consider realistic versus AI generated. And I thought one of the most interesting parts of what they shared with the technical breakdown was just levels of different compute impacted the end result. Just at scale, they're seeing these emergent properties whereas like the more compute that you had it just almost magically got better. They don't even understand where these emergent properties are coming from, but they showed one specific example of a video of a dog and it was like 1X compute. And the dog was all warbled and it didn't look like a dog. You're like, okay. And then at 4x compute, it still looked a little bit weird, but you could tell it was a dog. And then at 16x compute, it was like extremely high fidelity. You could tell that this looked like a real dog. And the implication was just that at higher scale with more compute, these AI models get significantly better. And that is even without any other breakthrough, just like more compute, the levels of intelligence start to multiply. And so hearing about generative music technology, it feels like it's gotta be pretty similar to other forms of media in the sense that: it's creative and it's like it's generative and that ultimately levels of compute are going to make a huge factor where you just want to have as much data as possible with this type of music. So especially with what you mentioned around copyright issue and training the data… this is an important conversation, important issue for that we're sort of figuring out right now is: what should be copyrightable and is it wrong for machine model to train based off of copywritten material or not, the same way that a human can be influenced by other material, but create something that's totally original and totally new.

Certainly OpenAI is completely trained up on a lot of data that is probably copy written, but it's just hard to draw the line of what's original versus what are the ethical boundaries. I'm kind of rambling here [laughing]. I'm talking too much here, but I'd be curious to hear your perspective on, on everything that I just shared in terms of where we're headed with a music model and your guys plans with Aimi to be able to create a model that ultimately is meant to serve the creative and allow them to express their vision with less friction. 

Edward Balassanian: Yeah, I mean, look, our vision is very simply to turn generative music into a service, and we're providing an SDK and API and a bunch of tools that allow anybody who wants to incorporate music as a service into their application or product or platform to be able to do so easily. And we want this service to be able to generate music on demand, meaning it's not recorded content that's just being served up. It's music that's made from scratch on demand to purpose. So if you've got a video that's five minutes and 30 seconds long, you'll get audio that's five minutes and 30 seconds long. That kind of service is difficult to conceive of in the context of copyright because every single time it generates something that's unique, it's different each time. And there's a big debate right now about whether you should even be able to copyright something that was machine generated. Copyright was originally created to protect humans, not machines. So this idea of a machine being able to create an infinite amount of music that's never repeating, always unique, and then copyrighting that seems kind of antithetical to the whole point of copyright from our perspective. We're not in the business of selling content. So when you use Aimi, we're not charging you for the music that's made. We're charging you for the resources that are consumed to make it. So it's a SAAS model and we don't care what you do with the music. You can download our consumer app right now, which is the first app that we built on this platform, free, and it's going to stay free. And we don't really mind if people use the audio for other purposes. The one thing that we don't want to encourage is people take the audio output of Aimi and copywriting it and making money off of it counterproductively from the original goal, which is to incorporate this in as a service. Having said that, there will always be a place for recorded music. If you look at the history of recorded music, it was to capture the essence of a human artist. We recorded it because the artist couldn't be everywhere performing at once. So we tried to record it, protect it, and then resell it again and again. You don't need to do that with machine created music, right? If you have a TikTok video and you need 30 seconds of music, that's the thing, let it generate the music for you. If you want a big hit, go talk to the publishers and the labels about getting Drake or The Weeknd into your TikTok video because they recorded an artist and that's a very different medium, I would say, than generative music needs to be.

Michael Walker: super interesting. Yeah, it definitely seems like the realm of music licensing and sync licensing is one that in particular, this type of generative music could be such a lifesaver, such an amazing resource for a supervisor or someone to be able to soundtrack quickly. How does this tool interface with existing artists? Like, is there any sort of integration for if someone wants to create their own AI model of that represents them as an artist and can create music in their style or in their genre, or: great music that's influenced by X, Y, Z artists… How do you play with that sort of integration? 

Edward Balassanian: Well, the service itself essentially mixes, masters, arranges and produces music in real time. So we've got this vault of samples that we've created and that we own wholly. And when you interface with this system, let's say it's over the API, you can pass it in text prompts. You can pass it in a video and you can also pass it in a sample or multiple samples. In the case of multiple samples, we’ll build a song around your samples specific to a genre that you pick. Now, we've already written the scripts for dozens of different genres, and that is growing quickly so we'll have over 100 genres supported which span everything from indie-pop and hip hop to reggaeton and the piano and house and techno, et cetera.

Now, if you're sufficiently motivated, you can use any script and create your own scripts and create your own articulations of the mixing, mastering and arranging. And the neat part for you as an artist is… Let's say you're a vocalist or let's say you're an instrumentalist and you really don't want to go on to hire a studio, hire a producer, get a bunch of accompanying material created. Now you can just have the generative music service fill in the blanks around your beautiful vocal or your amazing chord progression. That's essentially what you do when you hire a producer anyhow, right? It's kind of a black box in a room. They make a bunch of sounds for you around your melody and they give it back to you. Think of that as the generative music service now. So that technology is working today and we're going to be releasing that very soon to the world so that producers can use this as a service to augment their workflow.
Michael Walker: Wow! That's so cool. So you can upload your own vocal track. You could [scatting] skip bop scuba do bop and then we could upload that as a track and it would actually build a bunch of additional tracks around it, or how does it work exactly? Does it return like one audio wave audio file? Or is it like recurring stems that you can play around with? What does that process look like?

Edward Balassanian: So the bulk of our efforts with machine learning was really on the ingestion process. So when you upload that vocal, we'll analyze it and we'll extract a bunch of different properties from it and we'll determine that it's vocal, we'll determine it's male, we'll determine what frequency space it occupies, what the beat density is, what the tempo is. What instrument it is, if there is an instrument and all of that will get fed into a script that's specific to a genre. So let's say you said: I want this to be incorporated into a deep house mix or maybe indie pop or jazz or whatever. So that'll then determine the set of complementary sounds that we are going to use to arrange around your vocal, and we'll then be able to show you a plan file that essentially shows you the breakdown of what the song structure is going to look like. You can also choose how long it's going to be, say, 7 minutes long or 5 minutes long. You can see that there's going to be an intro, a verse, a chorus, a middle, a bridge, all these different sections. You can get in there and tweak them. You can change properties associated with them. Maybe you want some instruments. Maybe you don't want some instruments. Once you're satisfied with that, you hit produce and out comes the audio. And that audio can be delivered in stem format or just as a wav file. And you can take it and stick it in your DAW and do all the finishing touches on it. And this goes back to my point earlier which is, we're really here to enable creativity, not usurp it. So we want you to do those wonderful vocals all day long and let us do the hard lifting of adding in all the accompanying material, but you're still steering, you get a hand on that steering wheel and you can drive it wherever you want.

Michael Walker: This is so awesome. How about in terms of. As you're building these tracks, like how much of a say do you have in terms of guiding the direction? You mentioned that you can actually have text prompts and you can upload your own file. But could you say: I want an accordion with it, or I want a violin track or something like that. Would it be able to take prompts like that and iterate back and forth there? What does the process look like after you create a generation?

Edward Balassanian: That's the benefit of this being an imperative programming language, right? We can actually show you what we are planning. We actually call it a plan. So we can show you the plan that is going to be used to match what you've asked for and you can get in there and visually edit the plan. So you can specifically say in this bridge section: I would like a saxophone melody or a chord progression with a guitar. And now, you might be okay with any number of chord progressions that would be in that section. And we'll go and find hundreds of thousands of them and pick the right one algorithmically. And you can get even more specific if you want you could specify what key it's in. So the more specific you get, the more imperative you are, the more directly, you're going to determine what sounds get incorporated in there. If you want to just be surprised, then just back off a little bit on the dictation of what happens and you'll get some pleasant surprises, not just in the sounds themselves, but also in the arrangements. One of the things that we found working with a lot of artists is giving them back their music rearranged by Aimi, more often than not, what we hear is: I would not have thought of doing it that way. That just inspired me in a new way to think about arrangement. And we love hearing that. Taking the snowball, the Christmas ball, the Christmas ornament to see the snowfall again, it lands in a different spot, and it kind of inspires to do things differently. 

Michael Walker: Wow. I'm super excited about this tool. I can't wait to play around with it. So you said that Aimi has an API that is not currently available, but you're working on opening up so other people can build apps on top of it, or what does the landscape look like for that right now?

Edward Balassanian: Yeah. So the first couple of years was really getting in service to work.

And in the beginning it was like your neighbor kid in elementary school found the drum set [both laughing]. So it took a couple of years to get to the point where we felt like the music was good enough to release. And that was not our decision. We worked with almost 250 artists who are considered the sort of Preeminent producers in their genres. So we really wanted their thumbs up before we decided that this is something that we would be proud of. We wanted them to be proud of it as well. We also worked with them to provide a lot of the underlying samples in our initial data set. So that was step one. Step two, we wanted to build an app to kind of show off how this service works. And you can download that app today. It's in the app store. It's great. called the Aimi Player. And what we did there is each of the, we call them experiences, they're essentially like a never ending album, So each experience creates this multi-dimensional musical space full of all kinds of different samples specific to a genre. And then when you hit play, Aimi starts mixing, mastering, and arranging the music in real time on your device taking you on this musical journey. You can actually give it feedback by giving it thumbs up and thumbs down, and that'll help Aimi identify the combinations of sound that match your personal taste. All of this was built on the same API that we're going to be releasing. The API is going to be called Aimi Music Services, and this is essentially generative music as a service. It'll be available to anybody who wants to incorporate generative music into their product or platform.

Michael Walker: Very cool. Yeah. I would love to be connected when it comes out officially. I would love to develop on top of it both personally and, we have a service called StreetTeam that's like a CRM platform for artists. And this is the kind of thing that if we were able to build an integrationsWith Aimi… One thing that we talk about a lot is the strategy of creating these moments and bringing their fans along for the journey before they fully have the final song ready, but they’re creating snippets like melodies; hooks. And they're basically like sharing some early versions of demos before they come out. And I just imagine how valuable a tool like Aimi would be for the ideation process, especially just like starting out, creating those hooks and creating a moment that you could publish every single day on TikTok/Instagram reels. And yeah, really looking forward to hearing more about that when the API becomes available for it.

Edward Balassanian: We'll be demonstrating it at South by SouthWest. So if you happen to be down there. We'd love to host you. We have a dinner and I'm also going to be doing a talk at South by where I'll be showcasing… there's two versions of it: There's an API that's designed for applications, and we're also going to have a website that you can go to and you'll be able to have a conversation with Aimi and essentially have it generate music for you. That'll be free for a period of time. We'll let people just go up there and have fun with it. You can, you can prompt it. As I said, you can upload a video. We'll actually analyze the video. We'll identify scene boundaries in the video. And then we'll analyze each scene. Determine what the tempo is. And we'll find a matching genre for that scene, and so we can score an entire piece of video very precisely using the service. And then of course the samples, you can upload your own samples and have it build music around those samples.

Michael Walker: It's wild. Yeah. I mean, since ChatGPT first took the world by storm like a year ago, I've been waiting for, and I know that you've been working on this for not just since like ChatGPT came around, but I've just that kernel in my mind is like: I know it's not an if it's when are we going to have access to a tool like this that allows us to more creatively generate music as well. So to hear what you've built… I can't wait to try it out myself and have more of a taste for the exact you know, the, the way that you can iterate on it, but it just sounds like it checks all the boxes of the things that I was hoping that something like this would do and that you're part of the roadmap is building it in a way that we can actually integrate something like this into our app as well is extremely exciting. So super cool.

I would love to open up the floor here for our live audience, for any questions that people have that they'd like to ask Edward about Aimi or about any of the stuff that we're talking about today. So if you have any questions, feel free to click on your icon and raise your hand. You can come in here live.

Here's a great question from Steve Cuban. He is curious about whether Aimi concurrently or whether it's in the plans to be able to create music real time in a way that's dynamic that can like, imagine if we were playing a show right now or playing a song, could it learn to adapt and play with the band?

Edward Balassanian: Yeah, that's a great question. I'm so happy you asked that because the answer is yes. And it's actually the way our app works today. So if you download our app from the app store, it's making music in real time. And when you move the sliders in the app to add or remove vocals, or you thumb up and thumb down, or you hit shuffle, or you change the gain on different elements, that's interacting with the underlying code in real time. So your inputs are directly impacting the composition of the music in real time. That's one of the unique things about the way this platform works is unlike a monolithic neural net that takes a lot of energy to spool up and run, Aimi can run on your device. So it can be incorporated into a game. It could be incorporated into an environment where potentially you want the ambient noise level in a restaurant to not compete with the music and you can adjust the energy of the crowd based on the music that you're playing. Going back to the example I gave of the live producing DJ, that's what I saw him doing. He was in real time managing the energy of the audience. You bring them up, bring them down, bring them up, bring them down. And that kind of tension and release, as you musicians all know, is integral to music making, but it takes this interactivity. And that's a big part of our platform. And that's a big part of what manuscript does.

Michael Walker: Wow, man. So one idea that came to mind is just like imagining we're probably not that far off now from having like a Tesla bot or having some sort of humanoid form of robot. I'm imagining a band rehearsal back in the day. You know, there's like four or five of us and then the one person in the corner is a robot and it's Aimi. It's Aimi, the robot [both laughing]. Is it possible today for you to hypothetically have that situation where Aimi's like sitting in the corner, the rest of the band is jamming and kind of come up with some stuff. And like, Aimi is adapting and playing the song in real time or, or there's a bit of a prediction that happens. If you're all playing a song that you all know, like you all know when the verse is about to change into the chorus and you all know when you're about to fall in line with the next part. If you're in a band, sometimes it can be really tight. You can be so in sync with your brain waves that you can even just predict where you're going and go there together. I'm wondering, yeah: how would Aimi interact with a band making music like that.

Edward Balassanian: Well, a lot of, a lot of what you're talking about too, is beyond the oral part of it, if you will, it's also visual cues that band members might send each other. There's a lot to the interaction of human beings when they play music together, that is computationally difficult to describe. But if you can describe it, we have a full featured programming language, so you can build it. One of the demonstrations that we've built is having people jump up and down on different lit up areas of a floor and using a camera to see that and having that impact different elements of music coming in. So there's ways to create interactivity like that. One of the more important things that we've done also is: because we extract so much information from samples that we are using in real time to make music, we can actually have generative instruments play along with those samples using no recorded content. So we just basically use a sampler. We know the MIDI for this chord progression that's playing along, and we can have a bass line play along with it, just like you would if you were the bassist in a band, someone was playing a melody in a particular key or a scale. So you could play along that way programmatically as well. 

Michael Walker: Wow. Okay. I've got one more question because then after like, I want to just go dive into the tool and play around with it immediately. Cause this just seems so awesome. What exactly does Aimi output? Like if we are using it to create music, can you output like MIDI files from different instruments or different tracks or different stems? But yeah, could you remind me exactly, what does it return when you actually create it?

Edward Balassanian: Our output is meant to be music. So we're focused more on generating midi, but because we're making music the way a producer does, we actually have the full arrangement so we can output stems as well. And in most cases, we have the full midi for everything that went into making those stems with the mixed out audio as well. One of the things that we've worked really hard on is automating midi transcription. And you can imagine it's not trivial and even human beings screw it up a lot. So, we have best in class MIDI transcription right now, but we still need humans to correct it and make it better. We also do chroma key extraction. There's all kinds of different things that we're pulling out of audio to understand it because we have millions of these samples that we've got to pick from in real time to figure out which 10 will layer perfectly at this moment, given all the inputs that we're looking at. 

Michael Walker: Wow, man. I was just imagining… if you have that level of backbone or a programmatic way to express those sounds, then when you output this piece of music, you could actually output the different instruments like in the DAW. And then someone like myself, if I can play a keyboard, then you can actually play those instruments and add things or change things, even if they're things that traditionally you wouldn't even think of as instruments; this is just like a really cool sound that no instrument exists. There's no guitar, there's no violin that sounds like this, but we can play it like an instrument, man. Very cool.

Edward Balassanian: Just real quickly, one of the really cool things that you can do because of the way the music is made is we can move the different elements, the different instruments in three dimensional space. We actually have just been demonstrating this idea of the music video of the future, where all the different instruments can move around and in three dimensional space. If you download our app, you can actually solo out different elements. And we have people using our lo fi experience and they're jamming with it, or they're downloading our hip hop experience, and they're rapping along with it. So they can pull in and pull out these different elements 

Michael Walker: Dang! Yeah, the Apple vision pro just came out. Quest has been around for a while, but I've been playing around with the vision pro and it's pretty dang cool. And you know, the spatial computer…

Edward Balassanian: That’s actually what we built a demo for!

Michael Walker: Wow. Is the demo available on the vision pro app store? 

Edward Balassanian: No, not yet.

Michael Walker: Not yet? Okay. As soon as it is, I'm going to check that out as well. It's very cool. All right. Well, Hey man, this has been super interesting. Thank you for taking the time to come on here and share a little bit about what you've built with Aimi.

Edward Balassanian: My pleasure!

Michael Walker: And for anyone who's interested, it sounds like there's gonna be a South by Southwest speech that they'll be able to attend, which is awesome. For anyone else who's here right now, who is like me and wants to go play around with Aimi immediately, where can they go to dive deeper?

Edward Balassanian: You can go to Aimi.fm, which is our website. There's also a discord server. You can jump on in there. You can download our app from the app store and at South by Southwest, we'll be announcing, I guess I did already announce it, a new music service, which is this new web service that allows you to build applications using generative music, and I'll be demonstrating a bunch of those as well. So, if you're interested in getting kind of more hands on, we welcome you to come down. We have a really cool theater space in Austin that we've set up for daily shows and we're going to be doing these immersive visual and aural theater experiences as well using Aimi.

Michael Walker: Amazing. All right, well let's do a virtual round of applause here for Edward. Yeeaaahhhh

Hey, it’s Michael here. I hope that you got a ton of value out of this episode. Make sure to check out the show notes to learn more about our guest today, and if you want to support the podcast then there’s a few ways to help us grow.

First if you hit ‘subscribe’ then that’s make sure you don’t miss a new episode. Secondly if you share it with your friends, on social media, tag us - that really helps us out.

And third, best of all, if you leave us an honest review it’s going to help us reach more musicians like you take their music to the next level. The time to be a Modern Musician is now, and I look forward to seeing you on our next episode.