Bobby Owsinski: Artificial general intelligence basically means AI becomes so smart, it becomes human smart. Once it gets to that, then humanity's in trouble. So you have the alarmists on one side that are saying, oh, this is gonna be a problem. Then you have other people that are saying, eh, it's a while to get there. We don't, and we're gonna put guardrails on it and it's not gonna be a problem. So we'll see.

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 my friend, Bobby Owsinski. So Bobby is an OG in the music industry. He's a bestselling author with 24 books, ranging from the Mixing Engineer's Handbook, the Social Media Promotion for Musicians, the Music Business Advice Book. He's appeared on CNN, ABC News as a music branding audio expert. He recently produced and makes an album that appeared at number two on the Billboard blues chart. And he's had a podcast that's in its ninth year, over 400 episodes, focussing on the music industry. So it's awesome to have you here, Bobby, and I'm really excited to connect with you. And what I love about you too, Bobby, is the fact that you have this breadth of experience and you're still innovating and you're still learning.

You're keeping on the cutting edge and part of what we're about to talk about today is AI and music production. And yeah, I think it takes a mark of someone who has a high threshold for… You have a lot of perseverance to stay in the game for so long. And, in addition to that, you also have an open mindset to continue learning and growing and along the way. Thanks for taking the time to be here today. 

Bobby Owsinski: Thanks for having me, Michael. I have another explanation. It's ADD. 

Michael Walker: [Laughing] That's like the opposite of what I thought you were going to say.

Bobby Owsinski: No, I get bored with things quickly. So I have to keep on going to new things. 

Michael Walker: Oh, I hear you. So that's the part with keeping it fresh and new. I guess it's sort of a juxtaposition, because it's both you've been here, you've been in the music industry for, for a long time now, and, luckily enough for us, things evolve quickly enough that it gives you a thing to focus on next by next. 

Bobby Owsinski: There's always something changing. That's the cool part. I mean, you have to be a Luddite to not see that there's a new technology in the music business every four or five years. And sometimes even more than that, and the music business changes. So you have to. 

Michael Walker: Absolutely. So I think that actually leads us really nicely into kind of the core of this episode and where I would love to hear your perspective on someone who has been here and it's sort of witnessed these big moments of revolution, in the music industry where everything changed. And I'm curious, from your perspective, what does this revolution that's happening right now with AI, what does this have in common with some of the other revolutions and what are maybe some of the differences? Do you think this is a very similar thing that's happening and you're basically going to follow the same trajectory as some of these other ones, or do you think that AI, there's something uniquely different about it that could even lead to more of a breakthrough? Yeah, I'd love to hear your thoughts on AI versus the other big breakthroughs and what those breakthroughs were. 

Bobby Owsinski: I equated to when digital first came in: analog to digital and the pushback against digital was so severe. You had scientists that were saying that it would physically damage you if you listened to digital audio. And they had studies that would say that, which obviously, have been disproven. 

Michael Walker: [laughing] Guys, this is not good for all of us. Our ears are going to fall off.

Bobby Owsinski: Yeah. We have the same thing with AI now where there's a pushback and it's mostly from people that read headlines and don't really read. Or, actually clickbait headlines a lot of times and then think they know what's going on with it but really don't and as a result, they're scared about what might happen and I say might because there's no guarantee that something will happen. Everybody thinks this AI is going to make them less creative that it's gonna come and take their jobs their musical creativity, their will to work, you know all those things and none of that is true 

Michael Walker: Yeah, awesome. So it sounds like what you're saying is that one of the big fears that's come up in the past and in particular, the shift from analog to digital, there was a lot of resistance in people saying this is going to destroy everything with: i’s good for our ears. And something similar is happening with AI from a sense that people are afraid that we're going to lose the ability to create new things. What do you think is a more likely scenario for ai? Like how do you see it evolving for people? And as a musician who's maybe here listening to this right now, what do you think are some ways that they can swim along with the tsunami or swim along with the new technologies so they can get the best benefit from it? 

Bobby Owsinski: AI is just another tool, but it manifests in so many different ways for a musician. First of all, it can manifest in composition creation: Helping you create a song, giving you new ideas for chord changes, giving you ideas for lyrics. Now I’m saying, I'm giving you ideas, not writing it for you. And that's the differentiating part there. It can help you when it comes to mixing because there's some really good tools for automated EQ, for reverb; A lot of people have trouble with reverb, just dialing that in. Compression, limiting, all those things. Spectral balance. Noise suppression, noise reduction. Separating out tracks, all those things that are really fantastic tools. And there are use cases for them all. Now, that being said, for instance, with EQ, it's not going to beat an A-list mixer who will use experience more than anything in order to dial something in. It's not going to replace that, but if you're unsure, it's certainly better than fooling around for five hours still not being sure that you got it right. It does get you in the ballpark fast. I mean, really fast. All that stuff is really good. It's still really lagging way behind when it comes to mixing. There are several platforms out there that will do it for you. They do mostly stems, and stems would be, eight, ten at the most, and then they'll just balance it up. But mixing is so much more than that. It's more than balance. Mixing is adding, making sure you can hear everything distinctly. It's adding ambience. That's the secret sauce for a lot of mixers, and adding effects. It doesn't do that yet. It's very good at mastering, very good, provided you know how to do it, because to just say, “master my song”, it'll be okay. If you use a reference track, which is the real key, then you'll get fantastic results. So it's excellent at that. It's pretty good when it comes to music, and let's just talk about marketing and promotion. There's some wonderful platforms out there that will do music videos, lyric videos for you. For graphics. Obviously there's lots of platforms that will do graphics. It's not as instant as everybody thinks, but it will certainly get you there pretty fast. And then finally for marketing stuff.

Michael Walker: Sorry, real quick for the lyric video: I'm personally curious about that one. I've been waiting. I haven't checked recently, but I figured at some point there's going to be a really good auto lyric generator music video. Do you know if there's like the name of one or are you referencing one that maybe has come out recently that uses AI that's even better?

Bobby Owsinski: It doesn't use it in the way you'd think, but Videobolt is one, and Rotor Video is the other one. Rotor: this is a recent feature actually, where they added that on Lyric Videos, but what it will do is you'll still have to tweak it. Because it will misidentify words and then it might put it in the wrong place, so you might have to move things, here and there, but still, it's faster than doing it the old fashioned way, so it's okay. And actually generating backgrounds and everything, so you don't have to do that. Andmaybe doing it better than you could if you don't have those types of chops or artistic bent. Yeah, there's lots of really good ones out there. Kabre is another good one.

Michael Walker: So cool. Yeah, it's an exciting time to be alive and to be making music. Our mutual friend and mentor Jeff Walker, describes AI in this way that I know that you've referenced it before too, because it's a great way of putting it is that AI right now, at least, it can get you to that zero to 80 really quickly. And a single click is zero to 80, but then that last 20%, like you're talking about kind of fine tuning it or just getting it perfect, it’s really beneficial to have a human kind of come in who's a master who can take that and actually make it, a hundred out of a hundred instead of 80 out of a hundred. Man, what a cool gift to be able to take that 80%, which for a lot of people is the hardest part and get there really quickly.

Bobby Owsinski: Yeah, but here's a good way to think about it: Humans are really good at nuances. And the better we are, the more expert we are at whatever we're doing. AI is not good at nuances. AI gets you in the ballpark with fundamentals, let's say. But it's not good at nuances, and that's what humans bring to the table. So you're always going to be able to beat an AI just on that alone. Now people might say, yeah, but it's going to learn nuances. Yeah, but not creatively speaking, because a painter will look at a painting in the process and will do one thing one day. And we'll come back and do something else the next day and maybe something else the next day. It's different. A human manifests nuances. It's not something that AI is good at right now. 

Michael Walker: Yeah, that's a great point. And I've heard it described like this a lot that AI isn't going to replace artists or replace humans in the short term, but humans who are using AI are going to replace people who aren't using AI because it's just such a valuable tool. And I think that's totally true and it's already happening. It's just such an amazing tool. 

Bobby Owsinski: The thing about it is most people will say: Oh, I don't want to have anything to do with AI, but yet they have an iPhone and they unlock it with their face. And that's AI, right? It's facial recognition. You're using it all the time. If you use navigation in your car, it's AI, you're using it no matter what. For someone to say, I don't want to use this, it's going to be there whether you want to or not, it just will not be what we call AI. Now, we're calling them co-pilots and there's all sorts of different terminology that basically says this is an AI to do that's doing it. 

Michael Walker: That's a really good point. Yeah. I had never really thought about it that way, but yeah, basically AI has been around for so long just in the form of, it depends on how you classify AI, I guess. Like more recently, about a year ago to date, cause the birthday of ChatGPT was either today or yesterday. But just insane. Like it's within one year. But, it seems like that was sort of like AI’s moment in the sun, where it’s like: mind blown. ChatGPT is such a powerful tool. But to your point, a lot of tools have been using machine learning or, what we could call artificial intelligence, not artificial general intelligence, but artificial intelligence. And even like where my mind went to was James Bond, 007 goldeneye, like that game back in the day. You could play against computer characters and they were pretty dumb. Like they're not good. But technically that was a form of artificial intelligence because it was an artificial intelligence.

I never really thought about it that way until you mentioned that.

Bobby Owsinski: It goes back to the 50’s, Michael, but they used to call them expert systems back then. Most AI’s. Up until recently would concentrate on just one area of what they're doing. Think of robotics. Think about if you have a robot building a car, this robot does this one thing, and this one does that thing. It's an expert system, but it's AI that's doing it, but it's only in this very narrow function. And now what we're having is multi-function AI that will do lots of different things. So that's general AI basically, and that's the difference. 

Michael Walker: Yeah. I love this conversation. I'm so glad that we get to talk about this. Going a little bit deeper down the rabbit hole as it relates to AI and what we're touching on, general artificial intelligence, could you describe for anyone listening to this, you maybe aren't familiar with the term general artificial intelligence, what that means, and I'd love to hear your perspective on where we're headed towards that as well. 

Bobby Owsinski: Yeah. Let's just go back a little bit because there's a lot of terms that people misunderstand and they think they're all the same thing and they're not. First of all artificial intelligence is a segment of computer science and artificial intelligence can cover a lot of different things but from there, we get into machine learning. Machine learning is a subset of artificial intelligence. And what machine learning is it's facial recognition. You teach it what your face looks like and then it will do something for you. It will open up your phone. We've been doing this a lot and what would happen originally is it would take a human to take all of these pictures, let's just go with pictures, and would label them: this is a cat, this is a dog, this is another cat, this is a dog, and then it would be able to figure it out and see a picture, oh, cat, dog.

Michael Walker: There’s something I like about the fact that to prove we're not a robot, we have to do things that eventually we're training the robot on how to do things. It's pretty ironic.

Bobby Owsinski: So then we go another step below that and we go to neural networks. Or above that, there's another way to look at it. And a neural network is part of machine learning, which is part of our artificial intelligence. And a neural network is a different way of putting things together. It's based on the neural network of our brains. At its very basic, there's three levels, three layers to it: There's an input level, there's what they call a hidden layer, and then there's an output layer. Okay. So now what that can do is it can function on multiple different things instead of just doing this one robotic thing; just doing facial recognition. Now it can work on multiple problems at the same time. But if we go beyond three layers (input, output, hidden), all of a sudden it becomes deep learning. So now deep learning has a lot of different layers. For instance I've heard ChatGPT has 95 different layers. And they constantly talk against each other. And there's all these neurons within the layer. The problem is, and this is scary, I must admit. I'm not afraid of AI, but there is something here that bothers me. No one knows how these layers work beyond the third layer. There are actually research teams at various colleges that are researching various layers within a neural network to try to figure out what's actually happening here. That's scary. There's also large language models that we keep on hearing on ChatGPT. What that is, it looks at a lot of text-data language. Looks at a lot of it, and what's interesting here is, it doesn't store it, it just looks at it, and as a result, all these lawsuits that are saying: Oh, you know what, since you trained on my data, you're infringing on my copyright, but now the courts are saying: not really. Because it's as if you went to a museum, and you looked at all the paintings. You didn't reproduce any of them, you looked at them, and then it went, pshhh. And you remembered some of it, and then when you try to reproduce it, it's completely different. And that's what's happening with AI. It takes in huge amounts of data that it learns from, but it doesn't store it. That's what people don't quite get. Large language models are things like ChatGPT, Clod from Anthropic, BingAI, Google BARD, all those things. And they're just learning off big data sets. An interesting thing here is that there's something called ethical AI, and ethical AI basically means: okay, my company has paid for this data, so we licensed it, people have made money from it, and you'll find this for a lot of music AIs, where now they're going to say, okay, we want to learn something, we want to learn how you put your music together, so let us put pay you for the privilege of ingesting it, take a look at it, to train on it. So it's ethical AI which is happening. So as you can see, there's a lot of different things here that are going on, every time a question comes up, it's addressed, but it's addressed in a lot of different manners from a lot of different angles. And you have to keep on top of it to understand how it shakes out because sometimes it shakes out and you're not aware, what happened. This can go down the rabbit hole, as you would say. There's a lot of different things here. Especially when it comes to copyright. Copyright is unique, let's say. And the reason why is: our copyright laws are old. Last time in the United States, we changed copyright, we updated copyright was 1998. This is before the digital revolution and all of our copyright does not involve anything that's digital. So it really has to be updated. In the meantime, you get the copyright board that comes out with rulings. In March last year, they had a real big one that basically said, okay, anything that's 100%... so in other words, I can't go to Aiva or Boomi or anything like that, any of those platforms and say: okay, write me a song and then go and try to copyright it. No, it needs a human. What they didn't say is how much of the human has to be involved. So that's the arbitrary part. There's another thing that happens where you get something off of an AI platform and it could be music. It could be a video. It could be an image. It could… be those are the big three actually. And there are some platforms that say, okay, you know what? We own the copyright and we're going to grant you a license. If you pay us enough, we're going to grant you the copyright. But… I had long discussions with some of the best intellectual property attorneys in the world on this and the general consensus is: these companies don't have the right to grant a copyright. So what'll happen is somewhere down the line, someone's going to come back and say: Oh, you know what? We don't think you have the right for that, which is going to blow up their model. So look for that to happen down the road. 

Michael Walker: Remind me of exactly what we were referring to there. So there's providers that are creating and selling the copyright for what exactly? For AI generated 

music?

Bobby Owsinski: If you generate something, let's say you, you go to, like I say, Boomi, Loudly, and you say: write me a song. You massage it and you make it better so it's not 100 percent AI generated because you're actually working it. Then what happens is you get some options. The first option is: it's for free. So they say, yeah, no problem. You can have this. But really low resolution. You can only get an MP3 and it's only going to be 30 seconds instead of three minutes. So then the next option is let me pay for it. Yeah, sure. For $9.99, you can definitely take this. You can now download it. We’re going to give it to you in a little bit higher resolution. But you know what? You cannot use it for anything on YouTube. Or you can't put it on Spotify. You go: what good is it then? But so then you pay them a little more and they say now we're giving you the license to do that. Now, you take this and use it as you like on Spotify, YouTube, videos, whatever. But we own the copyright. We're just granting you a license and you're saying I think I want to own the copyright. No problem, for $99 a month you're gonna have the copyright. So then you have the copyright, but it turns out they may not have the right to grant that copyright, or, to grant licenses or anything. And we see this in some videos, for instance, some video platforms use that same methodology. And there are others that are hip to this and say: no, guess what? Whatever you generate is yours. It's your copyright. You generated it. You paid us to do this at a really high resolution and guess what? You own it. No problem.

Now I'm going to take this another step. This is what no one ever talks about. This is something that I found out after doing all this research for my book. I went to a copyright attorney and said what's up with this? And he said, yeah, you're right. Let's talk to some other people. And we started to talk with other copyright attorneys and they're all going: yeah you found something, here's what it is. So when you generate some AI music, the platform will say, yeah, no problem. We'll grant you the copyright or license or whatever. Only for the recording. Don't forget, there's two copyrights. There's the publishing and there's the recording. They never talk about the publishing. Who owns the publishing? Because you own it. They never talk about it. So the way around all this, you say, I don't need an audio file. Just give me the MIDI file. Download the MIDI file. Plug it into your DAW. You make it better because you'll get better virtual instruments. You can edit it. You can do whatever you want and you bypass all that crap. So these are secrets everyone's learning today. 

Michael Walker: Yeah, it's amazing. Yeah. And because we're here live right now, you get the secret awesome sauce from Bobby. And I know you're going to be doing deeper into it during your workshop, so I'm looking forward to talking more about that too. So interesting. So if I understand you I haven't seen companies doing that specifically where it's an AI tool that you can use to generate some new artwork. And then they say, you have to pay us to actually use this commercially. But what it reminds me of is with ChatGPT, they're very clear that Hey, like anything you generate, like you can use commercially, you can use it for whatever, and Oh, by the way, if you do get sued or anything, Here's our stamp of trust where we'll take care of all the legal stuff. That's pretty cool. So maybe, could you talk a little bit about that too? Like with some of the companies, they're literally going to defend you if something happens. 

Bobby Owsinski: Yeah. Okay. So what happens there is almost none of the companies that are working with large language models, almost all of them, I actually have all of them because I haven't seen one that said anything other than you own it. No, you're copyright. You generate it. You own it. The problem is where you have some people, book authors, for instance, there were a lot of lawsuits about this where people were not happy about ChatGPT using their material to train on. So the whole idea is he used my material to train on. I didn't get any money from it, but now the rulings are going against them where the judges are saying okay, if there's an infringement, let me see it. So Sarah Silverman, for instance, was in the forefront of this with her book and the judge said: okay, if there's infringement, then ask ChatGPT to print out your book and it can't.

Cause it just trained on it. It doesn't do it. That's why they can, that's why ChatGPT and all of those can say, we're not worried about this so we will indemnify you from this. And that's what's happening there is it's again, this is one of those things where it took the better part of a year to suss this out and actually get through the courts where there were court rulings on it. 

Michael Walker: Yeah, that's great. I actually hadn't heard the latest verdicts and in terms of that they had sort of made a decision that it is okay to be trained up on the data. Because, yeah, it's an interesting one. The way that we all learn is by, essentially, getting influenced by other people's music and artwork, and we're all basically, influenced by machines. And we have training data in our brains that turned it into something unique. And we formulate our own version of everything. But it sounds like the courts have decided that we're going to treat these machine learning or AI algorithms in a similar way where it's, as long as it doesn't directly infringe or just copy or clone something, that it's okay to, be influenced or have the DNA from some of these other pieces. 

Bobby Owsinski: Now, that being said, there are what they call guardrails on most AI platforms now. So if you were to go and say: write me a song like Elton John, it will most likely say, no, I'm sorry, I can't do that. Or, anybody that's famous, it will, it'll say: Oh, can't really do that. And that kind of eliminates a lot of that, unless there's something in the public domain, then it's not a problem. What I always thought was, this is actually going to be another revenue stream for popular artists anyway, you have to be popular, where, someone goes to Elton John and says, I want to use your catalog for training. And they'd say, sure, 20 million bucks. You give us a piece and we're good. But that hasn't happened yet. It has happened on some smaller scales. Like I was saying the whole thing with some platforms paying artists, especially for that, they've done ethical AI. So now I don't know how much they pay. That's the whole thing. I'm not sure if it's a lot. Probably isn't. 

Michael Walker: So interesting. I love as we're having this conversation, Bobby, every once in a while, I'll glance over at the chat and some of the things that I'm seeing, just like poking here. A said: free will might be an illusion. So that's the only context. I haven't read all the other stuff. I'm like, yes, this is the conversation that's happening right now. How about soon… maybe after this last question, we can actually migrate over to some audience participation and ask some questions and answer some things from folks who are here live. So if you're here live, then make sure to take the next minute or so to think about what's your number one burning question that you'd love to have answered here by, Bobby. And then afterwards, you can actually raise your hand. We'll bring you on live to ask your question. Maybe the last question for you, Bobby, as if we haven't gotten deep enough down the rabbit hole already. I think this is one of the defining questions or sort of movements of the next 10 to 20 years. I would love to hear your perspective on artificial general intelligence and where we're headed in the next 10 years or so, or maybe longer depending, obviously, depending on whether you even believe it's going to happen or how it's going to happen. I'd love to hear your perspective on exactly what artificial general intelligence is, what it means for humankind, and how that also might impact the music industry. 

Bobby Owsinski: Yeah. I don't think I answered your question from before about this. Artificial general intelligence basically means AI becomes so smart, it becomes human smart. It's as smart as a human. The thought is once it gets to that, then humanity's in trouble. So you have the alarmists on one side that are saying, oh, this is gonna be a problem. Then you have other people that are saying, eh, it's a while to get there. We're gonna put guardrails on it and it's not gonna be a problem. So we'll see. And then of course you can go a step beyond that where it's going to be smarter than a human. And that could be a problem, but we're not quite there yet.

Now, there's something new that may take this down the road faster and it's called adversarial networks. So now you use two AIs and they feed off of one another. So one AI, you ask a question and then it goes to the other AI and it goes back again and back again. And it becomes really precise; really good. So they're called adversarial networks. And this is the next realm, which may get us to general AI at that point. Generally AI, again it's now it's as smart as a human and now it's not even close. 

Michael Walker: Yeah it's interesting because human intelligence and digital intelligence have been so different for so long. And it seems like one of the reasons that ChatGPT was such a breakthrough was because for the first time it felt a lot more human because it did something that felt very human which was in a human-like way. It really is a different kind of intelligence, right? It can do machine intelligence and computers in some ways they're way superior to human intelligence and in some ways it's ways inferior and it's simple things that we can do, like moving our hands and grabbing things, have been really difficult for machines to do. But. Yeah, the idea that we have an artificial intelligence that can do those things that right now are sort of only able to be done by humans is interesting. The adversarial networks too. I hadn't heard about that idea, but what that reminds me of is as humans which it seems like neural pathways as well, with the neural nets with AI is just like looking at our brains and being like, huh, like maybe we could do a similar thing with computers. But similar with adversarial networks, that makes me think of capitalism and human society and how different viewpoints and how it really does come from this tension of the conversation we're having right now. There's two different agents here. There's two different perspectives and life stories and genealogies. Now we're both here having this conversation and there's 50 people here in the audience to have their own neural nets and their own networks. So to think that one of the things that might help AI to become even more intelligent is the same thing that it is for us which is creating these different viewpoints that in different agents that can communicate with each other is really interesting. 

Bobby Owsinski: Yeah. Yeah. By the way, all this stuff. Yeah, all the stuff we're talking about is in my book: “The Musicians AI Handbook” that just came out.

Michael Walker: that's actually a perfect segue. I would love to hear a little bit more about the book that you created. That's all about this where you go super deep into it. The musicians, AI handbook. Yes. Oh my gosh. So could you tell us a little bit about The Musicians AI handbook? And also I know that you have a workshop, some really like valuable resources that you've been gathering and creating for everyone that's listening to this live right now and anyone that might be listening to the podcast, could you share a little bit more about what that's all about? 

Bobby Owsinski:  Yeah. I do this twice a year whereI do a mixing workshop. And what it is: people send me mixes all the time, or just questions. And they all fall into roughly the same categories all the time. Balance; can't get my balance right. My mixes don't translate. My mixes are too bass heavy or there's not enough girth, or my mixes don't sound like… whatever, it's not distinct. There's all sorts of roughly the same questions that happen. So I tried to address this in my workshop. The next one is: The Awesome Mix Workshop. It's three days starting on Monday, and what'll happen is, each day we'll address one of these particular areas. So on Monday we're talking all about mixing, or all about balancing. And balancing, most people don't realize, also has to do with power in a mix. So people always say my mix isn't powerful, it's wimpy or whatever. Well, a lot of that has to do with balance, and I'll show you why. I'll also tell you some stories about how I got there. Because usually it wasn't that I know all this stuff. It was imparted to me, in very painful ways sometimes, that tend to be good teaching stories, which I'll also relate.

That's the Awesome Mix Workshop next week, three days, Monday, Tuesday, Wednesday. And it's free, less than an hour, 50 minutes, but I'll be around for any kind of Q&A afterwards. I'm looking forward to it. I always have fun doing these things. I enjoy the questions and enjoy presenting and hopefully opening up some minds. And also hearing the stories where people later will say: I tried this. It really worked. This was great. So that's what makes me feel good when that happens. 

Michael Walker: First of all, thank you for the work that you've done for literally decades now and bringing this value to our community. You're here live right now. We'll definitely make sure to throw the link in the show notes. If we can make sure Ari or Jared, if we could grab the link for the workshop and share it in the notes. 

In addition, if you're listening to this live on our pod, if you're listening to this on our podcast right now, then you can click on the link in the show notes for easy access to the workshop. And there's something so valuable about stories and being able to share stories as a vehicle for understanding and actually taking insights and wisdom and remembering them and actually using them in a practical way. I appreciate that you're able to share those stories in a way that also facilitates these lessons that you're teaching. I think if you don't have those, then a lot of it doesn't really stick. Yeah. So cool. Yeah. So with that… 

Bobby Owsinski: Yeah I'm up for Q& A if that's what you want to do. Yeah. 

That's where you're going.

Michael Walker: First of all, I just want to give a round of applause to Bobby. Let's give him a big thank-you round of applause in the chat. This is awesome. Really enjoyed this conversation. And we could probably be here for the next 10 hours talking about AI if we didn't pull things off and go to the Q and A. So let's get to folks in the chat. If you have any questions that you'd like to ask Bobby and I and come up here on stage, then you can actually raise your hand in the chat and we’ll bring you on here live. Otherwise, feel free to put your question in the chat. And we'll be happy to answer the questions as they come in. [laughing] Yeah, um, okay, he's asking the hard questions, Bobby. He asked, how do you teach AI to love?

Bobby Owsinski: There you go. It's human. It's nuanced. Isn’t love nuance? Maybe that's the greatest expression of nuance that you can come up with. 

Michael Walker: That's a great, that's a great question. Yeah. Yeah, it is. Maybe define… How about we all agree on what love means first?

Bobby Owsinski: Wasn't there a Star Trek about this? It was like, oh yeah, okay, teach a computer to love, or ask the computer what love is and the computer can't figure it out and it explodes. 

Michael Walker: That sounds like… at some point there'll be a big controversy in the world when we're trying to, we're trying to answer that question because it either the AI is going to be so convincing that it seems like it's that intelligent like it is giving loving that people can that people argue that they actually are experiencing true love. Or maybe it breaks this consciousness barrier and we actually define it as a living organism. 

Bobby Owsinski: I think we're there to some degree. There's a couple of AIs that you could try. One is character.ai, and character.ai allows you to either construct the character… So you can construct a lover if you want, but it also has other characters. So assume it's Elon Musk and ask Elon, have a conversation with Elon Musk or Napoleon or all these different… Steve jobs or whatever, but that's one and the other one would be replika. And as a matter of fact people were using that as an online lover, and it got to be so real that they had to shut some of it off because people were getting attached.

Michael Walker: There's a lot of lonely people. Yeah. Yeah. Wow. That's interesting. I mean, I know there's a little bit of controversy with therapy as well. ChatGPT if you try to go too deep, it's: Hey I'm not going to talk to you about this stuff. Like you need to get a therapist. But it also does seem like there's a big opportunity for a tool that is built in a way that is a 24/7 therapist and knows how to ask the right questions and knows how to help you look inside to process stuff.

Bobby Owsinski: In character.ai there is actually a therapist built in. So you can go to that and ask it questions and it works as if it's a therapist. I haven't tried that one yet, but I might after this call, I might. 

Michael Walker: I might try the same thing. I've definitely heard of it. I think it's really trending for young folks, maybe like teens right now. I feel like I heard of the top three maybe for like people on tiktok or something. But if I understand it it's what you can basically use it to create like any kind of care, like an AI character and you can train it up to, it's what ChatGPT is doing with their GPTs bots now, or you can create different GPTs. ChatGPT seems like they're trying to take a really clear stance that this is AI and it's not a human and don't get attached to us and don't treat us like a human. We're always going to correct you if you try to talk to us like a human. Whereas I'm assuming character.ai is the opposite where it's like, you can just create a character. 

Bobby Owsinski: Like I say, there are a number of characters already. Built in that you can just use that. But you can create your own as well. So yeah, there you go. Give it a try. 

Michael Walker: The world of video games and RPGs, I can imagine is just never going to be the same once some people really get going with creating an AI character model with a video game. 

Bobby Owsinski: Speaking of which, there's now an AI that they're beginning to use in video games, which is a music AI. And I can't think of the name of it off the top of my head, but it creates music on the spot. So, instead of having a number of music cues that it just pulls from, it creates original music for each situation. So that's cool. 

Michael Walker: Yeah, that's really cool. I mean, Music has always been about expressing your identity and for the people that listen to your music, it's not as much about you as it is about them it's like, them identifying with your music and what it says about them and it is interesting thinking about what if we all had our own personal soundtrack that was like literally designed just for us based on our exact situation. Certainly I'm sure some folks here who are listening to this have had this experience where maybe you're just goofing off and you wrote a song for someone when they were sitting next to you. They were like, Oh my gosh, that was amazing. That blew my mind. It really was not that good. It was like, I was just, like goofing around. But there is something about that personalization where, you know, if we were able to have a song that was literally written knowing exactly who you are and knowing exactly what your challenges are, what your goals are, what your feelings are, that it could really speak directly to you…. Huh.

Bobby Owsinski: A lot of the composition AIs are like that because you can go in and define all that. So you can define moods, you can define, you can go down the rabbit hole of what you're looking for. So to some degree you can sort of do that. But I know what you're talking about, you're talking about another level of that. 

Michael Walker: Yeah. And I guess to a certain extent, a lot of mainstream songs, like part of the reason that they catch fire because they're so relatable, like all of us can see ourselves in this song because we've all experienced love or we've all experienced this feeling or this idea that they're communicating. In a lot of cases, really great songwriters will write their songs intentionally to remove some personalization so it can relate to everybody, but speak things in a little bit abstract enough way to still be relatable to people.

I wonder if we'll ever reach a point where we go deeper and deeper into personalization, where we have these niches where people… this is totally a weird example that's probably never going to happen, but if there's like a dentist community, if it's like a song that was like for dentists, it's like [singing] I'm a dentist. I love teeth. But it was a great song, and it had all these inside jokes so all the dentists can relate to it, then there's a fair chance that's a song that's going to be shared within the dentist circles cause it's like they have the shared identity and they're like: Oh yeah, like we get it. Cause we're all dentists. And I guess that isn't a part of music.

Bobby Owsinski: Do you know Tiamo De Vettori? 

Michael Walker: Oh, yeah. Love Tiamo. Of course.

Bobby Owsinski: Okay. So one of the things that Tiamo teaches people is how to write custom songs just for those situations. So if there was a dental convention, for instance, then someone writes a custom song for the dentists for the dental convention. So that is exactly what you're talking about. 

Michael Walker: Yeah, it's cool. It's a cool opportunity. I think it's an interesting exercise for all of us to sort of think about how our soundtrack is really meant to give voice to these shared community traits. And ideally, if you want to create a strong community, then you take something that people feel that makes them weird or it gives them a sense of identity, but then you create a voice around it and then everyone that has that thing that makes them unique is able to have that shared identity around it, which again, the examples are like dentists. Oh, we're all dentists. We all get it. We get the inside jokes. We get the lingo. so It's an interesting thing for all of us to think about looking within and be like: what are the weird things that I think about or that I believe, or the things that I'm into that not everyone's into and how can I create something special around that? Cool. Hey, Bobby, thank you again so much for hopping on here and answering questions. It's been a lot of fun. It's rare that I get to have conversations that go this deep on this topic. Do you want to answer maybe one more question from the audience? Or do we want to wrap up now? I want to be respectful of your time.

Bobby Owsinski: Sure, Let's do one more.

Michael Walker: All right. So we've got time for one more question, folks. So I'm going to scroll up here and see if I can find a good one here. I love the conversation that's happening in the chat. This is awesome. A lot of people saying, thank you, Bobby. 

Bobby Owsinski: It's my pleasure. 

Michael Walker: Okay. Kay’s question: She was curious about if you have any experience with AI as it relates to marketing and what's one of the most valuable use cases that you've seen for musicians to help them with marketing their music. 

Bobby Owsinski: Oh, yeah, lots of good ones for instance: You can ask it to help you do a marketing plan for a music release. I've seen a traffic plan, for instance, here's all these gigs that we have figure out the best way to get there is always a good one. So what would be the best merch? What has the highest margins and what do you think that we should use and give me a plan how to put that together? Oh I'm having a release party for instance, and I have a budget of $300. How should I spend it? What should I spend it on? And it will tell you. That's another good one. There's a lot of different use cases that really work. Oh, here's a good on: Ask it to write a survey for your fans and followers. And it's good at that. I actually did that myself before I wrote the AI book. I asked ChatGPT to write a series of questions and I described my audience and it did a really good job. I mean, all this stuff you have to tweak a little bit, it can definitely get you in the ballpark and do it a lot faster than if you're staring at a blank piece of paper. Yeah.

Michael Walker: I hope everyone is paying attention and likes taking notes there because the list that you just rattled off there, I think could save every musician who's in here, literally hundreds, if not thousands of hours just by taking those things you're already doing and using those prompts specifically. It's so powerful. 

Bobby Owsinski: Oh, I'll give you another one. Social media posts. A lot of times people don't know what to post and just ask it for. Bunch of posts and be specific. I want it for X. I want it for Instagram. What should I do for tiktok? Give me 10 ideas for tiktok, stuff like that. But again, you can't just be superficial. You have to go deeper. And that's some of the stuff that's in my book where I show you how to engineer prompts, but you have to give it some background. You have to tell it who your audience is or who your followers are. And here’s another real good one and it's to ask them what. Ask ChatGBT or Clod or whatever: what are my, what are the fears that my fans have or what, why do my fans like this type of music or what type of music do they like besides my music. Things like that and you'll get some fantastic answers. 

Michael Walker: So good. Awesome. Hey on behalf of all of us again, Bobby, thank you so much. It's been a lot of fun. And again, I would highly recommend, if you are still here right now, and if you enjoy this conversation, then. Go ahead and check out Bobby's book. He literally documented and gathered all these resources and went even deeper and put them in a practical format so you can get the most value from them and have it in written format. Thank you for doing that. And I'll make sure that we get a link for the book and the show notes. 

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.