A songwriting battle with my AI clone | Jason "Poo Bear" Boyd, Elise Hu

(Singing) I’m the kind of brother Who’s been doing it my way Getting my way for years In my career And every lover In and out my life I hit love and left in tears Without a care Until I met this girl who turned the tables around Caught me by surprise I never thought I'd be the one breaking down I can't figure it out I'm so caught up Got me feeling it, caught up I don’t know what it is But it seems she’s got me twisted I'm so caught up Really feeling it, caught up I'm losing control This girl’s got a hold on me Thank you very much. Happy birthday. (Applause and cheers) Elise Hu: That was incredible. PB: Oh, thank you. EH: Thank you, thank you. What a treat. It was so great to be able to hear you in this theater. Like, the acoustics. Are you satisfied? PB: They did an amazing job, the acoustics are killer. EH: We're so glad. Oh, that was just so moving. I wish we could get a whole show from you. PB: Oh, man, maybe one day. EH: OK. (Laughter) Standing invitation. Before we get into the innovation part of this conversation, I just want to ask you what, to you, makes a great song. PB: To me, a great song is ... it’s a frequency that is simple but effective. So simple enough for a child to be able to remember and sing along with, but effective enough to fire off neurons in a brilliant mind. EH: OK, so do you just feel that? PB: I feel it, lots and lots of years of trial and error, and just having this balance of simple and effectiveness. EH: How, especially you mentioned lots and lots of years, technology is constantly changing, it's been improving your entire career, our entire adult lives. How have you, if you could just take us through, increasingly integrated digital technology into your work and your creativity? PB: So I've actually started off at my first success, my first record in the '90s. So I was working on reel-to-reel, ADATs and, you know, it evolved into Pro Tools. And that's when the digital era -- nd if you didn’t really evolve into that, you kind of were left behind. And every year, they've just added on cool plug-ins and things to help save time, because time is the most valuable part of our lives. So, you know, I think Pro Tools, Auto-Tune and all these things that were -- they always hate on it at first. And then you realize that it's only been created to help us. EH: Do you consider yourself an early adopter? Like when your friends or other producers introduce you to a new thing, are you like, "Great, I'm just going to go all in," or are you more skeptical and you kind of watch first? PB: Both. Like I watched a little bit, but once I used it and I understood that we're all being programmed by perfection, and that comes from, you know, the digital help. So I had to incorporate it, and it just helped save time. And it helps for people to enjoy the music as well. EH: Is there any tool, over the course of your career, that you wished you were sort of faster to jump on? PB: I would probably say AI. Honestly, yeah, I would say AI. EH: OK. How are you feeling right now about AI's growth in the creative industries? PB: I feel like once again, it's a tool that I feel like is absolutely going to be necessary. It's in the beginning of it. And I feel like there's kinks that need to be worked out, like everything in life. But I also understand the importance of it now. You know, at first, I saw a lot of people using it around me, and I was like, I didn't want to think lesser of myself for using it. But, you know, even now I understand, like, OK, I could probably use that. I could probably ask AI a question, and it can give me a cool answer. EH: Because I know about zero about producing a song or producing music. Can you talk a little bit, just practically, about how AI is being used, or AI tools, I should say, just not like, AI broadly, but AI tools are being used in your work, in your line of work. PB: So in my world, there's some people that are kind of completely relying on it, which I'm not going to knock it. But for me, it's just, you know, throwing out concepts or ideas. I haven't really integrated it musically because I still believe in the soul aspect of creating chords and having chords that we love. But I also use it, you know, if I have a question about something that I want to write about, I'll ask it and it'll give me a quick answer. And, you know, things that I probably would have to research a lot harder if AI didn't exist. EH: So essentially, you can use it as a starting point, or you have a starting point, you’re like, this is a musical phrase or here's a chord I really like. Walk me through. PB: Like I haven't integrated the music part, but more so, you know, just different questions. I love making songs that are questions. So not so much the chords, or I still believe in being able to -- I'm not the greatest musician, but being able to find chords that I love with somebody that is amazing at playing chords and be able to say, oh, I love that, or do a different second chord. But more so, like, just concepts and new ideas. And there’s just so much that hasn’t been written about. And running it through AI and making sure that, like, I'm not creating something that has been created already, which is important in the copyright world. EH: Yeah, for sure. That's already hard to avoid without AI. PB: Correct. EH: Do you feel like, and I guess we're going to -- I feel like I'm already teasing what's coming ahead, but do you feel as though there's still something ineffable about the human soul that gets worked into music that AI can't quite touch? PB: Yeah, you know, the AI can't -- I don't feel like it'll ever be able to quantify the human soul, you know? And I just feel like that's something, because AI is basically a math equation that kind of, you have this formula and it equals certain things. I just feel like, one day in the future, I won't say that it won't come close to the human soul, but I don't think it'll ever be able to emulate the human soul and being able to, you know, affect people's emotions the way that the human soul does. You know, it's just living, for me to be able to create something and it comes from a real place. AI has never had its heart broken, you know, AI's never been in love. So it's like those things. Of course, there will be a math equation that equals up to heartbreak or love or something like that. But I don't think, like, actually going through it. Like the fact that as humans, we go through these real emotions. I just don't think it'll ever be able to completely quantify it. EH: Does it matter that we can’t -- that soon, I mean, I know you can probably detect the difference, between artificial music and that which was created -- PB: Kinda sorta, yeah. Because it's kind of just not great yet. So when I hear something, I'm like, that sounds like AI because I wouldn't have said that. EH: But to be fair, there's also really bad pop songs that are written by humans. PB: It's the yin and the yang. There's a lot of bad pop music written by humans, correct. EH: Does it matter, though? Do you think that to the audiences or to the industry -- Wait, I'm answering my own question because it would matter to the industry. Because if it gets cheaper to write music, what happens to the artists? PB: Yeah, the artists, you know, I feel like the artists are safe. I feel like more for creatives. Just once they figure out all the new laws around the copyrights and stuff. You know, songwriters, up-and-coming songwriters. I feel like there's certain people who make music and they have a sound, and a lot of people, when they come to work with you, they're coming for that very specific sound, you know, that comes from years and years of programming frequencies into the universe, you know, by having a lot of radio success. So I do feel like for up-and-coming songwriters and producers that don't have that sound yet, I do feel like it's going to become more difficult for them to be able to, because they'll be battling AI literally, you know. EH: Yeah, and you mentioned regulation. Regulation, famously, is so many years behind how technology is developing. Are there certain guardrails or what would be certain guardrails that you would like to see to protect creatives and creative work? PB: You know, that's a great question. Just being able to protect the creatives that have spent a lot of time, a lot of years investing in creating a sound, just for AI not to be able to completely rip off our everything and take it and be like, “Oh, we’ll just do our own Poo Bear-style song” and pull from a thousand Poo Bear or different hits and create this new Poo Bear sound. So yes, I feel like the laws should encompass protecting the creative. So being able to detect and hear a song and say, you know what, that song didn't really come from the source that we think it's coming from. So I'm sure that somebody has to be thinking of this, to be able to scan the music and be able to say, oh, that's artificial. And at the same time, somehow be able to protect the creators that they're pulling from somehow. EH: You also kind of nodded to this already. What happens to the pipeline? You know, you're talking about musicians being able to create a certain sound who are really tied to what is distinct about them. I'm thinking of film directors who are similar as well. When you see a Wes Anderson movie, you know it's a Wes Anderson movie. So what happens to the future directors, the future musicians whose distinct eye or distinct sound hasn't been really out in the world or found an audience yet? PB: Yes, it hasn't programmed. That's the part that's scary. Like I look at it like it’s exciting and scary at the same time because I love the time factor, to be able to save time and be able to just prompt AI and say, make this movie or make this record. But I do feel like, I'm afraid for the new the directors, the creatives, just because there's really no way to really stop it from happening. It's inevitable, you know? So it's like, how do we get out in front of it? And we're able to still incorporate, like, include it, but still be able to have a sound or build a future sound like, for those new artists and new directors that haven't had any hits yet, still be able to use it as a tool, but still be able to build and develop a sound for themselves or a visual for themselves. EH: Can you give us a peek behind the curtain? Because you are friends with so many producers, so many people at the top of industry. What kind of conversations around this are happening right now? Because, as you mentioned, the cow's kind of already out of the barn. Sorry, I'm just always with my Texas references. PB: Yes, it's fine. You know, all my friends, they're kind of like, they're using it like, no different than Pro Tools or like, Auto-Tune. It's like, just using it for very niche things, you know, or like, oh, make this sound sound more like -- or make me a trumpet sound combined with a violin to create this new instrument that doesn't exist. Like cool things, but not to where it becomes the whole entire piece of art, you know? EH: Is this any different than, say, when digital synthesizers were first on the market and first came out? You know, they were also capable of making new sounds. AI is obviously, you know, exponentially stronger and faster and all of those things. But it sounds like the conversations around it are similar, right? PB: When you go back to the '80s, when they started making -- EH: Take us to the ’80s, Poo Bear. PB: Drum machines and the new keyboards, the Moogs, you know. Yes, we got to a place to where it was so exciting we just stopped hearing like, authentic, real instrumentation for a while. Like even the most amazing musicians like Stevie Wonder, they got into using -- EH: The keytar, is that what it's called? The keyboard guitar? PB: The keytar, yeah. It was like, wait, you know, they kind of forgot about. But then you realize that it always circles back around to the authenticity of like real instrumentation. But people just get excited, like, it's very similar to AI. It was like, oh, wow, we don't really have to think or do anything. We don't have to work. But then it comes a place where you miss it and you miss that analog sound, you miss the warmth and the human, the soul factor. EH: Do you think there's going to be kind of a backlash or a return to more analog sound -- PB: Everything's a circle. EH: Or even that like tangible media storage of yesteryear? What do you predict? PB: I predict the circle. It's an inevitable circle that happens, to where this new technology comes out. Everybody goes crazy over it, and then you just naturally start to miss you know, that thing that gave you butterflies, that real human being, the soul factor. And I think it's just a matter of time. I do feel like AI is inevitable, but I also feel like we're going to see a lot of artists that are going to be relying on it that ultimately have to go back to really performing and doing real live instrumentation. I think that's really important. EH: This live instrumentation question, I'm so glad you brought up, because it's so special to be here at TED to get to hear you perform just now, to have these experiences that are in real life. Gen Z, Gen Alpha, they're very into in-real-life get-togethers and convenings, largely because they were starved of that as they were coming of age during the pandemic. But live music is so expensive. I am finding that it is so much, it's like out of reach for so many people, especially young people, to go to concerts these days. How will we get together? How will we get to have these kinds of experiences where we are gathering and enjoying live music and enjoying sort of, the beauty of the craft? PB: It's a great question. So I feel like people are going to understand that there's ways to integrate concerts that's not purely based on ticket sales or merch. EH: That would be a dream. PB: There’s going to be ways, and I’m like actually working on it. And we've been testing some concerts, like, smaller concerts, but having brands come in and say, "You know what, I'm going to pay the artist, I'm going to pay for everything," and we're going to give a free concert and then see where that money that those fans might have spent on a ticket or -- really just on a ticket, maybe they’ll have a little bit more money to spend on a hoodie. So I think there's going to be cool new ways to give those experiences for free. But still people will be able to generate money, but just in another way. EH: Yeah, I'd love to see that. Are there examples that you're seeing of a return to either analog production, analog music, analog shows? PB: I think the examples are just, you know, if you do look at the concerts, the amount of money, the revenue that's being made, it still continues to grow. It's getting more and more. And I think that is because people do want to have that experience. You know, you can go to a movie, you can watch anything at home. And I think because of COVID, I think everybody just got so sick of having to watch stuff on their TV or laptops that, once the world opened back up, all the music, all the shows went through the roof because everybody was just so excited to be amongst each other and feel everybody's energy and frequency in the room. And I don't think you can really get that, you can have your VR set on and still be lonely and be alone, you know. So I think that the concert, like, the actual live performance aspect of music, I think it's going to get bigger and better. And I think that, you know, the revenue is going to continue to skyrocket. EH: And we just have to find a way to incentivize -- PB: Incentivize the concert so that they become free experiences where people still can make money and it makes sense for everybody. EH: OK. We’ve talked about live performances, and we're a little early to this, but I want to get to it because I'm so excited about what we're about to try. This is a Next Stage first. Y'all, we're going to enjoy a songwriting battle. But it's not between me and you, thank God. PB: Oh, man, I thought I was battling you. (Laughter) EH: It's not going to be between two musicians even. It is between Poo Bear, legendary music producer, and a machine. Well not really a machine, I guess and AI tool. PB: Yes. EH: Called Suno. PB: Yes, Suno. EH: It's called Suno. And just to be clear, Suno is a versatile tool. But as with all generative AI tools, this is my disclaimer to all y'all, as with all generative AI tools, the legalities for usage are still catching up to the technology. So do your own homework on there. Poo Bear, do you want to set up what you’re about to do? PB: I would love to set up. EH: Fantastic, I will leave the stage so that Poo Bear can do this challenge. Because you don't need me for this. PB: Oh, man, I thought you were staying. All right. Yes, great questions. Crushed that. So first I want to say happy birthday to everybody, because I feel like everybody should have a birthday every day. I don't feel like we should wait, you know. Like, "My birthday's in three weeks, ooh, I can't wait." I feel like when you wake up, you should be like, "It's an amazing day, I can't wait for today." So happy birthday to everybody. The whole world, TED Talk, happy birthday. So I'm excited to be here. This process is something I don't even think I ever dreamed of doing this. So it was like, I could say it's beyond my wildest dreams. That's Sasha, also my partner, hitmaker, producer, done a lot of records you know, as well. So we're actually, I'm going to battle Suno, the digital version of myself. So I'm going to get some concepts from you guys. You guys are going to actually kind of help me write the song a little bit. Maybe there's some publishing to give out somehow. I don't know, maybe like one percent or something like that. It could be life-changing. And from there, I’m going to create a hook and a post-hook, and then we're going to prompt Suno and say, once I get the concept, hey, make a Poo Bear-style hook, post-hook, with a song with a three- chord progressions. And we're going to see, once we listen, which one moves you guys the most, honestly. And if Suno beats me, it's totally fine. I'm OK with it, you know? So how are we looking, Sasha? Sasha Sirota: Looking good. PB: OK. So first we're going to find a nice three-chord progression. (Guitar music) There you go, I hear you and then I don't. Now I hear you again. I love that. Now, concepts. Can I get, yes, innovative ideas? Audience: Broke but not broken. PB: That’s deep. Can I get somebody from over here? Can I get a concept from anybody over here? Say it again. Audience: Voice for the voiceless. PB: Voice for the voiceless. These are like, you guys graduated from Harvard in here? This is heavy. Can I get some little, some less intelligent concepts? (Laughter) All right, wait. Let's go to middle school. Starting over. Ninja Turtles. I'm going to go with "starting over." I was that close from doing a Ninja Turtle. (Laughter) You almost had me. So, starting over. SS: We’re going to record one loop real quick. PB: Let's record the loop. SS: Here we go. PB: That's good. (Guitar music) So, starting over. I like to make ... Hello, hello, hello. There we go, starting over. So I believe in making hooks that are questions, but like rhetorical questions that we kind of know the answer to. So you say starting over and I'm going to say, should we be starting over? Alright? Were you able get a baby loop of that right there? Is the mic on? Yeah, OK. Am I double mic-ed right now? Sounds like I'm stacking myself without trying. OK. So my first step would be to come up with a melody. (Vocalizing) Are we starting over? Should we be starting over? OK, OK, OK. And then I will stop and I will say -- oh, let's do it one more time, I got it. So this is really how it happens in the recording studio, guys. This is like, real life. I like that, but I wasn't really feeling it. So now I'm going to try one more time. (Guitar music) OK. Yeah, yeah. Should we be starting over? Starting over. OK. All right, so we got that. I like that, everybody's OK with that first line? Wait, wait, you guys don't have to clap yet. Because there's going to be way too much clapping. I don't want you guys to be clapped out, you know what I'm saying? You can play that back, Sasha. Should we be starting over? OK Starting over OK Then I'll go again. So now let's punch right after that. Starting over Oh, yeah Yeah, yeah, yeah Starting over Keep it going. Should we be starting over? Oh Should we be starting over? OK. Let's check that out. You can start it right there, too. Just to hear, like, really close. Yeah. So when I go ... Should we be starting over? Should we be starting over? Oh, can I get that, can I push that last one? Should we be starting over? Should we be starting over? Cool, alright, so we got that. That's the first half of the hook. And then we say. Oh yeah, should we be starting over? Cause I don’t wanna be let down You don’t wanna be let down OK, stop. OK, let's listen back to that. Cause I don’t wanna be let down You don’t wanna be let down (Vocalizing) (Vocalizing) Should we take the time and turn this love around? OK. Let’s hear that back. I'm thinking of this, it's real-time, this is tough, guys. And everybody's looking at me. (Cheers and applause) Thanks. Should we take the time and turn this love around? Should we start over? Should we be starting over? (Vocalizing) Let's hear that. SS: Come on. (Cheers) SS: AI’s in trouble. PB: I don't know, man. I know Suno, the digital version of Poo Bear, is in there, like, warming up, doing push-ups right now. Like, yeah, shadowboxing. (Laughter) You don’t wanna be let down Should we take the time and turn this love around? Should we start over? Should we be starting over? OK, I accept that. So now, that's the hook, right? So that's the chorus, and now I’m going to do a post-hook. And this will be the part that will be the chant, kind of. So let's see what happens. Should we be starting over? (Vocalizing) Yeah Yeah Should we be starting over? OK. That might work, let's listen. (Music and vocalizing) Yeah. Don’t want to forget that. Let’s punch after that. This is after the second line. After the second “yeah.” Yeah. Should we be starting over? OK. Let’s hear that back, the post-hook. (Music) Should we be starting over? Yeah, OK, alright. That's my hook. That's it, but we're not even going to play that back yet. So now we prompt, let's prompt it. Poo bear, give me a hook. Should we be starting over? Three chords. This is it, guys. I'm a little nervous, a little bit. I'm sweating on the inside. I know you can't see my sweat. And this is -- SS: Generating. Watch this. It's going to be way quicker than what I just did. And this is the digital version of myself. What do you think? Eight seconds. SS: I’m going to put it on -- Five, four, three, two, one. (Laughter) Well, I did press play, it's ready. PB: OK. SS: See, this is why AI -- PB: Exactly, exactly. It's letting us down right now. (Laughter) SS: Oh. Suno: Should we be starting over? Chasing what we lost in the clover The cracks are wide but the heart’s still sober Should we try? Should we try? PB: OK, we're just having some internet problems. Suno: When the echoes don’t lie. Should we try? Should we be starting over? OK. Alright, OK, alright. That was the digital version of myself. (Laughter) I'm not sure if it made a lot of sense, but it's OK. So now, can we play it back the Poo Bear human soul experience? SS: Here we go. (Music) Should we be starting over? Starting over Should we be starting over? Should we be starting over? Cause I don’t want to be let down You don’t want to be let down Should we take the time and turn this love around? Should we start over? Should we be starting over? Should we starting over? OK. Sorry about the internet not having it. OK, so I'll let you guys decide which one ... The human Poo Bear. (Applause) Thank you, thank you so much. (Applause and cheers) Thank you so much. And that's my exercise. I hope this proves that we're going to always need human soul. But we will be able to use Suno, we should be able to adapt. We have to adapt. So let's not count it out, but let's incorporate human soul forever. Alright, happy birthday. (Applause and cheers) EH: Poo Bear, everybody.