The Syncreate Podcast: Empowering Creativity
Welcome to Syncreate, where we explore the intersections between creativity, psychology, and spirituality. Our goal is to demystify the creative process and expand the boundaries of what it means to be creative.
Creativity. It’s a word we throw around all the time, but what does it really mean? On the Syncreate Podcast, we share stories of the creative journey. We talk to changemakers, visionaries and everyday creatives working in a wide array of fields and disciplines. Our goal is to explore creativity in all its facets, and to gain a better understanding of the creative process – from imagination to innovation and everything in between.
The Syncreate Podcast is hosted by Melinda Rothouse, PhD. She helps individuals and organizations bring their creative dreams and visions to life through coaching, consulting, workshops, retreats, and now, this podcast. She's written two books on creativity, including Syncreate: A Guide to Navigating the Creative Process for Individuals, Teams, and Communities (winner of a Silver Nautilus Award for Creativity and Innovation), with Charlotte Gullick. She's also a musician (singer-songwriter and bass player) and photographer based in Austin, Texas.
The Syncreate Podcast: Empowering Creativity
Episode 101: Artificial Intelligence and Creativity - The End or Just the Beginning? with Douglas Eck, PhD
With the introduction of so many new AI tools for image, video, and music generation, is this the end of creativity as we know it, or the beginning of a new era? In this episode we explore the wild new world of creativity in the age of AI with Douglas Eck, PhD, Senior Research Director at Google DeepMind, and Research Co-Lead of Generative Media, which develops the AI tools Veo, Imagen and Lyria. He is the creator of Magenta, an open source research project exploring how machine learning can aid in music and art creation. His research focuses on machine learning and human-computer interaction. He holds a PhD in computer science and cognitive science from Indiana University.
For our Creativity Pro-Tip, whether you’re a skeptic, curious, or just don’t know where to begin, we encourage you to check out the Magenta Blog and try experimenting with one of the many new AI creativity tools now available and see for yourself how they might enhance your creative work.
Credits: The Syncreate podcast is created and hosted by Melinda Rothouse, and produced at Record ATX studios with in collaboration Michael Osborne and 14th Street Studios in Austin, Texas. Syncreate logo design by Dreux Carpenter.
If you enjoy this episode and want to learn more about the creative process in a variety of contexts, you might also like our conversations in Episode 91: Envisioning the Possible with Vlad Glăveanu, PhD, Episode 62: The Neuroscience of Creativity with Dr. Indre Viskontas, and Episode 58: Creative Problem Solving with BEST Robotics Executive Director Michael Steiner.
At Syncreate, we're here to support your creative endeavors. If you have an idea for a project or a new venture, and you’re not sure how to get it off the ground, find us at syncreate.org. Our book, also called Syncreate, walks you through the stages of the creative process so you can take action on your creative goals. We also offer resources, creative process tools, and coaching, including a monthly creativity coaching group, to help you bring your work to the world. You can find more information on our website. Find and connect with us on social media and YouTube under Syncreate. If you enjoy the show, please subscribe and leave us a review! We’d love to hear your feedback as well, so drop us a line at info@syncreate.org.
Episode-specific hyperlinks:
AI: Your New Creative Muse? Google DeepMind Podcast Episode with Douglas Eck
Show / permanent hyperlinks:
Melinda: Creativity and community are absolutely vital in challenging times. Creativity is also consistently named as one of the top skills of the 21st century, particularly with the advent of AI. Welcome to Syncreate, a show where we explore the intersections between creativity, psychology and spirituality. We believe everyone has the capacity to create. Our goal is to demystify the process and expand the boundaries of what it means to be creative.
We talk with visionaries and change makers, and everyday creatives working in a wide range of fields and media - from the arts to science, technology and business. We aim to illuminate the creative process, from imagination to innovation and everything in between. I'm Melinda Rothouse and I help individuals and organizations bring their dreams and visions to life.
At Syncreate, we’re here to support your creative endeavors. So if you have an idea for a project or a new venture, but you're not quite sure how to get it off the ground, find us at syncreate.org. Our book, also called Syncreate, walks you through the stages of the creative process so you can take action on your goals. We also offer resources, creative process tools and coaching to help bring your work to the world, including a monthly coaching group. You can find out more at syncreate.org.
My guest today is Douglas Eck. He's a senior research director at Google DeepMind and a research co-lead of Gen Media, which creates AI tools for creativity, including Veo, Imagen, and Lyria. He's also the creator of Magenta, a research project that explored how machine learning can aid in music and art creation.
His research focus is in machine learning and human computer interaction. Doug holds a PhD in Computer Science and Cognitive Science from Indiana University. So welcome, Doug. So great to finally have you on the show. And I feel like this is an exciting time to have this conversation.
Douglas: It's definitely an exciting time to have this conversation. And, Melinda, we tried to make this happen right at the end of Covid. I think it was on a boat on the Nile. (Laughter)
Melinda: On the Nile in Egypt.
Douglas: Which is a good place. We, so - Melinda and I had been on the Nile together with a number of other friends and cooked up this idea to do this podcast. And so, it's been a little while to make it land, but I'm really glad to get a chance to chat with you about creativity, and learn also more about Syncreate, which we talked about a lot then. But also, I want to see how this weaves together with the work that I've been doing and, you know, indirectly with creativity as well.
Melinda: Absolutely. So, I will say that this is our 101st episode. So, I’m excited to have you.
Douglas: I missed it by one.
Melinda: Well, we took a slight production pause in August. You know, just to take a little break and regroup, but I feel like we're kind of kicking off a new round of things here. So, excited to have you on. And so, I know you work at the intersection of artificial intelligence and creativity and technology, and all these kinds of things. So, I think, you know, there's a lot of questions on people's minds at the moment about like, what does the advent of all these AI tools mean for creativity as we know it? Like, is it the end? Is it just the beginning?
And of course, the answer is nuanced because there's a lot of jobs (including creative jobs), that are going to be sort of taken over by AI, right? And yet, there's these tools that make a lot of things possible that weren't possible before.
Douglas: Yeah. I think both are true. There's - I'm trying to add to what you just said, which was a brilliant summary of, I think, where we are. The thing that I would add is that, you know, I think we should be humble and realize that we don't really know how and when a new tool will give rise to new forms of creativity. I think that we'll see new forms of cinema and new forms of expression land. And we've seen little glimmers of that, but really, I think it's sitting squarely on us as creators to decide whether this technology helps us tell our stories and explore new spaces, rather than going the other way and saying, “Oh! Here’s this new technology that you have to use.”
It's the new thing. History is on my side there - for every new piece of technology that gives us a new way to communicate… like, for me, I'm a musician. So, you know, the electric guitar or maybe a bass guitar as, you know, you happen to play, as we all know.
Melinda: Totally.
Douglas: You know, that gave rise to at least 50 years. Had a good 50 year run maybe. Maybe more. You know, like, it's like, a really useful tool.
Melinda: It’s not over. (Laughter)
Douglas: Yeah. So, I'm here to tell you that it's not over. But for every one of these successful creative tools, there are thousands of failed attempts. And I think that's in some ways inspiring because it means that we are, as creators, we're the selectors. We’re the selection function for what gets used and what doesn't. If someone comes to you with technology that doesn't help you tell your story, you're just gonna throw it out.
And I find that genuinely inspiring and just raises the bar for people that are on the technology side of the story, to pick up their game and provide, you know, really useful and interesting new ways to express ourselves and to create new things.
Melinda: Yeah. For sure. So, I'm curious, kind of, you know, I was looking at the podcast you did last year with Professor Fry on the DeepMind podcast and, you know, you mentioned (which I thought was really interesting) that one of your first inspirations around the intersection of music and technology was a Player piano. And we had a Player piano at my house as well.
Douglas: No way.
Melinda: That at one point was so sort of severely neglected that it had gone deeply out of tune. And we used to play the rolls on it, and it sounded like a haunted carnival. That's a whole different story. But, you know, I'm curious, kind of, how are you using these AI tools in your own music?
Douglas: That's a great question. So, first - a caveat. Like, Professor Fry and I were speaking and I was wearing my sort of official hat as I'm a researcher at Google DeepMind - I'm getting the chance to talk with you as an old friend.
Melinda: Yes.
Douglas: So, this is great because, like, you know, I can talk about the way that we approach technology at work, but I think you can also go and watch other podcasts and read.
Melinda: Exactly.
Douglas: Honestly. My honest answer is that I struggle to use AI when I play music. I mean, I play acoustic guitar and I play piano, and I sing. I've had a blast doing some like, real time music generation, almost like, deejaying. Really interesting time exploring new sounds. And you know, for those of you that don't know, like what - forget about like, what’s happening now - I've been invested in AI and music since 2015. I created a project called Magenta, which is still rolling. It's an open source effort to connect technology to musicians. We started at the very early days of generative music, with neural networks and have been rolling since then.
We just recently released Magenta RealTime, which is this really cool, I think, like, open source way for you to play with like, kind of real time music generation, accompaniment DJ tools. And we'll keep strengthening that. That really speaks to me. I think for whatever reason, as a musician, what speaks to me is my voice and, like, really tightly playing along with what I sing. Acoustic guitar is just a great fit. I have a semi-hollow body guitar as well. And as I move towards electric guitar, I get less comfortable.
So like, as the professional I'm like, “Look at this electric guitar.” And as a professional I'm like, “Look at these AI tools.” But honestly, as a musician, I don't use them all that much. And you know what? I think that's fine.
Melinda: Absolutely. Absolutely.
Douglas: And honestly, I'm like, you know, what I do when I'm creating things and when I'm working on that part of myself, it's totally fine if I leave my work behind, even if my work is like, music or music adjacent.
Melinda: For sure. For sure. And of course, there are so many musicians working in so many different genres, and some of them use more electronic tools and some of them use more organic tools. And I mean, that's the full spectrum of what's possible, right?
Douglas: Totally. And also, I think, like, just to hammer the point home and then we can move on… I remember back maybe 2017 or so talking with a really good classical music conductor about this. And I realized that what she was presuming was that I wanted everyone to use AI. (Laughter) Like what? Let's just take one of the, like, first chair oboists in your symphony. It's fine. Like, not only is it fine, but why on earth would they use AI if they're so good at what they do?
Melinda: Of course.
Douglas: There’s no need. Like, then we need to invert that equation. Maybe in the next gen, Gen alpha kids will come along and they'll want to do something that we didn't think about. And I'd love to be the person that helps provide tools for them, but I think it's a very personal choice and a very like, interesting and deep choice, how you tie what you're trying to say as a creative act to the technology that you need. I mean, I think we need technology, even an oboe has technology. So like, except for your voice alone in a room, there's not much you can do musically.
Melinda: Sure.
Douglas: The first, you know, first human… you know this, right? I'm drifting a bit, but that the first known human artifact -
Melinda: - was a musical instrument?
Douglas: Bone flute.
Melinda: Uh-huh. That makes sense to me.
Douglas: We've been using repurposing, you know, doing technological advances to music since the Stone Age, right?
Melinda: Exactly.
Douglas: You can fact check me on that. I was told that's true. I did not find the bone flute to verify. (Laughter)
Melinda: Alright, well, we'll check it out. But it's interesting that you say that because, you know, in our kind of conversations, prior to this official recording, we were kind of talking about some of the differences you were noticing, between what kind of AI tools for music generation, versus AI tools for video generation. Right. And that, you know, there's something about playing music in a room with people in real time.
There's something organic, and that's part of the joy of it. Right? And yet, you know, you were suggesting there's been like, a much greater sort of interest or use in some of these video generation tools because they allow people to bring into being visions that they couldn't realize other ways.
Douglas: Yeah.
Melinda: Is that a good like, sort of -
Douglas: Totally true. And I think that is - that's exactly what we talked about. It’s, you know, one of the ways to look at this that is, I think, equally important for any artist or musician is, you know, how do you afford to do what you want to do? I think there's a fundamental model, which is, I have this idea clanking around in my head, and I want to get it out, right. And if my idea is that I want to do, you know, I want to film a car chase in the desert, you know, that looks like it's out of Mad Max. (Laughter) Turns out that's pretty expensive to do if I need to really go out in the desert. You know, it's a lot of money.
But, you know, what we're seeing is that particularly - I think there's at least some relationship to how old someone is, but really, I think it's about how progressed people are in their tooling, and where their journey has gotten them. Folks that are earlier in their journey are often looking for tools more actively because they haven't landed on a process that works for them to express. So we're seeing like, he's - and you know, you can find all of this in social media, just people commenting on, using videos - like, “I had this idea in my head and for the first time in my life, I can see some of these ideas that I've had in my head showing up in video.”
And, you know, realistically, on the music side, I mean, while it it's expensive to buy a good guitar, what it takes to set up a technology chain that looks like maybe what the Beatles had is not that expensive, compared to what it takes to film something in the desert. And so, people just don’t need AI to express themselves. They don't need to use it.
That's - I don't know if you want to go in this other direction, but you mentioned something else that strikes me as deeply true, which is, that the thing that strikes that I love the most about music is playing with other people. I want to quote a musicologist, Vijay Iyer, who really influenced my thinking on this, and he said crisply - had a chance to meet with him many years ago at the beginning of Magenta - I don't want to misstate - I believe he's still at Harvard, and he said, “Music is about co-presence.”
Melinda: Yes.
Douglas: It’s about all being in the same place together. And very fundamentally, our auditory systems work such that we're all immersed in the same auditory world. Our visual systems are different. I can continue this conversation, and I can look out the window, and now my visual world is different than yours, but our auditory systems, we’re all swimming in the same auditory soup, so to speak.
Melinda: Yeah. That’s a great point.
Douglas: Yeah. And so this idea that, like, we use music for co-presence, whether it's singing to a baby, which all mothers do, or at least, all cultures have exhibited this… or whether it's in another sense, using music or rhythm to keep people in the army synchronized as they march (maybe a less fun use case than singing to your baby)… or just jamming together, right? It's all about being in the moment. Music lives in time. I don't know how important it is, but it’s just, I think about that a lot, and how different that makes music from the visual arts.
Melinda: Yeah. Yeah. I think that's so true. I know for me personally, you know, a huge part of it - I mean, I can do stuff on my own, and it's important, you know, if you're writing music or whatever, sometimes - we talk about it a lot on the show - some aspects of creativity are quite solitary, and other parts are quite collaborative. But the real joy for me is playing together with other musicians and seeing what each person is bringing to the table, you know, with any given song or composition, and it's that synergy that gets me really excited.
Douglas: Is that synergy the same as the syn that I see in Syncreate? (Laughter)
Melinda: Yes, it is.
Douglas: I'm asking you a question I don’t know the answer to. (Laughter)
Melinda: Synergy and co-creation - that’s where Syncreate came from. Exactly. Yeah. Yeah. So, well, there's so many things I want to ask. But I'm curious a little bit more about your journey. Maybe you don't get to speak about that as much, but, you know, we've known each other since we were both graduate students at Indiana University back in the day. And, I'm curious, you know, how does a guy from South Bend, Indiana, get a PhD in Cognitive Science and Music Cognition, and then end up at Google developing AI tools? Like, could you ever have imagined when you were a kid where you would be now?
And I'm curious because, you know, for this show, we talk a lot about the creative process and the creative journey and like, how does somebody get from A to B to C? Like, did you have a vision in your mind as a young person of what you wanted to do, or were there different kind of synchronicities that happened along the way?
Douglas: Wow. Yeah. Where to begin? So first, my imposter syndrome runs high. (Laughter) Don't tell anybody. And I have to throw in that the PhD was in Computer Science and Cognitive Science.
Melinda: Okay.
Douglas: Colleagues at Google will lose all respect for me, you know? But no, I think, there's a bunch of general themes here. One is, I feel very grateful to have had parents who supported me but did not, for lack of a better term, guilt-trip me. I got to kind of roll my own adventure with my life. My father - both of my parents have passed away - my father was an electrician. He worked for the public utility company. Pulling meters and repairing them. He was high school educated. My mom was a secretary for insurance firms. They were both wonderful human beings, but they weren't the kind of parents who had mapped out my life for me. Right. “You have to be a doctor.”
And I've met so many people along the way who have built their careers based upon their parent’s expectations. I had the - I got very lucky - I had parents who were supportive, but not directing me quite so much. And so, I had to sort of set my own compass. I think I was, you know, kind of an odd bird. I think everyone knew that I was doing something creatively that other people weren't. And I think you see that in people. I don’t think it really reflects as being, any more than saying, “Oh, I was particularly athletic…” or I was particularly - which I wasn't particularly athletic.
You know, I think it's just an interesting trait. It's kind of how your brain is wired and what you're wired to do. My journey, I think, the short version of this is, I went to college. I was the first kid in my family to go to college, and I didn't know what I wanted to do. And I liked to write. So I was going to do a journalism degree, and it ended up being an English degree. I had no idea what I was going to do with it after I got out of school. But in parallel, I was fascinated with computers and I got a job working in the computer labs at IU (Indiana University).
You maybe went to some of those computer labs later. So, for the crowd here, Melinda and I met when we were both in grad school. (Laughter) This was the earlier version of me. So I sort of figured out the early internet. Very early internet, I mean, ARPANET and machines, like, before any of this had happened, and I just was watching this happen. And so, when I left school as an undergraduate in Computer Science, I was - sorry, strike that - with an undergraduate in English Lit. Creative writing. You know, I ended up getting a job as a database programmer, and and talk about fake it til you make it. (Laughter) Turns out you can learn to code. Coding is writing. I view coding as writing. Funnily. If you do one…. And if you're writers out there and you've never coded, it's a kind of very structured essay writing. Now there’s a level to it. You have to kind of understand what these new words mean -
Melinda: - And you have to learn the language.
Douglas: Yeah. Yeah. Totally. But it's absolutely accessible and even more accessible in the era of having ChatGPT and Gemini, and Claude, and all these tools to support that. And so, once I had that - I was also a musician, I was playing guitar - self-taught with some lessons, like… and I just ended up running that out until I got bored doing that. My musical career pinnacled in Albuquerque, New Mexico. (Laughter) Sang in coffeehouses for bagels and a cup of coffee, and writing my own songs and singing with other people. And I just realized that I could - I had to do something else. I just didn't know what.
And I was so fascinated with this nexus of music and and computing. It was just so self-driven. Like, I - for those of you, probably if you're listening to this podcast, you care about creativity - I was at that kind of like, you know, it's like that boiling over point where you, like, just have this, you have this thing you need to do and you don't quite know what it is, and you’re figuring it out. And so, what happens within these times that I guess - I'm sure it's a pattern - but something will unblock that for you. And that's, did you call it synchronicity?
Melinda: Yes.
Douglas: The right event at the right time. And my synchronicity moment was talking with my then girlfriend's mom and saying, you know, “I'd really like - I'm really fascinated in technology, and AI and music. And I’d just love to like, go back to grad school and do a PhD and actually do this as a researcher.” I think, and I was 24 at the time, and I knew it would take me six years because I was an undergrad in English Lit. So I have to do a bunch of, you know, basically have to do -
Melinda: - yes. Right. All the pre-reqs.
Douglas: Yeah. And she just looked at me and said, “You know what Doug? You're going to be 30 in six years. And you can be 30 with a PhD or not.” And it almost chokes me up now to remember that moment because it was just like, completely mundane advice. But that mundane advice. “Yeah. You’re going to be six years older whether you have a PhD or not, kid.”
Melinda: You might as well - (Laughter)
Douglas: I mean at that time, it was the thing that just sent me down this side of a ridge, and down the other side of a ridge.
Melinda: That's amazing.
Douglas: Yeah. And from there, it was all just iteration. Like, that set me on my path. And, you know, then you just keep hill climbing and keep trying out new things.
Melinda: Yeah. Yeah.
Douglas: I can keep going, but I think that's a good moment to stop.
Melinda: Yeah. No. And I think it's also a testament to, you know, just the profound effect that talking to someone about your ideas can have, you know? And the feedback we get, and even a casual seeming conversation can change the course of your life.
Douglas: Absolutely true. I'm not a mystic… like, I'm not a - I'm a fairly straightforward person. But this idea of like, synchronicity as I'm defining it, which is probably a shallow definition, but just where you are at this place, at this time… a quasi random event sends you in the right direction. It's just, it's so true. You see it all the time in life.
Melinda: Yeah. And that's part of the - I don't know - the sort of chaos of creativity in everyday life, right? Okay. So, I think probably a lot of people… kind of/sort of circling back a little bit, you know, to the original question… we keep hearing all these headlines about how AI is going to take away all these jobs, including creative jobs. So, like, how can people stay ahead of the curve? And you kind of hinted at it earlier when you were talking about kind of the generational thing. And like, you know, I think part of it is not being afraid to play with the tools and experiment with them, right?
Douglas: Yeah, I think that's - so, first, let me say something that I believe in firmly. There are problems in the world that we should look at as a society and address. And those problems, I think as - let me think like a tech, like a computer science person for a minute - let me walk you… it’s really important in framing - why AI. Like, I think the number one problem that we face right now is climate change. We could argue - maybe it's number two, and there's something else, political stability, may be number one, etc. - But like, this is where I'm, you know, this is me just talking as an individual here, just to stress that… but like, we could make that list, and we could fight about the ranking, but like, probably most of us would agree on the top five.
Now, there's another list that you can build. Which is the list of things that AI can solve. And that's going to be a very different list, because the first - number one, climate change is like, there's a lot, you know, maybe pushing battery technology, maybe, you know, changing how transportation works. There are things we could do. I so firmly believe that we should as a society, be focusing on that first list, and we should have the humility to say, whatever technologies you bring to the table, if they help with 5% or if they help with 10% of the problem, focus on the hard problems, right?
And if not, if we don't do that, then what we're doing is a classic error, which is designing for the technology at hand. Into cities, for cars in the 40s and 50s. So I don't want to design societies for AI. At the same time, I wouldn't be doing this if I didn't think there were very, very positive outcomes. Just look at what we're doing in biology with AlphaFold that can come from this, and that can help us build a better world.
So, I wanted to frame this because I'm not a techno-utopian, but I believe that what we're doing with AI can and will lead to socially, you know, beneficial, good. And I'm sure there are people on this podcast that disagree with me. I understand that. But, you know, I think if we're humble enough to focus on these hard problems - I think health is a huge place. Biology and basic science is a huge place where we can basically further our understanding of the world. Use AI as a kind of simulated microscope to better understand the world.
And that, from that, as a scientist, I believe we will find solutions to other problems that we're trying to address, that matter for us. And I would… I'm so glad to get a chance to say that on your podcast, because I believe it, and there's not always space to talk about that. And we'll come back to the AI and societal change in a minute. But I wanted to get that out of -
Melinda: - Well…and it makes me think of, you know, there's that adage which has been said in different ways by different people, that, you know, sort of “What got you here, won’t get you there.” Or, kind of like Einstein said, you know, “We’re not going to be able to solve the problems at the same kind of level that we created them with.” Like, we're going to have to go somewhere. And these tools offer potential opportunities. Like, you know, I've heard you speak about the potential for modeling things using AI tools that we wouldn't have necessarily been able to model before, right?
Douglas: Right. So, I just looked up a stat. I snuck in a quick... (Laughter)
Melinda: Okay. Sneaky devil.
Douglas: I don’t want to carry on this discussion. I want to like, if it's okay, we'll go back to the question. I'm not trying to dodge the question about, “Wait, wait. What about AI? And what do we do here?” So, I think about other large societal changes. One of them, and I think the scale of what AI could change in our society is probably at least proportional to what mobile did and what the internet did. But fun fact, I'm becoming a farmer… or maybe I should say, I married a wonderful woman who's becoming a farmer, (Laughter) who loves tractors.
Melinda: Shout out to Amy. (Laughter)
Douglas: That’s right. And it's quite it’s eye opening. So one of the reasons Amy wants to work in farming - we've - I'll give a shout out to our small farm in Sonoma County called Hopyard Farm. It’s just believing that maybe a few more people - there's room for a few more people to do farming and to have small scale farming having more impact on the food that we eat, than large scale farming. And I'm mentioning this because if you think about AI, there was a sea change in farming. I just pulled up this stat in the United States: about 83 to 90% of the population farmed in 1800, right? Most of us farmed, and now most of us don't farm.
Melinda: Right, and most of the farming is controlled by large corporations and not individual people or families, right?
Douglas: That's right. That's right. And you know, like, this is not a podcast about farming, and I'm not equipped to be interviewed on a podcast about farming. (Laughter) But fundamentally, if you think about, there was a lot of societal upheaval that happened when people had to move to the cities, during the Industrial Revolution. In some sense, the Industrial Revolution was the real draw here, pulling farmers out and then providing technologies to change things. There's - I have two minds about this - at the same time, you know, we have - child mortality rates fell from right around 50% across the globe to much, much, much, much lower.
I happen to be a city person. I am inspired and I love San Francisco, where I live, and I love New York, and I love the culture and what was created via this change, right? But at the same time, I'm very conscious of the fact that life became very, very difficult for a lot of people when they had to move to the city.
Melinda: Yes. Without a doubt.
Douglas: Right. I don't - my belief, or at least my informed hope, is that, despite what many people have been claiming, who are really kind of pumping up this moment - there will be job shifts and there will be adaptations to be made. You know, I think they're manageable, but they're real. Like, they're real. I don't want to understate - they're real. And even in the creative industries, there are jobs that won't exist in five years.
Melinda: A lot of writing jobs, design jobs. Things like that, yeah.
Douglas: And as a writer, you know, that's troubling. My belief is also that what we do… like, there's such a great and brilliant and lovely non - the way I call it - our non-zero sum game. Which is basically that if we solve one problem, we work on another one. And that these transition periods get minimized, and their impact is minimized by that. So by that, I mean that, like, when we eliminated - there used to be whole floors in office buildings called secretarial pools, and it might in, like, you can see, if you just Google for the secretarial pool from like, a 1940s office building.
It's these massive - you know, they invented the open workspace. A bunch of people, generally women, sitting at typewriters, having handwritten documents handed to them, and they type them. That whole world is gone, and that whole - but at the same time, I think, you know, people have worked on different things and skilled for different things. And I don't really want to go back to a world of secretarial pools, and I don't really want to go back to a world of the agrarian, pre-industrial society.
Melinda: Right. And change is always happening. It always has. It always will.
Douglas: So then, how do we ride out this one? Well, first, I mean, I think it's on everybody to lean in a little bit to the technology that's coming at you. Whether it's an automobile or it's AI. And the analogy that I love to use is an avalanche. (Laughter) It’s not a very good - it’s a dark analogy. But, you know, the reason I use avalanche is that it's a point, it's an event that happens in time and has a beginning and an end. So if you're in a particular mountain range -
Melinda: - and also a very forceful event.
Douglas: Yeah. That’s right. So things happen fast, and happen forcefully, powerfully. You know, if you're in the mountains… there's an avalanche on Wednesday. If you were there on Tuesday, you're fine. You ski right through it. If you’re there on Thursday. Well, okay, you know, at least it's over. Like, there's probably some debris left behind. But if you're there on Wednesday, look out. The folks that are there on Wednesday, in my analogy, are basically kind of late career people, middle to late career people. Everybody, but I think folks that have already built their tooling, and when that happens, the ground gets pulled out from underneath you.
And you can all, you know - forgive me for the corny analogy and pushing it too far - but you know, avalanche survival is about swimming and trying to keep your head above the snow. It’s all about just run, just fight, fight, fight, fight, fight. If you just ride it out, you get buried. And I think even though it's a corny analogy, there's a kernel of truth to that, that there are so many ways that people can augment what they're doing with technologies as they move through. And staying a little bit ahead of that curve puts you in so much of a stronger position than kind of sticking your head in the sand, you know.
And so, you say, you know, even if it's not something you love, it's on you to understand where technology is going and what that's doing for society. And then it's on us as a society to build out the policies. And I'm not here to comment on policy or basic income or all of these things, right? But, you know, build out the world we need to build out as technology supports us in solving some problems and, you know, creates new ones.
Melinda: Yeah. And I think, you know, as we were sort of talking about earlier, there's always technological tools that are being developed, and all these kind of things. And then at the same time, there's also these sort of like, old fashioned people-centric things, like playing music in a room, or, you know, human interaction that can't be replaced. Of course, we hear a lot of talk about, you know, AI therapists and people having AI relationships and things like that. (Laughter)
Douglas: I think that's a terrible idea. But whatever. (Laughter)
Melinda: Right? Right. So, I'm curious with some of these, like, AI tools specifically for creativity. You know, obviously there's tons of pros and tons of cons, and we could talk about a lot of these different things, but I'm curious, kind of, you know that old adage by Marshall McLuhan, “The medium is the message.” I'm curious what you think about, you know, as we see more and more AI generated, especially kind of images and video, you know…
I saw an article about how they have sort of AI [fashion] models now, and maybe modeling as a career is going to go by the wayside, or things like this… as we see or as we are exposed to more and more of this, AI generated visual material, and the visual is such a powerful perceptive field, right? Like you know, how will this affect our own understanding of ourselves and what it means to interact human to human?
Douglas: Absolutely brilliant question. So first, I wake up every morning, as I'm sure you do, and wonder if my chance to be a professional model is finally over. (Laughter)
Melinda: Right! Well, you know.
Douglas: It’s all about AI taking that chance away from me. It’s not me, it’s the technology, Okay, seriously. Answering this question. There’s like 10 questions in what you're saying.
Melinda: Sure. Pick whichever one feels most interesting.
Douglas: So, first, I love the comment about “The medium is the message”. And what that requires is that we experience the medium long enough to get the message. So, like, “The medium is the message…” couldn't happen with year one of television. It couldn't happen with year two. What you do is you go, “Oh, we have been experiencing television for long enough that in and of itself, it is influencing how we communicate.”
Melinda: Yes.
Douglas: Right? So, I have a very hopeful view of that, and which is that humans are very, very good at being selective. And in fact, a lot of what Marshall McLuhan was worried about, we've rather successfully rejected. Right? The way that the medium defined American humanity when I was a child is not happening anymore. Like, you remember this - like you memorized the TV shows that were on Thursday night, right? (Laughter) We had TV dinners, we had TV sets, we had TV Guide, right?
The whole world was configured around that and that allowed TV to shape the message. By the way, I'm aware to those of you, that I'm not really speaking to the heart of Marshall McLuhan's message, but it's the one that I want to run with here, which is that, as we experience AI, we both develop antibodies to what we're not liking about it. Hence, the term ‘AI slop’, which I'm very comfortable with. Dump a bunch of new crap out there. Great. And, you know, I think there's a version of AI slop - I hope you can make this connection with me - the mobile, the iPhone, and, you know, I happen to use an Android because it's where I work, but I actually like Android, for what it's worth.
But, you know, like, this created kind of ‘photo slop’. Like, the fact that you could just take 1000 photos at the beach, or you could take 1000 photos at a party, people are like, “Yeah. Where’s the curation in that?” So I think in some ways - even Instagram, in its earliest days, when Instagram was really about sharing your photos - kind of was pushing against this idea that you would want to just share a folder of 1000 photos. You know, it's like the modern version of that person when we were kids, who's like, “Do you want to see the slides of my vacation?” Hour later, they're like, clicking… “Please stop”. Like, that is so easy to do now with a mobile phone.
So, we've kind of self-regulated that. I think I would call that like, ‘iPhone slop’. You know, it's something that this technology gave rise to. By now you have, like, the next generation - not even like - kids now that are in their 20s, shooting on film, because they want that constraint. And at the very least, like, you watch the culture of how often you post, how often do you update your Insta, right? Like, it's all very metered.
Melinda: Well, and how do you choose what you want to actually put out there?
Douglas: Absolutely.
Melinda: I mean, there's still a curation aspect to it, and you're not just going to dump your entire phone full of photos on Instagram. Hopefully. (Laughter)
Douglas: I think this, like… the self-regulation that we get with curation is already hitting… you know, ours is Gemini, ChatGPT, Claude, in the sense that I think people are kind of going, “Oh, yeah. That was written by Chat.” I don't want to read it. Maybe I'll have to joke, like, “Maybe your chatbot will contact my chatbot, and then we'll, you know…” The point is, like, so… that is like, a societal immune system response that takes time to build, right? Because you have to experience this new world for a few months or a few years before you can build them. And I applaud those being built. Even though I am a big - I think AI has a great role in music making, I have no problem with somebody doing it. Like, I don't want a single shred of AI in my music.
I got my guitar. You want to sing, yeah. And I see that happening across the board. So that's only part of what you were saying, which was kind of the Marshall McLuhan messages that is the medium. And like, how does that… what’s happening with AI? I think we're already seeing great signs of us building up the immune response, so to speak, or just like, rejecting what we want to reject as a society. The voice is very strong about, you know, how and when we want this. Then the question is, what do we build and how do we pull forward? You can steer me differently, but I would switch and talk about education right now.
Melinda: Sure, let's do it.
Douglas: So, what I'm seeing… I'm very, very excited… my combination of fear and excitement about education with AI… I worry about where we are with the current tools. It's just really easy for like, lazy kids (by the way, that means all kids), sorry. Lazy kids -
Melinda: Including graduate students who I teach, you know? (Laughter)
Douglas: “I don't want to do my homework. You know, I’ll let Chat do it for me.” But you know, I think we're already seeing glimmers of like, very interesting, self-guided educational experiences be made possible. I believe - like, I'm always trying to advance as a guitarist. I'm not a particularly good guitarist. I mostly play rhythm guitar to sing, but, you know, over the years, I've been trying to get better at playing lead guitar. Kind of bluegrass and also country, and then also, you know, straightforward rock and jazz. And like, we're so close to me being able to go in and say, “Here are five books that interest me about how to learn the guitar. Help me build a curriculum. I want to practice two hours a night. Help me judge what I'm doing.”
Like, all of that, right? Like, we're just getting on top of this hill of understanding this technology. We can open up a bunch of, I think, really inspiring directions to go. And so, the reason I focus on education is it's definitely a double edged sword. We definitely see that like, kids overwhelmed with social media and having access to AI to solve problems for them may find themselves in a position where they haven't been appropriately tooled to be critical thinkers. Or even writers, which is really something to worry about.
At the same time, if we can get through that and build out the right tooling and the right constraints, then, you know, I think we can actually use AI as a really great accelerant for learning. And actually offer kids - especially kids that don't have quite the same opportunities - access to really high quality, personalized, in real time, training opportunities. And that's inspiring.
Melinda: No, it is. And it's interesting because it's sort of anecdotally, when I talk to people, just in my day to day world, I can't believe how many people are using these Chat platforms in really interesting everyday ways, as a tool of like, personally directed exploration, right? Like, “How can I refine my business model?” Or, you know, “I've got these ideas. I'm not quite sure how they all fit together. I'm gonna throw 'em in the Chat and like, see…” You know, and that they're getting inspiration and able to kind of synthesize and think about things in different ways through their sort of self-directed interactions.
Douglas: I love that. I love that. And I love the self-directed part of it. So, I predict that, you know, we'll see more and more ways in which (a) teachers can do their jobs. Part of their job is to be a critic and to say, “You did this right, and you didn't do this right, and I'm going to give you this grade.” And having seen systems where the kids don't have that sort of critical feedback, I think, you know, we need that. But as we get better and better systems for doing that, I think AI can be a really, really helpful goal.
Melinda: Well, this has been such a fascinating and far ranging conversation, even beyond what I was kind of imagining. (Laughter) And I want to not - to end it on like, a dark note - but I think this is an interesting topic that doesn't always get a lot of attention, because we, you know, (myself included), tend to talk about how wonderful and amazing creativity is, and how great… everyone’s creative, and you know, la, la, la. There's a dark side of creativity, also. And we can talk about this in terms of technology and just creativity more broadly, because creativity itself is kind of value neutral, right?
Douglas: Totally.
Melinda: We can use our creativity to do a lot of wonderful and amazing things. And as we were talking about earlier, to be, you know, a benefit in the world and solve complex problems. You know, we can also use our creativity and some of these tools that the technology is offering, you know, to create destructive things, right? And I know that some of these platforms or these tools have certain like, limits sort of built/programmed into them so that, you know, there's certain places, you sort of can't necessarily go with them, and then we get into questions of, like, censorship and things like that. So, I'm just curious what your thoughts are on the potential dark sides of some of this.
Douglas: Yeah. (Laughter) So, I'm a, you know, we’re… very sensitive question and very good question. So, first, as a parent, it's obvious to me that we need strong limits on the kinds of media that people can create and circulate, right? And there are already, you know, laws in place for this, and we may need refreshing. And I definitely don't want to go down a policy direction. It's too close to the hat that I wear at work, and there's just, and I'm not - it’s not what I do. And as a parent and as a technologist. And I’m just a human being on this planet with everyone else. Yeah. Of course. Of course.
You know, I think we need to, as a society, just set very, very strong social guidelines for what's acceptable and what's not for particularly video generation and image generation, but also for text generation, and the downsides of what it means to be able to do things like control robots. One thing I would say is that while AI opens new doors to this, both good and bad, we’ve been dealing with this for a long time. There are some shocking stories to all of this.
One of them that I just read yesterday, and it actually made me sick to my stomach. Child pornography was not outlawed in all states, even in 1970. You could legally buy magazines of child pornography. And I was alive in the 1970s and I think where we’re sitting right now, that is just like, are you kidding me, right? And so, our social norms are constantly moving. We're constantly correcting, and we need to absolutely be 100% solid on making sure that we don't create situations like child abuse downstream from these kinds of tools.
Melinda: Of course.
Douglas: I focus on this mostly because I read about it yesterday. There are many other problems, but it's a big one.
Melinda: Sure. Sure. Of course.
Douglas: And it's like, it's like a red line drawn in the sand. You know, we have to absolutely make it a felony to create these materials, and we should throw people in prison that create them. It's very simple to me on that side, and we should not be building tools that enable that.
Melinda: On the other hand, you know, I know of instances of creators who feel frustrated by some of these limits, not because they're trying to create child pornography or anything like that. But you know, what would you say to a creator who feels a little bit stifled by some of these imposed limits?
Douglas: I see this all the time myself. So, the filters… so, part of this is like, getting the technology to work. You know, the basic technology is you first try to train the models to do the right thing. Let's talk about next generation model. Don't generate obscenity, and if it does, you filter it. So, the model has generated something. Then there's a filter that looks at it that says, I'm not going to give that back to the user. Yeah. Like, when those filters, we call it ‘over firing’, when they generate false positives. Super frustrating. You're like, yeah, trying to make a picture of a mushroom, making a joke. And you, I don't know why… but it's super frustrating, right?
Look, we'll get better at that, and all companies will get better at that. I'm not here talking about my company. The world will get better at that. I'm perfectly comfortable erring on the side of caution here and communicating, like, as we get more and more sophisticated with building models… for example, we'll make it possible to make PG 13 content, or maybe even rate our content, without opening the door to worse things happening, you know. But there's a lot of work to do there. So, I guess what I would say is, overall, I'm happy. I would rather live with the frustration of over filtering and over censoring, than have the opposite problem. And this is an iterative game where we just get better and better at it. I also say that different companies will have different lines they'll draw in the sand. Be more or less conservative.
I'm actually openly - I've openly said this, and I'll say it now - I'm quite conservative about this. I'm very happy with content that could be damaging being very hard to generate, even at the expense of limiting the nearby content. Especially everyday users. but I'm also happy to keep improving things. But, yeah, it's really scary. Like, you know, we really need to do this right. It's on us to do this right. Because, you know, the potential damages are high.
Melinda: Yeah. Okay. Finally, I just have a personal curiosity. And I don't know if you've run up against this. You know, we hear all kinds of stories, again, on both sides of, like, you know, sort of AI platforms going off the rails, so to speak. But then, on the other hand, I've also noticed, is there a little bit of a positivity bias built into some of these platforms? For example, you know, giving us positive confirmation or telling us what they think we want to hear. Or -
Douglas: Yeah.
Melinda: Yes. Right?
Douglas: Great question. (Laughter) I love that question. You're a genius. (Laughter)
Melinda: Haha. Thank you!
Douglas: It's all too easy, right? Yeah. Getting like, I like that. I don't have a personal connection with any chat bots. Like, I don't interact with them in a way that I get that kind of feedback. And the way that we form empathetic relationships with other people that we're communicating with, makes me nervous that people form empathetic bonds with -
Melinda: Yes. Someone was just telling me yesterday they know someone who is in a romantic (active romantic) relationship with a chat bot.
Douglas: Right? (Laughter) Frankly, it makes me really nervous, and I am not - I don't understand the world well enough to be supportive of the… products are being built for doing this, too. It's not my world. I mean, I work in technology, but that's not my world. I won't go and say it's bad with a capital B, and never, ever, ever do it. But I just don't think we understand well enough how we connect to each other to try to replace that connection.
Because, you know, I could argue the other side with maybe there's a role for helping combat loneliness, especially amongst the elderly, by having empathetic bonds with chat bots. I just don't know that we've done the necessary research and vetting to do this well.
Melinda: Yes. It’s worth exploring. And of course, you know, when we engage in relationships with other flawed human beings, all kinds of things come up. And maybe there's a way of practicing positive relationality that we can then bring back into our in-person relationships.
Douglas: This brings me to one other related point that I wanted - I'd hoped - we'd talk about. This relationship that you form. There is a relationship that I form with technology that I think is really healthy, for me at least, which is a lot of what I do that I consider creativity is actually flow.
Melinda: Yes.
Douglas: Right? And flow for me is a kind of mindfulness. And so, what I've been… when I play piano, I'll have a glass of wine, I'll play piano, and I'll just improv. I’m almost always alone when this happens. (Laughter) And I almost always lose track of time, to the point where I've sometimes gone, “Oh, I think I should turn the lights on. It got dark”.
Melinda: Which is like, a beautiful feeling.
Douglas: And that is a kind of, it's not an empathetic bond, but it is a loop that I'm forming with this piano, which is a piece of technology, right? And like, that kind of creativity… I’m sure we didn't go in this direction, probably because, you know, this is what you talk about. This is what you do, is think about these… and so this is kind of about AI. But for me, I think there's a very interesting and hopeful role as we do more and more real time generative experiences where you are controlling it.
So, in some ways, the AI becomes not a generator of materials that you then edit, but, like, it's just this thing that you're playing. You're kind of playing it like a musical instrument. Kind of like an adaptive video game or something like that. I can imagine, like, really finding a lot of like, flow value in that. That's creativity but it's like, you know, as we've talked about - we talked about this in the past - I'm not going to record that, edit it and share it with others. It's me.
Melinda: It’s part of your process.
Douglas: That’s right. And to be completely honest, I think for me, that's like, at this point in my life, 90% of the game for me with creativity. I'm doing very little. I don't have… I'd like to do an album. You know, I've written a few songs. I'd like to do - you know, it's like, maybe I retire and I do some albums. Like, that's edited, that's like, craft, that's working to get an artifact out for other people to enjoy.
Melinda: Sure.
Douglas: I want to spend the rest of my life doing intentionality via flow-based creativity and improv for me. And I think there's maybe an interesting role for AI in that, if we get it right. And fundamentally Magenta, which I'll put another plug in for -
Melinda: - Yeah, sure. And we'll put a link in the show notes.
Douglas: We have a bunch of open source stuff for, you know, around, like, kind of real time treating AI as a musical instrument. A cool (maybe possibly coo), new instrument that we could play with and experience for a few hours at night, you know, with a glass of wine, watching it get dark, you know.
Melinda: I love that, and I think that's a great kind of point to end on. And I usually like to end each episode with what I call a Creativity Pro Tip. Something that people can go out and try on their own.
Douglas: Okay.
Melinda: So, I think this is a perfect example. Whether it's Magenta or, you know, any of these other AI tools for creativity, just check it out. Play around with it. Not with any goal in mind, right?
Douglas: Like, flow. Like, do you know… there's vibe coding, right? So it's like, vibe coding… if your listeners aren't familiar, is basically using a chat bot to help you write code. You just kind of keep vibing with it until you get what you want. Actually, largely doesn't work for complex software development projects. Hint. (Laughter) But the idea that you can just kind of be like, “I'm just going to go on this journey and with some AI tools that are built for that. And like, the stakes are low. I just want to see. Like, this might be fun, and I might get something…”
Melinda: What can I discover? Yeah.
Douglas: Exactly. I think that's a really good way. That may be a good entry point for some people to experience this technology in a way that might provide, at least… I mean, what's the worst case? You're like, “I'm an hour in. This just sucks.” Right? (Laughter) Like, it's not the end of the world. You've given one hour of your life to trying something.
Melinda: Yeah. It can't hurt.
Douglas: Like the trumpet, you know, right?
Melinda: Sure. Not everything sticks, right? Well, Doug, thank you so much. This has been such a fun conversation. And if people want to find out more about what you're up to, including Magenta, for example, what are the best ways to find you?
Douglas: Yeah. So we have, for those of you that want to look back at the history of how we have approached music making and creativity, please visit the Magenta blog. I think I posted the first “Welcome to Magenta” blog posts in June of 2016, so we're coming up on 10 years of it.
Melinda: Amazing.
Douglas: Very research focused, and kind of gives you an interesting history of how we and Magenta team have approached creativity. And we can put it - I'm sure you can put it on the out credits - but it's magenta.withgoogle.com. And the main thing you want to go after is that our open source offerings and our blog where we have all of our old blogs listed there for you, it's kind of a time capsule. Now we've been doing this for almost a decade. I'm quite proud of what we've done. We've been creator first. We've been musician first. We're not trying to replace anyone with anything. We're just trying to have a blast with this technology and see where it can go.
Melinda: Amazing. Alright.
Douglas: Thank you very much.
Melinda: Thanks Doug.
Douglas: I’m glad we got to do this! After the Nile.
Melinda: I am too! I know. I know. It’s a real thing. Yeah. So thank you so much, and I hope to connect with you again soon.
Douglas: Fantastic. Thank you.
Melinda: Alright. Take care.
Find and connect with us on YouTube and social media under @syncreate. If you enjoy the show, please subscribe and leave us a review. We're recording today at Record ATX Studios in Austin, Texas, with Doug joining us from the Bay Area. The podcast is produced in collaboration with Mike Osborne at 14th Street Studios. Thanks so much for being with us and see you next time.
