Shift podcast | Season 7

Jon: This is Shift. A show about how change actually happens – messy, human and full of surprises.

Mikaela: There’s the official story, and then there’s what really goes on in rooms where decisions get made. 

Jon: I’m Jon Finkelstein, Executive Creative Director here at PwC Canada. I’m interested in the human friction, the weird, the ‘why are we doing it this way?’ moments.

Mikaela: And I’m Mikaela McQuade, Partner in our Policy and Economic practice. I care about the connection between big ideas and real impact – how strategy meets reality.

Jon: Our guests are people who are in the thick of it – Canadian leaders, builders, decision makers – the ones you’d expect PwC to access. 

Mikaela: But we’re not here to ask the usual questions; we’re here to ask better ones. The type that make people pause, laugh, rethink and sometimes say, ‘I probably shouldn’t be telling you this…’

Jon: Mikaela and I don’t always agree…

Mikaela: …and that’s part of the fun!

Jon: This is Shift, and it starts now. 

Jon: All right. Well, welcome to another episode of Shift. Wow. We're back. And we have an amazing guest today, David Usher, the founder and CEO of Reimagine AI. David, welcome to the podcast.

David: Thank you very much.

Mikaela: We have to start with our warmup question, which is, ‘what is your most embarrassing AI story?’

David: My most embarrassing AI story.

Mikaela: I can tell you mine.

David: Tell me yours. Yeah.

Jon: Her’s is funny.

Mikaela: I'm, like, relentless in trying to get AI to like me as people pleaser. And I'm just always saying ‘please’ and ‘thank you’ and really wanting to impress that with my prompts.

David: You know what? AI likes you no matter what.

Jon: That’s what I said! It’s so sycophantic!

David: Totally.

Mikaela: I know, but I don't want it to just like me the regular sycophantic amount. I want it to really like me.

Jon: So, do you have an embarrassing AI story?

David: I don't really. I mean, I use it every day, all the time, 24/7, seems like. But I don't really have an embarrassing story now.

Mikaela: Oh, what about you?

Jon: I don't know if I have an embarrassing story, but I do like to critique my poetry. And then I argue with it, that it doesn't understand me. So there you go.

David: It's just supposed to like your poetry. 

Jon: I just wanted affirmation. You're right. I do want affirmation. But, I'm sure you get asked this question a lot, so forgive us, but people know you probably for your music career. And that's obviously a big part of your brand and who you are and your experience. Now you're kind of in the AI space, and we're kind of wondering how and why you made the switch from creative music to creative AI. Where - what was that, transitional moment for you?

David: Yeah, a lot of people wonder that. Yeah. I mean, I think the transition for me came a long, long time ago. You know, I was living in New York. The band was really successful, and it was right around when, Napster hit the music business. And I really got very interested in to why, digital was rising and, the music business was falling into the sea, essentially. And that led me down the path of really understanding the Internet and really exploring the Internet, led me my first startup, which was a social media aggregation startup. So I did my first project with, with Google Brain doing AIs that could, collaborate with humans writing song lyrics. And this was about ten years ago, and from that I really got the AI bug. So about, I guess it was, 2018, I started to reimagine AI, my company. That was at the time building creative AI's for like, interactive virtual beings, for, museums and science centers and things like that. And slowly over time, we've transitioned to becoming more of a venture studio, working in healthcare and in something what we call digital immortality. So we've transitioned a lot over the years as the technology has changed.

Jon: That's one of the horrifying things. As soon as you get somewhere, it's like the tech changes and suddenly you have to go, ‘now what do we do?’ Do you do you find that that you start something, you go somewhere and then, ‘Oh, wait.’

David: Yeah. I mean, when we started, we were building these, AI was really just an idea, you know? Google was using AI, but it was it was not programmatic at the time. It wasn't things that were released into the world. It was more, it was specialized code, but not programmatic, where anyone could use it. And now we're finding that, the AI train is moving so quickly that it's very, very difficult to stay on top of it. So when we were first building virtual beings like AI powered virtual beings, it was really like these mass dialog trees that we were building to try and make these beings seem intelligent so they could really answer not every question, but almost every question. And as the technology has changed, it's been really our job as a company to stay on top of every new bit of tech that comes out and understanding every new change in the environment and what it means in relation to your tech stack that you're building and you're basing your business on.

Mikaela: You mentioned creativity a few different times, and I'm curious how that stacks up against coding and programmatic language and really building kind of machinery around this. Has your perspective on creativity changed fundamentally from moving from that initial kind of artist space in music to where you are today?

David: For me, not at all, because I think of creativity as a methodology. It's a way of looking at a problem. It's a way of working a problem, essentially. And it's the same. I use the same creative thinking when I'm building a song or a book or presentation, like a keynote, or with my coders. It's exactly the same process. But what you need to do as you're moving across verticals, is you need to change the language that you're using. You need to understand the language of the new vertical. Because if you don't understand the language of the vertical, you can't communicate with the people that you're working with. So when I moved from the arts to, you know, speaking, doing keynotes, I had to understand the language that we were all using. When I'm talking to programmers, my coders, I have to understand what they're doing and the language. But the great thing about artificial intelligence right now is that, you know, for everyone out there that wants to get involved with AI, it's changed so much so quickly that if you can get involved now, you can really be at the forefront of what you're building in your in your vertical or your industry.

Mikaela: It's interesting you say that it hasn't changed at all, because my perception of creativity has changed fundamentally over the past couple of years

David: Oh really? 

Mikaela: Yeah, absolutely. Kind of coming into my life and seeing it in action, both in the workplace and in my own personal life. I don't think it would be a stretch to say that a lot of people view creativity as something that might be kind of uniquely human, that it's a good thing about us, that it may not be kind of replicable by machines and in any way. Are we redefining that boundary, that are we seeing creativity in technology in different ways?

David: Yeah. But, I'm not saying when you put a prompt into an LLM that you're being creative. That's not my definition of creativity. But what I'm talking about is the way you build a product, whether that product is a song, or a piece of code, whether it's a program. But prompt to LLM I don't think that's creative particularly, no, but the understanding of how you need to prompt is creative. So that's a subtle difference. And prompting isn’t, even if you're coding…. Like at my company, everyone uses AI agents. We're all using it from marketing to business development to coding. We're all using AI agents all the time. But you really have to understand those agents and how to prompt them and how to make them work for you, still. That won't always be the case, but right now it is.

Jon: As a creative person, I certainly find that the definition of creative and taste and judgment and all these kinds of things, there's a commoditization that's happening with a certain level of it. The great average is, they call it right. 

David: 100%. I mean, creative industries are about to are in the middle of a grand transition.

Jon: Yeah.

David: We're about to see, and it's not that human creativity is going to be necessarily different, but the business models that live underneath that creativity are going to be disrupted completely. So the model that we built these industries on is going to radically change what AI really means in creative industries is the end of scarcity. What was scarce is no longer scarce. So in that world, what is going to be scarce in the short term? It's going to be the rough edges of things, right? The humanness of things. But AI is going to be able to replicate that. So my belief is what is going to be scarce are human spaces. I think we're about to see the rise of human spaces where humans want to experience things with other humans, similar to chess, right? Chess is the great example of AI can beat humans at chess all the time. But chess has never been bigger. Online chess has never been bigger, with humans playing humans because humans want to play against other humans. They don't want to play against a machine that can beat them all the time. So that space of chess has risen up. And I think that's going to happen in all sorts of other different art spaces as well. Different creative spaces for people.

Mikaela: Often you think of AI as a threat to that kind of community. Those human spaces, we come together. And so that's an interesting point.

David: I think that there's going to be a reshuffling for sure, and we don't know how it's going to turn out.

Jon: Seems like a great place to transition I think to get more of the business context a little bit. I'd love to hear what you're saying about, you know, the business models underpinning creativity or product or advancement or whatever word, we're using right now. It's going to change. So as you think about disruption, what do you think is going to be, you know, put your prognostication hat on, the most significant changes that AI are going to bring to business in the next decade?

Mikaela: So, imagine that you are in control of your organization. Technology is changing every part of your business. Your competitors are adopting it at breakneck speed. Everything is changing around you. You have a choice, essentially, to adopt and to kind of go headfirst into the AI economy and, and adapt your business accordingly, or to not or to do it in a, a little bit more deliberate. How do organizations, if you were sitting in that seat, balance the pressure to innovate and the pressure to compete, which is extremely high, especially in the global economic circumstances that we find ourselves in, with the discipline to pause and ask, should we be doing this?

David: Yeah. I mean, you have to compete, right? Not only is not only as organizations or companies, but as individuals, right? The individual has to compete within the marketplace. And if you don't understand, if you're if you're an individual, if you're an employee, if you're a student and you don't understand the AI tools, it's going to be really hard to compete in the marketplace. Now, I always look at it as a micro/macro. In the micro, and I consider that to be to the company level, everyone has to figure out AI - how it's going to work within the company, how it's going to change their company, because that changes everything, not only for the executive, but for all the employees and whether they're going to survive as a company. But in the macro, that's more about how we are going to regulate this new technology and how we are going to prepare for this new technology on a cultural, social level, economic level. Those two things have to coexist and have to be worked on at the same time. But I always tell individuals, you got to understand this technology, even if you don't like it, you've got to understand it. Like that's the thing. It's like saying, I don't want to use a computer. Well, you don't have to use a computer, but you're going to be going to have a trouble competing without it, without knowing how to use Microsoft Word.

Jon: Totally. Well, that's a great segue also into this idea, a question about, how do leaders or organizations prepare for this, especially the ones that are, let's call them laggards, the ones that are lagging behind?

David: Well, I think that this is the perfect time because the technology is moving so fast that what is, you know, that everything's new now. You know, what existed six months ago is irrelevant. I think the main thing and this is where the marriage to creativity really starts to rear its head is that why creativity becomes the competitive advantage is because you need your curiosity now more than ever. If you are curious and you want to learn, it's right there. You can learn the language of this new landscape very quickly. And once you learn the language of the landscape and you start to become involved in it, then you can participate in it. Once you're participating, it leads to a whole lot of things. But it begins with this, this voracious curiosity that you have to have in order to participate. You have to say, okay, I don't know anything now, but I can learn this stuff really fast because it's a simple language and then I can participate in this. And that goes for leaders, that goes for employees, that goes for everyone. Because the wonderful thing right now is that everyone is sort of struggling, trying to figure it out. But those who do have a huge advantage over those who don't.

Jon: Curiosity for me is, like, one of the best words, because as a creative, as someone who's trying to explore technology, come up with different ideas for clients, if you're not curious, then what are you?

David: It's the cornerstone now. Because in this world where things change so fast, you have to be able to absorb ideas very quickly. You have to be able to look at the landscape and absorb things from different places that you never might have had to before, because everything is shifting on you. You have to be able to absorb everything and then do an analysis on it, do a synthesis on it, and then come up with an idea. Right? Ideas are everything in this world, but they have to be informed ideas. So that curiosity, that creativity is really a competitive advantage, right now, I think.

Jon: Yeah. For sure. And for me, like the corollary of that is judgment, right? It's all very well to have an idea. Whether it's AI generated or you're using it as, you know, your sounding partner and all that kind of stuff, but be able to then layer on expertise, judgment to be able to make a decision as to whether or not it's a good idea.

David: Most of our all of our ideas are terrible. Right? Mine, yours. They're all terrible. But the more you're in, the business of creativity, the more - the quicker you can identify terrible and kick it out. And the faster you can hold on to those great ideas and, and push those through to product.

Mikaela: Are we losing anything like with this, this kind of relentless emphasis on efficiency and pace? When you when you think about especially the intersection of creativity and your arts worlds and everything that we're seeing today, like what do you think is at risk of us losing in that process?

David: I think we're losing for sure. But you know what? In my life, I mean, it's been amazing, like the creativity. So I'm having a blast right now because I love the tech and I'm very interested and I'm engaged with trying because we're a creative company building all sorts of interesting things or I think are interesting, anyway. For me, the creativity is, it's like it's an amazing place to be, an amazing time to be. But it's different for sure. I'm not sure if it's better or worse, but it's definitely different.

Mikaela: And it's interesting that you press on that because I think, focusing on where AI can be incremental, where it can add meaning, where you can use it meaningfully. When you evaluate projects that you want to do, how do you decide whether it's kind of clever and a new, cool use of AI and consequential as an artist?

David: Well, I'll run through a couple of projects that we're working on. We're working on virtual beings on tablet for Alzheimer's and dementia patients. Right. So if you think my mother has Alzheimer's, she’s had it for seven years. And anyone, any person out there who has a parent with Alzheimer's understands the guilt, right? If we can make that work, that's a win. Our other big project is called Second Echo, and it's a platform where you can build, it's a new way of visualizing how we remember and memory. So if you think about how our, you know, our grandparents recorded their memory, there's almost nothing recorded. And for us living in this era, we have too much stuff. We have endless photos on a reel that we never look at. We have posts all over the place and journals maybe, and what we're using is modern AI technology to connect all those things, to build a digital echo of yourself. So it's a new way of remembering, a new way of journaling, and we think that that has value in this new world.

Jon: I'm on the waiting list.

David: Oh, you are? Okay. Well, we're just adding people to the beta all the time. 

Jon: I mean, come on, dude, how long do I have to wait? 

David: It’s only been like a week.

Jon: It feels a lot longer in internet time, man.

David: Yeah.

Jon: So, as organizations start to adopt AI kind of at the enterprise wide and make it an expectation.

David: Yeah, it's difficult, but I would say that you have to think of it as an investment. If you're going to try to get your workforce to understand and use artificial intelligence, you're going to build a skunkworks within your company that can experiment with all the new tools that figure out what programmatic AI should be used and what shouldn't be used. That's going to take investment, right? You have to invest in this over the long term. But if there was ever an investment to make, that's the investment to make, and now is the time when you have to make that investment over time to build the programs that, you know, build literacy within your company, to build the skunkworks that really can test everything that's coming. Because for most businesses, for most enterprise businesses - there's some small startup using AI, that is really coming to change the way business is done. And if you're not competing within that same framework, with artificial intelligence, there are going to be a lot of people coming at you from different directions, I think. We have to think of it as a brand new technology. The way I look at the stack, is this: We used to have the Internet as the base layer. And then we had all of these applications built on top of it, right? Whether you're talking about specific applications or phone apps or SaaS businesses, whatever. The way it is now is you have the Internet, then you have what I call the LMMOS, the large language model operating system. And then on top of that, you have a million wrapper companies that are going on top of that, building new things that utilize this new layer of AI that lives between the internet and their application, and they're building all kinds of crazy stuff that no one's ever thought of or have, or many things that people have thought of, but in different ways that take a lot less manpower, that are going to change the way people do business.

Mikaela: So then it’s really not just about what people do, but AI is changing how people learn and how they think, right? And I, like you talked about the lowest hanging fruit. AI eats the small pieces from the bottom up. What kinds of human skills become most valuable at a workplace that's being disrupted by this technology?

David: I really think it's those two pieces. It's the curiosity. People that want to learn this stuff and want to be the people that control the operations and people that can use their creativity to synthesize data. You can see the pieces, you can see these new pieces and how they fit together, the new technology, how it fits with the old technology. If we're building things, we're usually building in groups, right? What are the key parts of a group that still need to be utilized? What is what is now no longer needed? What pieces of technology that were old are needed? What are the new pieces of technology that we can merge together to build something new that works more efficiently that is better, faster, cheaper? And very soon everything that we see or do or buy is going to be mediated by artificial intelligence. It's already starting, right? The things that we see on the internet, like as we as we transition from Google to these AI interfaces to the internet, those are mediated by LLMs, right? What we see is what it serves us, the way it talks to us, what it believes. And so we have to be really conscious of how we control or how government controls or how the people control the mediation layer that is giving us the data that we all make decisions with. Because it used to be that we would search and then it was that we were served up things and then now it's being we’re being served up specific things. It's gone narrower and narrower and narrower. So there's much less exploration. And now we're just being served up what it thinks that we want or what it wants to serve us. It's two ways. We get much more direct answers. But those answers are being directed as well. So we have to be conscious of this idea that there is a mediation layer of the large language model, that is our interpretation of the world, especially if you think of about glasses. Suddenly the world is now being interpreted all the time by this layer of AI, and it serves us up ideas that may or may not be correct.

Mikaela: It’s interesting because I don't know that generations that come after us will necessarily have those skills or that literacy to discern what is AI and what isn't. I mean, even I sometimes am taken aback by it, and I often look at a piece of writing and think, I don't know how I know that this is AI, but I know it’s AI. When I when I think of my three year old son coming in behind me, are we doing anything? Are we doing enough to really build literacy, to be able to discern what's real and what's imagined?

David: No, not at all. Not in the slightest. I talk about this all the time in education, this crazy idea that, you know, professors are now using artificial intelligence to write their lesson plans. And then the students are using AI to write the essays, and then the professors are using the AIs to catch plagiarism. And then the professors are using the AIs to mark the papers. I wrote a post on this on LinkedIn the other day about Writing is thinking. And when I learned to write, I learned to think. So if you don't have to learn to write, how do you get that thinking piece? It's difficult. Or I think it's difficult. I’m putting out way more problems than I'm putting out solutions. 

Jon: This is this is the whole point of this conversation, though, is that we want to ask the questions that maybe don't even have the answers. And by the way, the thing that you were talking about before made me think of, you know, the more there is, the less choice we get because of scale. Think about cars: Cars in the 60s used to be myriad of colors. Now they're black and white and red, you know what I mean? Because of scale. So, what do you think about that in terms of, do you think it's a cross-purposes with this idea that AI can imagine anything? 

David: I absolutely do. I mean, AI is the auto tune of everything,

Jon: I love that.

David: It's sands off the rough edges of everything and makes everything similar. But that's in the short term. In the long term, there'll be the rough edges knob that you turn up, right. We're seeing it in music. I don't know if you guys have listened to much AI music. When AI music first came onto the scene, first, it was awful. It was just unlistenable. And then it was generic. And now there are some pieces that are kind of incredible where the voice cracks the way the voice is supposed to crack in the right place. I don't use AI to make music because I'm not just it's not that much fun for me. And that's what I mean by human spaces, you know. You know, I'm going to go back to my guitar because it's fun. Those things. I write songs with my friends because it's fun. I play for people because it's fun, and they come to the show because that's an experience. But maybe we'll gain that back. Look at chess. You know, more people than ever playing chess. They love that game. I don't know if you follow me. I follow a lot of chess vloggers. I don't know why because I'm a terrible chess player, but I'm sort of addicted to this idea of chess. But chess is a giant thing where people are playing, there're doing all these crazy - it's. It's exploding. So it is possible too we'll get the explosion of human spaces where people are just, you know, finding whole new ways to interact humanly.

Mikaela: It's interesting to me. We kind of think of that and think of it in combination with the future of work because, I think, especially when I think of my career and my trajectory and how I learned the things that I learned. There's so much meaning and purpose in those tasks that you develop and that you create either a disdain for or a passion for. And I think it's ultimately going to change how we have our careers, how we develop careers over time. But have you seen anybody that's kind of exemplifying how to keep that purpose and meaning in work and kind of at the collision of AI?

David: You know, everyone on my team that we're, you know, we all use AI all the time, but we still love the work, and we're passionate about what we're building. And it doesn't replace us. It augments us. It can be a great tool. It can be used in amazing ways. It can make you more efficient. It can make you faster at stuff you don't want to do. So you have more time to do the things that you do want to do. We're going to see smaller companies being able to do so much more. And we're seeing a lot of changes, I think, in the startup world as well, where companies now are really rethinking this idea that they need endless VC money to scale, because now maybe we can get, you know, some venture money and then we can scale with AI.

Jon: Especially when it comes down to things like rapid prototyping and testing ideas and that kind of stuff. But I think one of the traps that happens and this is for me, one of the things that I worry about, is that the companies and people who are taking AI at kind of its basic level and not doing the actual work to make something great.

David: We're going to see tons of things that come out that don't work. People try things, people experiment. People invest big, big, big wash outs. But then we're going to reach the next plateau, which is things that are working, things that are scaling, new companies coming in, understanding the technology better, understanding what the opportunities are better. This is the great wide open where no one really understands what we're doing. So there's you know, everyone is out here experimenting, trying different things. And I think that for me that's really fun. That's adventure. But there's going to be tons of tears, too.

Jon: So you talk about the importance of using AI meaningfully. I think the Alzheimer's one is a great example. And it's a big part of my family as well. So it really resonates with me. Do you have any specific criteria that you use to judge whether an application is valuable or responsible, etc.? Like how do you think about that way?

David: It's kind of like how you think of yourself as a person when you're thinking about whether you're using AI ethically. So for our project Second Echo, the way we're thinking about it right now and it's in its infancy, is that we're thinking that it's 18 plus because we don't think that kids should use these things. It's you can only build yourself because you need to give consent, and you should only use it for a certain period, a certain amount per day, because it's meant to be a way you record your memories, not you create your memories. You create your memories out in life and you record them with Second Echo. So those are the kind of ways we're thinking about that we can use AI ethically within the space that we're trying to build in. I think that's what ethics are. It's a choice. So you're never going to be successful all the time, but you're at least trying. If you’re really thinking about it and saying, you know, what are we going to try to do?

Mikaela: I don’t think all humans are willing to ask that question. Or, if they do ask it, maybe they don't like it. Or they don't like the answer and they go a different way anyway. So it sounds like so much of the ethics of this is about the humans behind the technology and the humans behind the algorithm. Is there any such thing as an ethical algorithm?

David: The interesting thing about these LLMs is, when we think about artificial intelligence, we often think that we're building them. Right? But really, we're growing them because we don't really understand them. If you look at a piece of code, you have a bunch of programmers and they're writing code and there's a bug in the code. They can find that place in the code and fix that bug. With AI, they create an algorithm, they run it on this data set, they build this thing, and then it acts and they can manipulate it a little bit here and there. But it's not the same way as building something and being able to identify specific place where this is. It's kind of like we're creating these creatures that we then use, and then we create another creature at the next level, and then we use that and we sort of have an understanding of it, but we don't really know what it is or how it's going to react.

Mikaela: David, thank you so much for being here with us. We really appreciate it. 

Jon: Thank you very, very much for doing this.

David: Appreciate it.

Jon: We'll see you in the next life.

David: Yeah. Exactly. See you in Toronto.

Jon: Okay, ciao. Bye bye.

• Navigating large-scale transformation as a business leader
• Building trust in telecom with a human-led, tech enabled approach
• Walmart: Valuing customer experience and helping Canadians make better choices