276: How AI is Reshuffling the Rules of Business

How many founders think AI is about faster automation?

What if the real opportunity lies in reimagining how entire systems work?

Sophia Matveeva sits down with Sangeet Paul Choudary, best-selling author of Platform Revolution and Re/Shuffle: Who Wins When AI Restacks the Knowledge Economy, to explore how AI isn’t just speeding up workflows.

It’s rewriting the rules of business models.

From TikTok’s behavior-driven growth to Shein’s algorithmic supply chain, Sangeet explains how AI coordination is quietly reshaping industries.

You’ll learn why being a “second mover” might now be an advantage, and how to design your company around AI’s strengths instead of bolting it on as a tool.

In this episode, you will hear:

• The difference between using AI for automation and building an AI-first business
• How Shein’s real-time supply chain sensing outpaces Nike’s legacy model
• Why “first mover advantage” is outdated in the age of AI
• The simple framework Sangeet uses to identify what won’t change—no matter how fast tech evolves

Free AI Mini-Workshop for Non-Technical Founders

Learn how to go from idea to a tested product using AI — in under 30 minutes.

Get free access here: techfornontechies.co/aiclass

Resources from this Episode

Book: Reshuffle: Who wins when AI restacks the knowledge economy https://amzn.to/4n7NgF2

Podcast: listen to the platform series on Tech for Non-Techies: 

• Tech for Non-Techies episode 90: What makes platform businesses SO successful https://www.techfornontechies.co/blog/platforms-episode
• Tech for Non-Techies episode 91. How to launch a platform when you've got no users https://www.techfornontechies.co/blog/chicken
• Tech for Non-Techies episode 92. How to get people to be nice to each other on your platform https://www.techfornontechies.co/blog/how-to-get-people-to-be-nice-to-each-other-on-your-platform
• Tech for Non-Techies episode 94. Learning effects: why getting more users isn't the only key to success https://www.techfornontechies.co/blog/learning-effects-why-getting-more-users-isnt-the-only-key-to-success

Follow and Review:

We’d love for you to follow us if you haven’t yet. Click that purple '+' in the top right corner of your Apple Podcasts app. We’d love it even more if you could drop a review or 5-star rating over on Apple Podcasts. Simply select “Ratings and Reviews” and “Write a Review” then a quick line with your favorite part of the episode. It only takes a second and it helps spread the word about the podcast.

Listen to Tech for Non-Techies on:

• Apple

• Spotify

• YouTube

• Audible

• Pandora

TRANSCRIPT

Sangeet Paul Choudary  00:00

Are you using AI as a tool that you're bolting onto existing workflows and just not changing the fundamental business model, not changing the fundamental product paradigm, you're just improving or using it to speed up something within the existing paradigm? Or are you re imagining your business model and your product around the capabilities of AI, and I'll give a very simple example to illustrate this. You know, social networks like Facebook, Instagram and YouTube have historically been built around the logic of the social graph, which is that you need to follow somebody in order to see content from them, and you need followers for for your content to have influence and spread. When Tiktok first came in, it used AI not just to create the social network on the same model, but it reimagined the product and the business model around the capabilities of AI.

Sophia Matveeva  00:57

Hello and welcome to the tech front and techie podcast. I'm your host, Sophia matveyeva, if you're a non technical founder, building a tech product or adding AI to your business, you're in the right place. Each week, you'll get practical strategies, step by step, playbooks and real world case studies to help you launch and scale a tech business without learning to code. And this is not another startup show full of jargon venture capital theater or tech bro bravado. Here we focus on building useful products that make money without hype and without code. I've written for the Harvard Business Review and lectured at Oxford, London Business School and Chicago Booth, so you are in safe hands. I've also helped hundreds of founders go from concept to scalable product, and now it's your turn. So let's dive in.

Sophia Matveeva  01:51

Hello, smart people. How are you today? By the time you hear this, I'm going to be in Saudi Arabia. I'm going to be hosting an event on tech fun, on technical founders, with our partners at Microsoft and the Saudi digital Academy. So our reach is going far and wide, and it's really, really exciting. And I'm also really excited about this episode because I have a guest for you whose work I have been obsessed with for years. His name is Sangeet Paul Choudary. Sangeet is one of the world's leading thinkers on platform economics and network effect. He's the co author of a book called Platform revolution, and that's an international bestseller, and it's a book I reference constantly, and have actually, and have actually done a series on how to build and scale a platform based on that book. So if you want to know how to build a platform or a marketplace from scratch, then listen to episodes 90 to 94 of the tech for non techies podcast, and the links to those are also in the show notes. Sangeet's work has appeared multiple times among Harvard business reviews, top 10 management ideas, alongside legends like Michael Porter and Clayton Christensen and Sangeet, also sits on more boards than we have time to discuss today. In this conversation, we discuss his new book, which is called reshuffle, who wins when ai restacks the knowledge economy, you will learn why the real opportunity in AI is not just about fast automation, which is what we're all hearing about, it's actually about coordinating entire systems. And if this sounds vague, then don't worry, because you will learn from real examples like why Shane can sense market demand in real time, and how that's different from Nike's business model and how Google Maps reshaped our city. So there are plenty of examples to make abstract concepts really, really clear. And you might be surprised, as I was, by what Sangeet says about the myth of the first mover advantage. So this is a really, really good episode, especially if you're a business owner or a non technical founder, trying to figure out how AI can genuinely reshape your product or reshape your business model. Listen to this, and if you enjoy this episode, then make sure that you are subscribed for more, because this is the best place to learn. If you're a business leader building a tech venture, you chair

Sophia Matveeva  04:24

Sangeet, tell me, what do maps and artificial intelligence have in common? 

Sangeet Paul Choudary  04:29

I'm glad you bring that up. Sophia, maps is one of the best examples that I think we have in terms of understanding how AI can actually change your world. And I use the idea of maps because the point that I want to make is that a technology doesn't have to, you know, beat human benchmarks to transform the world around us. A technology that is Natalie performant can also transform the world in very meaningful ways. Now, if you look at our experience of using maps before the. Sudden, GPS based navigation applications came out. Traditionally, we used to use paper maps. Paper maps were static, and while they helped us make sense of the world, there was only so much effect they had on the world around us. What happens with Maps is a GPS based navigation application today, which works on Google Maps or an equivalent, starts by creating a representation of the world, and I think that's the starting point for any AI system as well. If you look at any AI system, it creates a representation of the world. It models how it sees the world, whether it's a sensor based industrial system that's capturing data from industrial processes, or whether it's a large language model, it's essentially working off its understanding of what the world looks like. So map, first and foremost, creates a representation of the world on the basis of which it helps other users make better decisions and actions and take better actions. So the key decision that Google Maps, or ways helps us with today, is around navigation. How do I get from point A to point B most effectively. And on the basis of that, it then gives me turn by turn directions which guide my action, which, if you really think of it, replaces a job that would traditionally have been performed by a navigator sitting next to you with a paper map in hand. And so that's what the Google Maps lady essentially performs in substitution of that human role today, and in many ways, we may think of the effect of maps stopping over there. But if you really think about how our cities have transformed, how our traffic patterns have transformed when everybody starts using GPS based navigation applications, traffic patterns evolve in response so the larger ecosystem of traffic movements evolves in response to the information that maps are processing, or these applications are processing, and the decisions that they are helping users make about where to route and how so really, Maps is a great example to think about how very narrowly performing technology can have fundamentally large impact on how the overall ecosystem works. So today, you might have a quiet neighborhood, and tomorrow it might be part of, you know, thoroughfare, just because, because of changes in navigation patterns, Google Maps always redirected most users in that direction. And I'll just close by saying that you know, the five key things that a map is doing over here is, first of all, it's creating a representation of the world. That's what all AI systems do. On the basis of that, it's guiding users to make decisions and take action, which is again, how AI has a potential to the wire our knowledge workflows today. But most importantly, what the map does is it rewires how the various players in the ecosystem, in this case, every other driver, every other person who is using this application, how they interact with each other, how they coordinate their activities, even though they don't coordinate it explicitly, the map coordinates them implicitly to avoid traffic jams and bottlenecks from forming. And that's really where I believe the power of AI lies in Z orienting coordination the way maps does that in the real world.

Sophia Matveeva  08:14

But why is it that you talk about maps and you give this example of how you give examples of maps that you before Google Maps, how maps had changed lives and changed countries, and you give, you gave an example of the British Empire. And we can actually just look at the African map where, clearly, somebody just took a ruler and then just decided, Oh, these are countries, without any thought to the geography, to the history, to the villages there and so on, and that has resulted in lots and lots of strife. And then you talk about how Google Maps has changed quiet neighborhoods into places now with lots and lots of traffic, and you know, not entirely sure that residents are happy about this. So this is an example of technologies that we really take for granted today, but technologies that have really changed our systems. And what is it that you want people to know now that we're in this age of AI, what is it that you want people to learn from this map, example of how maps have changed our world to what AI can do for

Sangeet Paul Choudary  09:21

us? Yeah, absolutely. I would say that there are three key points that I was trying to make with the maps example in the book, which I believe are very important today. The first is that when we think about where we are in our current moment with AI, we are always focused on, when do we get to artificial general intelligence? You know, is AI today equivalent to a fifth grader, or is it a PhD in your pocket? So we're always thinking about AI in terms of intelligence, and one of the key points that I make is that AI, even in its current form, even if it has not reached, you know, one of those benchmarks of intelligence that we're hoping it will reach is still underutilized, and as a Natalie per. Performance Technology, which is quite the opposite of how we think about when AI becomes useful. When we think of AGI, we think of general intelligence. But more importantly, AI is valuable as a Natalie performing technology that can solve a particular problem and then reorient the entire ecosystem around it. And we haven't yet even unlocked AI's benefits in a big way, in these narrow domains itself. So my first point really is that even if you think about how GPS based navigation changed, how or had larger effects on things like traffic patterns and even city design and neighborhood design, AI can have really large second and third order effects, even with its capabilities today. So we don't really need to get to AGI to see all of those things getting unlocked. So that's the first real point. The second point, as you mentioned very clearly over here, was that it's not just GPS based maps. Maps, as technologies, have always had significant impact in terms of how regions were governed, and that is because maps are what we call socio cultural technologies. They are technologies that help us make sense of socio cultural systems, and hence help us govern those systems. Now the reason that's important is because AI is not just a technology for task execution, it's fundamentally a socio cultural technology as well, because what gets represented in AI, and this is why we have a lot of debate about bias, and, you know, the inherent bias of AI models and the ethical implications that emerge on that basis. The key point over here is that the impact of AI on our systems, on our you know, social and economic systems does not play out purely, or even primarily through task execution or through making anything faster, better cheaper. It plays out primarily through how it reorients how we think about the environment and the system around us. And that's what maps helped the British Empire, for example, to govern far flung locations using that technology. And the third point that I want to make over here building off the second point is that the primary value of AI, we often think of that as a technology that speeds up our workflows by automating individual tasks, but the primary value of AI is not in automating individual tasks, it's in helping us rethink our workflows, redesign our organizations, and redesign, or rethink, even how we differentiate ourselves as companies and as individuals, and that's the conversation that we are having. Relatively less when we talk about AI, we are still very focused on the automation idea, and that's why, again, I believe the idea of maps as a coordination technology is so

Sophia Matveeva  12:41

important, you know, I especially love the first point that you make, which is, okay, let's stop all of this intellectualizing about how smart it is, because, I mean, that's interesting. But if you have a business to run, or, you know, you've got work to do, you actually need some, you know, you've got some jobs to be done. So essentially, what we need to focus on is, can we use this tool to solve a particular problem? And then the next question is, okay, if we can use this tool to solve this particular problem, how does everything else around it? How does it affect the rest of our day or the rest of our team? So in the book, you also gave another example of what Nike was doing with the supply chain, and what she in the Chinese e commerce behemoth is doing in terms of their supply chain. And I would love to, I'd love for the audience to learn about that, because we've actually had an episode on Nike and on Phil Knight, the brilliant founder, and what they did was extremely revolutionary. And not that long ago. I mean, the guy is still very much alive and kicking, and sheen, they've just had their IPO and, you know, they have completely shaken up e commerce in a totally different way. So what's Nike doing? What what did Nike do? Why was that revolutionary? What's sheen doing, and why is that revolutionary?

Sangeet Paul Choudary  14:04

Yeah, the Nike versus sheen contrast is something that's interesting, because I primarily try to bring out the key idea of you know how the logic of the supply chain changed with the capabilities of the technology that these companies had at their point in time. So Nike, when it first invested in global supply chains and decoupled the act of design from the act of manufacturing and moved most of the value in the value chain to design and away from manufacturing, this was really built around the capabilities of the technology that was emerging at that time, shipping containers had transformed global trade and had allowed shipping to happen reliably, which meant that you could now manage global supply chains. And alongside that, ERP systems and communication technologies were improving so you could manage and govern supply chains globally, and so the ability

Sophia Matveeva  14:58

to just so. We remember what exactly was happening. Nike was being designed in the US and manufactured in Japan, if I'm correct, that was, yeah, that was really revolutionary, because basically where things were designed previously was where they were manufactured. Because you just didn't have this supply chain, you didn't have organized container shipping, you didn't have all of this coordination. But then when shipping changed, that basically meant that how everything got made got changed. And so Nike was one of the revolutionaries there.

Sangeet Paul Choudary  15:33

That's absolutely right. What really made that shift happen was that shipping became reliable so you could work and, you know, even go externally, govern activities that were happening in another part of the world. And I think that particular lesson is interesting when we look at sheen, because Sheen is really an outcome of a new technology of coordination, which is algorithmic coordination that's possible today. The way sheen works is that unlike a traditional player like Nike, which would work on large batches and bring out a whole new seasonal collection, and this applies even more to fashion than it does to sportswear, but unlike traditional players of this form, Sheen, first of all, works on constantly sensing the market by capturing demand signals from social media. And this particular piece is important, because this ability to constantly sense the market fundamentally changes the architecture of the of the supply chain. Because traditionally, the way you know fashion in particular worked was you would have merchandisers travel to Tokyo or to Milan, figure out the new styles and then curate a new collection for the next season. What sheen does is it constantly curates collections, and it does not finalize the collection until it's actually tested it. So what that means is it also changes the size of the batch that gets produced. A micro batch gets produced based on a micro trend that sheen identifies, and then it tests that in the market. And if it gets validated, it increases the production. If it doesn't get validated, it moves to the next micro trend. And the reason that's important is there's a very tight feedback loop between the demand side and the supply side that's created because a lot of this can be managed algorithmically. You don't have to have slow, manual processes in between, which was really, really the constraint that forced longer cycles of turnaround and hence larger batch production. So once that constraint goes away, smaller cycles, quicker cycles, with shorter batches come in. And the final point that's very interesting about Sheen is it actually changes the job of the designer as well. It does not automate the job of the designer, but the more it learns from all of this data, the more it modularizes The job of the designer so that a designer is reduced to micro tasks. They're given very specific tasks to perform on modifying a particular design so that, over time, who can become a designer also fundamentally changes in the sheen value chain.

Sophia Matveeva  18:05

And this is so interesting, because Sheen is obviously, you know, they have cheap prices, basically. But when you look at the luxury market, like, you know, a famous designer would never allow for this to happen, because the whole point of a famous designer is that they do get this inspiration. You know, they're an artist. They're more of an artist, and they work with a commercial team to make sure that the stuff that they dream about can actually be worn by a real human being, and, you know, also be purchased by a real human, by a very rich human being. And so this is an interesting division between the luxury market, which is still pretty traditional. You know, the way couturiers would think about designing a dress would still be kind of how couturiers would still think about designing like how they thought about it 200 years ago. It's still going to be the same how they think about it today, but on this very, very mass market side, as you were saying, the job of the designer is driven or largely dictated by algorithm. And so I wonder, how does this spread to other industries? Where is there this spread between, okay, the luxury creativity that's still human, but mass market is really algorithmically driven.

Sangeet Paul Choudary  19:24

Yeah, that's a really interesting point, because there are two different ways to think about it. The first is that you could have these markets exist completely independent of each other, which is that there's sheen, which is creating a fundamentally new market of people who would not have purchased items unless they were available at those prices. So it's creating a fundamentally new market which has its own characteristics, and it's not touching the luxury segment. That's one way to think about it. But if we look at what Airbnb did to accommodation, we also see that the other way to think about it is that you could use algorithm. Coordination. Airbnb essentially has an algorithmic rating system, which is how it determines what's a trustworthy place to stay at. So it's it's created its own algorithmic coordination system, and if you look at Airbnb, it started with the unorganized, fragmented side of the market, but as certain forms of supply, certain hosts become more popular, become more vetted, and the quality of their listing improves, they start serving the more organized B to B travel segment, corporate travel segment over time, or the more discerning traveler over time. So my key point is that when you start with creating or organizing a new, fragmented market, you may not directly be a substitute for the organized, higher luxury end of the market. But as you improve over time, there's a possibility for some level of substitution. You could argue that Airbnb still is not taking away sheraton's business. In fact, they're very in very different businesses, but there's still some level of substitution that happens across the two markets, and that's something that I believe is possible with an example like sheen as well. The more data it collects, the more accuracy and the liabilities it has on its collections. Over time, it can move towards creating some viable substitutes in the higher end of the market as well. And I think that's something we see across the board when we look at the power of today's technologies, whether it was the first way of digital transformation with platform business models coming in, or it's aI today, all of these technologies unlock the ability to create new value in unstructured, fragmented parts of the market, just like Airbnb did, Uber did and so on. But as their reliability improves, they do come up as credible substitutes for some parts of the luxury market as well. So I think that's something that we'll see increasingly happening in other sectors as well.

Sophia Matveeva  21:51

So when a non technical founder or a business leader at a corporate is thinking about, Okay, well, you know this AI, hype is everywhere, and I'm using chat GPT every day to figure out what to do with my vacation and what's in my fridge. So you know these people are they know that AI can be useful, but they also want to think about, okay, how do I use it in my work? Or how do I create a new product where this technology is going to be really fundamental. So what would you advise to this person, what are the some of the first things that they would need to think about?

Sangeet Paul Choudary  22:27

Yeah, I think the first thing that is worth thinking about is really asking yourself the question, are you using AI as a tool that you're bolting on to existing workflows and just not changing the fundamental business model, not changing the fundamental product paradigm, you're just improving or using it to speed up something within the existing paradigm. Or are you re imagining your business model and your product around the capabilities of AI? And I'll give a very simple example to illustrate this. You know, social networks like Facebook, Instagram and YouTube have historically been built around the logic of the social graph, which is that you need to follow somebody in order to see content from them, and you need followers for for your content to have influence and spread. When Tiktok first came in, it used AI not just to create the social network on the same model, but it reimagined the product and the business model around the capabilities of AI and around the capabilities of a mobile first interface. So on Tiktok, a typical video is less than or at that time was less than 60 seconds, and a typical session by a user involved going through multiple different videos, and Tiktok would capture every data point, what you were watching, where you were stopping, what you were pinching and zooming versus not. And using all of this data, it would infer a behavior graph, you know, what was the content you were interested in? So it sidestepped the logic of the social graph and created a behavior graph and started showing content on that basis. Now the reason this is important is because it applies to all founders today. You know, when you're thinking about AI and you're saying, here's how a CRM used to work, or here's how you know Travel Planner used to work, and you just use AI to speed that same logic. You're not re imagining an AI first business. That's what most social networks did when AI came in. They just improved the recommendation feeds by using the latest AI but they did not move away from the social graph logic. Initially, only after Tiktok showed them that this was a fundamentally new way to create a social network, did companies like meta move into what they call today the Open Graph. And so eventually the entire model of the industry was re designed around the logic of AI. And I think that's the real opportunity that is there for founders today, because if you're competing against incumbents, you know large players like traditional SaaS players, and you're trying to build AI first business workflows, or large travel players like Expedia, and you're trying to build that AI. First Travel Planner, you cannot just look at those models and see how to improve that using AI, because that's something those guys can Anyway, do themselves. You have to really think about what are the assumptions or constraints around which the previous model was built? So social graph versus behavior graph. What's the new assumption that I can create now that AI is available, and how can I the architect the business model in its favor as well? So that's really what I would encourage anybody thinking about AI today to think of.

Sophia Matveeva  25:31

And you know what you're saying. I think is good news for people at the earliest stages, because this week, I spent two days at the JP Morgan tech investor conference in London. And there were companies that were definitely late stage. I mean, you don't get to go and speak to public markets investors unless you are either you've had an IPO or you're about to have an IPO in the next few years. So these are pretty late stage companies. And it was every single found, every single leader was asked about, okay, how has AI changed your business? And I would say that 90% of the people, 90% of the company leaders, said, Well, we're using AI, really, for cost cutting inside that we are not yet. There were a couple of examples of founders who said, well, actually, we are creating a new customer experience. This is driving new revenue. We're creating new products. But that was very much an exception. The rule was, you know, we're basically, you know, making our content creators, our developers, like we're making our teams more efficient. And you know, this is definitely how I've been using AI is that I have become more efficient. I've therefore become more productive. This is great, but I wouldn't necessarily say that for tech put on tech cases, has completely changed the business model, although we have created new products based in Rome. But that's that's another thing, and so is this what you are seeing, too, that companies right now are still in the let's call it low hanging fruit stage, which is where can we trim costs? Which doesn't really take that much imagination, but few companies are moving into what you're talking about.

Sangeet Paul Choudary  27:14

Yeah, I think that's exactly where things are at this point. And I think there are two factors that are keeping things at that stage. The first is it doesn't take a lot of imagination while. At the same time, you can piggyback on constantly improving models. So all the capital is anywhere flowing into the models. Might as well just wait and see what else we can automate with the next model coming. And that's the dominant mindset that a lot of companies have. It takes a lot more imagination and a lot more risk to rethink and reorient your business around AI. So that's one big reason we are still stuck in that trap. The second big reason is really that when the piece of innovation is the way we see today, it's a breakneck piece of innovation. There is a lot of noise that comes with it. There's a lot of hype that comes with it. So it's difficult for business leaders to clearly and credibly differentiate between what is just hype and what is actually possible, and when that kind of uncertainty comes into any new technology, and we saw that in the early 2010s with digital as well. In such phases, you know, business leaders are a lot more hesitant, but once a few use cases play out, and a little bit of the hype gets corrected, we will hopefully see companies think more more strategically about what's possible with AI.

Sophia Matveeva  28:29

So a thing that actually was brought up at the conference yesterday, somebody was talking about the second mover advantage, or the third mover advantage, because they were saying, Well, yes, you know, this is something that we're really interested in, but exactly what you're saying, you know, there are so many issues right now, and frankly, the low hanging fruit is paying off. Our costs are getting lower, margins are higher, like this is life is getting good. And what he was saying is that actually, there are benefits to not being the first, because the first is the pioneer who basically gets hurt and makes all the mistakes. And this reminds me of Apple, because Apple is not a first mover, like they in general. In some ways they are, but in general, they take something that already exists but isn't super great, and then they take this existing thing and they make it super amazing, like the technology in the iPhone had already existed, but Steve Jobs put it into an amazing, beautiful design, and here we are today. And so is this what kind of you are seeing when you're speaking to corporate leaders, when you're seeing to founders, that people are kind of like, yeah, that I'm interested, but I'm going to be the second mover

Sangeet Paul Choudary  29:42

Yeah, I think it's important to make a distinction over here that the nomenclature or the parlance of the first mover advantage came out in the previous century when typically businesses expanded by playing by the same rules of the game, but moving into new markets. So when the structure of the business and the logic of competition. Is the same and you're exploiting new markets. That's when first more advantage makes sense in technological disruptions, the first more advantage never makes sense, because you're waiting for that disruption to play itself out, and somehow we've never corrected the use of that language, because that language really came from a different era, when expansion was about new customer segments, new geographies, new channels, but with the same logic. But when we talk about technological discontinuities today, when you know mobility happened or the cloud happened or AI is happening right now, you're not just taking an existing playbook and moving it into new channels, new markets, new geographies. You are trying to figure the playbook, and when you're dealing with that uncertainty in the structure of what you are trying to build itself, there is no first mover advantage. There's hardly ever any first move advantage. There is instead rapid iteration to validate different hypotheses and then first to proof advantage whoever can validate hypotheses in their favor first, they are the ones who have an advantage. So I think we need to make that distinction, because we still carry that balance, because it became so popular in the 90s, when we were going into globalization, but today, it does not really stand stand true anymore. So I agree with that framing interesting. And so

Sophia Matveeva  31:18

my last question would be about, how do people even think about frameworks for the age of AI? And what I mean is that the technology is rapidly evolving, and there are, you know, big changes that seem to just shake up the ecosystem. And the example I'm thinking about is deep seek, you know, so chat GPT came in lots and lots of downloads. It's this huge new thing that's changing everything. And then we kind of accept that these companies need to raise lots and lots of money, and then deep seat comes in with a completely different cost structure, and it's open source, and that just blows everybody's mind. And you know, I think especially people in the Western world weren't really paying attention to what was happening in China. I mean, they definitely are now. And so it's a very rapidly changing product. Last week, they were doing something that you know maybe they're not going to be doing in a couple of weeks time. And so when an author like you is publishing a book, you know you want people to write to read your book, and you want people to read your book this year, but also hopefully in a few years time. So how do you think about frameworks, and how to think about the age of AI in this world where we're constantly getting new updates, and in a world that's constantly changing?

Sangeet Paul Choudary  32:42

Yeah, that's a great question, because that is a question that I feel very passionately about, and that guides a lot of my work. Whether you look at my previous book platform revolution or the new book reshuffle, what I look at is not what's constantly changing, but what are the principles that are not going to change no matter what happens. And those principles typically lie in the socio technical system that surrounds a technology. How technologies are used, how you know, certain forms of value get commoditized when new technology comes in, how new forms of value emerge. There are very clear patterns and very clear drivers that stay consistent despite changes in technology, and so that's what I really focus on. That is a true line that you'll see across my books, because platform revolution was written in 2015 it's still as relevant today, with a few exceptions, because it focuses not on the technology, but on the socio technical system around it, the idea that's still relevant, and I do the same thing with reshuffle. In fact, you know, I made a choice which was a little risky, where, essentially I said, even though there's no single thing called AI, there's, you know, there are many different technologies that fall under that umbrella. I'm still going to talk about AI as a composite, not because everything, you know, all AI is similar, but because in terms of its impact on the socio technical system around it, in terms of how it changes work, how it changes how companies compete, all these technologies have two things in common. They a, improve the capacity to do work and B, they improve the ability to coordinate work. And if we can understand how those two factors interact with the system around it, that will you know those factors will then remain consistent even 10 years from now, even if the underlying models change. So I think focusing on what doesn't change is really important, while also rapidly updating your thesis on what is changing and where value will move next, because the combination of these two things, and you know, Jeff Peters talks about this a lot in 10 years, I don't care about what's going to be different. I care about what's not going to change, because that's what I can bet on. But the combination of the two things is important, because you need to see what's the latest thing that's changing that will help you identify our. Arbitrage opportunities, catch something, catch a wave, before others do, but then you need to have a clear thesis on what won't change, so that you can use that arbitrage to then strengthen your underlying thesis. You can't just keep moving from wave to wave without a core underlying thesis, and that's really why I believe it's important to have a framework where the underlying thesis is not going to change, or the underlying thesis is going to be relatively grounded and informed by how things change around it.

Sophia Matveeva  35:29

This is actually a really good and practical note for the innovators listening to this to to end on, because think about what will definitely not change in your industry. So I think in Jeff Bezos example, he says that people will always want lower prices. That is true. They wanted lower prices 2000 years ago. They will want lower prices 2000 years from now, if we're still here. So apart from you know fundamental truth like that, what else is true for your sector. And then once you've got that, then definitely read the book reshuffle, and then see, okay, how will technological advancements change the industry, or how can you use them in your favor whilst keeping true to that thing that is not changing? So Sangeet, thank you very much. As listeners will know I'm obsessed with your book platform revolution. We've actually had several episodes on it already, which we'll link in the notes. We'll also link your book reshuffle in the show notes, which is out today. So everybody go read it. And apart from reading the books, is there another way that people could keep in touch and keep updated with what you're working on.

Sangeet Paul Choudary  36:43

Yeah, absolutely. I write a very regular [email protected] I write weekly, and that's where you can find my latest thinking. I also have a new website coming up called the shuffle book.com which essentially is going to build on the frameworks that are there in the book and create show its application to different industries and customer segments, and that's going to launch in the coming weeks as well.

Sophia Matveeva  37:09

Awesome. Well, thank you very much. Thank you for joining the show and sharing your insights.

Sangeet Paul Choudary  37:15

Thank you so much. Sophia,

Sophia Matveeva  37:21

wasn't that interesting. And if this conversation got you thinking, I want to build a tech venture, but I'm not technical, well, then I've got something for you. It's actually my free class on how to use AI to make your first product idea real, all without learning to code. And this will help you test your idea and test its features with your target customers. So you'll learn how to turn an idea into something you can show and test using AI as your creative co founder. The whole class, it's free, and it takes literally 27 minutes to complete, so you can have your test product ready to show people for free in just an hour. So basically, you have to do it if you're serious about this. Go to tech fun on techies, dot tier, forward slash ei class, or click the link in the show notes to start the class. And on that note, thank you very much for listening, and I shall be back in your delightful smart ears next week. Ciao.