Recoding Business: The AI Transformation Flywheel | unDavos 2026

"I would rather overspend by $200 billion than be on the wrong side of that bet." — Mark Zuckerberg on AI infrastructure. Professor Vijay Gurbaxani argues we are at the dawn of the next industrial revolution — and most companies are approaching it myopically.

**WHAT THIS PANEL COVERS** → Why AI is not just another efficiency tool — it is a general purpose technology that shifts the entire production frontier, transforming what companies can do with the people and machines they already have → The AI Transformation Flywheel: a comprehensive framework linking vision, value proposition, technology ecosystem, talent, culture, and operating model into a self-reinforcing cycle → How Reuters uses AI to write first-draft financial news for speed, then reinvests savings into hiring original news reporters worldwide — competitive differentiation through AI, not cost-cutting → Dr. Treat: a two-clinic veterinary startup that used AI-assisted consultations and an insurance database to deliver superior care, earning acquisition by a 700-location private equity firm → Why data and proprietary know-how are becoming the primary competitive moat when all companies have access to the same AI platforms → The winner-takes-most dynamic in AI markets — not winner-takes-all, but consolidation toward two or three dominant players per segment

**PANELISTS**

• **PANELISTS**
• Stephan Balzer — Keynote Speaker, Host & Moderator; Managing Director, red onion *(Moderator)*
• Prof. Vijay Gurbaxani — Taco Bell Endowed Professor of Technology Management & Computer Science; Founding Director, Center for Digital Transformation, UC Irvine

**KEY QUOTES** "Every single job in our company is going to be affected by AI." — Doug McMillon, CEO of Walmart "This is not an efficiency tool. This thing is going to transform the world for better and for worse." — Prof. Vijay Gurbaxani "I would rather overspend by $200 billion than be on the wrong side of that bet." — Mark Zuckerberg

unDavos is a community-driven summit running during WEF week in Davos, democratizing the conversation around global challenges.

🌐 undavos.com

Tags: AI transformation, AI flywheel, digital transformation, AI strategy, competitive advantage, Reuters AI, Dr Treat veterinary AI, AI first company, AI intensive, value proposition, data moat, AI investment, production frontier, general purpose technology, UC Irvine, Vijay Gurbaxani, Davos 2026, unDavos, REWIRE, next industrial revolution, McKinsey AI, winner takes most

---

TRANSCRIPT

So, here we are back, welcome, I hope you were able to refresh in the break and we want to continue with a different format, we call this the fireside chat, that you know we don't have a fire behind us, but hopefully we need it, yes we need the fire in Switzerland, right? So, first, give a warm welcome to our guest, and I will start and you can stop if I talk about you too long, Vijay Gorbaksani is the Taco Bell Endowed Professor of Technology Management and the founding director of the Center of Digital Transformation at UC Irvine, and you're specializing in digital transformation? Correct. Like one of the key things we're discussing here, AI strategy and the economics of technology, which is a great combination, looking back to the panel we had before, and it says here that you're blending rigorous academic research with practical insights to guide business leaders and boards, so also the topic we just touched upon, into adopting AI, so there's more to talk with you about, but maybe you want to take it from there and share your first thoughts. You want me to, just to make it clear, so first of all, good afternoon to everybody, it's lovely to meet all of you, and the last panel was absolutely fantastic, and I hope to be able to add something beyond that, but you want me to talk for about five min

Transcript

SPEAKER_01 0:09

So here we are back. Welcome. I hope you were able to refresh in the break. And we want to continue with a different format. We want to we call this the high side chat that you know we don't have a fire behind us, but we'll hopefully need it in Switzerland. We need it. Yes, we need the fire in Switzerland, right? Absolutely. Um, so first um give a warm welcome to our guests. And I will start and you can stop if I talk about you too long. Uh Vijay Gurbaksani is the Taco Bell and Dell Professor of Technology Management and the founding director of the Center of Digital Transformation at UC Irvine. And you're specializing in digital transformation. Great. Um, like one of the key things we're discussing here, AI strategy and the economics of technology, which is a great combination, looking back to the panel we had before. And it says here that you're blending rigorous academic research with practical insights to guide business leaders and boards. So also the topic we just touched upon into adopting AI. So there's more to talk with you about, but maybe you want to take it from there and uh share your first thoughts.

SPEAKER_00 1:24

And you want me to just to make it clear. So, first of all, good afternoon to everybody. It's lovely to meet all of you. And the last panel is absolutely fantastic, and I hope to be able to add something beyond them. But you want me to talk for about five minutes, is what you said. You can. So I'm if you allow me to interrupt you, no. So I I'm gonna go ahead. I'm gonna talk for five minutes, no interruptions. You can be like one of my students. I tell them to be quiet for five minutes, which is very hard. Um, I teach executive MBA, but I I want to set the ground, and I've been sitting too long all day, so I need to sort of get my blood flowing again. So uh in five minutes, uh, this is a talk that's on YouTube, by the way, but I'm trying to do a 35-minute talk in five minutes, so hopefully we can summarize it. So I want to say a couple of things. One, I believe we are at the dawn of the next industrial revolution. I call this the digital tsunami in America. They know me as the digital tsunami guy, and because that's a metaphor I've been using for a long time. But I think it's really important to understand that we are at the literally in at a moment in history. We are watching an economy. I've been teaching for 40 years, and I have never ever said to my students or to executives that I speak to that we are facing the next industrial revolution. Now, why is that important? Industrial revolutions are always enabled by technologies with electricity, the steam engine, the internet, lots of other examples like that. But what's important to remember is they transform the economy and society over very long periods of time. So we went from the agricultural economy to the industrial economy to the digital economy, and now the age of AI. Why that's important for all of you in the room is because it's not just individuals and companies that will succeed and fail because of the technology. It is also uh countries and societies. And the world as we know it is changing very, very profoundly. So the most important city in the late 18th century, 19th century in the world was London, England, right? They had the British Empire. It was because they mastered the second industrial revolution. And then it was the US, now China, and in the age of AI, it's it's uh the US and Japan. I mean, sorry, US and China are widely considered to be so the leaders in AI. So I think it's important to all our futures and for the economic well-being and societal well-being of humankind that we focus on transformation. And I can give you example after example of uh some of the innovations that you probably know a lot about, where AI has been really, really important in solving very difficult problems. Drug discovery. Google's got a fantastic model for climate prediction, you know, which integrates atmospheric physics with AI. And there's more and more examples that I can give you of where AI enables discovery. So a lot of industrial revolutions have been about rethinking our businesses for scale, for scope, how we manage them. What is different about AI is it's also like so it has all of those features where you get productivity gains, you get efficiency gains, but it also acts like a microscope. So just like a microscope allowed us to see sort of further than the eye could see and discover these new sort of microorganisms and understand much more, for example, about blood. AI allows us to discover lots of new things in science, in history, in any field, marketing, advertising, it doesn't matter. We can discover new patterns because anywhere there is a pattern with data, deep learning in particular can discover that. So at that point, when you start thinking, so my talk is actually called Recoding Business Logic in the Aid of Recoding Logic, the logic of business in the age of AI. One of the things I want to emphasize in the few minutes I have here, now we hopefully get to more detail. I'm gonna be bold here and say most companies are approaching this somewhat myopically. So a lot of what I hear about is AI for software development and AI for customer service. You've all heard that. So the much more important question to be asking is what is the future of companies or the future of society? And I'm going to focus on companies, I'm a business school professor in the age of AI. What does it mean to be an AI-first company? What is that real what does that mean? What do you have to do? So I built this framework. It's sort of the what we call the flywheel. It begins by asking the question, what is your vision of the future? And I will so the technologies like AI are called a general purpose technology, which is an economics term. And what that basically says is there's something called the production frontier. What is the most you can do with the people you have and the machines that you have, given the state of technology today? So if you go five years ago, it was something. Now you go five years out and you say, with AI, how much more can we do with the people and the machines that we have? And the answer is a lot more, right? And so if you're not operating at the frontier, you're going to fall behind. So the first question you have to ask is what is my vision for the future? It's not about efficiency. It is about how much more can I be doing? What is my value proposition for the future that is better than ever before? Often that translates into what know-how and intellectual property. So companies that succeed are companies that know how to do something better than your competition. You have to be really, really good with Microsoft, very good at software. But you have to ask yourself in the age of AI, what more can I be doing? Then it comes down to technology ecosystem, which is where are you going to get these AI capabilities from? What are you going to do yourself? What are you going to partner with others to do? And that is a whole new, because the AI technology stack is quite different from the technology stacks that we have today. Biggest thing which sort of we alluded to is talent, work, and one more word that should be on the slide is culture. You know, the Walmart CEO, Doug McMillan, he just stepped down and said at best, every single job in our company is going to be affected by AI. So we have to think about. I gave a talk to the PeopleOps team, which is HR at Google. And you know, one of the things we talked a lot about was these career ladders that we've been talking about for years are gone, right? And you still talk to HR leaders and they're talking about career leaders. I'm like, what career, what ladders are not doing what what tracks are these people going to be on? So you have to think about jobs and tasks and what is AI going to be able to do better than humans? What will humans do better than AI? And I fundamentally believe that we should focus AI on allowing humans to be more productive than having humans be replaced by now. There will always be some of that in a capitalist society, but we really need to focus on building AI for humans to be better. Like I don't see the point of building AI to make machines happy and humans worse off, because machines totally don't feel happy, right? So you think about it. And finally, upskilling and educating the workforce, because that is a huge, huge challenge. I never thought when I started my career that I would be spending time with high schools and community colleges in the US, because these are the people who will educate the vast volumes of people coming to the system. And at the center of it all, there's an operating model, which is how do you get this wheel to spin? So these general purpose technologies are transformative. And if you approach it function by function, process by process, you are not going to get the Big Bang. You have to do all of these things simultaneously. It is a comprehensive transformation process. Yes, it will ultimately come down to improving each process, but you really have to think at the big at the macro level first, which is what does it mean to be sort of a company that is successful in the future? What is the nature of your value proposition and how do you deliver that with technology? And it takes all of these things to be invested in simultaneously. And if you get it right and you get your operating model, which includes the incentives, the cultures, your practices, you want this wheel to spin faster and faster. I can give you examples when Stefan asked me a few questions. So thank you for that. And if you I'm sure there'll be a lot of questions, I try to do it very, very quickly.

SPEAKER_01 9:41

You can come closer.

SPEAKER_00 9:42

Okay. Oh, sorry. He wants me on the red chair. I'm not quite sure what.

SPEAKER_01 9:45

The red one. Yes. The red pill. It's the hot, the hot seat here. So um so thank you for that. Thank you for that. Thank you for the flywheel. Um actually, we can start. Maybe you give some examples you want to do this, because I would have a question for the the value creation. Um, because if we think of AI first companies and all those companies have the same access to all those platforms, like what is actually where is the competitive advantage? Where is it? And where is it generated? And how great question, right?

SPEAKER_00 10:17

So I I so to increasingly from the data that you have and what and what makes your company do. I'll give you a great example. I'll give you a very simple example. Uh there's a company called Dr. Treat, which nobody knows about. That's why I like to use this example. It's a two-veterinary office clinic. So they see animals, and I'm an animal lover. And this guy was a software entrepreneur, and it's sort of interesting. He just saw one more space ripe for disruption. But what is powerful about that the company was uh he decided, I'll just give you a simple version. So in the US, animals are considered property, they have no rights, so they have none of the healthcare regulations that humans have. But but here's the thing. So when the veterinary doctor talks to the pet parent, the AI is listening and it is prompting the vet with ask this, ask that, uh, and you have a sort of a more informed conversation. Because as it uh so that's part one. And part two, they bought, because they're a software-focused AI-first company, they went out and bought an insurance company database. So it turns out that the average vet, so let's say you have a yellow lab, you know, a dog, uh, and the average vet has probably seen 10, 12, 20, depending on the scale of their uh how big their company is. And so they may not know about the unique peculiarities that that particular breed or subbreed has. And all of a sudden, AI is giving them much better things. And so they start, they got bought by a massive private equity company, which has 700 veterinary clinics, and they're now deploying their technology out in these 700 things. So that's an example of the new frontier. How much more can you do? Uh because if if you men, I don't know how many of you live in the US, but the veterinary experience in the US is quite broken. Uh and so this is an example of something that we can do a lot better. So that's one example.

SPEAKER_01 12:13

Which is more, would you say that that is the typical disrupting approach, or is it adding value where there was actually no markers?

SPEAKER_00 12:22

No, I think it's a disruptive approach. Okay. So so but here's the thing. I actually found out that I said, how is how are they doing with the rollout across the 700 clinics? And it's going very slowly because this is where the cultures don't match. And so even though we have a working solution proven in two clinics, if you don't have the right people and the right culture, you won't do it.

SPEAKER_01 12:43

Uh so we come to the culture of breakfast uh as always. Comment, always, always coming back to this. Um, I mean, you talked about the um, you know, the human human side as well. I would like you to touch shortly on on the on the human role. Like, and when it comes to leadership, human leadership, we we talked a lot about what is necessary here and leadership. Um what do you see as well? Because you you teach, you teach, you do executive MEA, so you have people from the industry coming in, you know, getting trained, getting maybe for first-time insights in you know, those topics, etc. Um, if you would comment on on the last panel, what do you see as like the most important thing that we need um you know in boardrooms for the at least say maybe not AI first, because I doubt that we, you know, we have like Germany has legacy companies going back family businesses like 150 years and more, right? So they are and then we have 10,000 of them. So um they will never be AI first, but they want to maybe adopt. So what is needed in those leadership circles? What would you say?

SPEAKER_00 13:50

So um it's a tough question to answer because you know, a lot of leaders are talking the AI.

SPEAKER_01 13:56

That's why you are here and you are the professor.

SPEAKER_00 13:58

Um I I think so. I just I've just completed a paper with a very prominent board member on what boards need to do, which is also focused on what CEOs. So boards are focused a lot, unfortunately, on making sure they don't screw up because they have a lot of liability associated with so a lot of it is about control, a lot of it is about making sure AI is safe, which you have to do. I mean, I would be an idiot to say that you don't do that. But on the other hand, they're not focused enough on growth and innovation around AI. And I think uh this is for better and for worse. And I know the last panel talked about being careful with being being fast, but at the same time, you can't be slow and you can't fall behind because this is a winner-take-all industry, right? So winner, you know, economics, we basically say first mover advantage, and the company that wins will can win quite big. Uh, and that's why you see the massive investments of money, you know, the whole thing that people talk about every day, the AI bubble.

SPEAKER_01 14:58

But Elvin in the morning, you know, when we started, was questioning exactly that. Is it really the winner-takes all? Because it was interesting to see him say that probably it's not happening. He doesn't see there's many examples in the history of technology in the last 30, 40 years where it's not the when it takes all. And there will be also, and I I also, you know, might would like to get your comment. Uh, you know, Scott Galloway from MWU Stern. And Scott, um, in a conversations we had in Hamburg like last year, it was interesting.

SPEAKER_00 15:27

It was indeed theater.

SPEAKER_01 15:28

Yeah. And and he he was asked uh what he sees, what what role AI would play? And he said that he thinks that AI will probably rather be embedded everywhere and it will be probably free, like a free good, but Alan also argued. Um, and there's not he thinks not one company will take it all. He doesn't he doesn't think that this will happen.

SPEAKER_00 15:49

It's would you agree here or would you just so let let's let's break it down to different segments. Right. So if you start talking about big tech and you start talking about the anthropics and the open AIs, so there's there's that world. Um I I think increase I don't believe that all of them will survive. Uh I just do not. Because ultimately, if you especially if you if you believe in artificial superintelligence and you have to ask, like, if it's a superintelligence, you only need one, right? But I mean, but so there will be, and if you look at all software markets, there's usually two or three. So we I'd rather use the phrase winner takes most than winner take all. Because winner takes most basically said this. So if you look at search, how many of you remember how many companies started off an internet search? Yeah, there was Ask Jeeves, uh, Yahoo actually started, you know, and and and any number of companies, and ultimately it became Google. But but the point of that is ultimately there was one. And there's still it, you know, there's still others, uh, but not that many. And they have much smaller market shares. And you look at operating systems for uh smartphones. I mean, that there used to be a lot of different companies, now we're down to two, right? And so I think these software markets have that kind of economic, probably high fixed cost, low marginal cost. Now, in AI, there's a lot of marginal cost, but the training is very, very expensive. But I but that said, there will be sort of smaller models built on top of some of these foundational models, without a doubt. But but the second part of the answer is within industries, and I suspect most of you don't work for big tech, you have something that makes you successful today. You are different and better than your competitors on some dimensions, maybe not all, right? And that is what you have to ask yourself the question: will AI erode that competitive advantage? Because you can copy some of the know-how pretty easily with AI. So what is going to make you distinct it? But there's plenty of physical world companies that will do just fine because if you're a hotel, location matters. If you're a pharmaceutical firm, if you have a patent on a drug, you don't need to be first. So you have to be AI intensive. I think AI first can be misleading unless because we say AI people like me say AI first, but in our head we have lots of nuances to AI first. But I think AI intensive is probably a better phrase. Okay, I like that.

SPEAKER_01 18:17

Um So striking examples, my question. Can you mention, or would you mention, some examples where you see like really well executed um restructuring or maybe changing organization model or using a part of your flywheel in the in the process?

SPEAKER_00 18:36

My favorite example these days, and I've got a lot, but I'll give you one because of the interest of time, is Reuters, uh, you know, the news agency. And uh I actually had the president of Reuters speak at my conference, and I was really impressed. So I actually didn't put him on, I had one example for every dimension, but I put him in the middle saying this guy's got the right operating log. So you've obviously heard a lot of companies saying, media companies saying, oh, we will not use uh AI to write because they want to protect the writers and so on. And Reuters said, look, we're basically a financial news or news reporting organization, and I need to be first because if you are first to market, it's proprietary. If you're second to market, it's you don't get the same advantage. And so he sort of invests in having AI write, and the human actually checks the writer because we all heard in the last minute, AI writes really well these days. He invests in hiring news reporters and putting them in different parts of the world. Because the one thing that makes Reuters different is they have more original news reporting than pretty much any other uh new uh news organization. So if you think about that for a second, that to me is what we're talking about, which is what makes you different. Well, I have news that nobody else does. And I'm using AI to make myself faster because if you're trading uh in the stock market, you want to know that the results that you're getting out of Reuters are super fast because you want to trade on that information. And Reuters, this is a funny aside, got its start with carrier pigeons. How many of you know this story? Can I tell it? Yeah, please. It's a very funny story, actually. So Mr. Paul Reuter figured out that financial information was actually conveyed by wire. And it stopped somewhere, I think, in Aachen, Germany, and it had to go to Brussels, but there was no connection because of the geopolitics of the time. There was no wires possible. So the wire was printed out in Aachen, sent by train to Brussels. So this guy figured out he bought carrier pigeons, put the physical wire in the carrier pigeons and flew them to Brussels so they could wire them across because he was trying to get that advantage. Now, here's the here's the here's the funny part. How did the pigeons come back to Germany by the same trend that was kept? Because I didn't know this homing pigeons can only go home in one direction. They cannot understand two homes.

SPEAKER_01 21:03

I didn't know that. That's funny. Yeah, that's a nice picture. Um the point is differentiation. Right. Um we would need more time because I have so many more questions. Um, but I like the the example. Um maybe last question. Um one advice to leaders um that you would give from the work that you've done so far, if they want to advance you know their teams, their companies.

SPEAKER_00 21:30

So I think the first thing is two pieces of advice. Number one, you have it, it's real, it's here. This McKinsey estimates$7 trillion being invested in AI infrastructure by 2030. It's not going away. It is going to change the world. Whether you like it or not, I have misgivings just like any of everybody in the room. But at this point, you cannot stop it. It is a, you know, Mark Zuckerberg had a phenomenal quote, and he said because he's spending so much money and he said, I would rather the exact not exact paraphrase quote. I would rather overspend by$200 billion than be on the wrong side of that bet, which is to be behind in the race for position. That's part number one. So this is serious. It's here. Part number two is this is not an efficiency tool. I've had so many CEOs tell me that this is just another technology in my toolkit. Initially I was polite and respectful, and I don't want to argue with the CEO. Then I finally said, look, this is ridiculous. I said, sure, you can look at it that way. It is another technology, but if you that is extremely myopic, this thing is going to transform the world for better and for worse. And you have to get with the program. So I would say embrace the charge and you cannot delegate it. The leadership absolutely matters. But you know, you have to do a lot of work. I mean, those of us who teach AI in business schools, we are reading and teaching. Like, I cannot give last week's lecture next week to a different class. It is that fast. So you have to take it seriously and embrace it.

SPEAKER_01 23:03

And it's a big leadership topic, right?

SPEAKER_00 23:05

It has to be in the So I run a conference with 100 leaders every year, and this is it's called the digital leadership agenda, and this is what they come for.

unknown 23:13

Okay.

SPEAKER_00 23:14

Thank you so much. The last thing I'll say is comprehensive view. If you you cannot just do any one of these, it has to be all of the above because this is not one function every time. Great.

SPEAKER_01 23:25

Thank you for coming. Thank you for sharing this. Pick up loss.