Transcrição completa de Recoding Business: The AI Transformation Flywheel | unDavos 2026

industrial revolution. 00:02:11.520 --> 00:02:13.520 I call this the digital tsunami in America. 00:02:13.599 --> 00:02:18.639 They know me as the digital tsunami guy, and because that's a metaphor I've been using for a long time. 00:02:18.800 --> 00:02:24.479 But I think it's really important to understand that we are at the literally in at a moment in history. 00:02:24.639 --> 00:02:26.000 We are watching an economy. 00:02:26.159 --> 00:02:34.479 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. 00:02:34.560 --> 00:02:35.680 Now, why is that important? 00:02:35.919 --> 00:02:43.439 Industrial revolutions are always enabled by technologies with electricity, the steam engine, the internet, lots of other examples like that. 00:02:43.680 --> 00:02:51.280 But what's important to remember is they transform the economy and society over very long periods of time. 00:02:51.520 --> 00:02:57.680 So we went from the agricultural economy to the industrial economy to the digital economy, and now the age of AI. 00:02:57.919 --> 00:03:08.879 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. 00:03:09.199 --> 00:03:13.120 It is also uh countries and societies. 00:03:13.360 --> 00:03:16.560 And the world as we know it is changing very, very profoundly. 00:03:16.800 --> 00:03:23.840 So the most important city in the late 18th century, 19th century in the world was London, England, right? 00:03:24.080 --> 00:03:25.680 They had the British Empire. 00:03:25.840 --> 00:03:29.759 It was because they mastered the second industrial revolution. 00:03:29.919 --> 00:03:35.759 And then it was the US, now China, and in the age of AI, it's it's uh the US and Japan. 00:03:35.840 --> 00:03:39.599 I mean, sorry, US and China are widely considered to be so the leaders in AI. 00:03:39.759 --> 00:03:48.080 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. 00:03:48.240 --> 00:04:01.520 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. 00:04:01.680 --> 00:04:02.800 Drug discovery. 00:04:03.039 --> 00:04:09.439 Google's got a fantastic model for climate prediction, you know, which integrates atmospheric physics with AI. 00:04:09.599 --> 00:04:15.439 And there's more and more examples that I can give you of where AI enables discovery. 00:04:15.680 --> 00:04:22.639 So a lot of industrial revolutions have been about rethinking our businesses for scale, for scope, how we manage them. 00:04:22.879 --> 00:04:32.879 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. 00:04:32.959 --> 00:04:43.519 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. 00:04:44.079 --> 00:04:52.959 AI allows us to discover lots of new things in science, in history, in any field, marketing, advertising, it doesn't matter. 00:04:53.120 --> 00:05:00.160 We can discover new patterns because anywhere there is a pattern with data, deep learning in particular can discover that. 00:05:00.480 --> 00:05:09.839 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. 00:05:10.720 --> 00:05:16.079 One of the things I want to emphasize in the few minutes I have here, now we hopefully get to more detail. 00:05:17.040 --> 00:05:23.360 I'm gonna be bold here and say most companies are approaching this somewhat myopically. 00:05:23.600 --> 00:05:30.720 So a lot of what I hear about is AI for software development and AI for customer service. 00:05:30.879 --> 00:05:31.920 You've all heard that. 00:05:32.160 --> 00:05:39.439 So the much more important question to be asking is what is the future of companies or the future of society? 00:05:39.519 --> 00:05:43.600 And I'm going to focus on companies, I'm a business school professor in the age of AI. 00:05:43.759 --> 00:05:46.639 What does it mean to be an AI-first company? 00:05:46.800 --> 00:05:48.240 What is that real what does that mean? 00:05:48.319 --> 00:05:49.360 What do you have to do? 00:05:49.519 --> 00:05:51.439 So I built this framework. 00:05:51.680 --> 00:05:54.319 It's sort of the what we call the flywheel. 00:05:54.480 --> 00:05:58.959 It begins by asking the question, what is your vision of the future? 00:05:59.439 --> 00:06:06.879 And I will so the technologies like AI are called a general purpose technology, which is an economics term. 00:06:07.040 --> 00:06:10.959 And what that basically says is there's something called the production frontier. 00:06:11.040 --> 00:06:17.680 What is the most you can do with the people you have and the machines that you have, given the state of technology today? 00:06:17.920 --> 00:06:20.399 So if you go five years ago, it was something. 00:06:20.639 --> 00:06:26.319 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? 00:06:26.480 --> 00:06:28.560 And the answer is a lot more, right? 00:06:28.720 --> 00:06:32.000 And so if you're not operating at the frontier, you're going to fall behind. 00:06:32.160 --> 00:06:35.279 So the first question you have to ask is what is my vision for the future? 00:06:35.360 --> 00:06:36.720 It's not about efficiency. 00:06:36.800 --> 00:06:38.879 It is about how much more can I be doing? 00:06:39.040 --> 00:06:44.000 What is my value proposition for the future that is better than ever

before? 00:06:44.399 --> 00:06:50.160 Often that translates into what know-how and intellectual property. 00:06:50.399 --> 00:06:57.040 So companies that succeed are companies that know how to do something better than your competition. 00:06:57.199 --> 00:07:01.600 You have to be really, really good with Microsoft, very good at software. 00:07:01.759 --> 00:07:06.639 But you have to ask yourself in the age of AI, what more can I be doing? 00:07:06.879 --> 00:07:12.800 Then it comes down to technology ecosystem, which is where are you going to get these AI capabilities from? 00:07:12.959 --> 00:07:14.240 What are you going to do yourself? 00:07:14.480 --> 00:07:16.639 What are you going to partner with others to do? 00:07:16.879 --> 00:07:23.120 And that is a whole new, because the AI technology stack is quite different from the technology stacks that we have today. 00:07:23.920 --> 00:07:29.360 Biggest thing which sort of we alluded to is talent, work, and one more word that should be on the slide is culture. 00:07:29.600 --> 00:07:36.800 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. 00:07:36.959 --> 00:07:38.079 So we have to think about. 00:07:38.240 --> 00:07:41.680 I gave a talk to the PeopleOps team, which is HR at Google. 00:07:41.759 --> 00:07:48.879 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? 00:07:49.120 --> 00:07:51.759 And you still talk to HR leaders and they're talking about career leaders. 00:07:51.839 --> 00:07:57.040 I'm like, what career, what ladders are not doing what what tracks are these people going to be on? 00:07:57.199 --> 00:08:02.000 So you have to think about jobs and tasks and what is AI going to be able to do better than humans? 00:08:02.160 --> 00:08:04.000 What will humans do better than AI? 00:08:04.240 --> 00:08:11.439 And I fundamentally believe that we should focus AI on allowing humans to be more productive than having humans be replaced by now. 00:08:11.519 --> 00:08:18.000 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. 00:08:18.160 --> 00:08:24.480 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? 00:08:24.560 --> 00:08:25.600 So you think about it. 00:08:25.680 --> 00:08:30.800 And finally, upskilling and educating the workforce, because that is a huge, huge challenge. 00:08:30.959 --> 00:08:41.919 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. 00:08:42.080 --> 00:08:46.399 And at the center of it all, there's an operating model, which is how do you get this wheel to spin? 00:08:46.639 -->

00:08:50.320 So these general purpose technologies are transformative. 00:08:50.879 --> 00:08:57.360 And if you approach it function by function, process by process, you are not going to get the Big Bang. 00:08:57.600 --> 00:09:00.399 You have to do all of these things simultaneously. 00:09:00.480 --> 00:09:02.879 It is a comprehensive transformation process. 00:09:03.120 --> 00:09:16.960 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? 00:09:17.120 --> 00:09:21.840 What is the nature of your value proposition and how do you deliver that with technology? 00:09:22.000 --> 00:09:25.440 And it takes all of these things to be invested in simultaneously. 00:09:25.519 --> 00:09:33.120 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. 00:09:33.279 --> 00:09:36.240 I can give you examples when Stefan asked me a few questions. 00:09:36.480 --> 00:09:37.279 So thank you for that. 00:09:37.360 --> 00:09:40.799 And if you I'm sure there'll be a lot of questions, I try to do it very, very quickly. 00:09:41.039 --> 00:09:41.919 You can come closer. 00:09:42.080 --> 00:09:42.240 Okay. 00:09:42.399 --> 00:09:42.799 Oh, sorry. 00:09:43.440 --> 00:09:44.399 He wants me on the red chair. 00:09:44.480 --> 00:09:45.279 I'm not quite sure what. 00:09:45.519 --> 00:09:46.000 The red one. 00:09:46.080 --> 00:09:46.240 Yes. 00:09:46.399 --> 00:09:47.120 The red pill. 00:09:47.440 --> 00:09:49.120 It's the hot, the hot seat here. 00:09:49.279 --> 00:09:50.879 So um so thank you for that. 00:09:51.039 --> 00:09:51.600 Thank you for that. 00:09:51.759 --> 00:09:53.200 Thank you for the flywheel. 00:09:53.440 --> 00:09:55.679 Um actually, we can start. 00:09:55.759 --> 00:10:00.399 Maybe you give some examples you want to do this, because I would have a question for the the value creation. 00:10:00.639 --> 00:10:12.159 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? 00:10:12.399 --> 00:10:13.120 Where is it? 00:10:13.279 --> 00:10:15.039 And where is it generated? 00:10:15.200 --> 00:10:16.960 And how great question, right? 00:10:17.200 --> 00:10:24.159 So I I so to increasingly from the data that you have and what and what makes your company do. 00:10:24.480 --> 00:10:25.360 I'll give you a great example. 00:10:25.440 --> 00:10:27.120 I'll give you a very simple example. 00:10:27.279 --> 00:10:28.799 Uh there's a company called Dr. 00:10:28.960 --> 00:10:30.720 Treat, which nobody knows about. 00:10:30.799 --> 00:10:32.480 That's why I like to use this example. 00:10:32.720 --> 00:10:34.960 It's a two-veterinary office clinic. 00:10:35.039 --> 00:10:37.440 So they see animals,

and I'm an animal lover. 00:10:37.600 --> 00:10:40.559 And this guy was a software entrepreneur, and it's sort of interesting. 00:10:40.639 --> 00:10:43.440 He just saw one more space ripe for disruption. 00:10:43.679 --> 00:10:50.799 But what is powerful about that the company was uh he decided, I'll just give you a simple version. 00:10:50.879 --> 00:10:59.200 So in the US, animals are considered property, they have no rights, so they have none of the healthcare regulations that humans have. 00:10:59.519 --> 00:11:01.200 But but here's the thing. 00:11:01.360 --> 00:11:14.960 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. 00:11:15.039 --> 00:11:17.279 Because as it uh so that's part one. 00:11:17.440 --> 00:11:24.639 And part two, they bought, because they're a software-focused AI-first company, they went out and bought an insurance company database. 00:11:24.799 --> 00:11:36.799 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. 00:11:37.039 --> 00:11:42.639 And so they may not know about the unique peculiarities that that particular breed or subbreed has. 00:11:42.879 --> 00:11:45.279 And all of a sudden, AI is giving them much better things. 00:11:45.360 --> 00:11:56.000 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. 00:11:56.159 --> 00:11:58.799 So that's an example of the new frontier. 00:11:58.960 --> 00:12:00.799 How much more can you do? 00:12:00.879 --> 00:12:07.279 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. 00:12:07.440 --> 00:12:11.440 Uh and so this is an example of something that we can do a lot better. 00:12:11.600 --> 00:12:12.799 So that's one example. 00:12:13.200 --> 00:12:21.600 Which is more, would you say that that is the typical disrupting approach, or is it adding value where there was actually no markers? 00:12:22.000 --> 00:12:23.919 No, I think it's a disruptive approach. 00:12:24.080 --> 00:12:24.240 Okay. 00:12:24.480 --> 00:12:25.759 So so but here's the thing. 00:12:25.840 --> 00:12:31.039 I actually found out that I said, how is how are they doing with the rollout across the 700 clinics? 00:12:31.200 --> 00:12:34.960 And it's going very slowly because this is where the cultures don't match. 00:12:35.200 --> 00:12:43.120 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. 00:12:43.279 --> 00:12:46.399 Uh so we come to the culture of breakfast uh as always. 00:12:46.799 --> 00:12:48.559 Comment, always, always coming back to this.

00:12:48.720 --> 00:12:53.840 Um, I mean, you talked about the um, you know, the human human side as well. 00:12:54.000 --> 00:12:57.519 I would like you to touch shortly on on the on the human role. 00:12:57.600 --> 00:13:03.440 Like, and when it comes to leadership, human leadership, we we talked a lot about what is necessary here and leadership. 00:13:03.679 --> 00:13:07.279 Um what do you see as well? 00:13:07.440 --> 00:13:18.399 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. 00:13:18.799 --> 00:13:39.279 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? 00:13:39.360 --> 00:13:41.039 So they are and then we have 10,000 of them. 00:13:41.200 --> 00:13:44.639 So um they will never be AI first, but they want to maybe adopt. 00:13:44.720 --> 00:13:47.840 So what is needed in those leadership circles? 00:13:48.000 --> 00:13:49.200 What would you say? 00:13:50.000 --> 00:13:56.080 So um it's a tough question to answer because you know, a lot of leaders are talking the AI. 00:13:56.559 --> 00:13:58.320 That's why you are here and you are the professor. 00:13:58.720 --> 00:14:01.759 Um I I think so. 00:14:01.840 --> 00:14:08.240 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. 00:14:09.039 --> 00:14:22.080 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. 00:14:22.159 --> 00:14:24.559 I mean, I would be an idiot to say that you don't do that. 00:14:25.120 --> 00:14:29.919 But on the other hand, they're not focused enough on growth and innovation around AI. 00:14:30.000 --> 00:14:33.440 And I think uh this is for better and for worse. 00:14:33.519 --> 00:14:44.559 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? 00:14:44.639 --> 00:14:51.440 So winner, you know, economics, we basically say first mover advantage, and the company that wins will can win quite big. 00:14:51.600 --> 00:14:57.840 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. 00:14:58.159 --> 00:15:01.840 But Elvin in the morning, you know, when we started, was questioning exactly that. 00:15:02.000 --> 00:15:03.600 Is it really the winner-takes all? 00:15:03.759 --> 00:15:08.080 Because it

was interesting to see him say that probably it's not happening. 00:15:08.320 --> 00:15:16.000 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. 00:15:16.159 --> 00:15:20.799 And there will be also, and I I also, you know, might would like to get your comment. 00:15:20.879 --> 00:15:23.519 Uh, you know, Scott Galloway from MWU Stern. 00:15:23.759 --> 00:15:27.440 And Scott, um, in a conversations we had in Hamburg like last year, it was interesting. 00:15:27.679 --> 00:15:28.480 It was indeed theater. 00:15:28.639 --> 00:15:28.879 Yeah. 00:15:29.039 --> 00:15:35.039 And and he he was asked uh what he sees, what what role AI would play? 00:15:35.120 --> 00:15:43.440 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. 00:15:43.600 --> 00:15:47.440 Um, and there's not he thinks not one company will take it all. 00:15:47.600 --> 00:15:49.679 He doesn't he doesn't think that this will happen. 00:15:49.919 --> 00:15:54.480 It's would you agree here or would you just so let let's let's break it down to different segments. 00:15:54.720 --> 00:15:54.799 Right. 00:15:54.960 --> 00:16:03.120 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. 00:16:03.279 --> 00:16:08.240 Um I I think increase I don't believe that all of them will survive. 00:16:08.399 --> 00:16:09.600 Uh I just do not. 00:16:09.840 --> 00:16:17.600 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? 00:16:17.679 --> 00:16:24.320 But I mean, but so there will be, and if you look at all software markets, there's usually two or three. 00:16:24.480 --> 00:16:28.080 So we I'd rather use the phrase winner takes most than winner take all. 00:16:28.399 --> 00:16:30.240 Because winner takes most basically said this. 00:16:30.320 --> 00:16:34.399 So if you look at search, how many of you remember how many companies started off an internet search? 00:16:34.960 --> 00:16:43.679 Yeah, there was Ask Jeeves, uh, Yahoo actually started, you know, and and and any number of companies, and ultimately it became Google. 00:16:43.840 --> 00:16:48.559 But but the point of that is ultimately there was one. 00:16:48.720 --> 00:16:53.200 And there's still it, you know, there's still others, uh, but not that many. 00:16:53.360 --> 00:16:55.120 And they have much smaller market shares. 00:16:55.279 --> 00:16:58.879 And you look at operating systems for uh smartphones. 00:16:59.120 --> 00:17:02.639 I mean, that there used to be a lot of different companies, now we're down to two, right? 00:17:02.799 --> 00:17:09.119 And so I think these software markets have that kind of economic, probably high fixed cost, low marginal cost. 00:17:09.200 --> 00:17:14.960 Now, in AI, there's a lot of marginal cost, but the training is very, very expensive. 00:17:15.279 --> 00:17:23.839 But I but that said, there will be sort of smaller models built on top of some of these foundational models, without a doubt. 00:17:24.160 --> 00:17:34.640 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. 00:17:34.880 --> 00:17:39.759 You are different and better than your competitors on some dimensions, maybe not all, right? 00:17:40.079 --> 00:17:46.880 And that is what you have to ask yourself the question: will AI erode that competitive advantage? 00:17:47.039 --> 00:17:50.400 Because you can copy some of the know-how pretty easily with AI. 00:17:50.480 --> 00:17:52.480 So what is going to make you distinct it? 00:17:52.640 --> 00:17:58.160 But there's plenty of physical world companies that will do just fine because if you're a hotel, location matters. 00:17:58.319 --> 00:18:02.640 If you're a pharmaceutical firm, if you have a patent on a drug, you don't need to be first. 00:18:02.720 --> 00:18:04.799 So you have to be AI intensive. 00:18:05.039 --> 00:18:13.519 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. 00:18:13.839 --> 00:18:16.319 But I think AI intensive is probably a better phrase. 00:18:16.400 --> 00:18:17.119 Okay, I like that. 00:18:17.440 --> 00:18:20.960 Um So striking examples, my question. 00:18:21.279 --> 00:18:35.680 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? 00:18:36.400 --> 00:18:43.039 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. 00:18:43.200 --> 00:18:47.920 And uh I actually had the president of Reuters speak at my conference, and I was really impressed. 00:18:48.000 --> 00:18:54.319 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. 00:18:54.720 --> 00:19:03.839 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. 00:19:03.920 --> 00:19:12.400 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. 00:19:12.559 --> 00:19:15.599 If you're second to market, it's you don't get the same advantage. 00:19:15.759 --> 00:19:24.880 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. 00:19:25.200 --> 00:19:31.519 He

invests in hiring news reporters and putting them in different parts of the world. 00:19:31.759 --> 00:19:40.799 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. 00:19:40.960 --> 00:19:46.640 So if you think about that for a second, that to me is what we're talking about, which is what makes you different. 00:19:46.720 --> 00:19:48.640 Well, I have news that nobody else does. 00:19:48.799 --> 00:20:01.279 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. 00:20:01.440 --> 00:20:05.839 And Reuters, this is a funny aside, got its start with carrier pigeons. 00:20:06.240 --> 00:20:07.839 How many of you know this story? 00:20:08.160 --> 00:20:08.960 Can I tell it? 00:20:09.039 --> 00:20:09.359 Yeah, please. 00:20:09.599 --> 00:20:10.880 It's a very funny story, actually. 00:20:11.039 --> 00:20:11.519 So Mr. 00:20:11.599 --> 00:20:19.599 Paul Reuter figured out that financial information was actually conveyed by wire. 00:20:19.839 --> 00:20:27.279 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. 00:20:27.359 --> 00:20:28.720 There was no wires possible. 00:20:28.960 --> 00:20:33.599 So the wire was printed out in Aachen, sent by train to Brussels. 00:20:33.839 --> 00:20:48.240 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. 00:20:48.400 --> 00:20:50.799 Now, here's the here's the here's the funny part. 00:20:51.119 --> 00:20:56.480 How did the pigeons come back to Germany by the same trend that was kept? 00:20:56.720 --> 00:21:00.480 Because I didn't know this homing pigeons can only go home in one direction. 00:21:00.559 --> 00:21:02.720 They cannot understand two homes. 00:21:03.039 --> 00:21:03.759 I didn't know that. 00:21:03.839 --> 00:21:04.559 That's funny. 00:21:05.519 --> 00:21:06.640 Yeah, that's a nice picture. 00:21:07.039 --> 00:21:08.559 Um the point is differentiation. 00:21:08.799 --> 00:21:09.119 Right. 00:21:09.359 --> 00:21:13.519 Um we would need more time because I have so many more questions. 00:21:13.599 --> 00:21:15.680 Um, but I like the the example. 00:21:15.839 --> 00:21:18.240 Um maybe last question. 00:21:18.480 --> 00:21:29.680 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. 00:21:30.640 --> 00:21:33.680 So I think the first thing is two pieces of advice. 00:21:33.920 --> 00:21:37.680 Number one, you have it, it's real, it's here. 00:21:38.079 --> 00:21:44.480 This McKinsey estimates$7 trillion being invested in AI infrastructure by 2030. 00:21:44.960 --> 00:21:46.240 It's not going away. 00:21:46.319 --> 00:21:47.359 It is going to change the world. 00:21:47.519 --> 00:21:51.359 Whether you like it or not, I have misgivings just like any of everybody in the room. 00:21:51.599 --> 00:21:53.680 But at this point, you cannot stop it. 00:21:53.920 --> 00:22:03.440 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. 00:22:03.599 --> 00:22:12.799 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. 00:22:12.960 --> 00:22:13.839 That's part number one. 00:22:14.000 --> 00:22:14.880 So this is serious. 00:22:14.960 --> 00:22:15.440 It's here. 00:22:15.680 --> 00:22:18.559 Part number two is this is not an efficiency tool. 00:22:18.799 --> 00:22:23.839 I've had so many CEOs tell me that this is just another technology in my toolkit. 00:22:24.079 --> 00:22:27.680 Initially I was polite and respectful, and I don't want to argue with the CEO. 00:22:27.839 --> 00:22:29.759 Then I finally said, look, this is ridiculous. 00:22:29.920 --> 00:22:31.839 I said, sure, you can look at it that way. 00:22:31.920 --> 00:22:38.880 It is another technology, but if you that is extremely myopic, this thing is going to transform the world for better and for worse. 00:22:39.119 --> 00:22:40.880 And you have to get with the program. 00:22:41.039 --> 00:22:44.960 So I would say embrace the charge and you cannot delegate it. 00:22:45.039 --> 00:22:46.960 The leadership absolutely matters. 00:22:47.119 --> 00:22:48.880 But you know, you have to do a lot of work. 00:22:49.039 --> 00:22:55.359 I mean, those of us who teach AI in business schools, we are reading and teaching. 00:22:55.519 --> 00:22:59.440 Like, I cannot give last week's lecture next week to a different class. 00:22:59.599 --> 00:23:00.640 It is that fast. 00:23:01.039 --> 00:23:03.279 So you have to take it seriously and embrace it. 00:23:03.519 --> 00:23:05.279 And it's a big leadership topic, right? 00:23:05.680 --> 00:23:13.119 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. 00:23:13.359 --> 00:23:13.680 Okay. 00:23:14.160 --> 00:23:14.960 Thank you so much. 00:23:15.279 --> 00:23:17.759 The last thing I'll say is comprehensive view. 00:23:18.079 --> 00:23:24.400 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. 00:23:24.960 --> 00:23:25.200 Great. 00:23:25.839 --> 00:23:26.240 Thank you for coming. 00:23:26.880 --> 00:23:27.599 Thank you for sharing this. 00:23:27.759 --> 00:23:28.960 Pick up loss.