Milken Institute Global Conference 2025, panel: Part 2: AI and the Economy: Transformation and Disruption

Speaker 1: Please welcome the panel on AI and the economy, transformation and disruption, moderated by news anchor and fintech founder Richard Louie.

Speaker 2: Well I was thinking we all should have worn our black leather jackets just to be consistent but again we got the note a little bit later. Thank you all for being here for this conversation. Really something else to follow Jensen and Mike, some super questions and conversations. I have the great honor of introducing our panelists to discuss AI and the economy. I'll start with Lorena Yee.

Speaker 2: She's a Senior Partner Director of McKinsey Global Institute McKinsey and Company. And as I introduce each and every one of you, tell me what brought you to AI, what's your interest, as well as where are we in the arc? So Lorena, you first.

Speaker 3: Well I've spent the last twenty five years in Silicon Valley helping companies with all of these technology transitions. And for the last five years, I've led all of our frontier research. So it's kind of a front row seat to what is one of the most exciting technology transitions that we will experience as humans. So where are we on the curve?

Speaker 2: Where are we?

Speaker 3: Yeah. Is it a curve?

Speaker 2: Is it a hockey stick? Is it an s? What what is it?

Speaker 3: So I think that there are two curves, not to complicate it. One is the innovation curve and one is the adoption curve. On the innovation curve, whether you make that, you know, a hockey stick or you make that just innings of a baseball game, I think we're about halfway there. Okay. And we have a separate debate about AGI.

Speaker 3: On the adoption curve, which is probably what most people here are focused on, we're very early innings. And we know that 1% of companies will say that they are at AI maturity, 11% are in production. Those are very low numbers for the last three years.

Speaker 2: Very interesting. Jason Allen Snyder, I'm just going to go down in order as you get farther from me in distance. Chief Artificial Intelligence Officer Momentum Worldwide. How did you get here? What brings you to AI and where are we in the arc?

Speaker 4: Sure. So I'm a technologist, a futurist and an inventor And I have spent the majority of my career at the intersection of culture, technology and business. And my background is in systems, deep systems engineering and embedded computer systems. So, have been inside of machine learning and AI for a very long time. To answer the question about kind of where we are in the curve, I'd say, it's a for me, it's a it's tell the story this way.

Speaker 4: I think I think it's apt that we're having the conversation here in Los Angeles, so close to Hollywood, right? Because we have people that aren't lawyers and they're not doctors and they're not software engineers, but they play them on TV and in the movies. And I think that's kind of where we are with AI today, right? Generative AI doesn't reason, it performs reasoning in the same way an actor would perform a part. And I think that that's very important to understand because I don't really believe we have artificial intelligence today.

Speaker 4: I think we have artificial inference. We have machines that are very good at predicting language. They have good manners and they're very, very good at rhetoric. And where that comes into play with business obviously is that whoever is closest to the consumer controls the conversation and they seem to control.

Speaker 2: It sounds like you're either talking about somebody at a dinner party last night or we're very early on with good manners. Mark Minowicz, strategic partner Mayfield, President Going Global Ventures. Tell us what brought you to this space and where are we in the arc?

Speaker 5: Thank you very much. It's an honor and pleasure to be here at this prestigious Milken Conference. I think this is the story right now today. This is probably one of the most important panels that we're seeing and we're sensing and we're feeling it. AI is our economy.

Speaker 5: AI is the world, and I'm passionate about this. I think we're going through massive transformation. I'm a visionary. I'm a leader in AI. I'm an investor.

Speaker 5: I've been involved with AI for over twenty years and various other technologies. I think we are experiencing something really incredible. The world of world of research, the world of innovation, the world of technology has seen AI scientists, people, they've been experiencing it since nineteen sixties, actually nineteen seventies. All of this machine learning, neural networks, that it's been around. But something has happened in 2022, November 2022.

Speaker 5: You all remember where you were at that moment. OpenAI announced shared GPT. It changed their lives forever. Suddenly, this machine is now helping you to solve problems. It's writing something.

Speaker 5: It's creating context. It's doing your marketing. It's researching. From the two three year period, Richard, we have really matured. Three year period, we are we are doing contextual understanding.

Speaker 5: We're doing some basic reasoning. We're doing some marketing and insights. But now, seems today, we are moving to mission critical stuff. And what I'm really excited and the reason I got involved is because I felt this is going to be a revolution which is gonna change the world, and I didn't wanna miss it. But most importantly, I didn't wanna be in the world of efficiency as corporations and financiers and funds want AI to be.

Speaker 5: I want AI to help us, humanity, people around the world. We all of us, we need help in health, in infrastructure, in smart cities. And I think that we have to take AI to this better planet, more sustainable planet. That's why I'm here. My heart is here, and I want to make sure that AI goes in the right direction.

Speaker 5: Thank you.

Speaker 2: The technical crew, his mic's not working.

Speaker 5: Mark, thank Maybe we'll fix it.

Speaker 6: No. I love it. I love it.

Speaker 2: Roger Ferguson, Chief Investment Officer and Director Red Salt Partners. Quite a career you've had as well. Tell us what brought you to this space and where are you where do you believe we are in the ARC and speak as loud or as quietly as you'd like.

Speaker 6: For those who know me, my background is in central banking, so we'd like to calm things down a little bit. So what brings me here? Three things, frankly. First, I'm on a number of boards, including the board of Alphabet. All of you know that company, obviously, and the incredible role that it is playing in driving forth.

Speaker 6: It created the the most important pay for and transformers through through our DeepMind subsidiary. And I'm on the board of another company called Klarna, which is is an active user. So I've learned a lot from the board perspective in terms of how AI is influencing the world and might influence the world going forward, though I am not a spokesperson for either of those two companies. Secondly, I am, as you said, the chief investment officer of Red Cell Partners among other things. That is an incubation firm that I cofounded with a young colleague of mine plus several other partners.

Speaker 6: And we have incubated now 13 or 14 different companies, many of which are built around using AI to do some of the transformations you're talking about, health care, cyber, defense, and other things. And then the third thing that brings me here is my background as an as an economist. We are talking about AI and the economy. You know, as an economist, I've been thinking a great deal about the history of technological transformation and what that means about the way we organize ourselves to provide goods and services, create jobs, etcetera, and that is critical to the way we think about AI. So those three things.

Speaker 6: Where are we on the curve? I think the answer is very early days. The technology is certainly moving, but we're not where we want to be in terms of the ultimate goal. I don't know what the ultimate goal may well be this mysterious thing called AGI, which is basically general intelligence. We'll find out if we're there.

Speaker 6: I think on on adoption, again, early days. There, though, I must say I'm surprised at how quickly we are moving. And the final point I make about the curve and let it go is unlike many other technological advances that we've looked back in history where there was an s curve of some sort, we don't actually know if this is gonna plateau in terms of the technology. And so, you know, where we are in the curve, early days. But I think the mystery here is what is the shape of this curve and how is it similar to or different from other technologies that we've seen?

Speaker 2: It's estimated that AI will contribute to the economy $20,000,000,000,000. As we know, our annual GDP here in The United States just under 30,000,000,000,000. Is that number too big, too small? And if that's not the number that we should be considering in understanding the impact of AI into our economy, what is it? 20,000,000,000,000, though is the one of the estimates out there right now.

Speaker 3: Go ahead. I think that's one of the more aggressive of that invest estimates. I mean, our own research, call it conservative, is 4,400,000,000,000 in terms of economic potential and that is without a time. That's probably one, two decades. So I'm not sure it's about getting a bigger number

Speaker 6: Yeah.

Speaker 3: Because actually in the near term, if you look at earnings that have come out, you're not seeing the millions to the billions. So I'm not sure it's really relevant if it's 4,400,000,000,000. Yeah. 9,000,000,000,000 is something I've seen

Speaker 2: Why is number lower? Why is it 5,000,000,000,000 roughly is what you're saying in your studies?

Speaker 3: So I'm not sure where your number is, but we did it bottoms up in terms of the impact of automation in actual actual work. Work. And so we've looked at two things, the impact on individual jobs and occupations as well as the impact on economy where you believe that generative AI would have a substantial difference in the outcome. I do think the more important question is not the headline of it's big. All of those numbers are more abundance than we can possibly

Speaker 2: For sure.

Speaker 3: Imagine. The question is the delta between the potential and the actual.

Speaker 2: Yeah.

Speaker 3: And what are both the innovators doing to bridge that gap and what are those of us who are business leaders doing to bridge that gap?

Speaker 2: And there's also the delta between that which is cost versus revenue, right? Because as you implement a new technology or anything into society, you will have a cost and you will have the revenue. And that's one of the considerations, what's the cost to implement? Jason, thoughts or

Speaker 4: Yeah. I mean, there's many domains inside of it, right? So automation is one part of it. But AI generally is a combinatorial innovation. So you have massive parts of this which are data and managing data decay and the information that's feeding the models.

Speaker 4: There's the model creation. And then there are all of the new and emerging technologies which are going to contribute to advances across AI beyond the chipsets, things like quantum biological computing that are going to change the dynamics. So I think it's a very complex environment with a lot of deltas.

Speaker 2: So five or 20?

Speaker 4: I'd probably say it's closer to 20.

Speaker 2: Closer to 20.

Speaker 4: Yes.

Speaker 5: And so I would say, I agree with you, Jason. I think it is a very complex environment and there are a number of pieces to it. If we look concretely, I think the digital labor is something which is going to generate trillions of dollars into the market space. Digital labor is already shifting, and we're going to talk about AgenTiKi. It's a big component of the digital labor.

Speaker 5: So I expect trillions and trillions of dollars to be accumulated, which is going to impact every industry in United States and the world. But we also have the new industries which are coming on, humanoid robots. You heard about all of this. Autonomous vehicles, robotics, in a health care field. So when you start combining the value, economic value creation and economic frameworks we're gonna which are gonna change around the world, we're talking about trillions of dollars.

Speaker 5: So probably more than 30,000,000,000,000. But in terms of our actual revenue, I think, Richard, you alluded to revenue. I think we are gonna get in a couple of years to about 200 or 300,000,000,000 of actual revenue. And we're already seeing with companies, consulting companies are claiming that the biggest impact that they're having right now with customers is AI. Some are even saying 2,000,000,000 or 3,000,000,000 or 4,000,000,000 is coming specifically from AI.

Speaker 5: So right now, we're seeing AI as a major driver of revenue around the world and AI is critically important. Again, I wanna emphasize this point. It has to do with economic value creation. If you're not going to focus on being a leader, AI first, you might as well give up. Leave the room.

Speaker 5: Leave the building. Go somewhere else. Okay? Because After panel. After the panel.

Speaker 5: Don't don't leave this building. This building is safe. Okay?

Speaker 2: Yeah. Yeah.

Speaker 5: We have great we have great mind, brilliant people, but but you have to focus on becoming AI first, whatever it takes. Because 2025 is the year. And by the way, we haven't even mentioned, Richard, data centers. The great amount of work that Jensen, who spoke eloquently, data center, the there's so much money that is going in in creation of the next generation compute infrastructure, chipsets, semiconductors. That's itself a multitrillion dollar industry, and it's only growing.

Speaker 5: So huge expansion. And I think the countries which are gonna capture this, the companies, whether it's financial institutions or governments are gonna do really well. The ones that are not gonna capture this are gonna fall behind and never come back. So, Roger, you handled post nine eleven. This was an absolute shock to the economic system.

Speaker 5: This would be a shock

Speaker 2: but the other way around potentially. Putting on one of the many hats that you have, how would you look at this on the policy side, on on the government side?

Speaker 6: So let me pick up exactly the the line of questioning that you're driving. I think it's sort of a fool's errand at this stage to try to put a number around the impact. So imagine if we were here, you know, a hundred and fifty years ago and somebody said, well, what's the impact of electricity? What's the impact of the industrial revolution? I think this is actually that transformative, to be very clear.

Speaker 6: In economic terms, we're going to say we're going to have a massive supply shock, which is potentially a good thing, but it is a double edged sword because, as we all know, supply shocks mean there'll be individuals who are put out of work and there'll be brand new jobs created. It's similar to what happened when we had an agricultural revolution in this country. We had about 85%, 90% of Americans at the time of the revolution were working in agriculture. We had the agricultural revolution, mechanized everything. We can now feed the world with 2% of our population.

Speaker 6: That was both good and bad. Right? If you were a person who was in the business of being what's called a farrier, I. E. Putting horseshoes on horses, you were out of work.

Speaker 6: If you were a person who was learning how to manage mechanisms, you know, a harvester or something like that, you had a new job. And so that's the challenge that we're confronting. And I think the ultimate answer here is when you see this kind of major shock, we're going to have to rethink some of the elements of capitalism. So I'm going to take this to a much higher level because every time we've had a shock, capitalism itself has been forced to rethink how it goes. So first, we had the industrial revolution that created capitalism 1.o.

Speaker 6: Then we discovered there are bad things about capitalism1.o, children working in mines, tainted food, tainted pharmaceuticals. So we had capitalism2.o in The United States. That's most closely associated with Teddy Roosevelt and progressivism. A genius of a man, very complex, created a whole new way of thinking about how The US capitalism system should work. And all of a sudden, we had the FDA.

Speaker 6: We had the SEC eventually under his cousin, Teddy rose Franklin Roosevelt. We had Social Security. So we created ways for capitalism to keep evolving as technology evolved. And now we then had capitalism two dot one. Jimmy Carter, then Ronald Reagan came in and said, we need to streamline the amount of regulation we have to allow The US economy to continue to grow and flourish.

Speaker 6: And that, I think, unleashed a massive amount of creativity in in The US economy. And now we're on the verge of capitalism 3.o, which is gonna respond to this set of questions, which is the big supply shock that's coming from AI will change the way jobs are done, create new jobs, but unfortunately, also eliminate other

Speaker 2: Roger, do you think we need a new cabinet secretary?

Speaker 6: Do we do we need a new cabinet

Speaker 2: I'm just staying in line with what you're saying is is possible. Right?

Speaker 6: And how big So I think to be fair, we don't know what we need. Just as, you know, in in when Teddy Roosevelt came in, he had this broad concept and we tried lots of things. So I think by the time we meet here in another five to ten years, we'll have a brand new way of organizing work. We'll have a very different view about things like Social Security. We'll be talking about things like universal basic income as perhaps perfectly normal or rejected.

Speaker 6: We'll be thinking about, you know, how we manage a world in which maybe, to your point, we no longer are driving cars. I'm a I'm a big believer in Waymo, to be very clear.

Speaker 5: We'll be flying.

Speaker 6: We flying possibly. And so I do think we're gonna have to really rethink our frame of reference for organizing the economy. So I think this supply shock from AI is actually about as big as what we saw at the agricultural revolution, the industrial revolution, information. Got it. And now this is a new one.

Speaker 2: Quick reaction from the panel quickly.

Speaker 3: Well, I mean, I just think if we're on the economy piece, one of the interesting things is where have we grown? Where have we gotten productivity? In the first instance, in that early stage, it was physical labor and how did you create basic machineries for physical labor. Then we had the transaction economy. You saw a lot of that in banking in particular with all the transactions and what you saw in the back office.

Speaker 3: What's really fascinating from an economic perspective is generative AI has tipped the scales on a class of labored knowledge work, which we actually haven't seen technology address at such scale. So the ability to address the full economy, I think, is pretty profound. But I think something Jensen said is really important as we think about how we design for the future, which is right now, and I actually usually am the most passionate person, but I think you win this contest and the most optimistic. I'm going to be a little bit more cautious. I think right now, the

Speaker 2: Have a thumb wrestle later on.

Speaker 3: I mean, I I think I'd have to have like a full body wrestle But I think what we see is that people are using the technology to do current work better, faster. That is not where you get any exponential gains. You will start to see the exponential gains when you say, if I were, as Jensen said, if I were to design a chip from scratch, if I were to think AI native and design how work gets done, how would that work? Not an activity, but the actual full work. And that is where we will start to see a more dramatic shift.

Speaker 3: Absent that, you've got a lot of technology and not the type of usage that brings in any trillions of dollars.

Speaker 4: Yeah. I I think the the only other point I would add is, you know, we're we're we're at a place where our technology has surpassed our legislation and on morality. And we've we're entering new territory where generative AI is actually creating a new understanding of our reality, some of it which is based on fabricated information, right? So it is no longer fake news. It is kind of fake truth.

Speaker 4: And we have to deal with this complexity in ways that we haven't before and it's going to take us a minute to unwind and understand what this means.

Speaker 5: Yes. So I just want to comment and I agree a lot with my colleagues and what they're saying, but I think the exponential growth of AI is so unprecedented that we haven't seen it during agrarian transformation or the industrial transformation and services. It is just unbelievable. Every seven months, you have massive improvements on compute. Every seven months, there are improvements in algorithms.

Speaker 5: Every seven months, we have new massive amount of data. We have new types of roles, new capabilities, new ways of solving problems, new business cases exponentially every seventh month or eight months. It is just growing by leaps and bounds. So nobody's prepared And I think you have to work together, the public sector and private sector. In order to get it right, you need large companies, small businesses, and governments working together.

Speaker 5: I just come back from meeting with the prime minister of Estonia. I spent some time in Estonia. And Estonia doesn't have actual people. We're very few, if any, in the federal government. All of that is automated.

Speaker 5: All of that is digitalized. And yet they are extremely efficient, yet they are somehow been able to protect the population, making sure the population is not only aware of AI, but are using AI everyday life. So I think for us to be successful, to make sure that everything that we spoke about today, it actually comes to fruition, we have to improve our education both in universities and enterprises. And this is not just a job for private sector. This is a job for private sector and the government.

Speaker 5: And we haven't done we haven't done our job. We are failing our population. Right now, population and students are failing. They don't know what the next job is. I don't know what to tell my daughter, 16 year old, what should she be doing, what kind of a job, And that puts a lot of stress, anxiety in the system.

Speaker 5: We have to spend the time now, all of us, in this room, throughout the country and the world and figure out where the next opportunities of jobs are gonna come from, space industry, quantum technologies, what's gonna happen? Because if we don't have the right formula, economic framework, we will definitely falter and we'll fall behind. And and even though the companies are gonna do great Right. Maybe some governments, people are going

Speaker 2: to fall behind. So since we're in the anxiety space for a moment and you rarely see Jensen flummoxed or stutter or not get his answer absolutely clear and clean. But when he was headed to DC and he was sitting there and the few times that he was wearing a tie, he was asked, so are we ahead or behind China? Mhmm. And whatever that answer is from all of you, so what?

Speaker 2: What's the next step? What should be done? Ahead or behind China?

Speaker 6: The United Can I jump in and and pick up on one of the points here that's really important, which is the speed of transformation? So I introduced the agrarian revolution, industrial revolution. All those things, to your point, took twenty five to fifty years to play out. And I agree with you. This one seems to be playing out in a matter of quarters and a much smaller number of years.

Speaker 6: I've seen on some of the boards, I've gone just massive adoption to this technology and so Are they doing it well? Very good point. Should think about

Speaker 2: Are they doing it well, Roger? I'm sorry? The massive adoption you're saying? Is it

Speaker 6: It's the speed of adoption. It hasn't been massive. But what I've seen is this is a first technology that can be done, as Jensen said, by anyone in this room. It does not have to be intermediated through the IT department, bringing out hardware and software and networking, all of these things. We already have the networking capability.

Speaker 6: We already have smartphones. We have our our Pixel phone or whatever the phone may be. And so now what's very interesting about this one is the ability for the relatively average person to start to be a programmer, as Jensen said, to use AI. And so the process of adoption is moving much more

Speaker 2: So quickly, have to ask you very that hang hang hang on one second. So So the question I've got for you since you're talking to boards here, this is on the board level, the word cyber is used. Now cyber is from what year?

Speaker 6: Cyber is from what year?

Speaker 2: From a year. In the nineteen sixties. Nineteen sixties. So are are they reacting well is my point when you're saying they're reacting quickly?

Speaker 6: Well, look, cyber is a slightly different thing.

Speaker 2: Yes. It is a different thing. I understand The answer

Speaker 6: is yes to to some degree. Right. The thing that's different about cyber is there is a there's an an enemy force that is attempting to use it to penetrate It's what I think of as the good more the term and So there, it's a combination of offensive and defensive, Now for let me pick up on two more points.

Speaker 2: Is Quickly, please. Yep.

Speaker 6: Here we are. Where are we vis a vis China? I think it is, one, unknown. But two, if you think about a couple of things that are really important here. First is the amount of data that's available.

Speaker 6: It's arguable that in some ways they're ahead on data because of facial recognition and other things that perhaps are done more easily there. Question of where we have the largest number of scientists right now, The US, I think, has the largest number of AI capable scientists, but that is not going to stay forever. And obviously, we've done a great deal in terms of chips here that they are catching up. So I think it's a close call.

Speaker 2: It's a close call.

Speaker 3: I also think that pace and you're both all picking this up. The pace of the technology innovation is so fast. It's also incredibly transparent as we all saw with DeepSeek. And I think one of the terms that Dario Amade uses a lot from Anthropic is race to the top. And so I think that's a very different paradigm when you are racing to the top in terms of how can I do this?

Speaker 3: If we look at the internet adoption, if we look cloud adoption, it was absolutely slow. It was years, it was planned releases, it wasn't sort of, I mean I don't even think it's seven months to be quite frank, it's down to weeks where things change and there's fundamental innovation not just in the parts that we see but all the way down the stack. But I think the most important thing is if you had a democratization of access and that's not a political term, that's a people can use this. They don't actually have to have a PhD in machine learning to actually be users, not creators of the technology. And if we looked at labor and we looked at shortages that we had, we have always been saying for the last decade, oh, if only I had more software engineers.

Speaker 3: So now let me put that positive question back. If you have 30% more software engineers in your company, how can you compete differently? You have to ask the question a slightly different way, not of scarcity but of increased abundance that you have at your disposal now.

Speaker 2: Ahead or behind China?

Speaker 6: Ahead. Ahead. Okay.

Speaker 3: But that's not a staying position forever. So race to the

Speaker 2: top. Jason?

Speaker 4: Ahead today for sure. But I I think the the point the data point is so important because without the data, these systems don't work. They're meaningless. Right? And we have tremendous data issues.

Speaker 4: Right? Whether it's about consent and how that data is collected, whether it's about how that data is decaying and then being hallucinated and amplified in negative ways inside of these models. And we really need to start to codify practices both really from a business perspective, right? Because we are focused on innovation, not regulation. And so, it's going to be the responsibility of everyone in this room and everyone endeavoring inside of this work to make sure that we are maniacally focused on making sure we have the best possible data that's based on consent.

Speaker 2: So hey Jason, if we need 100, in terms of the data that you're saying we need, that we may not have, just focusing on one point, if we need 100 and you're saying we don't have enough, do we have 50 right now? Do we have 40 right now? Do we have 75?

Speaker 4: It's difficult to answer that question because the veracity of the data that we have is unclear. Right?

Speaker 2: But just give me a general sense. I mean, are

Speaker 4: we how empty is this glass? It's pretty empty in my opinion. And the real risk that we have is the synthetic data that's being created by these models. When we're looking at the big tech companies, we're looking at the meddas and the open AIs, it's the research that they're doing which is being largely automated and largely use synthetic data based on training data, which is bad.

Speaker 2: Mark, quickly, do you agree? Are we glass half empty on on what we need to do it right on the data side when we're thinking about AI? What do we

Speaker 5: have more? We're capturing a lot of information from public data, and I think the most interesting data is going to come from private sources, and that's where sovereign AI comes in Yes. Down the line. And there's gonna be a line right now, a red line that companies and countries and institutions cannot enter. There's gonna be a firewall.

Speaker 5: So to to that point. But I wanna go back to The US and China. If you were to ask this question, Richard, about two two or three years ago when OpenAI released shared GPT, we were so far ahead. It wasn't even it wasn't even a question. We let our guard down.

Speaker 5: So today, China is very, very close. We're still winning on the large language models. We still have the major concentration of LLMs and foundation models across the world. We're winning by far in in that category. We're also leader in AgenTek AI, this new set of agents or digital workers that are doing actionable things.

Speaker 5: We're winning there. We have hundreds and hundreds of companies and industries in the beginning stages, but we have a lot. No one else in the world even comes close. But China is getting much further ahead in robotics. It's much further ahead in drone technologies and it's and it's also evening up in some maybe less sophisticated foundation models and LLMs.

Speaker 5: So it's challenging us. And you and you basically said in terms of researchers, Jensen mentioned in Washington DC, and it really shook me. He said 50% of the researchers in AI are now coming from China. If you look at the patents worldwide, intellectual property, majority is now not coming from the Western world. It's coming from China.

Speaker 5: So it's a serious contender. But also, there there are some interesting hotspots, Richard. Yeah. You were just there in United Arab Emirates, in KSA. We have many members, colleagues here from the MENA region.

Speaker 5: They're doing some amazing stuff. They're saying, well, we're gonna compete because we have energy. We have abundance of power. We have surplus. We have a lot of capital.

Speaker 5: We have we are gonna be able to get a lot of bright people from across the world. So they're becoming regional and, in many senses, global hotspots on innovation and AI, especially at The Kingdom Of Saudi Arabia and UAE. That's the region you have to watch for. There's lots of investments. And not only investments, Richard, we come down to this point is, do we have a vision in this country?

Speaker 5: I think now with president Trump and the people that he brought over, we have part we have that vision. We have vision coming from Silicon Valley. But for many years, we didn't have a vision. In in MENA Region, they have a vision, vision 2020. They are building smart cities.

Speaker 5: They're building new hub they're building new hubs. They're building new logistics. So right now, what I would do is I would watch the MENA Region. We're still the powerhouse. China is slowly catching up.

Speaker 5: The rest of the world is asleep, sleeping. Thank you.

Speaker 6: I put one more point? So let me let me

Speaker 2: Build on that, but I do wanna ask you one question. This is in your wheelhouse related to this, and that is lawmaking. So The UAE, for instance, is now bringing AI in into their lawmaking, And this brings up the issue of government and efficiency. Do react to what was said and I also want to get your thought on that.

Speaker 6: Sure. So let me react to a couple of points that were said. One is this levelling of the global playing field that is emerging. So you just came back from UAE. I'm getting ready to go.

Speaker 6: But I just came back from Singapore. Just across the border from Singapore and Malaysia, they are building a massive number of data centers, all for basically training these models. And what we're going to discover, I think, is jobs around new jobs, new capabilities around the world because of the ability to network very quickly

Speaker 5: Yeah.

Speaker 6: To pick up and do things in a greenfield. So this is actually a very exciting potential where we will have, for the first time, a technology that may bring the world along together. Now to your bigger point or your other point on regulation and government, two or three points. First, having been a regulator for ten years at the Federal Reserve, I learned that first, have to know what you're regulating. You cannot regulate the models themselves.

Speaker 6: I think what you want to regulate is the interface between the models, call it AI, and individuals, where harm can be done. Because the goal of regulation, I think, is to incent innovation but reduce the amount of harm. So go back to my initial analogy about capitalism one dot o, the goal of f of the FDA was to make drug development safe. And so we have to figure out where is the harm potential in this job, in this AI. I don't think we know the answer to that question yet.

Speaker 6: So I'd be pretty cautious about regulating. The second point I'd make about regulation is you have to do it, I believe, when you're in these new areas, if you can do it in cooperation with the industry. The major AI drivers in this country, Alphabet, Jensen's company, you name

Speaker 2: it,

Speaker 6: want AI to be widely adopted and consequently, they have to want it to be safe. And they need to be in the room when we talk about what we're regulating because they are I think the experts on where the risks are versus the safety.

Speaker 2: Great points made so far by our panelists and I will get to some of the comments that all of you are making on the app on the QR code very soon. Have about eight minutes left. We'll

Speaker 5: get Okay. To that Stay last

Speaker 1: three

Speaker 2: with me. Want to will.

Speaker 6: We'll let Richard do his job as modeling.

Speaker 2: Yeah. It into the next one is what I'll say. I want to get into hardware, the stack. Right? The CapEx conversation.

Speaker 2: There seems to be one clear leader when we're looking at chips and hardware and even as was said on this stage about thirty minutes ago, well if you want a job, what how are you going to help me build the next chip? Right? So react to that what is the next chip? React to the, if you will, almost single source in terms of the leading hardware piece you need to do well. And three, the stack overall when it we're looking at because we're not just talking to LLMs, we're not just talking to databases, we're talking about the data centers as well.

Speaker 2: Who wants to go first?

Speaker 3: We talked about it a little bit before. So there's a lot of focus on the chip. According to Jensen, we should all just check out YouTube for understanding I mean, I think that the chip is getting a lot of the spotlight, but what you really have to look at is the full infrastructure stack. And the reason why I look at that is because right now the vast majority of companies who are deploying AI have not yet done it at scale. If you are using it at scale where like every single person is an active user and using a massive amount for work and changing that, the infrastructure costs

Speaker 2: Right.

Speaker 3: Will be enormously high. That's a bill that most people can't afford. So what we need to be also looking at is the innovation across the infrastructure stack to take the total unit cost down. And I think in there, we're gonna find these hidden gems. And so you were talking about the data center build out and how that is actually gonna be helpful for a lot of sovereign economies and also places in The United States where data centers are being built.

Speaker 3: And there are going to be not just sort of this simplistic paradigm of jobs gains, jobs lost. There are also going to be jobs transformed, jobs renewed. So for example, in rural parts of The United States where we're building data centers, we're finding we have a shortage of what job? Electricians. Why?

Speaker 3: Because electricians are incredibly important for data centers. We are also seeing what I think of as like the completely cool basic innovations, cooling technology. We could go on and on and on. Like, that's actually really interesting. Or I met a start up and they said, you know what?

Speaker 3: We're putting our data centers near wind farms where energy actually trades below market because they haven't actually been able to extract that and use Those are investments we made ten, twenty years ago. And what's really cool is if they put their data center there, they could potentially get the price of that wind energy up to market. So I think there's like a lot, like, the big headlines that is really, really exciting for us. Quickly. Maybe not as exciting as

Speaker 2: you constraint? We don't seem to have a capital constraint at the moment. One word answer. What's the constraint right

Speaker 3: now? Pace.

Speaker 2: Pace. Mark, I knew one of you know this.

Speaker 5: I I just wanna comment. I think there are really great comments that just came. I wanna say right now, when you look at the data centers, they're gonna be they're being built differently right now. And the way they're gonna be built, it's not just a chip, Richard. Chips are very, very important, I think.

Speaker 5: NVIDIA does amazing job in terms of innovation, breakthrough innovation, in terms of speed, the processing power, but we're also approaching more slow, and there's so much we could do with silicon. We are gonna go to quantum in the next few years. There's a lot of amazing quantum technology. We're gonna be able to do more complex cryptography using those new type of quantum designs, which is coming up. But if we stick to the existing silicon technology, I think there's a lot of stuff on the horizon, neuromorphic technologies, different types of proprietary chipsets that are coming from Alphabet, coming from AWS, different companies.

Speaker 5: So, you know, right now, you have data training, which is still very important. You wanna train the data for foundation models. But even more important is ability to actually inference, ability to ask very complex questions and be able to assist them for processing. And there are some amazing companies now on the horizon, Gruk, Cerebras, and many others across the world. So big organizations are doing it.

Speaker 5: Small organizations are doing it. But more importantly, I and I want the the audience to leave of this, if we don't design the chipsets and data centers with the new paradigm of cognitive as a service, cognition as a service, everything, the whole technology stack has to be revamped. It's not a technology stack of yesterday. It's a stack based on cognition technology, cognition networking, cognition agents, cognition operating system, all based on cognitive capabilities inside the stack.

Speaker 2: So Roger, I am I'm failing at this, but you wanted to get

Speaker 6: I I did I was attempted to get you in as moderator, so Please go ahead. Me speak to something Jensen said that I hope you heard. He talked about, quote, the chip as basically being a multi ton activity across $3,000,000,000 It's not $3,000,000,000 It's not the chip that we think of. And the thing that's important about that, it's called a server, is once you look inside, it's the connectivity across that stack of chips that makes the difference. Here, there's going to be another breakthrough, and it's going be moving from copper, which is how all these chips are now intertwined, to using glass and fiber optics.

Speaker 6: Right? And that is, I think, the next breakthrough. And here, fortunately, we do have and I happen to be on the board of Corning now. We also have companies that are driving that finding. So this whole thing about data centers, it's all about interconnectivity and moving from copper to glass.

Speaker 2: So quickly, one word, what's the bottleneck right now? What will the bottleneck be in five years?

Speaker 6: From my standpoint, the bottleneck I think ultimately is going to be consistent energy nearby, Jason, which means

Speaker 2: to you quickly and then closing comments.

Speaker 4: Sure. I think with chips, chips are just one way of solving the problem. To the point, right, quantum computing is a thing, biological computing, protein based computing. When we have good brain computer interface, right, there'll be shared compute inside of that. So I think that chips for right now are just a moment in time and I think we need to remember that with the exponential growth of this technology, we're going to see new ways of solving this problem

Speaker 2: computationally. What's the bottleneck today? What will the bottleneck be in five years? One word each please.

Speaker 5: Ethics. I

Speaker 2: didn't expect that. Okay. I want to read some of the comments here. Regarding US versus China. What do we think of velocity of this changing and lead in this AI race?

Speaker 2: Another comment, how do you reconcile AI's precision use of data and privacy vis a vis an example like 23andMe that came up throughout the conference? Is capitalism three point o some sort of tech feudalism? It was said that AI drives most of the revenue nowadays. Do you have an idea about that globally? Okay.

Speaker 2: React to any of those and your final remarks in, well, in thirty seconds if you can.

Speaker 6: So let me let me someone raised capitalism three hour. Let me be right

Speaker 2: And your closing and your closing statement, please.

Speaker 6: So I will close with this. One is, certainly, we're the very beginning of an AI revolution. Two, history tells us whenever we have revolutions of this magnitude, we have to rethink how we organize the way we generate wealth, how it's distributed, etcetera, etcetera, which is what's leading me to think about capitalism 3.0 as, I think, an inevitable outcome of this kind of revolution. It is moving much more quickly than we've seen before. It has these international dimensions, which is why you asked about The US versus China.

Speaker 6: It has these fundamental job destruction and job creation dimensions. And so I think all of us have to be thinking about how do we want our economic society to be organized so the great wealth that's about to be created through a supply shock is spread evenly across society to keep all of us feeling as though we are gonna be winners out of this new revolution.

Speaker 2: Thank you for that. Mark, thirty seconds as well as thirty seconds. Hang on as well as reacting to any of the creation,

Speaker 5: I agree. I think this was gonna be the biggest wealth creation in the world. I worry about talent. I worry about countries companies and universities not putting up the energy to make sure that we have the right talent for the future. Without talent, we're doomed.

Speaker 5: Most importantly, right now, I wanna follow Jensen who outlined the strategy and the road map into the future. If you were at Jensen's conference about two months ago, he said we are coming from LLM generative AI to agentic AI right now to agents' workloads, solving complex problem, and we're going to physical spatial world, edge computing, robots, humanoid robots. So we have an exciting future. We haven't even scratched the surface.

Speaker 2: Alright. Jason?

Speaker 4: If the service is free, you're the product. Be careful what you consent to.

Speaker 2: Oh, less than thirty seconds. You're giving time back as they say.

Speaker 3: Well, thank you for that. I would say that this is a moment of leadership. It is more important to know what questions to ask than what answers to give and hopefully if we did our job on this panel, we're asking harder and better questions.

Speaker 2: You have done your job and made this conversation I have like another 20 and I'm sure all of you do as well. Thank you Roger. Thank you, Mark. Thank you, Jason. Thank you, Lorena, and thank all of you.