Transcription
Spoke to at the UBS event in China just a few months ago. A great conversation about AI, and, obviously, you have a lot of thoughts on it as well. Is AI gonna be the great job destroyer that many people in the banking industry are talking about, hedging their comments, and maybe not communicating as well to the rank and file?
So, Jerome, and I work on this a lot. We have a research center together at MIT. We've written a book, Power and Progress, on this issue. And I think it's there's a danger that that AI in its automating form, in the form that displaces workers, will predominate, whereas what we're arguing for is using AI to to strengthen human expertise, to increase the demand for human expertise, and to allow you to expand into new areas because that will actually strengthen employment.
You said many good jobs, though, are gonna be lost. You say it's looking increasingly dark. How do we stay away from that worst case scenario when, I guess, the complaint is that the larger companies are gonna be able to get richer because of it and the the worker's gonna be left behind.
Well, this is this is exactly where we are right now, which is the automation part of AI is is moving ahead pretty much at full steam. And the new job creation part, the part that creates new tasks, the part that expands what your company can do is being somewhat neglected. Now we think that's mostly about corporate executive decisions. It's easy to replace people. It requires more work to think about the the new frontiers, but you need to think about the new frontiers. And there's tremendous value to be had there. There's a lot of profit to be had if you can bring the latest ideas out of universities, commercialize them, turn that into manufacturing. And Hong Kong and China have got a great opportunity to do this, but so does The US.
Yeah. I was strumming through some of the book in 2023 that you talked about, power and progress. Right? Yeah. Over a thousand year struggle over technology and prosperity. You argued that tech does not automatically create good. Right? It's it benefits the benefits go to a narrow elite. Right? China seems to have a different approach to AI as well. They're worried about that kind of, progress going to the corporate elite. So they would like to see a more, disbursement, if you will, especially in younger unemployed people who can kinda do one person companies through agentic AI. Do you see a completely different bifurcated AI development and philosophical approach to AI between The US and China?
No. Actually, I think there's convergence. Really? Maybe the rhetoric is different. The US doesn't quite talk about as much about the the opportunity of young people yet, but it's it's coming. And I I think both teams, if we we can call them that, are pursuing the technology, the frontier. They're realizing the automation is is real and and and dramatic, but what are the people what where are the good jobs gonna be? So I think both US and China have got their own political grappling with the same issue.
Have you come up with that calculus? Where are the good jobs gonna be? Science. New innovations. But not necessarily coding and software. Right. So, obviously, some of the tasks within existing businesses and also new businesses are being taken over quite effectively by AI, and coding is at the forefront of that. But what you do with coding, how you use computers, there's still a huge amount of work to to to be done there. And and there's great value if you're a computer scientist, but not if you're, let's say, a relatively lower level programmer.
So, you know, with this agentic AI, how transformative could this potentially be in China, which doesn't have access to the most powerful NVIDIA accelerators? They have lots of young engineering talent. This seems to be giving a a new platform for people to grow their businesses on. Have you explored that in-depth in how China is approaching this?
Well, absolutely. This this is the the latest topic, and and every the researchers like myself are running pretty hard to keep up with the realities. I I think that, with safeguards and and and and with appropriate management, agentic AI can be immensely powerful. And I think China's ability to build really meaningful technology with, let's say, the less than fully powerful chips, that's actually gonna matter a lot for the world. Because when it comes to sharing the technology and disseminating it, there's a lot of other countries that can't afford NVIDIA chips or can't get access to them. So China's got an edge a little bit on the export side there perhaps.
I don't know if you saw the story. We had a Bloomberg scoop today saying that now China's basically expanded their travel curbs to private companies. I mean, we've always known that state owned enterprises in key areas of national security interest, maybe the top level scientists or the top executives have their passports held by maybe the the the party apparatchik in that company. But now we're hearing that private companies, what is are having the same kind of travel restrictions. What does that tell you about how strategically important this is to the Chinese?
Well, I think it tells you that everywhere this has become a matter of national security, including where is the talent, can the talent move, is the talent sharing its ideas in the ways that the authorities feel are inappropriate. I do think a big advantage of The US model and what we're building in The UK, for example, is we're very open. We allow people to come. We allow people to go. We're we're totally comfortable with ideas being shared, and and that's what I would recommend for for everyone. But, clearly, the the stakes are the stakes are high.
You've been appointed recently as an a as an AI ambassador in The UK. What what are the how do you see the priorities in The UK on how to harness and maybe put the guardrails in AI as well, but also not necessarily see that job destruction that there's a lot of concern that is gonna happen in The US. Right.
So the the the The UK is attempting to position itself between The US, which is sort of anything goes model, and the Europeans, which is like, the answer is no. Now what's the question? Yeah. Yeah. Yeah. So The UK is is got a is is recognizing that it's not at the frontier of Silicon Valley, and it's not big enough to be like China. So where does it play in that global space? Where where can they create value? Where can they be useful? What's the public private partnership that will generate jobs? They're asking all the right questions. Now these are not easy questions. Let me be clear. But I I welcome any anyone and everyone to come visit us in The UK and look at what we're doing, kick the tires, because we wanna bring in private investment as as a catalyzed by the right public policy decisions.
Where are the areas that we're not necessarily looking at that you are looking at the kind of significant social and economic impact that AI will have? You told me you're writing a new book, who rules the world. Right? We've seen AI as well be implemented on the battlefield because we happen to have two simultaneous wars going on right now, and the rules of war are changing because of AI as well.
Yeah. So drones have changed the rules of war, and now we have drones that are powered by AI. So it's really quite a different future, and these drones are cheap. So they're easy for relatively smaller countries, less well off countries to afford. So Ukraine and Iran are both using drones. Russia is, of course, producing a lot of drones that it's probably at this point supplying to to Iran and feeding into that conflict. These these are these are massive changes. And I think the the key point in who rules the world is innovation. If you are innovative, if you can use the innovation that's available off the shelf, move the frontier yourself, apply it to your own problems, which include national defense, you are gonna come out relatively well compared to other countries. Now I don't want there to be more war. I think we're better off when there's less war. But then innovation is still gonna be critical in terms of can you create prosperity? Can you share prosperity in your country at different income levels?
So how is AI changing power dynamics? As the the great, you know, g two as people call between China and The United States? Right.
So, obviously, we have the g two two countries pushing the frontier very hard. Then we have another a set of countries that have scientific capacity and the ability to innovate and and invent new things. That would be in The UK, that would be in Europe, that would be in Russia, that would be in in some other parts of the world. They are figuring out or or need to get their skates on and and speed up figuring out where they position themselves relative to g two. But then there's a lot of other countries that don't have that kind innate technological capacity. They have a problem. They have to face up to the fact that many people in their economies will be adopting this technology that will displace workers where are the new jobs gonna come from in those countries. That's a really hard question.