Thank you, sir. All right. So we are going to start with your amazing work—the work that won you the Nobel. So most people in the room are familiar with Schumpeter's work back in the '40s—the concept of creative destruction. I think what's interesting is his words are described as poetic. Some have called it like a tragic drama. And now, you've taken this concept to an extraordinary level, formalized it, put rigorous quantitative analysis behind it. Can you explain for all of us here your journey? What did you do? What is the breakthrough? Why were you awarded the Nobel?
Thanks. So I will say there is a lot of luck in all this and failing. So in a nutshell, in my students' days, the leading growth model was the so-called Neoclassical Growth Model, Robert Solow. And it was a model based on capital accumulation. You produce flow of wealth with a stock of capital, and you save part of the flow to increase the stock of capital, which allows you to produce more. And that was the leading growth model. But Robert Solow knew himself, was aware of the fact that you cannot go on forever just based on capital accumulation, because there are decreasing returns to capital accumulation. You run out of steam at some point. You need something else, and the something else is technical progress. But Robert Solow could not explain where technical progress was coming from. And we talk about the Solow residual, which was everything you cannot explain. Was this in black box? And I wanted to know what was inside the black box. And there was Schumpeter, of course, that talked about creative destruction, the notion whereby new innovation displace old technologies. So that's the concept that he introduced in his writings. But in my students' days, Schumpeter was a modernity. He was a curiosity. He was not mainstream at all. You would hear about creative destruction in a course on history of economic thought, or you would hear about creative destruction on the [? IU ?] class. But [FRENCH], there was no model. There was no empirics. So with my co-author Peter Howitt, who won the prize with me, we decided to write a new growth model which would encompass creative destruction, but would be a model that you can solve and you can confront with data, which is a way to put industrial organization into macro growth model. So it was really the beginning of putting micro into macro. And the model would revolve around three main ideas. First, long-run growth is driven by cumulative innovation. You always stand on giant shoulders. You build upon each other. The second idea is that innovations do not come from heaven. They result from entrepreneurial activities that are motivated by the prospect of innovation. When you innovate, you get around for a while until you are being superseded by better technologies. And the third idea is creative destruction. New innovations displace old technologies, and at the heart of this theory, there is the firm. There is enterprises. So it's very different from Solow. And there is a contradiction. On the one hand, you need innovation rents to generate innovation. But on the other hand, yesterday's innovators are tempted to use their rents to prevent subsequent innovations because they don't want themselves to be subject to creative destruction. And managing a market economy is how you manage this contradiction. And you always have this phenomenon that yesterday's innovators, they grow, and you want them to grow. But eventually, they will use their rents to prevent newcomers. And that's where competition policy is important. And you get into this and that. And so fast-growing economy, modern economy is one where all the time you have new talents coming in. You let them grow. You let them create firms. But you want to make sure that once they've grown, they won't prevent new talents from coming in. And that's the difficulty. That's the challenge. A bit long, I'm sorry.
I will be short, I promise.
We're going to come back to this, but I wanted to set the table with the Nobel and the content around it. Now, just a quick one on you. So amazing background—I found one of the best ways to understand someone is to go back to lessons in their history, where they grew up, some of the stories from childhood. And your childhood is absolutely fascinating—the son of Gaby Aghion, the visionary founder of Chloé, who completely revolutionized fashion, going from haute couture and how we were all shackled as women to pret-a-porter empowering women. You spent your childhood surrounded by amazing artists like Karl Lagerfeld. Your father was a prominent intellectual and an art gallery owner. And from those early days, you became a Nobel Prize winner in economics. Weave that together. What happened in your childhood that explains who you are?
That's right. Well, no, it's true. So one thing is that Karl Lagerfeld would do my German homework, and that's why I didn't learn any German. Yeah. No, I mean, it's true to be with artists, with innovators, there is something there. My father wanted to do social revolution in Egypt. He failed. My mother wanted to do revolution in fashion. She succeeded. And it's true. It's a mixture what got me into economics. It's true to be surrounded by inventors. I'm sure that that must have. There is something there. I think there was the thing that I was very politically motivated because I wanted to make the world a better place. I was a post-'68 French, good left wing. I wanted to make the world a better place. And I was not too bad in math. And the combination of not being too bad in math and wanted to improve, to make the world a better place, that combination got me into economics. So the father wanted to show he failed. Maybe that's where he failed. Maybe I will succeed a bit more. And I think it's that mixture that, well—
All right. So going back to the problem that you solved. So you've said already economies eventually stall when they only focus on perfecting old technologies. And you said they reach this point of diminishing returns. You have a very intriguing concept, I think, of resetting the innovation clock. And can you explain to us what does that mean—this resetting the clock to 0 and then link it, if you would, to how it relates to AI.
Yeah, so first there is a big debate between Robert Gordon, who is a fantastic colleague at Northwestern, and he believes that great ideas have been found already, that it's like a fruit tree. We got the good fruits. The steam engine revolution was a big deal. The electricity revolution was still a big deal. IT, he believes, was less of a big deal. AI, I think, now he thinks there, no. But it's maybe he's resetting his own clock, I think. But that's Robert Gordon, who is a fantastic colleague, and Joel Mokyr is more like an optimist. And there is still an ongoing debate. Are ideas harder to find? Did we get the big ideas, and now we're running out of steam? Or is there a reset? And my view is that within each field, you end up running out into decreasing returns. But you keep finding new topics, new ways. We saw that with the two previous sessions, that you always have a reset. So for example, I look into economics. I use LLM with my friends to identify topics using word recognition and abstracts. And you see within topics. It's kind of inverted U. You have a rising phase and then a declining phase. But you see there all the time new topics emerging. So if you take the upper envelope, it's clearly going up even before AI or with AI even more because AI, you see, new ideas are often recombination of old ideas. With AI, you can recombine much more, and you can select ideas you want to experiment much more quickly. And that's hence the huge growth potential of AI. So less than ever, even before AI, ideas were not harder to find overall, even less with AI.
Now, you've also described Europe as a Sleeping Beauty, a place with immense talent—
Yeah, Sleeping Beauty place. Yeah.
Sleeping Beauty place, immense talent resting on the laurels of incremental, safe improvements. We've talked about Mario Draghi's report and the warning that he's had that if Europe doesn't actually clear away the shackles it needs, regulation, build capital markets, we're not going to get the benefit of growth from technology. And Draghi now warns that this stagnation has reached a point where it is existential. So if AI is the kiss that finally wakes that Sleeping Beauty, how do we ensure the destruction of old jobs leads to a golden era?
So there is a huge thing. So first Europe, and when I say Europe, it has to include the UK. And I think now that's the great thing, is that the new world, that the way that people talked about this morning, Trump, Putin, the nice part of it is that the UK is back for us, Europe, and UK is back with us. And that, I think, is very important.
I was born in England. I like that.
I was born here. So I'm with you.
Yeah. And so the idea is that we have, in Europe, we have fantastic scientists, engineers, mathematicians. And people were pioneers in AI, like Yann LeCun, who was with us in the AI Commission that was mentioned before. Yann LeCun wants to do the next AI revolution after LLM, and he wants to do it in Europe. So Europe has enormous potential. And we carry values. We talk about the soft power for rule of law, freedom more than ever. Where do we have freedom now in the world? In Europe, more than anywhere else. And so we have fantastic potential. And it's incumbent on us to get the right institution to harness AI. The growth potential—AI has enormous growth potential for the reason I mentioned, mainly easier to get new ideas. You need to get a mixture of competition policy and industrial policy. In Europe, in the name of competition policy, we precluded industrial policy, unlike in the US, unlike in China. We need to have full steam industrial policy, the equivalent of the DARPA in defense and AI. That's on the one hand. And on jobs, we are closer to the good flexicurity system. We talk about—
Explain that, if you will. Flexicurity—
Flexicurity system—in Denmark, when you lose your job, for two years, you get 90% of your salary up to a certain level of salary, and the state retrains you and helps you find a new job. And we know that in Denmark, losing your job has no negative effects on health, unlike in other countries. So there will be job destructions with AI, of course. AI will replace tasks. And for some jobs, most of the tasks are replaced by AI, and those jobs are in jeopardy. So it will destroy jobs, but it will create new jobs, AI. Why? Because first, firms that adopt AI become more productive, therefore more competitive, and therefore the world demand for their products increases. And that's what we see, that sales go up. Firms that adopt AI, including generative AI, sales go up because they have more demand. So that productivity effect will be good for employment. And the second thing is that, as I told you, it's much easier to have new ideas with AI, new ideas or new activities and therefore new jobs. So AI will destroy jobs, but it will create jobs. So you need a system that smooths out the transition from old jobs to new jobs. That's where the Danish system must be part of the response.
I want to stay on this topic because there was so much content in what you just said. If I try and break what you said into two buckets, one is jobs will be destroyed, but jobs will be created, new ideas, et cetera. Help to mention that. My concern, we've talked about it, is that a lot of the narrative coming out of Silicon Valley is not tethered to history and is not tethered to economics. We now have somebody who can actually help us understand that a bit better. And then, no matter what that ratio is, there's going to be a transition. Take us through the Danish model and the policy. So let's first start with jobs. What's your view on this whole question of creation, augmentation tasks, and destruction.
Yeah, so there will be some jobs. That is a collection of tasks. Some jobs, most of the tasks will be replaced by AI. And so those jobs will have to be transformed. But there will be firms that adopt AI. They create jobs because they expand. And there will be new ideas, and new ideas are new jobs. So there will be some job destruction, but there will be for sure job creation. For example, what will change with the jobs? Soft skills, for example. Jobs that enhance the ability of leadership, of coordinating with others, of taking initiative, that's the soft skills. Soft skills will be a very important element where humans are important there. You see, there is also in having to do with health care and care—nurses, psychiatrists, geriatrics. There will need for humans. It cannot be entirely done by machines. The human element is very important there. Plumbers, electricians, people who help maintain the robotics, maintenance—all these jobs also, more than ever, will need a human component—so apprenticeship. So I think it's a set of policies that have been put in place. My view of technology is that technological revolutions are like a horse. A horse can take you to the wall, but it can lead you everywhere you want. If you know how to steer the horse, steering the horse is to put the right policies. So I talk about flexicurity. I also say encourage firms that enhance soft skills of their workers. Push apprenticeships and change the schooling system. At school, we should learn to learn. So there should be part of the school with AI, but part of the school without AI where you put emphasis on reading. You should calculate. You should master reading and mathematics, the basic knowledge because at school you learn to learn. So if you have a good schooling system with good test scores, it's much easier to put a flexicurity system under. So you see, the thing is that the AI has enormous potential. But where I'm less optimistic is that shall we be able to put the right policies and institutions in place to fully harness the growth and employment creation potential of AI.
