Simon Johnson, the Nobel Prize winner in economics, had a dialogue on how AI affects economies with Jany Hejuan Zhao, the founder and CEO of TMTPost.

AsianFin –  Simon Johnson, the Nobel Prize laureate in economics in 2024, pointed out that artificial intelligence-automated digital advertising make it easier to manipulate people, which is not beneficial for inclusive discourse on social issues. 

Johnson made the comments during his dialogue with Jany Hejuan Zhao, the founder and CEO of TMTPost, at the recently-concluded 2024 T-EDGE Conference.   

Johnson, a professor of economics at the Massachusetts Institute of Technology, believes the application of AI in digital advertising is going in the “wrong direction.” 

“The way digital advertising works is by grabbing your attention, by making you upset, by getting you emotionally engaged. And of course, what it learns from all our behavior, and it is learning that how to make us angry and how to make us not listen to each other, “ said the professor of economics and former IMF chief economist.

“That part of the technology is making it easier to manipulate people. What you want is people to calm down, spend more time talking, meet more in person. People don't shout at each other anywhere near as much in person as they do on the internet,” he said, deploring that the way to harness AI to help people engage in a rational dialogue has not been figured out.

Johnson also said the adoption of AI in low-wage countries may cause a bigger shock to employment, compared with that in high-wage economies.

He argued that Japan needs AI to supplement their abilities to support the senior people given the aging population and a limited number of young people.   

“But if you take it (AI) to a lower wage economy where people need those jobs and where the ability to create new jobs is limited, other countries may struggle,” said Johnson.

As a professor at MIT, he is not against AI or the application of AI. “I think it will come to some extent, driverless vehicles. I'm not opposed to driverless vehicles, per se, in some context. But I think if it comes very quickly, we're all going to have a lot of trouble,” he said. 

The Nobel laureate said many problems in the United States are wrongly attributed to globalization while automation is the primary driver for the widening income gap. “The best approach is to double down on invention and investments in science and technology in the United States and move the frontier in a lot of positive ways, because you create a lot of good new jobs there, and spread those jobs around the country.” 

The following is the transcript of the Q&A between TMTPost founder and CEO Jany Hejuan Zhao and Simon Johnson, the Nobel Prize winning economist of the MIT.

Zhao: Welcome to the T-Edge conference by TMTPost. Following this, I will have the pleasure of engaging in a special dialogue with Professor Johnson. And then congratulations for the Nobel Prize in Economics.  My first question would also be about your studies that were awarded the Nobel Prize in Economics. In your brilliant study of how institutions are formed and affect prosperity, you have demonstrated that inclusive institutions are essential for the long-term economic growth. But what is the new tech law in such an inclusive institutions at different times? And today, how to see the AI technology? And is AI technology more beneficial for promoting the inclusive systems or the opposite?

Johnson: So, Hejuan, I think that is the key question of the day. To be absolutely honest, because it could go either way. So AI could be very good for many people. It could enhance all our capabilities. And that would tend to reduce gaps within countries and across countries.

But the danger is that AI could become a technology that is controlled just by a few people, that benefits just a few people. And in that case, you'll see much more polarization of outcomes. So I'd say there's a fork in the road. And it isn't like today we make the decision, but this year, next year, these are really pivotal years. Because what we develop in terms of the technology, what becomes available, what's offered to people, has some reinforcing feature, particularly because we're talking about machine learning, where the technology is getting good at learning from the humans who use it.  So we could go heavily down the pro-worker path or heavily down the elite path. And we favor the pro-worker path.

Zhao: So in your book, Power and Progress, and you also mentioned in your speech, and also in your speech right now, you have one point that left a deep impression on me. It's that the technology advancement does not inherently bring prosperity. And in fact, it may even exacerbate social inequality. And you just talk about that a lot. So if the control of new technology is concentrated in the hands of a few, the benefits of technology will also be highly concentrated among the elite, while the situation for ordinary workers may become increasingly unstable. So my question about that, do you think we will see similar scenarios in the age of AI? I know that you mentioned a lot in your speech, just about some negative impact. So what measures should be taken to improve the conditions for those ordinary workers? And the next question is related that you also mentioned in your speech that it will threaten to imagined middle class. So not only the ordinary people, but also the middle class. So can you explain more about what will happen to the middle class and what we can do to solve this problem?

Johnson: Yes, again, very, very good and, of course, extremely difficult question. So Hejuan, I always look at economic history to find some parallels or some experiences that we can draw on. And to me, the early British Industrial Revolution is very relevant. So I'm from the north of England. I'm from the town of Sheffield. And Sheffield and Manchester and other cities were at the forefront of the adoption of new technology from the 1780s, particularly the textile towns, which are a little bit further west than where I grew up. And what we know is that from the 1780s to the 1830s or 1840s, the new textile factories were built with incredible levels of productivity and machines that could produce -- that could first spin and then weave cotton in a way that had never been done before. And their productivity was 1,000, 10,000, 100,000 times higher than the most skilled artisans, for example, in the UK or in India. So what happened was that the previous, let's say, middle class or people with those strong artisanal skills, they got wiped out economically. They got pushed down to be factory workers. But factory workers, you didn't need a lot of them, because it was mostly about automation. So the demand for labor was not that strong. A few people, people who owned the factories, did very well. But most people did not benefit until the 1830s or 1840s.

So what was the change? What made the difference? Well, the answer is broader industrialization, including the creation of the railways, which was not like a magic single innovation. But the railways unlocked a lot of other potential. A lot of other industries began to modernize and make investments. And the railways reduced transportation costs and created a huge number of new tasks. And when you were running the railway, of course, the enine driver didn't have a lot of formal education. But you wanted to pay him good money, because he's taking care of this very expensive engine. You want the man who is in charge of the signaling system to really pay attention the whole time, because, again, terrible things happen if you mess that up, and so on.  So this very productive new enterprise and all these enterprises required responsibility. They required expertise. And they paid people for that expertise. And from that time, real wages in the UK take off.

That's the technology that spreads around the world, where sometimes it's used very productively in some context. And I think, Hejuan, that's what we need to do again, which is find new ways to harness technology, creating new tasks for people, including things that humans have never done before. But they need to be productive things.

They need to be things that are compensated by the market. That's how most of the world works these days. And I think pushing the technology to move quicker in that direction globally is really the key to harnessing this technology and making sure it helps everyone and not just a few people.

Zhao: So what should the government do? Because there are different countries that have the different policies. So what should those different governments can do to-- you know that you mentioned a lot about that also in your study. The key point is the inclusive institution. So if it can use AI or to make AI to help us to establish a more powerful, inclusive institution, if we can do that. So I think-- is it a good idea?

 Johnson: Yeah, it's a great idea. Believe me, that is one of the best ideas ever.

The problem is, of course, that the-- so again, I'm just speaking about the United States, which is what we follow most closely. The problem is that the AI has an immediate application and immediate value in digital advertising. But the way digital advertising works is by grabbing your attention, by making you upset, by getting you emotionally engaged.

And of course, what it learns from all our behavior, and it is learning that how to make us angry and how to make us not listen to each other, honestly. So that's going in the wrong direction. I think it makes it currently that that part of the technology is making it easier to manipulate people. What you want is people to calm down, spend more time talking, meet more in person.

People don't shout at each other anywhere near as much in person as they do on the internet. But unfortunately, we haven't yet figured out how to harness AI to do that.

There is some good work along these lines. My colleague, Deb Roy at MIT, for example, has a very interesting lab that focuses on using AI to facilitate conversations and support conversations in the non-profit sector, for example, in the U.S. I think that's great work and very important. 

But the industry is not focused on that. It's racing ahead with the digital advertising, which is, I think, not having good effects at all on inclusivity.

Zhao: So I have an idea that it is possible to have the different solutions for the developing countries' government and the developed countries' government. So in China-- China is a developing country. And China also has some companies. They are doing their driverless taxis business. But many people in China, they will be worried about that. The drivers will lose their jobs to AI. So generally speaking, what do you think of the impact of AI on employment? Also in the different areas, different countries, they will have the different situation. So some people will think that in China, such a developing country, we should not use AI so quickly. But in the developed country like the U.S., they can do it. So do you think it's different? 

Johnson: Yeah, this is a very important point. So for example, let's talk about Japan.

Because Japan is interesting as a high-wage country and has a very rapidly aging population, as you know. So in Japan, the perspective on these issues is a little bit different. They say, well, look, we don't have that many young people. We need AI to supplement their abilities in order to support the older people. So that's the Japanese labor market and the Japanese perspective.

In the United States, we have high wages. Population is aging, although not quite as fast as in Japan. And the perspective of the people inventing the technology-- let's call them Silicon Valley inventors, although obviously they're not just in Silicon Valley, literally. But they're mostly on the West Coast of the United States. Their perspective is, let's invent it and see what happens.

And if you take the same technology to a lower income country, so where the wages are lower-- and we know this technology spreads very fast. I mean, ChatGPT was being used within 24 hours, 48 hours of its introduction. Every corner of the world that I've ever encountered, they were already playing with it. So if you take this technology that's built for a high wage economy, it could be useful for Japan, another high wage economy with the aging population, maybe more useful for them.

But if you take it to a lower wage economy where people need those jobs and where the ability to create new jobs at this very intense speed with new tasks is limited-- I mean, I don't know if we can do it fast enough in the United States. So other countries may well struggle.

And if you could suddenly make all taxis driverless-- and if customers wanted that, of course, because customers will choose-- that would be a big shock to labor markets and to employment across a very wide range of countries. Not everyone could afford to adopt driverless taxis. Of course, there's some capital investment.

 But if you have a country like China, which is very dynamic and has plenty of capital, you could see a very big shock there. And yeah, I would worry a lot about whether any country-- not specific to China-- but whether any country could manage all the consequences of that being very quick.

I think it will come to some extent, driverless vehicles. I'm not opposed to driverless vehicles, per se, in some context. But I think if it comes very quickly, we're all going to have a lot of trouble.

Zhao: So it will make another problem that may think about. We mentioned or we talked a lot about the technology may widen the gap between people. But we seldom to talk about AI technology. Also, we are also widening the gap between countries and the regions. So what can be done to make AI benefit most nations and the regions, not only some developed countries or the powerful countries who can control the AI technology?

Johnson: Well, I think that's a very important question. What I would suggest is that countries get together and decide the problems they want to solve with AI and then present themselves to the global market and say, look, we're a billion people in this part of the world or with these issues. And if you build a technology that does this, this, and this for agriculture or for transportation or whatever, then we will buy it.

So if you could figure out how to organize this sort of pre-purchase agreement where you specify what you want, that will get the attention of the inventors. Right now, the inventors are very focused on high-income countries. That's where the money is. That's where the market is. And they're very focused on digital advertising in the United States, for example, because that's very easy, quick money. And you can see how AI can be used. 

So what you really want to do, Hejuan, is change the vision. Provide a new vision and say, look, this is the problem we want to solve. This is what we think AI can do. Bring it to us. And if you really do solve the problem, we will be a very willing and large market.  

That conversation does not currently exist. It requires, of course, communication and cooperation across governments. But it's a very easy, I think, form of cooperation because all you're doing is agreeing on what it is you would like to buy and specifying what that market can look like. So it's a win-win type of conversation and personally, I try to encourage it every opportunity. 

Zhao: Yes, I hope that we can have a win-win future, but you know that in reality, is that it's another question also related about it. Is that I want to ask about the competition. Yes, you mentioned the competition, the cooperation among nations and the regions between different giants, like between China and the U.S. So does an inclusive economic institution tend to favor globalization and oppose trade protectionism? So how to find the balance between national prosperity and the common prosperity of different countries along the world? So I think it's a challenge for all the governments in different countries. To get them together is not so easy.

Johnson: Well, that's obviously the question that is gonna be before us all as the Trump administration rolls out its policies. Look, I think the experience of the last 40 or 50 years is clearly that trade can be a win-win. It can benefit lower income countries. It can allow them to rise.

But honestly, in the United States, we've not done a good job of helping people

move to new opportunities, to higher productivity opportunities. We don't have a strong social safety net, as you know, in the U.S.  So the effects of this globalization is, I think, not the primary driver of polarization in the U.S. I think the primary driver is automation. But the globalization has also put pressure on the middle in the United States, reinforcing that picture I showed at the beginning of the dispersed paths of earnings growth since the 1960s. And I'm afraid there is a bit of a backlash, and that hopefully will not become too destructive, because it could be destructive to the U.S. as well as to the world, right?

So we're in a difficult phase. And I think a lot of problems that have come from automation are being attributed to globalization in an inappropriate manner. The best approach, the approach that I prefer is to double down on invention and investments in science and technology in the United States and move the frontier in a lot of positive ways, because you create a lot of good new jobs there, and spread those jobs around the country.

So that's an alternate strategy that has attracted some attention. The 2022 Chips and Science Act reflected some of our ideas and proposals, but that has not become the mainstream dominant narrative. 

The dominant narrative at the moment is, I think very unfortunately, Mr. Trump's narrative, in which he thinks that the U.S. is somehow losing and everyone else is cheating, and he can change the playing field by unilaterally exercising American power. I think he's about to encounter some realities that are rather different from that. We'll see.

And I fear for the consequences for the United States, definitely. But mostly for the world, because I think disrupting trade in some abrupt manner will hurt a lot of people.

Zhao: So do you believe that Donald Trump will share the AI technologies to other countries,or Donald Trump will have willing to cooperate with other countries to develop AI technology together?

Johnson: Well, AI technology development is mostly in the private sector. It's not in the hands of government. And as long as the private sector complies with all the rules and regulations, they can do what they want. This private sector is very powerful, as you've noticed, and that power includes a very strong representation of their views on Capitol Hill in Congress. So there are things that Mr. Trump can do on trade. He can also specify export control restrictions, and I think you should expect that on high-end computer chips, for example.

So the U.S. will hold the hardware very tightly in the United States for national security reasons primarily, but also for some commercial reasons. But the US is very open to investment in the United States. They're welcoming companies to come into the US. And I think with regard to the software and the building the applications, there will not be any restriction. There's certainly nothing on the horizon. The key issue there is that the private sector builds for the market that it sees today and the money that it sees, follow the money, always a good line in the United States, and that money is in the hands of rich consumers.

This may be, the technology we develop may be entirely suitable and helpful to a place like Japan. I don't think it'll be so helpful and appropriate for a place like Indonesia.

Zhao:  Because of the time limit, I have the last question. The last question is, we talked about many technology around AI, so in addition to AI technology, there's another technology program, it's also very important, that has been widely discussed recently, which is blockchain and cryptocurrency. You know that recently, Trump has been advocating for the establishment of a Bitcoin national reserve. So is it possible for him to succeed in this endeavor? If a national Bitcoin reserve were to be established? And also, I think the further question about the, if it can be succeeded to establish it, what impact would it have on the current gold reserve system, national financial systems, and the global financial system, yeah.

Johnson: Yeah, that's a great question. I wish we had one or two more hours to discuss it. It's something I've worked on for a long time, ever since I was the chief economist at the IMF, but obviously cryptocurrency took off shortly after that, in the early 2010s.

Look, I think, first of all, it's a very strange policy to pursue, because the position of the U.S. economy in the world is very much linked to the dominant role of the U.S. dollar. So why would any American government want to promote an alternative currency, including a currency that's not controlled by any country?

This flies in the face of all economic and geopolitical historical reasoning doesn't mean it won't happen. And the cryptocurrency industry, as I'm sure you've also noted, is making a very big push in Washington. When FTX failed, there was a realization that the industry is not run on very impressive lines with regard to governance. And it went quiet for a while, but they were very active in this election cycle. And they're offering a lot of money to politicians for the next election cycle, which is in, just remember that's in two years, for the entire House of Representatives.

So we're gonna see a lot of cryptocurrency deregulation. And I think that will be wild.

And I advise everyone to be extremely careful because governance matters and weak governance always leads to trouble. You can have a boom, but then you have a very nasty bust.

So be careful, everyone, please. Now on Bitcoin per se, the government does own and does acquire a quarter of Bitcoin, of course, through its law enforcement activities. Bitcoin is used for illicit purposes in some instances, the federal government seizes it. They have some Bitcoin.

To create a strategic reserve is, I think, not a sensible policy. Could it happen? Sure, absolutely. I think this sort of thing where Mr. Trump with his rather unorthodox approach to policy could take the plunge.

How much effect would it have globally? I think that remains to be seen. I think that the bigger issue is the deregulation wave for cryptocurrency and a lot of the mad and inappropriate schemes that we're going to see. And I just remind everyone what happened in the 1920s, which is the U.S. had what seemed like a terrific boom in financial markets. We had a transformation of technology in the real economy that was real in the automobile sector, for example, and other parts of the economy. But the financial froth landed very badly  with the crash of 1929. And a lot of the rules that we put in place in the 1930s were to prevent that from happening again.

Now, those rules have lasted almost 100 years. My good friend and former colleague, Gary Gensler, was chair of the SEC for the past four or so years. And he has repeatedly articulated that strong investor protection is an important part of American national competitiveness and frankly, our national security.

Unfortunately, Mr. Gensler has been forced out on day one, I think by Mr. Trump and whoever replaces Gary Gensler will be favoring the views of the 1920s. The 1920s ended badly. Financial froth can really damage the real economy. Financial excess and crazy deregulation is bad for innovation, terrible for innovation.

Look what happened in 2008. I mean, again, we need another hour to go through the details, but that is not what we should be looking for in the United States or globally. But we may well be heading into another crazy credit-driven, deregulation-driven cycle.

So what I say to everybody is be very, very careful, buckle your seatbelts and focus on what's really important, which is people's lives and how we can harness and direct technology to improve the lives of more people around the world while also obviously maintaining and sustaining our planet.

Zhao: Yeah, so it seems that not so easy to do such things. And Trump said that after he has the opportunity to be the president of US, he will file General Gensler, Mr. Gensler, the first. So I think he's crazy, right? 

Johnson: Oh yeah, Mr. Trump has made it very clear that he favors a lot of financial deregulation, both for existing banks, for let's say the hedge fund non-banking sector, but also for cryptocurrency. And the history of deregulation is that you get a sugar high that everyone says, "Oh, wow, this is great. We can do all these fantastic new things that we weren't allowed to do before." And then it blows up in your face. So, and the U.S., look, if we do it to ourselves, that's our problem.

And I feel bad about that for Americans, but unfortunately what we do in the U.S., including with regard to financial markets has big global implications, as we all were reminded in 2007, 2008. So that I think is unfortunately very important context for anyone listening and anybody interested in real innovation, really changing people's lives, you've got to be super careful about madness emerging in the financial sector.

Zhao: And at the end of our conversation, can you also give us some, or give China some suggestions about the national strategy about AI and cryptocurrency like that, or the tech strategy to the Chinese government and the Chinese entrepreneurs.

Johnson:  So I used to be the chief economist of the IMF, which is all about giving advice to other countries. I try not to do that anymore, I think. But what I would say is to everybody, to all countries, don't focus on what we can invent, focus on the problems you want to solve.

What is the problem that you have in your society?

Who needs help?

What is it you want to advance?

And then come to the question about how technology can be used and harnessed to do that.

So I think all countries like the United States should approach technology differently and should think about your own specific issues, which might be about particular people in the labor market? It might be about aging across the entire population. It might be about poverty, you can decide for yourselves. But once you've articulated and understood and focused on that problem, then look for the technology that will help. And if that technology is available off the shelf from other countries, great, you can do a deal to bring it in. But a lot of times that technology will not be available because the people who currently dominate the commanding heights of technology innovation have different priorities, they live in a different country, they don't see your problems like you do. So in that case, you should develop your own cadre of people who are very deeply embedded in your own context and your own problems and empower them and enable them and connect them to the extent possible with experts around the world. And I think it might sound obvious, it's not generally what we do.

So my suggestion to everyone, I mean, not only to China, but to all countries, is to focus on the problems that you want to solve. Because the people who currently drive the technology

live in the United States, they see the world through a particular lens. And I'm not saying that's a bad lens or the wrong lens, but it's their lens. You have your own problems, you have your own issues, you need to decide for yourselves. What it is that you want to do? Do you want to focus on poverty? Do you want to focus on global health? Do you want to boost the middle-class? Do you want to develop certain industries? And I think when you have that articulation of the vision, you can then ask the question, where can we get the technology to pursue that vision? And if that technology is available from other people already, great, you can buy it or bring it in, in whatever form works for you.

But if it's not available, then you should develop your own cadre of people who can focus on local solutions using globally empowered technology.

But I think putting the technology first and then saying, well, look, we can invent this, what should we do with it? That's putting the horse before the cart. What you want to do is say, What is the problem to be solved? And how can we deploy and develop technology to solve that problem? So that is my recommendation to absolutely everyone across the entire technology space around the world.

Zhao: Thanks so much for your insightful opinions and the speech.

Johnson: Well, thank you for the opportunity. Thanks for the great questions and for the engagement. It's very encouraging to have all of these conversations around the world and to know that people like yourself and your colleagues and everyone attending this conference

are so interested in exactly the same questions everybody is grappling with around the world.

This is a very important fact and I think we should make the most of it.