Download PDF

When controlled by a select few, tech innovation can be self-serving and undermine the institutions that make it possible, says Simon Johnson

Eastern Europe in the late 1980s and ’90s proved an interesting case study for an aspiring economist who had just written his thesis on the hyperinflation and economic chaos in Germany and the Soviet Union in the 1920s.

After completing his PhD at the Massachusetts Institute of Technology (MIT) and starting a postdoctoral position at Harvard, Simon Johnson found himself working with Poland’s first noncommunist government and studying the emergence of the private sector there and in neighboring countries following the fall of the Iron Curtain. Johnson’s astute study of private enterprise’s successes and failures formed the basis for his enduring research on the role of institutions in economic development, which won him the 2024 Nobel Prize in economics.

Johnson has recently turned his attention to how technology is making its mark on today’s economy and the potential impact, of artificial intelligence especially, on the institutions he believes are so crucial for equitable growth. His latest book, with coauthor Daron Acemoglu, Power and Progress, examines the close relationship between technology and prosperity and cautions against allowing too few innovators to control technology’s strategic direction.

Johnson was chief economist at the IMF in 2007–08 and is now the Ronald A. Kurtz Professor of Entrepreneurship at the MIT Sloan School of Management. He spoke with F&D’s Bruce Edwards about technology, inequality, and democracy.

F&D: In Power and Progress, you challenge the assumption that technology always brings progress. Why was this a topic worth exploring?

SJ: Well, this is obviously the age of artificial intelligence, and there are great claims being made for the improvements that will permeate all human societies by making computers and algorithms more potent and able to do more thinking for us. While that could happen, we think, based on our reading of history and economic theory, that it’s not necessarily the case. Improving technology and expanding the capabilities of some people may not necessarily translate into improved living standards for everyone. A lot of Big Tech bosses are more intensely focused on improving the capabilities of people like themselves. These are highly educated people, mostly white, and mostly men. They have a certain view of the world, of what they want technology to do for them, and of where there’s money to be made. And it’s quite natural that they are pulled toward inventing things that favor that vision.

Our book is an attempt to propose some alternative visions. Why don’t we think about other ways to develop and use technology, including AI? Let’s look at what’s happened in the past when we’ve either had technology that’s been more tilted toward raising the productivity of less educated people or more tilted toward boosting the productivity of highly educated people. Because that decides whether there’s a divergence of job market outcomes, with higher-income, higher-educated people doing a lot better, or whether there’s more convergence in outcomes, with people on lower incomes doing better at the same time as the economy overall.

F&D: You warn against the risks of allowing a select few to drive tech. What are the consequences? Is a Big Tech oligarchy a real concern?

SJ: Perhaps not oligarchy in the traditional sense. But in the sense of who controls the vision for what technology can and should be, what we call the “vision oligarchy.” We’re in the middle of an AI boom. When you talk to people about contrasts between, for example, the US and Europe, they say, “Well, the US is inventing all this technology, and a lot of investment, capital, and talent is going there. Europe doesn’t have this.” So AI is driving the conversation, but what is AI? What is being built with AI? That’s a vision. And visions at the leading edge of rapidly changing technologies are incredibly important. I think that terrain should be contested. People should understand the stakes. They should realize that it’s not necessarily a good idea to put all the big decisions in the hands of a few people with their own individualistic perspectives. There’s nothing ad hominem here. We all have our own perspectives, but do we want 1 or 2 or 10 people to drive the discussion, or do we want more engagement and a broader conversation?

F&D: You studied the role of institutions in economic development long before technology. How do institutions play into the evolution of Big Tech?

SJ: First, you need good institutions to be a player. Why is the technology being driven by the US? Because it built really good institutions. Second, institutions shape the way democracy operates and how we should deliberate. But recently digital technology has undermined our ability to have debates. Shouting at each other over social media is not the same as getting together and finding common ground. Digital technology has to some degree begun to undermine institutions.

The big concern if we continue down the road of widening inequality, particularly a version of inequality in which less educated people feel left behind, is that greater anger fuels forms of populism, as we’ve seen in many countries. We didn’t have that in the US in the first two-thirds of the 20th century, primarily because a lot of people’s wages rose and the middle class expanded. Inequality was not the defining characteristic of America’s post–World War II economy. That’s changed since 1980.

Our concern is that AI, which is made possible by our institutions, is pulled in a direction that undermines democracy. That this causes some sort of systemic issue for our institutions or just tilts them toward being relatively or even extremely extractive. A few people get all the value, all the income, all the power, while everyone else is pushed backward in terms of their opportunities, incomes, and how much they can provide for their families.

F&D: And with so few countries having skin in the game, do you worry that AI will increase economic inequalities between countries?

SJ: Yes. Since the advent of industrial technology, it’s been the case that a few places have led the way inventing new machines, and everyone else in the world becomes a taker in that market. A country can go off and invent its own technology. The US did this in the 19th century when it moved from being a country that received technology from Britain to being a country that invented technology. Think of railroads or the telegraph. The US shifted its position; it’s possible.

China has shifted its position too. It was a recipient of Western technology in the 1980s, but now it’s pushing into global markets with sophisticated products, such as consumer electronics, electric vehicles, and of course AI itself. So you can change your place in the global division of labor, but it doesn’t happen very often. Typically you’re receiving technology and adopting it.

This winner-take-all dynamic is even more extreme at present than in previous modern technological revolutions. Now it looks like 95 percent of the money being spent on AI development is in the US, 3 percent in Europe, and 2 percent in the rest of the world. (This calculation does not include China because we don’t know how much it spends on AI.) 

F&D: How do we instill some democracy into our technological evolution to ensure it works for the betterment of society?

SJ: The key points are to recognize the situation and then find alternative paths to push technology in a pro-worker direction. Boosting the productivity of people without a lot of education is key in the US and around the world. The global tech industry, so-called Big Tech, is having a moment of unparalleled power, prestige, and access. Hopefully that comes with a sense of responsibility, a sense of “if you break it, you own it.” But some guardrails around Big Tech’s activities may also be necessary.

There are clear parallels to what we saw with finance in the early 2000s. I had a ringside seat as IMF chief economist in the buildup to the 2008 crisis. A lot of deference was shown to the “smartest guys in the room,” and bad things happened. I want to prevent bad things from happening again. We should persuade people that they must be more careful and be ready with policies and safeguards.

This interview has been edited for length and clarity. Visit IMF podcasts or listen below to hear the full interview.

Opinions expressed in articles and other materials are those of the authors; they do not necessarily reflect IMF policy.