Andrew McAfee on AI, jobs, and ‘permissionless innovation’

In this episode of the Century of Plenty podcast, Chris Bradley and Marc Canal speak with MIT’s Andrew McAfee about technological progress, AI, inequality, and the prospects for continued growth. McAfee reflects on AI’s accelerated progress, makes the case for “permissionless innovation,” and explains why he believes rising prosperity can be decoupled from ever-growing resource use.

In this episode, they discuss:

• Why exponential technological progress is so difficult to predict—and why McAfee believes efforts to steer innovations risk constraining their benefits.
• What the past 15 years have taught us about automation, jobs, and inequality, including why McAfee has become less concerned about mass technological unemployment.
• How technology, markets, public awareness, and effective government action can enable continued economic growth while reducing resource intensity and environmental impact.

The following transcript has been edited for clarity and length.

How far have we come?

Chris Bradley: Hello everyone and welcome to the Century of Plenty podcast. I’m your host today, Chris Bradley, one of the directors of the McKinsey Global Institute. I’m joined by my coauthor, Marc Canal, and a very special guest today, Andrew McAfee. Andrew’s a principal research scientist at the MIT Sloan School of Management, and he’s the cofounder and codirector of the MIT Initiative on the Digital Economy.

He’s widely recognized as a leading thinker, for many years, on how digital technologies and innovation will reshape business, work, and society. I first came across Andrew, it must have been 15 years ago, when I read his book Race Against the Machine. We’re going to come back to that.

So welcome, Andrew. Why don’t we start the way we start our book, which is reminiscing about what progress looked like over the last century. The way we did that in our book was by asking, “What was life like for our grandparents?” You might choose a grandparent, a grandmother, or even a parent—but talk about where you came from and your roots, and maybe relate that to what the world was like back then and how much the world has changed. So what’s your origin story, Andrew?

Andrew McAfee: I did a 23andMe a few years back, and they basically said, “You are the most boring genome that we’ve ever seen. You’re just like a Northern European guy.” My people tend to be Welsh, Irish, English, and German. So my grandparents were all born in the United States. I think if you go one generation back, there were some great-grandparents who were overseas.

In 1910, when my grandparents were born, the child mortality rate in the United States—which was already an industrializing country and maybe the wealthiest country in the world—was about 10 percent. One in ten children died before their first birthday. About one in eight died before they were five. The maternal mortality rate was a little below 1 percent. Around 1 percent of women died in childbirth. These are unimaginably dire statistics, but that was the state of the world in America 100 years ago.

I could keep tossing out examples like that. I don’t want to say we lived in the Dark Ages when my grandparents were born, but this nostalgia for a simpler, easier time—Holy Toledo, I do not share that. When I look at the evidence, I think our lives were much riskier, much shorter, and in a lot of ways demonstrably less pleasant.

Marc Canal: You mentioned that it’s hard to predict how technology will evolve. I wanted to bring up a small paragraph that you wrote. It’s from The Second Machine Age, published in 2015:

“Maybe we’ll see a program that can scan the business landscape, spot an opportunity, and write up a business plan so good that it’ll have venture capitalists ready to invest. Maybe we’ll see a computer that can write a thoughtful and insightful report on a complicated topic. Maybe we’ll see an automatic medical diagnostician with all the different kinds of knowledge and awareness of a human doctor. And maybe we’ll see a computer that can walk up the stairs to an elderly woman’s apartment, take her blood pressure, draw blood, and ask if she’s been taking her medication, all while putting her at ease instead of terrifying her. We don’t think any of these advances is likely to come anytime soon, but we’ve also learned that it’s very easy to underestimate the power of digital, exponential, and combinatorial innovation, so never say never.”

That’s quite interesting. Quite prescient. But you probably didn’t think it would happen so fast, right? How do you react to that? Does that scare you? Does it give you hope?

Andrew McAfee: Thank heaven that Erik Brynjolfsson and I included that final clause: “Never say never.” As you just quoted back to us, we thought all of those advances were unlikely—on the not-crazy-distant horizon. Well, all of those things have happened except the final one. Building robots that can navigate the physical world is extremely difficult, especially if you also want them not to terrify Grandma.

Aside from that, all these things we thought sounded like science fiction have happened. It reinforces to me not only that technological progress and innovation are hard to predict, but that it’s easy to underestimate exponential progress. Our minds aren’t well designed to understand what exponential growth looks like, because it looks boring until it looks crazy.

Linear growth is straightforward—we project it forward. Exponential growth, like we’ve seen with Moore’s Law and now with AI, is deeply counterintuitive. Erik and I underestimated exponential technological progress while writing a book about exponential technological progress.

Chris Bradley: You talked about the PlayStation 3 being more powerful than the most powerful computer in the world just ten years earlier. You understood the concept of exponential growth, but if you’d projected further, you’d have sounded crazy.

Andrew McAfee: Exactly.

Should we try to steer AI?

Chris Bradley: A lot of economists recently signed a short declaration saying, essentially, “AI is coming. It’s important. It could have a lot of bad effects, and we need to get ahead of controlling it.”

You wrote an edited version of the letter that rejected part of its premise. Talk to us about that, because it gets to one of the biggest debates we have when discussing our book. We all agree technology has generally been good. We understand that moving people off farms created prosperity and that if we’d frozen society to preserve jobs, we’d all be poorer. But we struggle to apply that same thinking to AI because it feels bigger, scarier, and faster. What motivated you to take a less pessimistic position?

Andrew McAfee: If you Google “We Must Act Now,” you’ll find it pretty quickly. As you pointed out, it’s only three short paragraphs. They did a beautiful job of making their point concisely. We believe exponential progress is astonishing. It’s one of the deepest forces in the world. We also agree that new technologies bring new risks, new challenges, and new harms. Economists often say technological progress is the only free lunch they believe in, but they don’t believe it’s an unambiguous good for everybody all the time. Great, I agree with all of that.

The letter says some of those things, but I had two objections—one moderate and one pretty significant.

The moderate objection was that it leads with negativity. It has an overall gloomy tone. It says AI might raise our living standards and could be great, but it could also be really bad, and we have to act now to minimize the bad outcomes. I think that’s a framing mistake.

The larger objection was the third paragraph, which said economists, policymakers, and technology leaders must act now to build guardrails. Right now there are debates across America about whether cities should allow Waymo and other autonomous vehicles to operate. It’s almost impossible to argue meaningfully that Waymo is less safe than human drivers. The data overwhelmingly show the opposite.

Chris Bradley: It’s about 90 percent safer.

Andrew McAfee: At least 90 percent safer. One analysis concluded we’d have around 30,000 fewer traffic fatalities every year in America if we adopted autonomous driving.

Chris Bradley: I think that’s ultimately the debate that will prevail. It’s hard to argue that we should keep more people dying in car accidents.

Andrew McAfee: I agree. But what opponents are saying instead is, “There’s going to be job loss. We need to get ahead of that by making sure taxi drivers, rideshare drivers, and truck drivers can keep their jobs.” The job preservation argument is now becoming an explicit argument.

One of my worries—and I don’t think it’s a crazy worry—is that people making this explicitly anti-progress, welfare-reducing argument will start waving around the “We Must Act Now” letter to support their case.

As I argued in the Substack response I wrote, it’s a clear endorsement of upstream governance: the idea that bad things could happen in the future, so we need to anticipate them now and put guardrails in place. Very often, “guardrails” becomes another word for legislation or regulation intended to prevent outcomes like job losses.

No. I’m on Team Permissionless Innovation. I’m not on Team Upstream Governance. If you look across the Atlantic from where I’m sitting in Massachusetts, you’ll see an example of a rich, prosperous, technologically sophisticated region that’s embraced upstream governance more enthusiastically: Europe.

Do I want the American technology sector or the European one? I want the American technology sector. We’ve taken a more laissez-faire approach. We’ve been more supportive of permissionless innovation, and the thing I really disliked about the letter was that it seemed to call for upstream governance of AI. I’m not on board with that.

Marc Canal: I have faith in Europe, though. One word in the statement stood out to me. It says we should “steer” AI in a direction that complements humans and benefits society. I sympathize with the outcome. Of course we want AI to benefit people.

What I struggle with is the idea—I’ve seen it in papers about pro-worker AI—that technology is like a menu you can examine in advance and choose from. You can’t say, “This application is pro-worker, this one isn’t, so we’ll choose A but not B.” That’s not how reality works. Technology develops as it develops. You can’t predict every future technology and choose among them ahead of time.

Andrew McAfee: I completely agree. In addition to “guardrail,” the word “steer” really activated me because it reflects the philosophy that some group of elites can accurately foresee the path of technological progress and guide us along it while preserving all the benefits and avoiding all the downsides. I simply don’t believe that’s possible. I also see relatively little evidence that this kind of technological dirigisme works particularly well. We can’t stop global AI, we could absolutely handicap American AI, and I think that would be terrible for economic, geopolitical, and national-security reasons.

Chris Bradley: If you look at your books or at Century of Plenty, you’ve got to approach AI with a few prior beliefs. The first is that exponential growth is real. And by the way, the numbers don’t need to be big. In our book, the path to universal prosperity is about adding 30 basis points to growth.

The second is that technologies that help people create wealth generally do exactly that. Usually it’s because they create entirely new things. The replacement effect becomes pretty much meaningless in the end. Every piece of data we have suggests that’s what’s happening with AI today: more new things are being created, and previously constrained supply is being unlocked.

The third prior is that markets and firms, when supported by good institutions, generally do a better job than central planners.

Andrew McAfee: I violently agree with all three of those premises. In particular, this idea that we can direct technological progress, get ahead of it, and have the foresight and wisdom to minimize harms while preserving all the benefits—that’s incredibly difficult. All three of us understand that progress involves trade-offs. Disruption isn’t fun for the entities being disrupted.

But if you want another century of plenty, you’re going to have to accept some disruption. I can abstractly worry about companies going bankrupt, but the real concern is that when companies fail, people lose their jobs. Losing your job is a deeply unpleasant experience. It feels terrible, and it can damage your economic future. I don’t want to minimize that. At the same time, I come down firmly on the side of autonomous driving rather than preserving the status quo and the jobs of people who currently drive us around cities.

What technology means for jobs and inequality

Chris Bradley: About half of the productivity gap between the United States and Europe comes down to faster creative destruction in the United States, and that ultimately shows up in wages.

Rather than skating past this topic, let’s dive into jobs, inequality, and superstar effects. When I read Race Against the Machine—it really opened my eyes. The big ideas were that exponential technology would fundamentally change the world of work. It would increase returns to certain skills, create superstar effects, and have broad societal implications. Inequality wasn’t necessarily the product of some ruling class. It was often an outcome of technological progress. Much of the inequality in the United States simply reflects productivity growing faster in some places than in others. Let’s dig into that.

As Marc mentioned, the book came out after the global financial crisis. Those with long memories will remember the phrase “jobless recovery.” That’s what was happening in the United States. Economic growth was recovering, but employment wasn’t. People talked about decoupling and even the end of work. Looking back at Race Against the Machine and your predictions about superstar effects, increased inequality, and joblessness, what have we learned from the past 15 years?

Andrew McAfee: We did talk about inequality, and the framing of the follow-up on that, The Second Machine Age, was that technological progress has two big effects: bounty—your century of plenty—but also spread, meaning greater inequality. We’ll come back to that.

After the global financial crisis, especially once we started understanding how powerful modern machine learning was around 2012, I started thinking, “We’re just going to need human beings to do a lot fewer things.” I worried there would be significant implications for jobs and labor markets. I don’t think I ever said I was certain we were entering an era of mass technological unemployment, but it definitely worried me.

OK. Here we are today, facing historically low unemployment across most of the rich world. I didn’t even look up the latest unemployment rate in the United States because I know it’s low. Mass joblessness should not be at the top of anyone’s list of concerns.

The mental mistake I made was, I think, a forgivable one. When creative destruction happens, it’s always easier to see the destruction—and foresee more destruction—than the creation.

Chris Bradley: That’s a great way of putting it. We can easily picture the destruction because we can see what’s disappearing. The creation is much harder because it doesn’t exist yet, so we can’t imagine it.

Andrew McAfee: Right. The destruction is very salient. People are losing their jobs. The creation can be more nebulous, and the destruction is often easier to predict. The creation is much harder to predict, and I think I was guilty of that. I love the John Maynard Keynes quote: “When the facts change, I change my opinions. What do you do?”

Over the past 15 years since Race Against the Machine came out, we have not seen massive technological unemployment. My opinions have changed. I don’t expect to see it going forward, even though we have another incredibly powerful bundle of technologies with generative AI today. I got that wrong, and I’m trying to internalize Keynes’s advice and adjust my views and my priors.

The one thing we got right, though, was the increase in inequality—the spread. There’s a huge ongoing debate about how bad inequality is per se. There’s a saying that I despise: “Every billionaire is a policy failure.” Well, we’ve now had our first glimmer of a trillionaire.

So the growth in inequality is real, but I’m going to quote another economist—actually a Scottish economist from about 250 years ago. Adam Smith said, “Whenever there is great property, there is great inequality.” He was absolutely right. Periods of great flourishing bring rises in inequality, Chris, for exactly the reason you pointed out.

Again, there’s an active debate about this. I don’t have a problem with very large wealth inequality. I have a much bigger problem with economic stagnation and with slowly moving backward. I’d also have more of a problem if the bottom weren’t being lifted.

In general, especially in rich countries—and around the world—people in the middle and lower parts of the income distribution are seeing their economic lives improve quite quickly by historical standards.

So, this idea that every billionaire, let alone every trillionaire, is a policy failure has a couple of real problems.

One is that it assumes fixed-pie thinking or zero-sum thinking in the background—that billionaires must be taking money from someone else. No. They’re creating value and capturing a slice of that value. As a result, they’ve improved our lives. They’ve created enormous numbers of jobs.

Marc Canal: That’s very interesting because it’s actually quite similar to what we write in the book. Even if you take the 100-year picture—and, by the way, we focus on income inequality rather than wealth inequality—globally we’re at the lowest point of income inequality in the past 100 years. Most importantly, as you were saying, even while income inequality in the United States increased from the 1980s onward, the bottom was rising as well.

One more thing: since around 2010 or 2015, post-tax income inequality in the United States has actually been quite flat.

That doesn’t mean there aren’t problems. Maybe I want to focus on one specific topic. Even if the aggregate picture looks good, there are specific places that get disrupted all at once. It could be the China shock. It could be technological change. There are many hypotheses. How do you respond to people who live through concentrated stagnation or even decline in a particular place?

Andrew McAfee: It’s real, and it’s a problem for several reasons. One is that if you cross-examine economists hard enough—and economists tend to believe they have a good tool kit for many interventions—a lot of them will admit that our tool kit for place-based interventions isn’t very strong. The theory in the background is that if aggregate growth exists, people will move to where the opportunities are.

Well, not everybody wants to move to a superstar city. Leaving aside the fact that superstar cities become expensive because we don’t build enough housing, some people simply want to stay in their communities. When jobs and economic activity leave those communities, that’s very bad news. It affects people’s lives and the health of communities.

I also think it changes politics because when your livelihood disappears and a populist or demagogue comes along and says, “Whether or not I have a solution, I do have an enemy: The enemy is billionaires, globalists, or some other group,” people become more receptive.

We have a lot of evidence that voters become more receptive to populists and demagogues when their economic lives and their communities have genuinely taken a turn for the worse. Those are legitimate concerns, and I wish our tool kit for place-based intervention was bigger and richer than it is.

Is AI breaking the first rung of the career ladder?

Marc Canal: I want to come back to young people, though. Alongside housing, another important issue is entering the labor market. We’ve heard a lot about this. Your friend Erik Brynjolfsson and others have written about whether AI is breaking the first rung of the career ladder. I have some reservations about that argument, but what’s your view?

Andrew McAfee: As you say, some of Erik’s data point in that direction. We also have another colleague, Matt Beane, who just wrote a really good book called The Skill Code about exactly this issue. He focuses not only on entering the workforce but also on apprenticeship in the AI era.

Matt documents that in occupations ranging from surgery to accounting and auditing, entry-level people traditionally learn by doing low-level work. They gain experience because companies need someone to do that work. I’ve spoken with leaders at some of the Big Four auditing firms, and they tell me the same thing. You learn to become an auditor by spending your early years doing a lot of relatively basic auditing work. Now we can automate much of that, especially with AI. They’re genuinely struggling with the question: What do skill acquisition and apprenticeship look like in an AI-powered world?

As we discussed earlier, episodes of very fast and profound technological progress bring new challenges. This is one of them. It’s happening right now, so we should be thinking about interventions now rather than sitting in our armchairs trying to anticipate what we might need five years from today.

Chris Bradley: Did you see the paper by Lambert from the London School of Economics challenging the Brynjolfsson hypothesis?

He argued that working from home is a much stronger statistical explanation for declining entry-level opportunities than AI. That fits my intuition. For people like us, who are well established, have good judgment, and strong networks, working from home is almost a superpower. I’m at home tonight—it’s quite late where I am. But if you’re a young person who doesn’t know anyone, doesn’t yet understand how things work, and is still learning, working from home can be a huge drag on productivity. I thought that was a really interesting perspective.

Andrew McAfee: It raises an important point. We’re arguing about what’s causing the decline in job entry for younger workers. That means the causality isn’t clear and the effect probably isn’t enormous. If we looked at labor-force participation rates for people aged roughly 25 to 40 in the United States, we’d see they’re close to an all-time high.

We have vivid anecdotes, but I don’t think the idea that recent college graduates are facing the worst job market in generations really holds up.

Chris Bradley: One more observation from the work Marc and I are doing on generations. Thirty years ago, if you graduated from college, you were part of a relatively small elite. There was a significant gap between college graduates and everyone else. Today, roughly two-thirds of young people go to college. So the “elite” has become much more normal.

Andrew McAfee: That’s basically the elite-overproduction theory, isn’t it? I wish we’d make being a plumber feel just as normal—in fact, prestigious. I mean that completely seriously. We live in the physical world, and there’s no shortage of important physical work that needs to be done. Think about the energy transition.

I think all three of us have recent stories about how difficult it is to get a plumber or an air-conditioning repair technician to come to the house. We need this work. As you pointed out, Chris, these can be very good, well-paid jobs. But nobody proudly says, “My kid’s a plumber.” I think we’re getting it wrong. Think about all the physical work needed to build AI infrastructure and complete the energy transition during the 21st century.

Whether it’s team nuclear or team renewables to ultimately lead the transition, it’s going to require an enormous amount of work. Think about our decaying infrastructure. Schools in America aren’t in great shape. We have an absolute metric ton of work to do just to modernize our infrastructure. Let’s make those jobs not just good jobs but elite, prestigious jobs.

Can we get more from less?

Marc Canal: That’s a perfect segue. Andrew, you’ve written many other books, and we obviously can’t cover all of them. Another one is More from Less.

In a nutshell, More from Less argues exactly what the title suggests: that we can continue growing the economy while using fewer and fewer inputs—whether materials, CO₂ emissions, or other resources. Maybe start by explaining the core argument of the book, and then we can dive into some questions and discussion.

Andrew McAfee: The explanation is actually pretty simple. First, why are we taking less from the planet every year? It’s not because we’ve renounced growth. It’s not because we’ve renounced consumption or higher living standards. Instead, it’s something much simpler. Materials cost money. Companies generally pay by the molecule. Profit-maximizing businesses want to reduce those costs, and the technological tool kit that allows them to do so keeps improving at an exponential rate thanks to Moore’s Law and related advances.

The book tells the story of how market forces and technological opportunity combine collectively—not because people suddenly became more altruistic, but because businesses want to save money—to help us lighten up on the planet.

Marc, you also mentioned pollution levels are going down. Even though I’m a strong believer in economics, my introductory economics textbook was right about one thing: markets alone don’t solve externalities like pollution. We need something else.

The book really tells two stories. Using fewer resources is a story about technology, capitalism, and market-oriented economies. Reducing pollution is a story about citizens demanding change and governments responding.

We’ve passed effective legislation around the world to reduce pollution. Waiting for markets alone to solve that problem would have been a bad idea. Instead, people said, “We’d like our children to breathe clean air,” and elected officials responded with measures like the Clean Air Act and the Clean Water Act. Those are exactly the kinds of guardrails in this case because the problem was real and obvious. That’s the broader story of More from Less.

It’s essentially a story about decoupling rising prosperity from resource consumption so we can continue the century of plenty while becoming better stewards of the planet.

Chris Bradley: I do want to push on the “less” part of More from Less, though. In the scenario we describe in our book, where the entire world reaches universal prosperity and even the poorest country achieves today’s Swiss standard of living, the global economy becomes about 8.5 times larger. That still requires materials and energy.

For example, we’d need roughly five times more copper and about twice as much steel as has been extracted since 1950. We’d need roughly three times more total energy and perhaps 12 times more electricity. We actually built a bill of materials to ask whether that’s physically possible. Can we produce enough copper? Can we generate enough energy?

You’ve done a great job explaining this remarkable long-term trend of improvement. But if we genuinely want universal prosperity, people need roads, bridges, buildings, and infrastructure. At the moment, someone in India has about one ton of steel surrounding them, while an American has around 11 tons. At some point, achieving universal prosperity means building more physical infrastructure. How do you think about that?

Marc Canal: Especially in emerging economies. Correct me if I’m wrong, Andrew, but most of the evidence you’ve discussed comes from advanced economies. We’re starting to see some decoupling in places like China, but not yet in countries like India. How do you think about the countries that will become rich in the future?

Andrew McAfee: I love your premise. What would it take to raise the poorest countries in the world to Switzerland’s standard of living?

I think we absolutely can and should do that. You’re both pointing out that doing so will have a material footprint—probably a big one in a lot of cases. I think it’s a profound moral mistake to tell countries like Bangladesh, “We understand you’d like the standard of living people enjoy in Santa Monica or New York City, but the planet can’t handle it, so you can’t have it.”

I know you two are not making that argument. Your point is simply that cities like Dhaka will require much more steel, and can the planet tolerate that?

There are two interesting things going on. Let’s take Dhaka as an example, and take it as given that we accomplish this century of plenty. I’m quite confident that future Dhaka won’t be as materially intensive as present-day Zurich. For example, will everyone in future Dhaka have their own cars? Maybe a few people will, but I think it would be fairly silly because transportation on demand will be widely available. Will they build copper telephone networks like Zurich once did? Absolutely not. They wouldn’t build that today. Will we learn to build skyscrapers using materials that are lighter and less intensive than we know how to do today? Certainly. Future Dhaka will absolutely use more materials and more energy than today’s Dhaka. But it won’t follow the same development path that Zurich followed, and it will almost certainly require fewer raw materials than Zurich does today.

Chris Bradley: In the same way that India and China are already following less carbon-intensive development paths than the United States did.

Andrew McAfee: Future Dhaka certainly won’t emit as much greenhouse gas per unit of GDP as Zurich does today. Absolutely not.

Chris Bradley: They definitely do need more steel, though. When people live in informal settlements, building proper housing requires materials.

We talk a lot about the energy transition, which is a very important topic. The phrase itself almost suggests we’re simply replacing one source of energy with another. In reality, a prosperous future requires much more energy overall—it will simply come from different sources.

Take the United States. You argued in your book that the country had reached peak fossil-fuel use. It may have risen slightly since then because of natural gas, but the shift from coal to gas has helped reduce emissions. Since the Paris Agreement, US emissions have fallen by about 17 percent.

Andrew McAfee: Exactly. Not just per capita. The total amount my country puts into the atmosphere has been trending downward.

Chris Bradley: It’s actually been a remarkable success story.

Andrew McAfee: Right. Your comparison between Dhaka and Sydney makes an important point. If you care about pollution, you’d much rather live in the rich city than the poor one.

Indira Gandhi captured this perfectly in a speech, I think in 1972, when she asked, “Are not poverty and need the greatest polluters?” When you’re struggling simply to feed your children, it’s difficult to prioritize something as abstract as the environment. As societies become wealthier, people start demanding cleaner air and cleaner water. That process will happen in Dhaka as well.

Chris, you’re right that future Dhaka will contain vastly more steel than today’s Dhaka. Absolutely. The question then becomes whether there’s enough steel. Or, put another way, whether there’s enough iron ore in the Earth’s crust to satisfy that demand. Yes.

Chris Bradley: That’s exactly what we calculated in the book.

Andrew McAfee: A thousand times yes. It’s a trivial question. I hope your book avoids repeating the mistake made by the authors of The Limits to Growth. They built models of the global economy in 1972, projected economic and population growth forward, and concluded that we’d literally run out of steel, copper, aluminum, fossil fuels, and other resources.

Chris Bradley: They forgot the single most important variable in economics: improvement.

Andrew McAfee: Exactly. And they forgot innovation.

Chris Bradley: You only need modest annual improvements for the long-term effects to become extraordinary.

Andrew McAfee: They looked at known reserves of different materials in 1972 and assumed those were fixed. And then they ran every model they could, and every model runs out of materials. What they ignored was that when demand for copper or steel increases, people search for new deposits, develop better extraction methods, and improve technology. I went back and compared the known reserves from 1972 with today’s figures. They’re several times larger. Even after extracting half a century’s worth of resources.

Chris Bradley: I call it the magic cookie jar. You keep taking cookies out, but somehow there are always more.

Andrew McAfee: Exactly. We absolutely need to avoid polluting the planet. But is our planet abundant enough to support the century of plenty you’re describing? Hell yeah.

Chris Bradley: With continued technological progress.

The case for optimism

Chris Bradley: Before we finish, though, our book argues that the real constraint isn’t the physics of growth but the mindset around growth. According to the Edelman survey, only about one in five Americans believes the next generation will be better off than the last. There’s actually a real crisis of hope. In countries like France, it’s even lower.

Our real problem is a crisis of hope. That brings us to your idea of the Four Horsemen of Optimist. Let’s ride those horses and get optimistic. Tell us about them.

Andrew McAfee: More from Less is organized around what I call the Four Horsemen of the Optimist, rather than the Four Horsemen of the Apocalypse.

We talked about them, they come in two pairs. The first pair is technological progress and capitalism. I deliberately use the word capitalism because it’s a triggering word for a lot of people. What I really mean is a decentralized market economy in which profit-seeking firms constantly try to reduce material costs.

The second pair is public awareness of problems like pollution and governments that respond. Citizens demand change, and governments enact policies that reduce pollution over time.

If we have those four horsemen working together, then I agree with Julian Simon, who remains one of the most underappreciated thinkers I’ve encountered. He argued that the planet is abundant enough for all of us, for our children, and for the children of today’s poorest people. He made an astonishing prediction that I think is exactly right.

He said, “The material conditions of life will improve for most people around the world without end.” That’s crazily optimistic but I think it’s also accurate.

But you’re pointing to something that genuinely worries me. We’re living in anxious times. We’re living in negative times. I don’t know what’s going on but I think periods of rapid, profound change naturally generate anxiety because human beings like the status quo. We don’t like deep profound change. The last time we experienced change this deep and this fast was during the Industrial Revolution.

Marc, on your home continent of Europe, the degrowth movement has become surprisingly influential, particularly among younger people. Voluntarily making ourselves poorer in pursuit of benefits we’re already achieving is a terrible idea. The same goes for proposals to create some vast global bureaucracy that simply redistributes existing wealth because supposedly we can no longer grow the pie.

Those of us who believe in progress, prosperity, and human flourishing—and who believe we can have it all—have our work cut out for us.

Marc Canal: To me, that’s actually an optimistic note to finish on. I wholeheartedly agree. Much of my own work pushes back against the idea of degrowth. If you think about Europe’s future obligations—aging populations, pensions, healthcare, higher defense spending, energy—a zero-sum world is one full of very nasty trade-offs. Choosing between those priorities would become extremely difficult.

Andrew McAfee: That’s exactly the point Mario Draghi made in his report that came out a couple of years ago. He made that on page two and essentially said that if Europe can’t restart its growth engine, it’s going to face some really nasty trade-offs.

Chris Bradley: Then let’s restart the growth engine.

Andrew McAfee: Before we finish, let me just say one thing. Thank you for writing this book. This note of optimism and grounded, evidence-based argument for optimism and for what the future can look like, we need more of that.

Chris Bradley: Thank you. Our basic message is simple. Growth is good. Growth is for everyone. Growth is possible. That’s become our mantra. Marc and I are really just numbers people. We just looked at the spreadsheets. That’s where these ideas came from. They’re not ideological.

Andrew McAfee: Whenever I start feeling depressed, I go to Our World in Data and click around. I’m serious. It’s like nerd therapy.

Chris Bradley: Try finding a long-term trend that’s actually getting worse. It’s not that easy.

Andrew, thank you for joining us on the podcast. Some small celebration of how good life is—and how much better it can become—feels appropriate.

Andrew McAfee: Amen. I love it. Thank you.