Welcome to Faster, Please! — The Podcast. Several times a month, host Jim Pethokoukis will feature a lively conversation with a fascinating and provocative guest about how to make the world a better place by accelerating scientific discovery, technological innovation, and economic growth.
☀️ My chat (+transcript) with economist Noah Smith on technological progress
Thursday, September 26, 2024 • Duration 32:27
Some signs of tech progress are obvious: the moon landing, the internet, the smartphone, and now generative AI. For most of us who live in rich countries, improvements to our day-to-day lives seem to come gradually. We might (might), then, forgive some of those who claim that our society has not progressed, that our lives have not improved, and that a tech-optimist outlook is even naïve.
Today on Faster, Please! — The Podcast, I talk with economist Noah Smith about pushing the limits in areas like energy technology, how geopolitical threats spur innovation, and why a more fragmented industrial policy might actually be an advantage.
Below is a lightly edited transcript of our conversation
Recognizing progress (1:43)
Pethokoukis: Noah, welcome to the podcast.
Smith: Great to be here!
Not to talk about other podcast guests, but I will very briefly — Last year I did one with Marc Andreessen and I asked him just how tech optimistic he was, and he said, “I'm not sure I'm an optimist at all,” that the most reasonable expectation is to expect the future to be like the past, where we have a problem building things in the real world, that some of our best ideas don't necessarily become everything they could be, and I think a perfectly reasonable baseline forecast is that, for all our talk about optimism, and “let's go,” and “let's accelerate,” that none of that happens. Does that sound reasonable to you or are you more optimistic?
I'm optimistic. You know, a few years ago we didn't have mRNA vaccines. Now we do. And now we have a magical weight loss drug that will not only make you lose weight, but will solve half your other health problems for reasons we don't even understand yet.
So much inflammation.
Right. We didn't even have that a few years ago. That did not exist. If you told someone that would exist, they would laugh at you. A magic pill that not only makes you thin, but also just solves all these other health issues: They would laugh at you, Scott Alexander would laugh at you, everyone would laugh at you. Now it's real. That's cool.
If you had told someone a few years ago that batteries would be as insanely cheap as they are, they would've been like, “What? No. There's all these reasons why they can't be,” but none of those reasons were true. I remember because they did actually say that, and then batteries got insanely cheap, to the point where now Texas is adding ridiculous amounts of batteries for grid storage. Did I predict that was going to happen? No, that surprised me on the upside. The forecasters keep forecasting sort of a leveling off for things like solar and battery, and they keep being wrong.
There's a lot of other things like reusable rockets. Did you think they'd get this good? Did you think we'd have this many satellites in the low-earth orbit?
AI just came out of nowhere. Now everyone has this little personal assistant that's intelligent and can tell them stuff. That didn't exist three years ago.
So is that, perhaps, growing cluster of technologies, that's not just a short-term thing. Do you think all these technologies — and let's say particularly AI, but the healthcare-related stuff as well — that these taken together are a game-changer? Because people always say, “Boy, our lives 30 years ago didn't look much different than our lives today,” and some people say 40 years ago.
But that’s wrong!
Yes, I do think that is wrong, but that people's perception.
When I was a kid, people didn't spend all day looking at a little screen and talking to people around the world through a little screen. Now they do. That's like all they do all day.
But they say that those aren't significant, for some reason, they treat that as a kind of a triviality.
Like me, you're old enough to remember a thing called “getting bored.” Do you remember that? You’d just sit around and you're like, “Man, I’ve got nothing to do. I’m bored.” That emotion just doesn't exist anymore — I mean, very fleetingly for some people, but we've banished boredom from the world.
Remember “getting lost?” If you walk into that forest, you might get lost? That doesn't happen unless you want to get lost, unless you don't take your phone. But the idea that, “Oh my God, I'm lost! I'm lost!” No, just look at Google Maps and navigate your way back.
Being lost and being bored are fundamental human experiences that have been with us for literally millions of years, and now they're just gone in a few years, just gone!
Remember when you didn't know what other places looked like? You would think, “Oh, the Matterhorn, that’s some mountain in Switzerland, I can only imagine what that looks like.” And then maybe you'd look it up in an encyclopedia and see a picture of it or something. Now you just type it into Google Images, or Street View, or look at YouTube, look at a walking tour or something.
Remember not knowing how to fix things? You just had no idea how to fix it. You could try to make it up, but really what you'd do is you'd call someone who was handy with stuff who had this arcane knowledge, and this wizard would fix your cabinet, or your dresser, or whatever, your stereo.
Being lost and being bored are fundamental human experiences that have been with us for literally millions of years, and now they're just gone in a few years, just gone!
So why does that perception persist? I mean, it's not hard to find people — both of us are probably online too much — who just will say that we've had complete and utter stagnation. I don't believe that, yet that still seems to be the perception, and I don't know if things haven't moved fast enough, if there are particular visions of what today should look like that haven’t happened, and people got hung up on the flying-car, space-colony vision, so compared to that, GPS isn’t significant, but I think what you have just described, not everybody gets that.
Because I think they don't often stop to think about it. People don't often stop to think about how much the world has changed since they were young. It's like a gradual change that you don't notice day-to-day, but that adds up over years. It's like boiling the frog: You don't notice things getting better, just like the frog doesn't notice the water getting hotter.
Do you think it's going to get hotter going forward, though? Do you think it's going to boil faster? Do you think that AI is such a powerful technology that it'll be indisputable to everybody that something is happening in the economy, in their everyday lives, and they look a lot different now than they did 10 years ago, and they're going to look a whole lot different 10 years from now?
Utility, remember — back to econ class — utility is concave. A utility of wealth, utility of consumption, is concave, which means that if you get 10,000 more dollars of annual income and you're poor, that makes a hell of a lot of difference. That makes a world of difference to you. But if you're rich, it makes no difference to you. And I think that Americans are getting rich to the point where the new things that happen don't necessarily increase our utility as much, simply because utility is concave. That's how things work.
In the 20th century, people escaped material poverty. They started out the century with horses and buggies, and wood-burning stoves, and freezing in the winter, and having to repair their own clothes, and having food be super expensive, and having to work 60-hour weeks, 80-hour weeks at some sweatshop, or just some horrible thing, and horrible conditions with coal smoke blackening the skies; and then they ended in nice, clean suburbia with computers and HDTVs —I guess maybe we didn't get those till the 2000s — but anyway, we ended the 20th century so much richer.
Basically, material poverty in rich countries was banished except for a very few people with extreme mental health or drug problems. But then for regular people, material want was just banished. That was a huge increment. But if you took the same increment of wealth and did that again in the next century, people wouldn't notice as much. They'd notice a little bit, but they wouldn't notice as much, and I think that it's the concavity of utility that we're really working against here.
In the 20th century, people escaped material poverty. They started out the century. . . having to work 60-hour weeks, 80-hour weeks at some sweatshop. . . and then they ended in nice, clean suburbia with computers and HDTVs . . .
So is economic growth overrated then? That kind of sounds like economic growth is overrated.
Well, no. I don't know that it's overrated. It's good, but I don't know who overrates it. Obviously it's more important for poor countries to grow than for rich countries to grow. Growth is going to make a huge difference to the people of Bangladesh. It's going to be life-changing, just as it was life-changing for us in the 20th century. They're going to have their 20th century now, and that's amazing.
And, to some extent, our growth sustains their growth by buying their products; so that helps, and contributing to innovations that help them, those countries will be able to get energy more easily than we were because they're going to have this super-cheap solar power, and batteries, and all this stuff that we didn't have back in the day. They're going to have protections against diseases, against malaria, and dengue fever, and everything. We didn't have those when we were developing, we had to hack our way through the jungle.
So growth is great. Growth is great, and it's better for the people in the poor countries than for us because of concavity of utility, but it's still good for us. It's better to be advancing incrementally. It's better to be feeling like things are getting better slowly than to be feeling like things aren't getting better at all.
So many things have gotten better, like food. Food has gotten immeasurably better in our society than it was in the ’90s. The food you can eat at a regular restaurant is just so much tastier. I don't know if it's more nutritious, but it's so much tastier, and so much more interesting and varied than it was in the ’90s, and people who are in their 40s or 50s remember that. And if they stop to think about it, they'll be like, “You know what? That is better.” We don't always stop to remember what the past was. We don't remember what food was like in the ’90s — I don't. When I'm going out to a restaurant to eat, I don't think about what a restaurant was like in 1994, when I was a kid. I don't think about that. It just doesn't come to mind. It's been a long time.
In Japan I noticed it a lot, because Japan had, honestly, fairly bland and boring food up until about 2010 or so. And then there was just this revolution where they just got the most amazing food. Now Japan is the most amazing place to go eat in the world. Every restaurant's amazing and people don't understand how recent that is. People don't understand how 20 years ago, 25 years ago, it was like an egg in a bowl of rice and sort of bland little fried things. People don't remember how mediocre it was, because how often did they go to Japan back in 2005?
It's better to be feeling like things are getting better slowly than to be feeling like things aren't getting better at all.
Redrawing the boundaries of energy tech (12:39)
Your answer raised several questions: One, you were talking about solar energy and batteries. Is that enough? Is solar and batteries enough? Obviously I read about nuclear power maybe too much, and you see a lot of countries trying to build new reactors, or restart old reactors, or keep old nuclear reactors, but over the long run, do we need any of that other stuff or can it really just be solar and batteries almost entirely?
Jesse Jenkins has done a lot of modeling of this and what would be the best solutions. And of course those models change as costs change. As battery costs go down and battery capabilities improve, those models change, and we can do more with solar and batteries without having to get these other things. But the current models that the best modelers are making right now of energy systems, it says that we're probably looking at over half solar and batteries, maybe two thirds, or something like that. And then we'll have a bunch of other solutions: nuclear, wind, geothermal, and then a little bit of gas, we'll probably never completely get rid of it.
But then those things will all be kind of marginal solutions because they all have a lot of downsides. Nuclear is very expensive to build and there's not much of a learning curve because it gets built in-place instead of in a factory (unless it's on a submarine nuclear plant, but that's a different thing). And then wind takes too much land, really, and also the learning curve is slower. Geothermal is only certain areas. It's great, but it's only certain areas. And then gas, fossil fuel, whatever.
But the point is that those will all be probably part of our mix unless batteries continue to get better past where we even have expected them to. But it's possible they will, because new battery chemistries are always being experimented with, and the question is just: Can we get the production cost cheap enough? We have sodium ion batteries, iron flow batteries, all these other things, and the question is, can we get the cost cheap enough?
Fortunately, China has decided that it is going to pour untold amounts of capital and resources and whatever into being the Saudi Arabia of batteries, and they're doing a lot of our work for us on this. They're really pushing forward the envelope. They're trying to scale every single one of these battery chemistries up, and whether or not they succeed, I don't know. They might be wasting capital on a lot of these, or maybe not, but they're trying to do it at a very large scale, and so we could get batteries that are even better than we expect. And in that case, I would say the share of solar and batteries would be even higher than Jesse Jenkins and the other best modelers now predict.
But you don't know the future of technology. You don't know whether Moore's Law will stop tomorrow. You don't know these things. You can trace historical curves and forecast them out, and maybe come up with some hand-wavy principles about why this would continue, but ultimately, you don't really know. There's no laws of the universe for technological progress. I wish there were, that'd be cool. But think solar and batteries are on their way to being a majority of our total energy, not just electricity, but total energy.
Racing China in research (15:59)
Does it concern you, in that scenario, that it's China doing that research? I understand the point about, “Hey, if they want to plow lots of money and lose lots of money,” but, given geopolitical relations, and perhaps more tariffs, or war in the South China Sea, does that concern you that that innovation is happening there?
It absolutely does concern me. We don't want to get cut off from our main sources of energy supply. That's why I favor policies like the Inflation Reduction Act. Basically, industrial policy is to say, “Okay, we need some battery manufacturing here, we need some solar panel manufacturing here in the country as a security measure.” Politicians always sell it in terms of, “We created this many jobs.” I don't care. We can create jobs anyway. Anything we do will create jobs. I don't care about creating specific kinds of jobs. It is just a political marketing tactic: “Green jobs, yes!” Okay, cool, cool. Maybe you can market it that way, good for you.
But what I do care about is what you talked about, which is the strategic aspect of it. I want to have some of that manufacturing in the country, even if it's a little inefficient. I don't want to sacrifice everything at the altar of a few points of GDP, or a few tenths of a percent of points of GDP at most, honestly. Or sacrifice everything in the altar of perfect efficiency. Obviously the strategic considerations are important, but, that said, what China's doing with all this investment is it's improving the state of technology, and then we can just copy that. That's what they did to us for decades and decades. We invented the stuff, and then they would just copy it. We can do that on batteries: They invent the stuff, we will copy it, and that's cool. It means they're doing some of our work, just the way we did a lot of their work to develop all this technology that they somehow begged, borrowed, or stole.
. . . what China's doing with all this investment is it's improving the state of technology, and then we can just copy that. That's what they did to us for decades and decades. We invented the stuff, and then they would just copy it. We can do that on batteries. . .
The original question I asked about: Why should we think the future will be different than the recent past? Why should we think that, in the future, America will spend more on research? Why do we think that perhaps we'll look at some of the regulations that make it hard to do things? Why would any of that change?
And to me, the most compelling reason is, it's quite simple just to say, “Well, what about China? Do you want to lose this race to China? Do you want China to have this technology? Do you want them to be the leaders in AI?” And that sort of geopolitical consideration, to me, ends up being a simple but yet very persuasive argument if you're trying to argue for things which very loosely might be called “pro-progress” or “pro-abundance” or what have you.
I don't want to whip up any international conflict in order to stimulate people to embrace progress for national security concerns. That wouldn't be worth it, that’s like wagging the dog. But, given that international conflict has found us — we didn't want it, but given the fact that it found us — we should do what we did during the Cold War, during World War II, even during the Civil War, and use that problem to push progress forward.
If you look at when the United States has really spent a lot of money on research, has built a lot of infrastructure, has done all the things we now retrospectively associate with progress, it was for international competition. We built the interstates as part of the Cold War. We funded the modern university system as part of the Cold War. And a lot of these things, the NIH [National Institutes of Health], and the NSF [National Science Foundation], and all these things, of course those came from World War II programs, sort of crash-research programs during and just before World War II. And then, in the Civil War, of course, we built the railroads.
So, like it or not, that's how these things have gotten done. So now that we see that China and Russia have just decided, “Okay, we don't like American power, we want to diminish these guys in whatever way we can,” that's a threat to us, and we have to respond to that threat, or else just exceed to the loss of wealth and freedom that would come with China getting to do what it wants to us. I don't think we should exceed to that.
I don't want to whip up any international conflict in order to stimulate people to embrace progress. . . But, given that international conflict has found us. . . we should do what we did during the Cold War, during World War II, even during the Civil War, and use that problem to push progress forward.
Recalling Japanese economic history (20:32)
You write a lot about Japan. What is the thing you find that most people misunderstand about the last 30 years of Japanese economic history? I think the popular version is: Boom, in the ’80s, they looked like they were ahead in all these technologies, they had this huge property bubble, the economy slowed down, and they've been in a funk ever since — the lost decades. I think that might be the popular economic history. How accurate is that?
I would say that there was one lost decade, the ’90s, during which they had a very protracted slowdown, they ameliorated many of the effects of it, but they were very slow to get rid of the root cause of it, which was bad bank debts and a broken banking system. Eventually, they mostly cleaned it up in the 2000s, and then growth resumed. By the time per capita growth resumed, by the time productivity growth and all that resumed, Japan was aging very, very rapidly, more rapidly than any country has ever aged in the world, and that masked much of the increase in GDP per worker. So Japan was increasing its GDP per worker in the 2000s, but it was aging so fast that you couldn't really see it. It looked like another lost decade, but what was really happening is aging.
And now, with fertility falling all around the world right now in the wake of the pandemic, probably from some sort of effect of social media, smartphones, new technology, whatever, I don't know why, but fertility's falling everywhere — again, it looked like it had bottomed out, and then now it's falling again. We're all headed for what happened to Japan, and I think what people need to understand is that that's our future. What happened to Japan in the 2000s where they were able to increase productivity, but living standards stagnated because there were more and more old people to take care of. That is something that we need to expect to happen to us, because it is. And, of course, immigration can allay that somewhat, and it will, and it should. And so we're not because of immigration
Will it in this country? In this country, the United States, it seems like that should be something, a major advantage going forward, but it seems like it's an advantage we seem eager to throw away.
Well, I don't know about eager to throw away, but I think it is in danger. Obviously, dumb policies can wreck a country at any time. There's no country whose economy and whose progress cannot be wrecked by dumb policies. There's no country that's dumb-proof, it doesn't exist, and it can't exist. And so if we turn off immigration, we're in trouble. Maybe that's trouble that people are willing to accept if people buy the Trumpist idea that immigrants are polluting our culture, and bringing all kinds of social ills, and eating the pets, and whatever the hell, if people buy that and they elect Trump and Trump cracks down hard on immigration, it will be a massive own-goal from America. It will be a self-inflicted wound, and I really hope that doesn't happen, but it could happen. It could happen to the best of us.
There's no country whose economy and whose progress cannot be wrecked by dumb policies. There's no country that's dumb-proof, it doesn't exist, and it can't exist.
Regulating AI well (23:49)
Do you think what we're seeing now with AI, do you think it is an important enough technology that it is almost impossible, realistically, to screw it up through a bad regulation, through a regulatory bill in California, or something on the national level? When you look at what's going on, that if it's really as important as what perhaps the most bullish technologists think it is, it's going to happen, it's going to change businesses, it's going to change our lives, and unless you somehow try to prohibit the entire use of the technology, there's going to be an Age of AI?
Do people like me worry too much about regulation?
I can't say, actually. This is not something I'm really an expert on, the potential impact of regulation on AI. I would never underestimate the Europeans' ability to block new technologies from being used, they seem to be very, very good at it, but I don't think we'll completely block it, it could hamper it. I would say that this is just one that I don't know.
But I will say, I do think what's going to happen is that AI capabilities will outrun use cases for AI, and there will be a bust relatively soon, where people find out that they built so many data centers that, temporarily, no one needs them because people haven't figured out what to do with AI that's worth paying a lot of money for. And I have thoughts on why people haven't thought of those things yet, but I'll get to that in a second. But I think that eventually you'll have one of those Gartner Hype Cycles where eventually we figure out what to do with it, and then those data centers that we built at that time become useful. Like, “Oh, we have all these GPUs [graphics processing units] sitting around from that big bust a few years ago,” and then it starts accelerating again.
So I predict that that will happen, and I think that during the bust, people will say, just like they did after the Dot-com bust, people will say, “Oh, AI was a fake. It was all a mirage. It was all useless. Look at this wasted investment. The tech bros have lied to us. Where's your future now?” And it's just because excitement about capabilities outruns end-use cases, not all the time, obviously not every technology obeys this cycle, for sure . . . but then many do, you can see this happen a lot. You can see this happen with the internet. You can see this happen with railroads, and electricity. A lot of these things, you've seen this pattern. I think this will happen with AI. I think that there's going to be a bust and everyone's going to say, “AI sucks!” and then five, six years later, they'll say, “Oh, actually AI is pretty good,” when someone builds the Google of AI.
Rethinking growth strategy in the EU (26:46)
To me, this always gets a lot of good attention on social media, if you compare the US and Europe and you say, the US, it's richer, or we have all the technology companies, or we're leading in all the technology areas, and we can kind of gloat over Europe. But then I think, well, that's kind of bad. We should want Europe to be better, especially if you think we are engaged in this geopolitical competition with these authoritarian countries. We should want another big region of liberal democracy and market capitalism to be successful.
Can Europe turn it around? Mario Draghi just put out this big competitiveness report, things Europe can do, they need to be more like America in this way or that way. Can Europe become like a high-productivity region?
In general, European elites’ answer to all their problems is “more Europe,” more centralization, make Europe more like a country. . . But I think that Europe's strength is really in fragmentation . . .
I think it can. I wrote a post about this today, actually, about Mario Draghi's report. My bet for what Europe would have to do is actually very different than what the European elites think they have to do. In general, European elites’ answer to all their problems is “more Europe,” more centralization, make Europe more like a country. You know, Europe has a history of international competition. France, and Germany, and the UK, and all these powers would fight each other. That's their history. And for hundreds of years, it's very difficult to change that mindset, and Mario Draghi's report is written entirely in terms of competitiveness. And so I think the mindset now is “Okay, now there's these really big countries that we're competing with: America, China, whatever. We need to get bigger so we're a big country too.” And so the idea is to centralize so that Europe can be one big country competing with the other big countries.
But I think that Europe's strength is really in fragmentation, the way that some European countries experiment with different institutions, different policies. You've seen, for example, the Scandinavian countries, by and large, have very pro-business policies combined with very strong welfare states. That's a combination you don't see that in Italy, France, and Germany. In Italy, France, and Germany, you see policies that specifically restrict a lot of what business can do, who you can hire and fire, blah, blah, blah. Sweden, and Denmark, and Finland, and Norway make it very easy for businesses to do anything they want to do, and then they just redistribute. It's what we in America might even call “neoliberalism.”
Then they have very high taxes and they provide healthcare and blah, blah, and then they basically encourage businesses to do business-y things. And Sweden is more entrepreneurial than America. Sweden has more billionaires per capita, more unicorns per capita, more high-growth startups per capita than America does. And so many people fall into the lazy trap of thinking of this in terms of cultural essentialism: “The Swedes, they're just an entrepreneurial bunch of Vikings,” or something. But then I think you should look at those pro-business policies.
Europeans should use Sweden as a laboratory, use Denmark, use Norway. Look at these countries that are about as rich as the United States and have higher quality of life by some metrics. Look at these places and don't just assume that the Swedes have some magic sauce that nobody else has, that Italy and Greece and Spain have nothing to learn from Sweden and from Denmark. So I think Europe should use its fragmentation.
Also, individual countries in Europe can compete with their own local industrial policies. Draghi talks about the need to have a Europe-wide industrial policy to combat the industrial policies of China and America, but, often, when you see the most effective industrial policy regimes, they're often fragmented.
So for example, China until around 2006, didn't really have a national industrial policy at all. At the national level, all they did was basically Milton Friedman stuff, they just privatized and deregulated. That's what they did. And then all the industrial policy was at the provincial and city levels. They went all out to build infrastructure, to attract FDI [foreign direct investment], to train workers, all the kinds of things like that. They did all these industrial policies at the local level that were very effective, and they all competed with each other, because whichever provincial officials got the highest growth rate, you'd get promoted, and so they were competing with each other.
Now, obviously, you don't want to go for growth at the expense of anything else. Obviously you'd want to have things like the environment, and equality, and all those things, especially in Europe, it's a rich country, they don't just want to go for growth, growth, growth only. But if you did something like that where you gave the member states of the EU more latitude to do their local policies and to set their local regulations of things like the internet and AI, and then you use them as laboratories and copy and try to disseminate best practice, so that if Sweden figures something out, Greece can do it too, I think that would play to Europe's strength, because Draghi can write a million reports, but Europe is never going to become the “United States of Europe.” Its history and ethno-nationalism is too fragmented. You'll just break it apart if you try.
The European elites will just keep grousing, “We need more Europe! More Europe!” but they won't get it. They'll get marginally more, a little bit more. Instead, they should consider playing to Europe's natural strengths and using the interstate competitive effects, and also laboratory effects like policy experimentation, to create a new development strategy, something a little bit different than what they're thinking now. So that's my instinct of what they should do.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit fasterplease.substack.com/subscribe
📖 My chat (+transcript) with Mentava founder Niels Hoven on accelerating kids’ education
Thursday, September 12, 2024 • Duration 22:40
When it comes to sports, everybody is basically aligned that the goal here is helping every kid reach their potential. We celebrate talent, we give athletes the resources and personalized support they each need to develop their skills. We have varsity leagues, we have junior varsity leagues. We make sure that kids are challenged at the appropriate level for their current level abilities. And for some reason, when it comes to academics, we throw all of that out the window.
Our progress as a society depends a lot on the brilliant ideas of our greatest thinkers. To improve our way of life, we should be promoting our best and brightest to the highest heights of their potential. Instead, we seem to be stemming the flow of great minds at the source: in our public schools. With a one-size-fits-all, equality-of-outcome model, we rob our kids, and our society, of their potential.
Today on Faster, Please — The Podcast, I talk with Niels Hoven, founder and CEO of Mentava, an educational software company. Hoven’s goal: to help kids learn at their own pace, whether that includes additional support, or simply the resources to excel beyond expectations.
Hoven is the father of four, former product manager at Cloudflare, and was VP of product development at Pocket Gems.
In This Episode
* Treating academics like athletics (1:35)
* School as childcare and instruction (5:44)
* The role of parents (8:04)
* Mentava’s mission (10:04)
* Reframing the public school (15:20)
*
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
⚡ My chat (+transcript) with venture capitalist Katherine Boyle on 'American Dynamism'
Thursday, September 5, 2024 • Duration 27:01
American global leadership is due in great part to its innovators — visionaries who drive society beyond the preconceived limits. Historically, government-led initiatives like the Manhattan Project or the Apollo Project pushed boundaries. Today, too often, government lags behind technologically.
Today on Faster, Please! — The Podcast, I talk with Katherine Boyle about American Dynamism, the spirit of pro-progress innovation, and how a new generation of Silicon Valley startups is spurring government to break out of its old habits.
Boyle is a general partner at VC giant Andreessen Horowitz, having previously been a partner at General Catalyst and a general assignment reporter at The Washington Post. She primarily invests in national security, aerospace and defense, and public safety companies, among others.
In This Episode
* American Dynamism (1:25)
* From software to the physical world (7:23)
* Government collaboration: challenges & opportunities (11:29)
* Playing the long-game in Washington (21:16)
* Building the American Dream (24:35)
Below is a lightly edited transcript of our conversation
American Dynamism (1:25)
Let's just start with a little bit of definition about American Dynamism. Broadly, what challenges or problems is this effort directed toward?
It's a bit of a long story as to what American Dynamism is, how it arrived to be a category of innovation, but the short definition is American Dynamism is built for companies that support the national interest. So a very broad category of companies, everything from aerospace, defense, national security, companies that sell directly to the US government and to our allies, but also things like housing, education, transportation, infrastructure, things that are built in the physical world where Washington or states usually like to regulate those things.
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
⚠ My chat (+transcript) with BCG economist Philipp Carlsson-Szlezak on dealing with macroeconomic risk
Thursday, August 15, 2024 • Duration 29:29
In our highly globalized economy, exogenous shocks and unsettling headlines are everywhere. It makes sense that market forecasters should be biting their nails, but so often their prophecies of doom prove completely false. Philipp Carlsson-Szlezak is a proponent of “rational optimism.” He believes there’s a calmer, more measured way of going about financial and economic analysis that sets us up to be more flexible to the highs and lows of the economic events. Today onFaster, Please! — The Podcast, I talk with Carlsson-Szlezak about why an overreliance on models — and a tendency to assume the worst — can impair our ability to roll with unexpected events and make the best of them.
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
🤖🧠 My chat (+transcript) with Google DeepMind's Séb Krier on AGI and public policy
Friday, August 9, 2024 • Duration 23:33
In a world of Artificial General Intelligence, machines would be able to match, and even exceed, human cognitive abilities. AGI might still be science fiction, but Séb Krier sees this technology as not only possible, but inevitable. Today on Faster, Please! — The Podcast, I chat with Krier about how our public policy should facilitate AGI’s arrival and flourishing.
Below is a lightly edited transcript of our conversation
Pethokoukis: Séb, welcome to the podcast.
Krier: Thank you. Great to be here.
The AGI vision (1:24)
Let's start with a bit of context that may influence the rest of the conversation. What is the vision or image of the future regarding AI — you can define it as machine learning or generative AI — that excites you, that gets you going in the day, that you feel like you're part of something important? What is that vision?
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
⤵ My chat (+transcript) with investor Ruchir Sharma on where capitalist economies went wrong
Friday, July 19, 2024 • Duration 23:27
✈ A quick note: I will be traveling through the middle of the month and will be posting a bit less than usual and perhaps a bit shorter than usual.
These days, it seems that critics of capitalism are more prevalent and more vocal than ever. But Ruchir Sharma, author ofWhat Went Wrong with Capitalism, argues that the free market never let us down; our government did. Today onFaster, Please! — The Podcast, Sharma and I discuss the American addiction to “pain management” — unnecessary economic intervention aimed at dulling the effects of the natural ups and downs of a free market, and how it crippled American capitalism.
Below is a lightly edited transcript of our conversation
Disillusionment (1:20)
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
⚛ My chat (+transcript) with the US Energy Department's Mike Goff on powering the US with more nuclear energy
Thursday, July 11, 2024 • Duration 25:46
✈ A quick note: I will be traveling through the middle of the month and will be posting a bit less than usual and perhaps a bit shorter than usual.
After decades of resistance to nuclear power, growing concern over climate change, rising electricity needs, and a desire for greater energy independence are spurring renewed public interest in a future powered by atomic fission (perhaps fusion, too). Today onFaster, Please! — The Podcast, I talk to Dr. Mike Goff about the state of US nuclear power, the developing advancements in nuclear technology, and what it will take to reach our vast potential.
Goff is the acting assistant secretary and the principal deputy assistant secretary for the Department of Energy’s Office of Nuclear Energy. He previously spent over 30 years at Idaho National Laboratory, including a major advisory and management role. He has written over 70 publications on the nuclear fuel cycle.
In This Episode
* Atomic Age 2.0 (1:31)
* Major concerns (7:37)
* Out of practice (11:04)
* Next-generation policy (17:38)
* Human capital (21:48)
* Fusion forecast (23:12)
Below is a lightly edited transcript of our conversation
Atomic Age 2.0 (1:31)
The Energy Secretary recently spoke about adding a lot more nuclear capacity, tripling it, I think, by 2050 or so. And before we get into whether that's possible, I wanted to ask you: As you understand it, what is the current consensus explanation for why the Nuclear and Atomic Age of the ’50s and ’60s, why that kind of ended? Because when the secretary spoke about building more capacity, I thought about the — and this is something maybe a lot of people are unaware of, that
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
🤖 My chat (+transcript) with Google economist Guy Ben-Ishai on seizing the historic AI moment
Friday, June 28, 2024 • Duration 31:32
Artificial intelligence may revolutionize the American economy, but whether we see that potential actualized depends on a few key factors: whether generative AI is a general purpose technology, whether the labor force makes a smooth pivot, how employers prioritize their resources, and whether the US chooses to take the lead in AI’s deployment. These are just a few of the topics I cover on the podcast today with Guy Ben-Ishai.
Below is a lightly edited transcript of our conversation
Is GenAI a general purpose tech? (1:22)
Pethokoukis: Do you have any doubt that generative AI, and perhaps machine learning more broadly, is an important general purpose technology that will eventually make a substantial and measurable impact in the economic statistics and productivity and economic growth?
Ben-Ishai; The immediate response is absolutely, but let me unpack that: Do I have doubts about the immense potential of the technology? No, and I'm saying that very confidently, which is uncommon for an economist. We put together the paper that you've initially cited at Google to look exactly at that question: When we say that AI marks a pivotal moment in human history, what does that actually mean for an economist? And I think the conclusion there that we're looking not in an ordinary technology, but rather at a general purpose technology, that is immense. That means that we're probably looking at the most transformative economic development of our generation. And to think, Jim, that the two of us are having a conversation about that today, that is historic. I feel incredibly privileged and lucky to think and work of these issues in our day and age.
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
🏙 My chat (+transcript) with ... economist Jason Barr on the skyscrapers of tomorrow
Friday, June 14, 2024 • Duration 29:43
The image of the skyscraper is the hallmark of the modern city. Futuristic depictions of urban landscapes nearly always feature towering structures high above the clouds. Today, however, developing countries seem to be putting the greatest effort into building the most impressive skyscrapers, from the Burj Khalifa in the UAE, to the future Jeddah Tower in Saudi Arabia. Whether you love them or hate them, it’s worth asking why we build skyscrapers and what their role will be in future cities. Today on Faster, Please! — The Podcast, I sit down with Jason Barr, author ofCities in the Sky: The Quest to Build the World’s Tallest Skyscrapers.
Below is a lightly edited transcript of our conversation
The image of the skyscraper is the hallmark of the modern city. Futuristic depictions of urban landscapes nearly always feature towering structures high above the clouds. Today, however, developing countries seem to be putting the greatest effort into building the most impressive skyscrapers, from the in the UAE, to the future in Saudi Arabia. Whether you love them or hate them, it’s worth asking why we build skyscrapers and what their role will be in future cities. Today on , I sit down with, author of.
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
🤖🌈 My chat (+transcript) with Nick Bostrom on life in an AI utopia
Thursday, June 6, 2024 • Duration 30:17
The media is full of dystopian depictions of artificial intelligence, such asThe Terminator andThe Matrix, yet few have dared to dream up the image of an AI utopia. Nick Bostrom’s most recent book,Deep Utopia: Life and Meaning in a Solved World attempts to do exactly that. Bostrom explores what it would mean to live in a post-work world, where human labor is vastly outperformed by AI, or even made obsolete. When all of our problems have been solved in an AI utopia . . . well, what’s next for us humans?
Below is a lightly edited transcript of our conversation
This is a public episode. If you’d like to discuss this with other subscribers or get access to bonus episodes, visit
The San Francisco algebra ban (17:50)
* Investing in our future (20:05)
Below is a lightly edited transcript of our conversation
Pethokoukis: Niels, welcome to the podcast.
Hoven: Thank you so much. I appreciate you having me here.
Treating academics like athletics (1:35)
You argue that the current American education system is fundamentally flawed.
I do think it has some issues.
How does closing achievement gaps hurt our education system? How does it hold students back?
So obviously my problem is not with closing achievement gaps, my problem is what happens when you set up policies with that as the only goal. I think what we've seen is that the goal of today's modern education policy is closing the gaps between high achievers and low achievers, which is, of course, wonderful, but the way that has actually manifested in schools is by slowing down high achievers and not giving them the opportunity to achieve their potential. In San Francisco, you're literally not allowed to teach material above grade level, which I think is crazy.
Most school systems have gifted programs. Doesn't that meet your concern?
So those gifted programs, I think they don't go far enough to support the learning needs of students who are really capable of achieving dramatically more, and, in a lot of places they're very, very hard to get into. So in our school district right now, in order to qualify for the gifted program, you have to take a series of tests and you basically have to score 99th percentile on all of those tests. All of those tests are basically grade-level tests, so they're not really teaching seeing how far above grade level you are, so it's really, “Are you really, really good at taking the tests, so well that three times in a row you can score 99th percentile on grade level stuff?” That's not really getting the kids who need their learning needs supported by these special programs, and these programs really only operate a single grade above grade level. What about the kids who could be doing calculus in middle school, or want to be moving much faster than that: Two years of math a year, every single year — we aren't supporting them.
You've proposed treating academics more like sports. What does that look like in practice and how might that change how we approach education and how we think about education more broadly?
When it comes to sports, everybody is basically aligned that the goal here is helping every kid reach their potential. We celebrate talent, we give athletes the resources and personalized support they each need to develop their skills. We have varsity leagues, we have junior varsity leagues. We make sure that kids are challenged at the appropriate level for their current level abilities.
And for some reason, when it comes to academics, we throw all of that out the window. We just say, “Okay, everybody must progress at the same speed, learn the same thing at the same time.” To me that's like saying, “Okay LeBron, you are not allowed to dunk until everybody else can dunk also.” And so I want to see us treat academics more like sports, where we encourage students to pursue their interests, to develop their talents to the fullest potential, and respect the diversity of kids' ability and motivations.
To what do you attribute the staying power of this — I don’t know if it's a one-size-fits-all system, but of a system that, in many key ways, isn't different than it was a hundred years ago?
It is a government-sponsored monopoly, so I guess that would be my answer. How did the taxi cab medallion system last so long, even though it was dramatically underserving everybody who wanted to take a taxi? There's no competition.
What does that more sports-like environment look like? It sounds like there'd be more freedom, there'd be less regimentation. What does that world look like?
What I'm really pushing for is I would like to see students receiving instruction appropriate for their current level. I talk a lot about high-achieving students, but this is also true for struggling students. Right now we have a very one-size-fits-all model of education, and that means students who are struggling and need extra attention to get caught up aren’t given the opportunity that they need to perhaps move at a slower pace or get extra support, and kids who want to be moving faster and maybe learning two years of math a year, every single year, so that they can be doing college-level math in middle school, they're also not getting that support. We managed to do that in sports, we have lots of different leagues so that kids can find the level of competition that is appropriate for them, but for some reason, when it comes to academics, we refuse to allow that amount of differentiation.
School as childcare and instruction (5:44)
You advocate reducing instruction time to two hours a day. One, is that enough? And two, what are the kids doing for the other . . . are they getting into mischief? What are they doing for the rest of the day if they're not studying?
I think we've really conflated the role of school, and I think an important question to ask is: Is school as we provide it now, is it childcare or is it academics? And I think it is both. An interesting fact about school is, despite all of the problems that we all understand our schools have, schools have like an 80 percent approval rating from parents, and that's because the job that schools do for most parents is actually childcare. It is free childcare for while the parents are at work, it is finding a place where your children are entertained and loved, and that is super important.
But somehow we have also layered this layer of academic theater on top of that childcare instead of saying, “Okay, these kids can play in the woods for eight hours a day, or they can play dodgeball or grow their social-emotional skills and build their friendships with a friend.” We had to say, “No, they have to be learning something – but not too fast at this very, very slow pace.” And if you look at things like homeschoolers, you see most homeschoolers do two hours of academics a day, and they have the same outcomes as kids who are going to public schools, so we really don't need that much more time doing academics as long as that time is being spent efficiently.
Is this new world possible within a mostly public school system as it exists today? Can you do this, or are you talking about private school, homeschooling, but does this have anything to do with the public school system, which seems to me fairly resilient? Certainly, I think the changes of the sort of magnitude you're talking about.
I like the public school system. I went to public school, I had a really positive experience in public school. My own kids go to public school. And I think the difference is that when I was in public school, people were much more accepting of the idea of kids who wanted to move at their own pace. And so Mentava, certainly we're happy to support kids who are homeschoolers, who are in private school, but the real vision is to allow kids to be part of, essentially, their local public school community, go to school with friends from the neighborhood, but still have the opportunity to progress at their own pace
The role of parents (8:04)
Tell me a little bit about your personal educational experience and how that shaped your views and how it eventually led to your company.
Education has always been very important to my family. My dad taught me to read early, when I arrived at kindergarten, I could already read, I was roughly a year ahead in math. And so he negotiated with my school to just let me, during math class, just for an hour a day, could I just go to the next grade up and sit in on their math class and then come back to my own class for the rest of the day. And we did that, and that worked great until third grade, because my school only went up to third grade, so there wasn't a class for me. So at that point, I just started doing independent study. Just during math class, for an hour a day, I would go to the back of the classroom, I would study out of a math book, and then at the end of that hour I would come back and rejoin my friends for the rest of the day.
And I did that for the next four years, and basically, thanks to that accelerated support, I ended up taking calculus in eighth grade. There are kids who can be moving that fast if you just kind of get out of their way. My own son — he goes to a public school — we also got permission for him to do independent study last year, and now in fourth grade he'll probably be ready to start pre-algebra.
This is doable now. This was doable when I was a kid with textbooks, this is doable now with off-the-shelf software, but it's harder than it needs to be. And so our vision is: We can make this easier. I think a lot of kids could have done what I did, but they weren't given the opportunity. We want to make sure that more kids have this opportunity to have their learning needs supported.
Do you think parents underestimate what their kids are capable of doing?
Parents have no idea what their kids are capable of doing, especially parents of high-achieving kids. We've seen this over and over again with the families who are entering in Mentava’s learn-to-read software now. We target our software at kids as young as two, but often age three and four, we're trying to teach them to read, trying to get them to about a second grade reading level in maybe six to 12 months. We just had a three-year-old complete our entire curriculum, which gets us close to a second grade reading level, in about six months. So it is doable, it can move fast, and we have parents who say, “I had no idea that my kid was capable of doing this at this point!”
Mentava’s mission (10:04)
So walk me through what your company does, the service it provides, how it all works.
The long-term vision for our company is to support the learning needs of kids who are not being supported in school. If you have a child who wants to learn two years of math in a year, the real gating factor of that is, a lot of times it's teacher availability, or it's school policy that says there's no one available to give them that instruction. But imagine that they had the opportunity to just go open a math book.
It's a resource issue. We'd love to do it, but we don't have the resources.
We don’t have the resources. Sometimes that's true, sometimes that's not true, sometimes it's policy, but whatever. But they could go get a math book, they could just study that book and go as fast as they wanted — but that's boring. Not every kid is going to have that motivation. And so, to some extent, we're not really solving for curriculum, we are solving for motivation. We want to build software that can deliver that same curriculum — we know how to teach math, we know how to teach reading — deliver it in a more sort of fun, entertaining, motivating way, and allow kids to essentially continue to progress at their own pace without being gate-kept by the availability of teachers to essentially unlock that knowledge for them. And so we are starting at a very young age by teaching kids to read with software.
What I really want to teach is math. I want to get kids to learn math as fast as possible, but in order for kids to be able to teach themselves math, they have to be able to read, and so that is our first piece of software: learn-to-read software for preschoolers.
And obviously preschoolers, these are young kids, so is your expectation that software will be done at home? Are there schools trying to incorporate in some way? How's that working?
We've started talking to schools about pilots, but I think, right now, we get a lot of attention from parents. Incentives are just better aligned that way. Schools right now are not particularly concerned with, “Are we supporting our kids achieving their fullest potential or are we ensuring our kids can learn as fast as possible?” But parents really care about that. And so right now we have a lot of customers who are basically parents at home who realize, “Oh, my three- or four-year-old is ready to start reading, what can I do to best support them now?”
How long has the company been in business?
We kind of accidentally launched about six months ago.
Was this a pandemic-related idea?
This was. I have four kids. I had three kids during the pandemic and the fourth one arrived during the pandemic. They were at home, doing school at home, and I also had a job at the time, so did my wife, we had two working parents trying to take care of three kids at home, we were trying to figure out how to help them learn, and really the only way to make that work was to give them the skills that they needed to teach themselves. And at that time, my kids were five and three. And so how can I get my five-year-old teaching themselves math? How do I get my three-year-old teaching themselves to read? And the solution to that is software. We know the curriculum, we know if you want to teach reading, it's phonics, but how do you get the kid to sit down and memorize the 44 sounds in the English language? Well, turns out that software and games are really, really good at solving motivation, so we just needed to package that all together and that was how Mentava was started.
So during this exact period that you've thought of this idea, you're putting together a company, putting together the software, we have sort of a new stage in software happening with chatbots and large language models. Are those technologies that compliment what you're doing? Are you going to have to do something different to use those technologies? How's that going to work out for you?
It's very complimentary. So we're not using AI right now, but we see it coming. There is kind of this perfect storm of timing right now where, I think because of Covid, parents started to realize that, “Oh, my kid is not learning as much in school as I thought. This is what they're doing in school?” We had all that visibility when our kids were doing homeschool in front of screens at home.
Technology has gotten to a point where we can give screens to every kid, and iPads, and other tablets. Touchscreens make learning much more accessible. We're seeing the effectiveness of some learning software — a lot of learning software is really, really bad, but some of it is good, and people are seeing that. And then, at the same time, like you see AI coming out and getting people very excited about the potential of software to affect education.
It’s funny, when we were raising money, the idea that software could be a teacher was a very contrarian perspective. Everybody said, “How could software possibly be a teacher? You’re going to need a human there.” And then about 12 months later, AI came out, people said, “Oh, of course you're going to have software teachers. We've always believed that.”
But my take on AI is that the power of AI is really in its adaptability, and you actually don't need that much adaptation for teaching reading or teaching math. You memorized the 26 letters, the 44 sounds in the English language, you learn addition, and then you learn subtraction, then you learn multiplication, then you learn division. It's pretty linear. It's pretty sequential. And so my belief is that there's actually this core learning pathway that you can really, really optimize, and we should focus on that. And it's fairly sequential, and it's fairly deterministic. And then the power of AI is to catch the kids who fall off of that and get confused and ask, “Okay, what are you confused about? I see you're confused about this thing. Let me give you some custom instruction and then get you back on that main pathway.”
Reframing the public school (15:20)
In an ideal world — and let's just stick with, I think it's reasonable to think that, for the time being, most kids are going to be educated by public schools. That's a lot of kids, a magnitude difference in how many kids are in private school or are homeschooled. What should that public school day look like, ideally, given what you've learned going through this process?
The biggest challenge for public school is that there's such a diversity of student needs there. Public schools are simultaneously academics, but they're also childcare, and they're also a social support network. They're a safety net for a lot of kids, and they're trying to provide all those services to all these different kids by giving them all the exact same thing. To me that makes no sense, and what I would really love to see in our public schools is just more differentiation, more acceptance of diversity of needs, diversity of motivations, diversity of abilities, and saying, “Okay, these children need this particular service from our public schools. Let's make sure that they're in a place where they can get those services. But we have these other kids who want to learn two years of math every year. They can do that in two hours a day, and then they want to spend the rest of the day playing in the forest.” That would be amazing.
Should that actual classroom time look markedly different? I'm sure that if I went into most classrooms — I had kids, one currently in high school and ones who were in high school not that long ago — that those classrooms, blackboards, teachers, lecturing: That's the classroom experience. That's in 2024, that was the classroom experience in 1924. Should that classroom experience look fundamentally different?
I think that's an interesting question. I think it's going to look different for different kids. I think there is a sense that some of the rigor I would say of the old days has been lost, and I think that there's good and bad to that. I think a lot of that is a result of conflating childcare with academics. You can't do rigorous academics for eight hours a day. It’s sort of like weightlifting; you can't do squats for eight hours a day, but you can do them very effectively for half an hour. But if you want to pretend that you're exercising for eight hours a day, then you, by definition, have to remove a lot of that rigor. So I would like to get rid of the academic theater and be very clear about, “Okay, this time is play time, this is childcare time, and this time is academic time, and we're really going to buckle down and focus here.”
The San Francisco algebra ban (17:50)
A few years back, there was a ban on teaching middle school algebra in San Francisco. Can you give me some background on that?
So this was passed about 10 years ago. The way it used to work is that most kids took algebra in eighth grade. If you were ready earlier, you could take algebra in seventh grade, but essentially in San Francisco, because some kids were not prepared to take algebra in eighth grade, they said across the board, all kids must take algebra in ninth grade. So even the kids who used to take it in seventh grade, the kids who used to take it in eighth grade, “Sorry, we're not doing it in middle school anymore. You all have to take it in ninth grade.”
Usually these sorts of educational decisions, they're just lost in the noise, parents don't have time to focus on the nitty and gritty of curriculum, but this was a big problem for parents because this meant that you could not get through calculus in high school without essentially taking summer school or getting private tutoring. And for a lot of competitive colleges, you need to be in calculus, have taken calculus, in high school. And so parents had a real problem with this particular curriculum change.
The irony of all of this is that this was enacted with the hope of increasing equity, of driving more equal outcomes, and it had the opposite effect because now it’s just the parents with the resources who are able to go out and do summer school, and private tutors, and then get their kids the math support that they needed. So this happened a decade ago, and it has been a battle for 10 years to get algebra back into middle schools in San Francisco, and it actually went the other way: California statewide nearly got rid of algebra in their statewide middle school curriculum because of the quote-unquote “success of San Francisco,” which is basically, if you look into it, it's just San Francisco cooking the books, literally lying about their outcomes.
And so finally, this past year, parents essentially had enough and they put it on the ballot and said, “We're going to take a vote about, should middle schoolers be allowed to learn algebra.” It's funny because a lot of times people think that this cuts along party lines: Conservatives versus Democrats, red versus blue, but even in San Francisco, the most progressive city in the United States, 80 percent of families were like, “Yes, we should support all kids learn these. Yes, if a kid is ready to take algebra in middle school, we should allow them to do that.”
Investing in our future (20:05)
I'll tell you the one thing I kept thinking of as I was learning more about your company and your outlook was it seems to me like it'd be really important, as a country, that every kid can reach their potential, but especially the very smartest kids, that we get everything out of them that we can, right? That's pretty important. These are people who are going to be designing the next stage of AI, they're going to be designing the new computer chips, they're going to be in biotechnology. If we can get more out of those kids, there's a huge multiplier there.
I believe that very deeply. I believe leaders are important. I believe in the power of single individuals to create huge amounts of change, but not everybody agrees with that. I was at a school the other day, and a bench outside the school literally has carved into it —a bench that every student sees as they go into the school — it says, “Strong people do not need strong leaders.” And I fundamentally disagree with that. I think we need strong people and we also need strong leaders, and the way we get both of those is ensuring that every student has the opportunity to have their learning needs supported and have the opportunity to achieve their potential.
What's the direction of the company? Where are you going to be in five years? What's the dream?
Right now, we are in the process of officially launching our learn-to-read app, targeted at preschoolers, and then what I really want to do is start transitioning into math. So once we have taught kids to read, we have essentially unlocked their ability to teach themselves. And so our goal is to keep up with this earliest cohort of kids who are learning to read and support them as they continue through their K–12 career. If they want to learn two years of math a year, then I would love to build two years of math curriculum each year so that they continue using Mentava to support their K–12 experience, and then they discovered that, I don't know, they're done with math in middle school, and then they get to figure out what's next after that. Do I go start a company? Do I do internships? Do I go learn marine biology? I don't know.
What about computer science? Does that play a role here?
When I say math, I would say specifically math and computer science are what I'm most passionate about. I think of it almost as vocational school. Those are the skills that we can teach that directly contribute to, okay, this person is able to create more value in the world because they know these two fundamental skills now.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
So one of the things that we saw in our own portfolio is that there are a lot of companies that we used to be classifying as “enterprise” or “consumer,” and really what they were were government companies because they had to interface with a regulator much earlier in their trajectory, or they saw government as a potential buyer of the product. So in cases of things like aerospace and defense, those are very obvious government buyers, but things like public safety, where we have companies like Flock Safety, for example, that started out selling to homeowners associations thinking they were a consumer company, but ultimately got extraordinary pull from local governments and from public safety officials because of how good the technology was. So the companies, in some ways, they were these N-of-One companies, really solving really important civic problems, but over time it became very clear that this was a growing category of technology.
But the broader underlying thesis, I'd say, of where the movement came from, and when we really started seeing this as an area where founders, in particular, were excited to build, I think it did come out of “It’s Time to Build,” my partner Mark Andreessen's canonical post where he basically said during Covid that we have to be able to build things in the physical world. And there was sort of this realization that technology has solved many, many problems in the digital realm that I think, in some ways, the last 15 years of the Silicon Valley technology story has really been about changes in consumer technology or changes in the workplace, but now we're finally seeing the need for changes in government and civic goods, and there's just an extraordinary amount of momentum from young founders who really want to build for their country, build for the needs of the citizenry.
Does it change what you do, or maybe the kinds of expertise that are needed, to think about these things as a category, rather than different companies scattered in these other kinds of categories. Thinking of them as like, “Oh, there's some sort of commonality,” how is that helpful for you?
The thing that's interesting is that there's sort of a “yes” and “no” part to that question. The yes is that the founders are coming from different places. So the companies that have led to this sort of, I would say, extraordinary wealth of engineering talent where people are not afraid to tackle these problems, there's a handful of the companies that have scaled: it's companies like SpaceX, companies like Palantir, where, 20 years ago, they were banging their heads against the wall trying to figure out, “How do we sell to government?” In many cases, they had to sue the government in order to be able to sell and compete against the larger incumbents that have been around for, in many cases, 50, 70 years. But now you have these talented engineers who've sort of seen those playbooks, both in terms of, they understand what good engineering looks like, they understand the pace of innovation, how quickly you have to bring new products to market, and they also understand that you have to be in touch with your customer, constantly iterating.
And so you now have companies that have scaled in these categories where there is this nice thing that happens in Silicon Valley, and I always say it's a mark of a really successful company when three, four, five years into the journey, you start seeing the early people at that company say, “Well, I want to solve this problem,” or “I want to go be a founder, myself,” and they start building more companies. So I think that, in some ways, the natural order of how Silicon Valley progresses, in terms of, do you need to have different expertise, or are there different talent pools? Yes, they're coming from different companies, but it's the same story of Silicon Valley Dynamism, which is, someone comes in, I always joke, they go to the University of Elon Musk and they learn how to manufacture, and then they say, “Well, actually, I don't want to just work on rockets anymore, I'd like to work on nuclear.” And so then you have companies like Radiant Nuclear that have spun out of SpaceX several years ago that are building in a totally different category for the built world, but have that sort of manufacturing expertise, that engineering expertise, and also know what it's like to work in a highly regulated environment.
Does it require a different expertise, then, to advise these companies because of that government interface?
I think in some ways it does, yes, the types of people who are investing in this category, maybe there's a number of investors where they got their start at Palantir, for example, or they understand the early journey of SpaceX. But at the same time, the thing that I think has been most surprising to us is just how quickly this movement caught on among the broader Silicon Valley ecosystem. And I think that's a very good thing, because, at its core, these are software companies in many cases. Yes, they're building hardware, but software is the lever that's allowing these companies to scale. So you are seeing the traditional venture capital firms that used to say, “Oh, I would never touch anything that is operating in the physical world,” or the meme you had five years ago, which is, “You'll never be able to sell anything to the US government, I'm not wasting my money there.” You've seen a complete 180 in the Silicon Valley ecosystem in terms of venture firms where they're now willing to take bets on these types of companies.
And you're also seeing, there's a number of founders where their first company, for example, might've done very well, and it might've just been pure software, or in a consumer enterprise, sort of a more classical Silicon Valley domain, and now you're seeing those founders say, “Actually, I want to build for the civic need. I want to build for the national interests. These are issues I care about.” And so you're seeing those founders actually decide to build in the category and team up with founders who maybe have a little bit more experience in government, or maybe have a little more experience in terms of how they're building in the physical world.
From software to the physical world (7:23)
That period you referred to, which seems like a lot of what Silicon Valley was doing with the first 15 years or so of this century: they're doing internet, social media, very consumer-facing. How valuable was that period? Because that is a period that, here in Washington, is much criticized as trivial, “Why wasn't Silicon Valley solving these huge problems like we did in the ’60s?” Again, there were some critics who just looked at it as a waste of brainpower. To what extent is that a fair criticism, and do you think, is that unfair? That stuff was valuable, people valued the kinds of products that were produced
You would actually be better able to speak to this than me, but I'll say, the graph or the chart that's going viral today, as we speak, is the comparison of 2009 US GDP versus Eurozone GDP, which were roughly equal in 2009, coming right off the Great Recession, to today, which I actually think it was tweeted something like, I think it's. . . the US is 77 percent greater in terms of GDP than the Eurozone countries, which means that, for some reason, the Silicon Valley ecosystem — and it is largely attributed to Silicon Valley. When I first wrote the thesis on American Dynamism, I looked actually at 1996, because it was 25 years when I published it, but 1996, if you looked at the top US companies by market cap, all six of them were outside of technology as an ecosystem. They were energy companies, I would say almost archaic industries that had grown over a long period of time, but if you look at those six companies today, they are all tech companies. And so something has happened in the 21st century. You could say the new American Century is actually built off the back of software. It's built off of these large tech companies that were built in California, in many cases. And so the 15-year period that you're talking about, which is this sort of, it was a zero-interest-rate environment, cost of capital was very low, there was a lot of experimentation going on, it was, in many ways, the canonical example of American Dynamism broadly, that you had risk capital going after many new ideas in many different areas, but they were particularly really focused on the areas that government was not interested in regulating.
And that's always been the theme of innovation in Silicon Valley is, “Let's go where they're not necessarily paying attention.” Maybe you had some one-offs in terms of, you'll always have to meet with your regulator at some point—in the case of Uber or Airbnb—but these companies were really born of the needs that founders understood. They were built off the back of a platform shift in terms of, 2007, 2008, the iPhone becomes the thing that everyone wants to build on, it becomes the mobile era. And so you really did have this focus of software, and enterprise software, and consumer, and companies were able to grow to extraordinary heights. And if you just look at what it has done for US GDP in comparison to even something like Europe, it is really extraordinary. So that is a story that I think we should be celebrating and telling.
But what has happened, I think, since Covid is this new shift, which is, we've explored many of the digital frontiers that we can. And of course there's always a new digital frontier. Every time we think it's over, we get hit with a new one — in the case of AI. But the thing that I think has really changed is that entrepreneurs now are not afraid of the physical world, and they are realizing — and I hate to use the word “inevitable,” but in some ways this is an inevitability — that you are going to have to interface with government at a certain point if you are going to build in the physical world. And there are so many opportunities, there are so many different places where founders can build, that that really did take on new meeting post- this slew of black swan events, in the case of Covid, and then of course Russia invaded Ukraine, where I think it did wake up a lot of founders who said, “I want to work on these really hard problems.” And thankfully we have companies that have scaled during that time, that have trained these manufacturing capabilities, they've trained engineers how to do these things. So it is our view that that 15-year period was extraordinary for software, but the next 10 years are going to be extraordinary for these American Dynamism companies, as well.
Government collaboration: challenges & opportunities (11:29)
When you talk about interfacing with government, what popped into my head was a bit of video of a congressional hearing, and they were trying to decide, do we want to bring the private sector and SpaceX into the space program, and so not just have it be a government effort by NASA? And I just remember these senators just lambasting the idea. And I think they might've brought in some astronauts, too. And if I was interested in interfacing with government and I had seen that video, I'm like, “Boy, oh boy, I hope the attitude of government has changed since then, because it seems like that's a wall.” What is the attitude on the other side? You said the attitude of the entrepreneurs has changed, of funders, but what about on the other side? What is the openness to the kinds of solutions that your companies are presenting?
I think it's changed because it has to, and I always point to the late former Secretary of Defense Ash Carter, who in 2015 started DIUx [Defense Innovation Unit Experimental] as an innovation unit for the DoD [Department of Defense], recognizing that there's a talent problem that US government has had; and it didn't start in the last 10 years. When I was exploring this talent problem, I actually realized that there had been a commission on the lack of talent going into the bureaucracy at the federal level.
In the 1990s, Paul Volcker actually chaired the commission, it was called a “Quiet Crisis.” Basically that young people today, unlike in the ’50s and the ’60s where government was seen as this extraordinary job that you could have or that you could go into one of these companies and work in a company for 30 years and then draw a pension, that young people today want to go work in the private sector, and the growth and dynamism of the private sector has actually been an issue for government, and that is not a new issue. It was explored in the late ’80s, early ’90s, and it has gotten precipitously worse because of tech. I would say that the technological innovation risk capital going to Silicon Valley and saying, “We're going to fund young people as they come out of college because they understand this new type of engineering.” You're seeing some of the best and brightest young people decide not to go to traditional companies, which has been a huge issue for the prime contractors that supply 40 percent of the government programs for the DoD, but what has happened is you're seeing this extraordinary engineering talent go to startups. And so I think what even — this is 10 years ago now — DIU saw, if all of our best and brightest software talent is avoiding government, or much of it is avoiding government, they're avoiding traditional companies that we work with, we have to meet them where they are.
And so the DoD I actually think was the first government organization to really recognize this crisis and to decide, we're going to need to have new interfaces. Now, whether that means new procurement, that's always going to be a debate, and that's a Washington issue that I think we've been fighting for several years now in order to change how these companies are able to work with the DoD on these big programs, but I think even just recognizing that this was an issue 10 years ago was a huge step for government.
And over time now, we've seen a handful of what we would call “defense 1.0” companies, in terms of startups, many of them kind of built off the back of a company like SpaceX, now realizing that you can build for USG [US Government], you can build hardware-software hybrid that you can then sell into production contracts, and it's companies like Anduril that were started in 2017 when people said, “This is impossible to do. You're never going to be able to sell to USG,” and this year was chosen for a massive program, the CCA [Collaborative Combat Aircraft] program with the Air Force, over many prime contractors.
And so I think that is the story now that Silicon Valley has seen, and I always joke that, particularly investors and founders, they really only need to see a handful of winners to know that something is a category, and so you're going to see more and more of these companies being founded, scaling, and I think that circuitous cycle and that virtuous cycle actually leads more to the DoD saying, “Okay, this is a real ecosystem now.” It's not as risky to take a chance on a startup, which is what government is always worried about, if we take a chance on anything innovative, are we going to look foolish? And so I think, in some ways, you are seeing the government respond to what's happened in the private sector, but this is not something that's a year old, two years old, or three years old, this is something that's been talked about for almost 10 years now, and of course SpaceX now is an over-20-year-old company.
Is this still primarily a Defense Department-focused effort? Are there other areas of government who are looking at what's happening with DoD and they're drawing lessons? How diverse of an effort has this become?
We see this across every sector that government cares about. So it's not just defense, it's aerospace, it's energy, it's logistics, it's transportation. We always joke, if there is a department in Washington that exists to regulate a sector, that is American Dynamism, and you are seeing innovation in those sectors. But it's happening at different rates. I'd say the DoD is one of the largest spenders. The largest private US company right now is SpaceX, so there's success in those categories, so you're seeing a lot of interest in it now, but then there's companies in public safety. That's an area where I think there's just been an extraordinary explosion of innovation in the last few years, largely driven by the fact that there is a labor crisis happening in public safety across America, but it is a different sale, it's not selling to federal government, selling to state and local.
One of our companies, Flock Safety, which I mentioned at the beginning of our chat, they now are involved in solving 10 percent of vehicular crimes in America.
What do they do?
So, it's a great story about a company that was founded in Atlanta in 2017, and they built a very small modular license plate reader that only tracks cars, not people, and started building for homeowners associations with the recognition that most crimes in America are committed with a car, and so if you can put these in areas of high traffic, areas to augment the work of law enforcement, crime will go down. And they started selling to homeowners associations and immediately got pull around Atlanta and suburban Atlanta from police chiefs who said, “I need 10 of these in areas where we don't have enough people who can look at different areas.”
So now this company is operating across America, they're in all 50 states, and what's extraordinary about what they've been able to do as a technology company, just putting up cameras in different sectors and following cars, is one of the hardest problems for law enforcement is when a car that has committed at theft or — one of the most extraordinary stories they've told us recently was there was a young girl kidnapped, a young child kidnapped in Atlanta, and the car went into a different county. And so when that happens, for law enforcement it's often one of the most difficult things, if a car goes into a different county, to do data sharing across these places. But if you have a network of cameras that can track the car, you find that kidnapped child, or you find that stolen vehicle much, much faster. In many ways, catching the cars at the moment where they've moved from county to county has actually solved one of the bigger data issues that law enforcement has.
What's interesting about this example — and it provides a nice lead-in to my next question — is, in that situation, the solution wasn't to help the various databases communicate better, it was a completely different sort of solution. So, are what these companies doing — it seems like what they're not doing is taking existing operations and improving efficiency, but providing a new way to approach the problem that they're trying to tackle.
Yes. And what's incredible about that story is it was not started as a company that was supposed to support law enforcement. It was started for homeowners associations, it was a consumerization of a civic problem. And I think that's what's really interesting is, one of the biggest issues, and this is why I think you're now seeing really interesting technology companies enter government at all levels is, you have a population that has grown up with consumer technology now. So as the boomers retire, the boomers remember what it was like to be in government, or to be in office places without Zoom, without the consumer internet, and without the things that make life much easier and tangible, as those people retire, you have young people demanding, “We have to use better technology.”
And so the solutions are not, “Okay, let's iterate on the existing systems that we've used for the last 10 or 20 years,” it's, “Why can't my experience when I walk into my job in government feel exactly the same way that it does when I walk into my home and I experience the consumerization of everything around me?”
So I think that is part of it, that you have this millennial generation that's now coming into leadership. In many cases, you have people who don't necessarily remember the world before the internet or didn't have formative experiences in the workplace or in government before the internet. And that is shaping and reshaping all of how government functions, and likely will for the next 20 years. The thing that, especially when we talk about the Department of Defense and the warfighter, the thing that has always been tragic is that you have more technology in your phone than you do when you go onto the battlefield. And so I think there's this understanding that young people are demanding to have the same level of technology and the same ease-of-use in all aspects of governance, in all aspects of civic goods.
Playing the long-game in Washington (21:16)
You seem like a very upbeat, positive person. My experience as people from Silicon Valley — or now, in your case, from Miami, a new startup hotbed — they come through Washington, they bring that optimism with them, then after a few days of dealing with people on Capitol Hill, the optimism is drained out of them, they go back shells of their former selves, because if you've dealt with a lot of people on Capitol Hill and staffers, what they're really good at saying is, “That will never pass . . . that will take 20 years . . . three of my predecessors worked on it, it didn't work . . .” How have you been able to maintain a fairly upbeat attitude, given that this is the world that your companies have to deal with?
I agree with you that one of the biggest problems that we see, and which we joke about, is that the only reason why people in Silicon Valley 10 years ago were going to Washington was to apologize for the things that they did. They would get hauled in front of Congress, say they're sorry, and so I think what we've seen in these sectors, in particular, is it's a specific type of founder and person who knows that this is very mission-driven. They are called to build these companies. They care about these companies. They're passionate about the national interest. And so they know they have to go to Washington repeatedly, and I think some of the mistakes that, say, founders who had no exposure to Washington, or have no exposure to regulated industries, when they would go to Washington, they'd say, okay, maybe I go once a year, shake some hands, it's kind of fun, and then I go back and I build, and they would be surprised when they got nowhere. And of course, I think that the most sophisticated companies recognize that they have to learn to play the game that Washington cares about. And there is a totally different culture in Washington, there's a totally different set of incentives. I say it's really the difference between, Silicon Valley is a positive-sum culture: Everyone helps everyone, knowing that the pie can always get bigger, and you always want a piece of that bigger pie as it's growing, and so the more things that you're doing, the better. It's why we have this beautiful angel investing network. It's why we have all of these things that make no sense to people in Washington where it's elections, where 10,000 votes in a state could decide the election, and it's a zero-sum game, and that is what decides who is in office and who is in think tanks. And so it's a very different way of thinking about things.
The thing that I think has changed the most about Silicon Valley is recognizing, we might not be good at zero-sum games and zero-sum thinking, but that is the people that we are interfacing with, and we need to understand their incentive systems when they decide to make a purchasing decision, when they decide whether they're going to vote on a bill in a certain way, when they think about, what do their constituents care about back home in a place that has nothing to do with Silicon Valley or California. So I think that level of empathy for what Washington does, which is very different than what Silicon Valley does, is important.
Is it hard to stay optimistic? There are times where you're banging your head against the wall, we're on very short time horizons, Washington can go in perpetuity doing what it does without necessarily seeing much change. But having those points of connection, and constantly having the conversation, and recognizing that it is a long game and not a short game, I think has been very beneficial, and now there are success stories: Palantir, Anduril, Shield AI, these companies that have been around for 10 years now, that have really shown that it is possible to do good work and to support the needs of the DoD, and to speak the language of the DoD, as well, I think has really led to this next generation of founders understanding what they need to do to be successful as well.
Building the American Dream (24:35)
What kind of world are you trying to create? I'm sure it's intellectually challenging, I'm sure it's well-paying, but, fundamentally, why are you doing this? And I would think it's to create some world that is better than the one we're currently living. What is the world you're trying to create?
I think there is a recognition post-Covid, in particular, for a lot of young people, a lot of engineers, that things were broken, things are broken in this country. The physical world has not kept up with the digital world, and there's been extraordinary changes, technology is moving as fast as it possibly can, and a lot of the things that people care the most about have been left out of that story: Education, which is something we haven't necessarily talked about yet, but education needs to be completely transformed in an era where technology is at our fingertips and where people who are good at learning learn faster than they ever possibly could, and people who are not good at learning don’t, and so you have a disparity between those people.
But there's an extraordinary amount of change that has happened in the last 25 years where the things that American citizens care most about have not changed in the way that they need to keep up with, again, the changes in the consumer internet and what we've seen in the enterprise.
And so the story of, how do we make America strong? How do we continue to be the most dynamic country in the world? How do we make sure that all American citizens and the things they care about most in terms of the American Dream are part of that story? I think that is something that the founders who work in American Dynamism care deeply about. They recognize, and I always point this out, but there are so many founders now who are working at companies like Anduril, like Saronic, where they don't necessarily even remember September 11th — they weren't old enough — but they care deeply about the idea that America needs to be a strong country, and that we need to have a mode of deterrence, and we need to have a strong national defense that keeps America the most dynamic country so that people can build inside of it. The same thing with recognizing that there needs to be changes in housing, needs to be changes in education, these are things that were part of the American Dream when our parents were growing up and feel a little bit distant for a lot of other young people growing up today. So I think there is a recognition that technology has to be a part of those big sectors in order to support the American Dream that many of us grew up with and that many of us aspire to.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Below is a lightly edited transcript of our conversation
Pethokoukis: Philipp, welcome to the podcast.
Carlsson-Szlezak: Thanks for having me.
Optimism during a polycrisis (1:39)
It seems as if there are multiple challenges facing the world and the global economy simultaneously. People described it as a “polycrisis,” and it could be everything from trying to navigate economies to a soft landing after a bout of inflation; part of it, I guess, is the big rise in debt; we have war in Europe; maybe, eventually, war in Asia over Taiwan; and, of course, climate change; aging population; falling birth rates; and some people seem to view AI as more of a threat than a positive, it's going to take jobs, and perhaps other bad things. If I've given a sensible description of the world, how can one be a “rational optimist” in a world of polycrisis?
Just taking maybe two of your examples: The soft landing that was supposed to be impossible, remember that? We need, what was it, six percent unemployment for how many years to bring inflation down? So clearly that didn't pan out like the pessimists said. Or think about the war in Ukraine that you mentioned, which, of course, is the tragedy, but the fact is that there is no recession in the Eurozone so far. The fact is that industrial production has held up rather well, even in the heartland of industrial production in Germany. So real industrial output is actually remarkably resilient. Overall, there's little doubt that there are many, many risks. There are crises, but, more often than not, we're focusing on the tail ends of the distribution and pretend that those risks are at the very center of the risk distribution.
It sort of reminds me — it's not a perfect analogy — of where we were in the early 1990s. I suppose you can always point to, and maybe this is part of your point, if you want to focus on bad news, the world will give you plenty of bad news to focus on. But I remember at the beginning of the 1990s, very bad recession here in the United States, a lot of concerns about the ability for rich economies to grow quickly. Again, debt was an issue, and though, looking back, it may seem like, wow, people should have been really excited: end of the Cold War, there was a lot of uncertainty what would happen to the former Soviet Union, a lot of talk back then about suitcase nukes, who knew what was going to happen in the world? And of course, all of this kind of concern and uncertainty led right into a big economic boom.
So to kind of get back to what you were saying, it seems to me that, rather than being rational optimists, we’re sort of naturally irrational pessimists.
Yeah, but we shouldn't be. I think your example of the ’90s and what the mid- and late ’90s delivered, I think is not a bad analogy for what I think will play out in the 2020s, the rest of the decade. We're in an “era of tightness,” is what we call it in the book, which is really a structural condition of the labor market. Lots of people think that shortages in the labor market are a byproduct of Covid, but that's not true. The labor market turned tight already in 2017, so, in technical language, that's when unemployment dropped below U*. Covid was an interruption of this tightness. Unemployment went to almost 15 percent, and then it came down just as fast, but we are in this era of tightness, and I think it will persist.
To be clear, even if and when you get another recession, you will return to a tight stance. And there are a lot of silver linings that come with eras of tightness: It translates into better real wage growth, it nudges and forces firms do capital-for-labor substitution, it pushes them towards the technological frontier in their respective areas, and all of that should lead into some boost of productivity growth. I'm not of the kind where we're predicting this big jump in productivity growth, I think that's too hyped, but I do believe that the structural tightness, which is almost like a spark to the fuel of technology, I think that will push gradually and measurably over the years to come.
And the driving forces of that tightness are what?
We have a number of things going. Essentially, you had a mismatch of demand of supply already in the late ’10s, as I described, so there's a supply issue, we don't have enough labor supply. You have certain forces on the demographic side that constrained that. And I think often we hear the story that AI is going to produce so much unemployment that there will be mass unemployment, and I don't believe any of this. I think that will play out very differently. Historically, technology has never given us structural or technological unemployment. On the contrary, technology is the deflationary force in the medium- and long-run. Firms that can deliver cost savings, they can lower prices, they will do so to grab market share. That is a real income boost for consumers. So when their real incomes grow, they redeploy that gained real income to new services, new goods purchase and consumption, and that leads to new employment. And so I think, essentially, you will remain with a story where labor is tight, and that is the defining underpinnings of what's coming.
Today people don't remember that even in the 2010s, I think it was Bill Gates, he wanted a robo-tax, a robot tax. Because why? Because automation was taking over the assembly lines and we're going to have to provide for all these people who are going to lose their jobs! Well, where are we today? Near record-low unemployment. And this is in a line of a long tradition where Nobel Prize winners, and technologists, and politicians, they've all predicted technological unemployment.
AI employment panic (7:47)
I wanted to talk a bit about AI labor, since you brought it up. If you don't think mass-unemployment is a valid concern, could there be other downsides from deployment of generative AI in an economy? Could it be, instead of higher employment, might it just be greater inequality? Maybe, before, we had technology hurting blue-collar workers; certainly there's a lot of white collars worried about it. I was just reading a story in the New York Times — all of these people in Hollywood are just terrified, whether they're doing special effects, or they're sound editors, they're all terrified that their white-collar jobs are going away. So do you see, in the near term, any downside from AI?
In macro there's always this tension between the aggregate, which is what macroeconomics is about, and then the distribution of experiences under the hood of macro, if you will. So there will be winners and losers, and there will be those that are harder hit than others, but I think when you look at the aggregate, you add it all up the net-net, I don't anticipate this being a structural or technological unemployment situation.
To go to the micro level, you can take the other side of that argument, too. This is not a big area of research for me, so I'm straying outside of my field of expertise here, but plenty of people have argued that perhaps AI will give a lift to those least-skilled. Why? Because AI is a companion for them that makes them more productive and allows them to create more value, and therefore to be paid better. So I think the jury is out on that.
I don't anticipate a smooth ride where everyone will be a winner and everything will be just plain sailing. Of course this is disruptive, of course there will be gyrations, but the story about technological unemployment's been told for so long. Today people don't remember that even in the 2010s, I think it was Bill Gates, he wanted a robo-tax, a robot tax. Because why? Because automation was taking over the assembly lines and we're going to have to provide for all these people who are going to lose their jobs!
Well, where are we today? Near record-low unemployment. And this is in a line of a long tradition where Nobel Prize winners, and technologists, and politicians, they've all predicted technological unemployment.
There's a nice story with Wassily Leontief, a Nobel Prize winner in economics. He said in the ’70s that human labor was going to go the way of the horse after the introduction of the automobile. Well, 50 years on, we're here with very tight labor markets. And Kennedy was worried about it, too, and others before him. And so I think we have to point to something that's truly different about AI to tell that story. I think we can potentially find some reasons that are genuinely different about AI, but before we all become hysterical about it, I think we should take a deep breath.
Does it strike you as odd that, in a world of low unemployment and, if you're correct, perhaps a longer-term structural tightness, that it's at this moment that people are very worried about technological unemployment, it's at this very moment that they're very worried about immigration coming in and taking jobs. You would think these would be concerns at periods of very high unemployment: people standing in line around the block to get a job, but that's not where we're at. Yet the public mood seems to not be in sync with that.
I think that's a good observation, I don't have a great explanation for it. The technology that's on display is impressive, it is novel, and what's different, generally, is that it makes a credible promise to impact the service economy. In our book, the way we position the slump of productivity growth has little to do with high debt and all those explanations that are occasionally fashionable. It has a lot to do with the fact that the US economy transitioned from being a physical economy to a service economy. And in the physical economy, the production of goods, you always had pretty respectable productivity growth, including the last few decades. But because of this mix shift into services where you did not have the technology to make progress on the productivity side, this mix shift dragged down aggregate productivity growth. If you have zero or very little productivity growth in services, which is like 65, 70 percent of the economy, and you have very high productivity growth and the part that is 30 or 35 percent of the economy, well the blended average is going to be low. And so, as we now have productivity growth promises from AI and services, I think a lot can change there. Again, not in a step change way, this is not like flipping a switch, it's going to be hard slog, and incremental, and cumulative, but something will happen there.
There's plenty of risk out there. New crises will come and happen, but for every true crisis, there are many false alarms, and that is something that I think we need to internalize more, and that is something that can help us see risk a little more calmly and in a measured way.
Risk-assessment strategy (13:08)
Just to take a quick step back: Describe your process for assessing risk. Where do you begin? What are the factors? Do you have a fundamental baseline model of the way the world works? How does that process start and work for you?
The way we go about risk in the book is to say macroeconomic risk should be viewed both as the downside, which is how we commonly view risk, like a recession, or even a structural downside like a deflationary depression — these are downsides. But for practitioners, risk is also hidden in potential upside if you miss out on it. And both the downside and the upside, they come in two flavors: tactical, short-term stuff, cyclical stuff; and the more structural strategic kind, like shifts that happen over longer periods of time. And so we like to think of macro risk in those four flavors: the short, the long term, and the upside and the downside, if you will
Generally when we look at risk, it's very seductive and tempting to focus on bad outcomes and then start analyzing how bad will it be and how quickly they're going to happen. In most situations, it pays off to take a step back, and take a deep breath, and say, “Well, how is the system constructed? What are the drivers? What is the history of this thing?” and an important question, “What would it take to get that outcome from the edges of the risk distribution? What does it take to get there?” Too often in public discourse, we jump straight to the tails of the risk distribution. We're immediately obsessing with the cliff edge and the fall into economic death, and then we're pretending that that risk outcome, which is part of the distribution, so we can't ignore it, but we're pretending that the edges of the distribution are the very center.
And so what we do in the book for a number of areas of risk in the real economy, the financial economy, and the global economy, we go over and over again into these approaches of asking, “How is this thing constructed? What are the drivers? What do you have to believe for the truly bad outcome?” There's plenty of risk out there. New crises will come and happen, but for every true crisis, there are many false alarms, and that is something that I think we need to internalize more, and that is something that can help us see risk a little more calmly and in a measured way.
How well did markets, investors, economists — how well was their risk-assessment process in 2020, given where we are today? My guess is that the global economy is better in 2024 than people thought in February, or March, or April of 2020 as the pandemic was kicking in. Did we do a good job assessing risk and reward back then?
No. Public discourse did a terrible job at that. The conventional wisdom and received wisdom in March and April, May, even June, July, and even August, the summer, when you had the first signs of recovery, the conventional wisdom was: This is worse than 2008, and this could be as bad as the Great Depression. And we have a great collection of headlines that we keep, there are lots of them in the book as well. It was, in my mind, a prediction failure. Why? Because what happens is that a lot of the commentary, a lot of the thinking, was too model-based, what we call “master-model mentality” in the book.
So how do you project a recovery, typically? You look at the unemployment rate as a proxy for the health of the economy, and if you have a high unemployment rate and a recession, well, it can take a long time to bring that down. After 2008, it took the better part of the 2010s to bring unemployment down, and you had “only” (in quotation marks) unemployment of 10 percent after the Global Financial Crisis. Now with Covid, you almost went to 15 percent. So the models extrapolated outside their empirical range. They said, “Well, if it took almost a decade to bring down this unemployment rate after 10 percent unemployment, then after 15 percent unemployment, well it's going to take even longer.” Hence, the narrative of “worse than 2008,” “as bad as the 1930s,” and blah, blah, blah.
Even at the time, you could ask exactly these questions, and I'm not saying this with hindsight bias. We did a piece on March 28th, 2020 in Harvard Business Review where we did exactly that thought experiment. We said, “What does it take for this to be a structural downgrade for the US economy? What does it take for to be worse than 2008?” You're going to need to see damage on the supply side of the economy. You're going to need to see the downgrading of the labor market of a window of capital investment that isn't happening, and the loss of skills, et cetera. And we asked, “Well, how likely is that?” And, of course, it comes down to stimulus. Of course it does, and it comes down to how innovative, fast, and big are we in backstopping the real economy? And we were. And so even in March — and this is shelter-in-place phase, right? This is not even sort of the full lockdown. Even then, you could ask sober questions about very bad risks. And if you did that, you arrived at answers that weren't predictive in “this will happen” at a point forecast level of accuracy. But there was clearly a path and a narrative in March 2020 that was consistent with what actually happened: the tightest recovery on record, and a US economy that was not pushed off its trend path. So after 2008, the US economy was actually pushed off its prior trend path, never made it back in terms of what the trend was, did make it back in terms of growth rates after 2008, but it never made this levels recovery in that sense. All of that was avoided in 2020, and you didn't have to be a magician to at least entertain the possibility that that was a meaningful part of the outcome distribution.
Federal Reserve predictions (19:44)
Speaking of models, how has the Fed's model performed? Which is another way of saying, how has the Fed performed, and continues to perform?
In the recent inflation surge, et cetera? I think the Fed has taken too much flak for what happened in the inflation spike. The inflation spike, by the way, again, was immediately spun into a structural inflection point, the 1970s narrative, off to the races, wage price spirals, all that nonsense. It was an idiosyncratic mismatch of demand and supply. It was an overshoot in consumption following stimulus, and you had the supply chain crunch, and then you had a number of exogenous shocks that nobody could foresee, and those who like to take credit for having predicted the spike, they didn't foresee the Ukraine war, the shock in oil that followed, and many other of the things that played into the spike. The bigger story, though, is quite simply that, as this mismatch of demand and supply unwound, inflation also came down pretty fast.
Now, back to the Fed: Yes, there were slow in responding. Would it have been better for them to go early, perhaps, yes, let's say yes. But at the same time, the idea that they would fail to step up and act, and reign this in, and stand by and watch this whole thing go to hell. I mean that was never credible. And they did step up, and they did what they had to do, and I think also attempting the soft landing was the right thing. People at the time did say, “We need a draconian recession right now to remove all the risk of a regime break in inflation!” Well, you would've cut short a really tight labor market that has a lot of real wage gain that delivers a lot of good things for particularly the bottom of the labor market. So it was the right call to attempt a soft landing rather than saying, “Look, we're going to cut this cycle short right now to remove any risk of inflation spiraling out of control.” They did the right thing. They're successful at it. The soft landing is a success. We're well into it. And so I give them more credit.
Is that what you think is the biggest mistake people are making, that they're still asking, “What about the soft landing?” And what you're saying is, “We're already into it. You sort of missed it. It happened. You're still looking for it.”
Every so often you still see the headline, “Are we going to get a soft landing?” And I'm like, “Well, let's just take a step back.” What is a soft landing? The task was to cool down the labor market, best seen through the eyes of job openings, to cool that down without pushing up the unemployment rate. These two are mirror images of each other. When firms stop hiring, they usually also start firing, so we had to pull off this trick: You stop hiring, but you don't start firing.
That was the soft landing. That's the definition of a soft landing, nothing else. And that is what happened: Job openings are down more than three-and-a-half million or so — don't nail me on the decimal — and the unemployment rate is up a little, but, as you and I know, the unemployment rate is not up because of firings. The unemployment rate is up for compositional and participation reasons. So if that is not a soft landing, at least, I would call it the second of three stages, if you will, I don't know what it is. And back to the topic of headlines, most of them are just confusing people more than they're helping them. It's always nice to write something clickbaity that people will be scared of and think this is the cliff edge. How about a headline: “Wow, this is a really great soft landing! This is remarkably good!” Why don't we acknowledge that for a moment?
You can always speculate. You can speculate on, for example, exogenous shocks. Covid is an exogenous shock, and there are others: There are solar flares, and there are new pandemics, and there are things that can do immediate damage, and you can spend millions of dollars on models, and they simply won't capture that exogenous risk.
Impending shocks? (23:41)
Yeah, I mean, the name of the book isShocks, Crises and False Alarms. As I look over the rest of this decade, if you take the most bullish and extravagant predictions about AI, it’s not clear to me what the economy looks like a decade from now. Again, if you take the most bullish kind of [view]: we get the human-level AI and all that. So there's that.
I also am not quite sure what the world looks like if some of these worst-case scenarios with Taiwan, and the US, and China, because that seems to me to be so potentially bad, I don't want to think about it. I don't know what the global economy looks like on the other side.
What are the big risks, or the things which you believe pose the greatest risk of disruption? Disruptions are going to be good and bad. If you're really worried about AI, that must mean AI is very powerful, it can do a lot of good things, too. So what out there do you really worry about that the disruption will be just bad and have far more downside?
You can always speculate. You can speculate on, for example, exogenous shocks. Covid is an exogenous shock, and there are others: There are solar flares, and there are new pandemics, and there are things that can do immediate damage, and you can spend millions of dollars on models, and they simply won't capture that exogenous risk. So that's one story.
Geopolitics, since you mentioned it, a third of the book, roughly, the third part is about those types of risks, and they can be devastating, there's no doubt about it; but would you build a central case around this and make that the base case and expectation of how to view the future? Geopolitics is extremely treacherous when it comes to translating its impact on the economy. It's fascinating to me often how little the complexity is acknowledged and understood, and, in the book, we use an example juxtaposing a start of World War I and World War II.
So when World War I breaks out, the Dow is down 10 percent, they close it for 136 days, and when they reopen it, it's down another 20 percent. Exactly as you would expect, right? It makes a lot of sense, a world war and the market's in the gutter. Yet, when World War II breaks out in ’39, the market jumps 10 percent and stays up. Why? It ends the Great Depression, it puts to use labor, it has capital expenditure and investment, and it singlehandedly ends a decade of malaise.
And there is a silver lining to this, and all of this doesn't sit well with how we want to think and should also think about geopolitics, which is in humanitarian terms, and also values and idealistic views. All of this is true and correct, but if we are to assess the impact on the economy, we are going to have to restrain some of these instincts, and we're going to have to say, what are the transmission channels from geopolitics to the real economy, to the financial economy, and to the institutions that we have in place that govern our economy? What are those transmission channels? And often, the bar is higher than you think.
Just think about it, the Ukraine war. It's left no mark on the US economy at all, virtually. Why? Because the real linkages weren't there; virtually no trade into this part of the world, either Ukraine or Russia. The financial linkages weren't really there; it's not like balance sheets of US banks were impaired by shutting off that part of the world. And on the institutional side, we can discuss sanctions, and we can discuss using the US dollar as a means of punishing Russia, and all that. But essentially, once you think soberly about, well, how is that shock supposed to transmit to the economy? Well, it looks a lot different.
The same could be said about the tragedy in the Middle East. The oil price is lower than before the attack on Israel, right? If you look at futures and forward pricing, or the price of insurance against swings in the oil price, it’s lower today than before the attack on Israel and before the retaliation that Israel enacted. So any of these geopolitical risks and hotspots, they are to be taken seriously. I'm not saying they don't matter, but when we extrapolate from them straight to the economy, it more often goes wrong than it goes right.
And a final thought: I’m not a Taiwan and China watcher, but one thing that's also clear to me, since you mentioned the unthinkable and the worst-case scenario, my next question would be, okay, what shape would that take? Would that be a blockade? Would that be an actual invasion? Would it be something that involves airlifting semiconductors out of the island? Would that be . . . There's a myriad ways of how such a thing could play out. It'd be a big shock, and terrible, but I can't say with confidence what it would do linearly to an economy like the US economy. It would come down to the details of that.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
I think that's a great question. In my mind, I think AI has been going on for quite a long time, but I think the aim has always been artificial general intelligence. And in a sense, I think of this as a huge deal, and the vision I have for the future is being able to have a very, very large supply of cognitive resources that you can allocate to quite a wide range of different problems, whether that's energy, healthcare, governance, there's many, many ways in which this technology can be applied as a general purpose technology. And so I guess my vision is seeing that being used to solve quite a wide range of problems that humans have had for decades, centuries, millennia. And I think you could go into so many different directions with that, whether it's curing diseases, or optimizing energy grids, and more. But I think, broadly, that’s the way I think about it. So the objective, in a sense, is safe AGI [Artificial General Intelligence], and from that I think it can go even further. And I think in many ways, this can be hugely beneficial to science, R&D, and humanity as a whole. But of course, that also comes with ways in which this could be misused, or accidents, and so on. And so huge emphasis on the safe development of AGI.
So you're viewing it as a tool, as a way to apply intelligence across a variety of fields, a variety of problems, to solve those problems, and of course, the word in there doing a lot of lifting is “safely.” Given the discussion over the past 18 months about that word, “safely,” is, one, I think someone who maybe only pays passing attention to this issue might think that it's almost impossible to do it safely without jeopardizing all those upside benefits, but you're confident that those two things can ultimately be in harmony?
Yeah, absolutely, otherwise I wouldn't be necessarily working on an AGI policy. So I think I'm very confident this can be done well. I think it also depends what we mean by “safety” and what kind of safety we have in mind. Any technology, we will have costs and trade-offs, but of course the upside here is enormous, and, in my mind, very much outweighs potential downsides.
However, I think for certain risks, things like potentially catastrophic risks and so on, there is an argument in treading some careful path and making sure this is done scientifically with a scientific method in mind, and doing that well. But I don't think there's fundamentally a necessary tension, and I think, in fact, what many people sometimes underestimate is how AI itself, as a technology, will be helpful in mitigating a lot of the risks we're foreseeing and thinking about. There's obviously ways in which AI can be used for cyber offense, but many ways in which you can also use that for defense, for example. I'm cautiously optimistic about how this can be developed and used in the long run
The risk conversation (5:15)
Since these large language models and chatbots were rolled out to public awareness in late 2022, has the safety regulatory debate changed in any way? It seems to me that there was a lot of talk early on about these existential risks. Now I seem to hearing less about that and more about issues about, maybe it's disinformation or bias. From your perspective, has that debate changed and has it changed for the better, or worse?
I think it has evolved quite a lot over the past — I've been working in AI policy since 2017 and there's been different phases, and at first a lot of skepticism around AI even being useful, or hype, and so on, and then seeing more and more of what these general models could do, and I think, initially, a lot of the concerns were around things like bias, and discrimination, and errors. So even things like, early-on, facial-recognition technologies were very problematic in many ways: not just ways in which they were applied, but they would be prone to a lot of errors and biases that could be unfair, whereas they're much better now, and therefore the concern now is more on misuse than it accidentally misidentifying someone, I would say. So I think, in that sense, these things have changed. And then a lot of the discourse around existential risk and so on, there was a bit of a peak a bit last year, and then this switched a bit towards more catastrophic risks and misuse.
There's a few different things. Broadly, I think it's good that these risks are taken seriously. So, in some sense, I'm happy that these have taken more space, in a way, but I think there's also been a lot of alarmism and unnecessary doomerism, of crying wolf a little bit too early. I think what happens is that sometimes people also conflate a capability of a system and how that fits within a wider risk or threat model, or something; and the latter is often under-defined, and there's a tendency for people to often see the worst in technology, particularly in certain regions of the world, so I think sometimes a lot has been a little bit exaggerated or overhyped.
But, having said that, I think it’s very good there's lots of research going on on the many ways in which this could potentially be harmful, certainly on the research side, the evaluation side, there’s a lot of great work. We've published some papers on sociotechnical evaluations, dangerous capabilities, and so on. All of that is great, but I think there has also been some more polarized parts calling for excessive measures, whether regulatory, or pausing AI, and so on, that I think have been a little bit too trigger-happy. So I'm less happy about these bits, but there's been a lot of good as well.
And much of the debate about policy has been about the right sort of policy to prevent bad things from happening. How should we think about policy that maximizes the odds of good things happening? What should policymakers do to help promote AI to reshape science, to help promote AI diffusing as efficiently as possible throughout an economy? How do we optimize the upside through policy rather than just focusing on making sure the bad things don't happen?
I think the very first thing is not having rushed regulation. I'm not personally a huge fan of the Precautionary Principle, and I think that, very often, regulations can cause quite a lot of harm downstream, and they're very sticky, hard to remove.
The other thing that you can do beyond avoiding bad policy is I think a lot of the levers to making sure that the development goes well aren't necessarily all directly AI-related. So it'll be things like immigration: attracting a lot of talent, for example, I think will be very important, so immigration is a big one. Power and energy: you want there to be a lot more — I'm a big fan of nuclear, so I think that kind of thing is also very helpful in terms of the expected needs for AI development in the future. And then there are certain things governments could potentially do with some narrow domains like Advance Market Commitments, for example, although that's not a panacea.
Commitments to do what?
Oh, Advance Market Commitments like pull mechanisms to create a market for a particular solution. So like Operation Warp Speed, but you could have an AI equivalent for certain applications, but of course there's a lot of parameters in doing that well, and I wouldn't want a large industrial-policy-type approach to AI. But I think generally it's around ensuring that all the enablers, all the different ingredients and factors of a rich research and development ecosystem continue to thrive. And so I think, to a large extent, avoiding bad regulation and ensuring that a lot of things like energy, immigration, and so on go well is already a huge part of the battle.
How serious of a potential bottleneck is the energy issue? It seems to me like it's a serious issue that's coming fast, but the solutions seem like they'll take more time, and I'm worried about the mismatch between the problem and finding a solution to the problem.
I suspect that, over the coming years, we will see more and more of these AI systems being increasingly useful, capable, and then integrated into economic systems, and I think as you start seeing these benefits more and more, I think it'll be easier to make the case for why you need to solve some of these kind of policy issues a bit faster.
And I also think these solutions aren't that difficult, ultimately. So I think there’s a lot that can be done around nuclear, and wind, and solar, and so on, and many regulatory processes that could be simplified, and accelerated, and improved to avoid the vetocracy system we're in at the moment. So I don't think the solutions are that difficult, I think mustering the political will might be right now, but I expect that to be less of a challenge in the coming years with AI showing more and more promise, I think.
Policy strategy (11:25)
Speaking of vetocracy, whatever the exact substance of the regulation, I am concerned, at least in the United States, that we have 50 states, and perhaps even more bodies if you look at cities, who all have a lot of ideas about AI regulation, and I'm extremely concerned that that sort of fractured policy landscape will create a bottleneck.
Can we get to where we need to go if that's the regulatory environment we are looking at, at least in the United States? And does, ultimately, there need to be a federal . . . I think the technical word is “preemption” of all those efforts? So there's a federal approach, and there aren't a federal approach, plus a 50-state approach, plus a 175-city approach to regulation. Because if it's going to be what I just described, that seems like a very difficult environment to deal with.
I'm not wildly optimistic around a patchwork of different state-level regulatory systems. I think that will come with various externalities, you'll have distortionary effects. It will be a very difficult environment, from a commercial perspective, to operate in smoothly. I think I'm a lot more open to something at a federal level at some point, rather than a big patchwork of city-level or state-level regulation. Now, it depends on exactly what we're talking about. There might be specific domain, and context, and application-specific regulations that might make sense in some state and not another, but in general, from a first principles level at least, I think that would probably not be desirable.
A second regulatory concern — and maybe this is dissipating as policy makers learn more, especially at the federal level, maybe, learn more about AI — is that, at least initially, it seems to me that whatever your policy idea was for social media, or about content moderation or what have you, you just kind of took that policy framework and applied it to AI because that was what you had. You pulled that baby right off the shelf. Are we still seeing that, or are people beginning to think, “This is its own thing, and my ideas for social media may be fine for social media, but I need to think differently about AI”? Obviously the technology is different; also, I think both the risks and potential rewards are very different.
Yeah, totally. I think that has been an issue. Now, I wouldn't say that's the case for everyone. There's been some groups and some institutions doing some very careful work that really think about AI, and AGI, and so on in careful, more calibrated ways; but also I’ve seen quite a lot of reports where you could have easily imagined the same text being about social media platforms, or some kind of other policy issue, or blockchain, or something just being repurposed for AI. And there's a lot of stuff out there that's just very high level, and it's hard to disagree with at a high level, but it’s far harder to apply and look at from an operational or practical perspective.
So I've been seeing quite a lot of that; however, I think over time, the field is maturing more and more, and you're seeing better thinking around AI, what it really is, what's appropriate at the model level versus at the application level and the existing landscape of laws and regulation and how these might apply as well, which is often that's something that's forgotten, or you have lots of academics coming in and just trying to re-regulate everything from first principles, and then you're like, “Well, there's tort law, and there's this and that over there.” You got to do your gap analysis first before coming out with all this stuff.
But I think we are seeing the field of AI governance and policy maturing in that space, and I expect it to continue, but I still, of course, see a lot of bad heuristics and poor thinking here, and particularly an underestimation of the benefits of AI and AGI. I think there's a tendency to always think of the worst for everything, and it's necessary, you need to do that too, but few are really internalizing how significant AGI would be for growth, for welfare, and for solving a lot of the issues that we've been talking about in the first place.
Is AGI an “if” issue, or is it a “when” issue, and if it's a “when,” when? And I say this with the caveat that predictions are difficult, especially about the future.
In my mind, it's definitely a “when” question. I see no real strong reason why it would be an “if,” and that being completely impossible. And there's been many, many, many examples over the last 10 years of people saying, “Well, this is not possible with neural networks,” and then 10 minutes later, it is proven to be possible. So that's a recurring theme, and that may not be sufficient to think that AGI is feasible and possible, but I'm pretty confident for a variety of reasons. About AGI, by the way, I think there's an excellent paper by Morris and others on Levels of AGI [for] Operationalizing Progress on the Path of AGI, and I think it's a very good paper to [frame one’s thinking about AGI].
And that goes back to one point I made earlier in that, at some point, you'll have systems that will be capable of quite a lot of things and can do probably anything that your average human can do, starting at least virtually, remotely, to start with, and eventually to the physical world, but I think they'll be capable in that sense. Now, there's a difference between these systems being capable in an individual lab setting or something and then them being actually deployed and used in industrial processes, commercial applications, in ways that are productive, add value, create profits, and so on, and I think there's a bit of a gap here. So I don't think we'll have a day where we'll wake up and say, “Oh, that's it, today we have AGI.” I think it'll be more of a kind of blurry spectrum, but gradually I think it'll be harder and harder people to deny that we have reached AGI, and as this stuff gets integrated into production systems, I think the effects on growth and the economy will speak for themselves.
As to when exactly, I would think that, at least the capabilities, I would expect that in the next five years you could easily see a point where people could make a very confident claim that, yeah, we've got systems now that are AGI-level. They’re generally capable, and they are pretty competent, or even expert-level to at least 90th percentile of skilled adults, and then the challenge will be then operationalizing that and integrating that into a lot of systems. But in my mind, it's definitely not an “if,” and I would say the next five to 10 years is the kind of relevant period I have in mind, at least. It could be longer, and I think the tech community has a tendency to sometimes over-index, particularly on the R&D side.
AI and national security (18:21)
Do you have any thoughts, and maybe you don't have any thoughts, about the notion that, as perhaps AGI seems closer, and maybe the geopolitical race intensifies, that this becomes more of a national security issue, and the government takes a greater role, and maybe the government makes itself a not-so-silent partner with tech companies, and it really becomes almost like a Manhattan Project kind of deal to get there first. Leopold Aschenbrenner wrote this very long, very long paper — is that an issue that you have any thoughts on? Is it something that you discuss, or does it seem just science fictional to you?
Yeah, I do do a lot of thinking on that, and I've read Leopold’s report, and I think there's a lot of good things in there. I don't necessarily agree with everything. I think things like security are really critical, I think thinking about things like alignment, and so on, is important. One thing I really agree with with Leopold’s report that I'm glad he emphasized was the need to secure and cement liberal democracy, “the free world must prevail” kind of thing. I think that is indeed true, and people tend to underestimate the implication on that front. Now, what that looks like, what that means and requires in practice is not fully clear to me yet. I think people talk about a Manhattan Project, but there are many other potential blueprints or ways to think about that. There could be just normal procurement partnerships, there could be different models for this. At some point, something like that could be defensible, but it's very hard to predict that in advance, given particularly. . . well, how hard it is to predict anything with AI to start with. And secondly, there's loads of trade-offs with all these different options, and some might be a lot better than others, so I think certainly more work might be needed there. But, in principle, the idea doesn't seem completely crazy or science fiction to me.
Chatbot advice (20:15)
You recently posted on X that you were baffled at how many people don't use these language models or chatbots daily. I think a lot of people don't know what they would use them for. Do you have any recommendations for ways that people who are not in your line, who are not coders, that people can use them? Do you use them in ways that are applicable to the way regular people might use them?
Yeah, I think so, and under the post, I gave a few examples of how I use it. Now admittedly, most of these wouldn't be something that anyone would do, but I thought about this last weekend when I was seeing my parents and I was trying to get them to understand what Claude or Gemini is and how to think about it, what kind of questions are worth asking, and what kind of questions are not worth asking, and it's very hard to come up with a very crisp way of sharing these intuitions. I think the first piece of advice I'd give is probably to just take one of these models and have a very long conversation with it about some sort of topic, like try to poke holes, try to contradict, and I think that starts giving you maybe a few better intuitions about what this can do, as opposed to just treating it as some sort of question-and-answer Oracle-type search engine, which I think is not the right use case.
That is probably the most unsatisfying way to look at it, and just treat it as a better Google search engine. I mean really that sort of conversational, curious aspect, rather than saying like, “Find me a link.” “Find me a link” isn't a great use.
Exactly, and people will often do that. We'll do a thing, we'll get some incorrect answer, or hallucination, or whatever, and then we'll say, “Oh, these things are not good, they're not accurate,” and we'll stop using it, and to me, that is just crazy. It is very fundamentally incurious, and I think there's ways of using them and thinking of them that is very useful. So what have I done recently? I'm trying to think of an example. . .
I had some papers that I couldn't understand very well, and I would just ask it for better analogies, explanations, try to dig into certain concepts and ideas and just play around with them until the insights and intuitions were easier for me to internalize and understand. And I think you could do that at different levels, and regular people also want to understand things, so I think that might be potentially an example. But the very first thing I would do is simply long, protracted conversations to really get a sense of how far the model can really go, and then, as you do that, you'll find things that are a bit more creative than, “Can you please rewrite this email for me? Can you find typos?” or “Can you fill in my tax report?” or something. I think one way a friend used it — and of course, there are obvious limitations to that, get a lawyer and everything — but he had a legal contract that someone sent to him, and he couldn't afford a lawyer straight away, so he just said, “Can you help me find potential issues and errors in here? Here's who I am in this kind of contract. Here's what I'm concerned with.” And it's a first starting point. It can be useful. It gives you interesting insights. It doesn't mean it replaces a lawyer straight away, but it is one potential interesting way that everyday people could use.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Pethokoukis: In the book, you write with some concern about the declining faith in capitalism, really among all Americans, especially Democrats and the young. It may be worrisome, but is it really surprising, given we had a financial crisis . . . hard to believe it was, what, 14 years ago? 15? Well, I guess 16, 2008 . . . Financial crisis, slow recovery. So, for a lot of people, there's a pretty good chunk of their lives where the economy didn't seem to be really zipping along very quickly and making a big change in their lives, so if people are skeptical of capitalism, can you really blame them?
Sharma: Well, as I argue in the book, Jim, that the current system we have in place is very far from capitalism. It's a very distorted form of capitalism that we have in place, and the surprise, I think, has to be the fact that, at the headline level, the numbers from the American economy look pretty good, which are the fact that the economy is growing at above two percent, the stock market is booming, America is seen as the center of all the tech innovation, AI, and, compared to its major rivals such as Europe, America seems to be in a much better place. And I think that a lot of people in the Biden administration try to put that out, which is that, “What's the problem, given how well the American economy is doing?”
And I think that the polling data is obviously very different. It shows a persistent and consistent decline in faith in the American economy, that the voters and people have, and, as you pointed out at the outset, that a lot of young people, in fact, say that now they prefer socialism to capitalism. So I guess that's the surprise, which is the fact that, at the headline level, the numbers look fine, and especially when you compare it to other countries, and yet, if you look beneath the hood, both the numbers, in terms of polling numbers and then a deeper look at capitalism, which is what I've done in the book, reveal that something is wrong with the system. The general feeling that the average American has, that the system is almost rigged against them, and rigged in favor of Big Business.
Clearly one reason people are sort of unhappy is because we had a big surge of inflation, and, even though the rate of inflation has come down, prices are still a lot higher than they remember. But that seems to me to be a temporary aberration. As every day, and month, and year goes by, we'll be a little further from this inflation surge. And then you mentioned all those positive things: in every sort of emerging technology, America seems to be the leader. Is there really a deep problem that will be more or less solved on its own the further we get away from the pandemic and that pandemic-era economy?
Well, anything can happen, but I wouldn't bet on that because, as you said, that the decline in the faith in America's brand of capitalism and the number of people who feel that the country's moving in the wrong direction, all that data predates the pandemic. So it's not as if there was a surge in inflation and that suddenly changed people's thinking about the economy and they're feeling much worse off because their real wages got wiped out by inflation. This happened even before that. Through the last 10 or 20 years, you've seen a consistent decline in American faith in the economy, in American faith in government. So this is not just about the near-term inflation data, or even in terms of what's happened in the post-pandemic world, and to bet, therefore, that, with the passage of time, as the pandemic sort of becomes more and more of a distant memory that this is going to change. I think the problems are much deeper, and it shows up in the fact that, as I argue in the book, that economic and social mobility in America today is close to record lows. Only 35 percent of Americans feel today that they're going to be better off than their parents, and when the American Dream was really flowering, that number used to be 70 to 80 percent of people felt that they'd be better off than their parents. So there's a whole host of data to show here that the problem and the disaffection with the economic system is much deeper than just the pandemic-driven surge in inflation.
Economic booms (6:12)
Is there anything about this economy that four or five years of above-average economic growth won't solve? I sort of recall that in the early ’90s, you could have made a very similar case that we had a nasty recession in the early ’90s. Some people have forgotten about it, but it was a bad recession. And then we had the Gulf War, and there were a lot of newspaper articles saying that the era of fast growth was over, America just couldn't grow fast anymore, and just as people were convinced that the good days were over, the economy took over, internet boom, tech boom, and we had very rapid growth—and, interestingly, it was also a period of high inequality, but people didn't seem to care so much because the economy was cooking and real wages were rising. So is there anything wrong that a little bit of faster growth couldn't solve?
In terms of the fact that I remember living through that era, and I think that if you compare the polling data, it shows the fact that the problems are much deeper now, and the disaffection is much deeper than what was there, in terms of the fact that what solved it back then — even back then, the basic faith in American capitalism was never lost. I think that what's happened now is a feeling that we don't have an equality of opportunity anymore, and that inequality levels now have risen much further than back then. So I think that it's always possible for some hopeful turn to take place, but I'd say that the problems this time are much deeper, and that's what I try and say: Why has this happened? The book is a deep investigation of why has this happened systematically over time. We've gotten to a point now where, across the Western world, leaders are universally, almost, unpopular, and they are also struggling to get reelected. This is happening in Europe, and I argue in the book that capitalism is in worse shape in Europe: much more statist, much more bureaucratic, much more intrusive, and Europe's an even greater regulatory hegemon than America is. So something which is going on across Western societies for this disaffection and feeling that the American government is more pro-business than it is pro-competition, which should be the essence of capitalism.
Pain management (8:49)
But where do you think it all went wrong?
That’s the crux of the argument. As the tagline of the book goes, that capitalism did not fail, it was ruined. What ruined it? As I show, that it is the suite of government habits, that how the government's role in American society has come to resemble that of pain management, which is the fact that every time there is the slightest hint of pain, we administer opiates. That's one of the reasons we have the opiate crisis, where people are just hooked to opiates, because at the slightest pain, you give them opiate to relieve the pain without quite solving the underlying cause of the problem. I think, in a similar way, what's happened with American capitalism is the fact that the government has been trying to socialize risk and take risk out of the system to try and mitigate pain, and in doing so, it's got the economy hooked to constant stimulus, constant government intervention, which is leading to a lot of perverse consequences.
What are those purpose consequences? One of them, as I've argued in the book, is that productivity has declined. Now remember, capitalism is supposed to generate lots of competition, lead to more creative destruction, lead to an increase in productivity, and productivity is the key to economic growth. But in the last 30 to 40 years, we have seen a big decline in productivity growth across the western world, including in the United States.
On the other hand, we have seen a lot of deadwood being kept alive due to all these interventions. The culture of bailouts, the culture of regulation, has kept a lot of deadwood alive in the system, which is not only dampening productivity growth, but it's preventing the entry of new businesses and new firms to come. As a result, the pace of startups in this country today has gone down significantly. It's picked up a bit after the pandemic, but for the last 30 to 40 years, the rate of new startups in this country has declined.
So I think that the systematic increase in the government's involvement in the economy has led to these perverse consequences, and those changes are quite recent. The American government was never this involved at a very basic level. The share of government spending in the economy was just three percent a hundred years ago. It has gone up over time, now we're closing in on 40 percent, the government spending, the share of the economy.
But it's not just that. The culture of bailouts: America never believed that it should be bailing out private sector enterprises. And yet, from the 1980s onwards, it's been an increased culture of bailouts. Before that, America did not do bailouts of private-sector companies, but once the precedent was set with the large financial sector bailout of 1984 of Continental Illinois, you've seen one bailout after another, and get bigger and bigger with each crisis.
And then, of course, you have the entire role that the Fed has played the, US Central Bank, in the way it has tried to micromanage the cycle and always try to act with policies where, on the upside, it's fine if markets rise, but on the downside, it's there to protect and socialize risk, which, once again, has undermined productivity and kept a lot of deadwood and possibly zombie companies alive.
So it's this suite of government habits that I think has corroded capitalism and brought us to this position today where capitalism is so distorted.
Famously, during the financial crisis, President Bush said, in his own colloquial way, “This sucker is going down.” Would you just have “let this sucker go down” rather than bail out banks back in 2008?
Each bailout seems justifiable because, at that point in time, you're in the midst of a crisis, and there's always this fear that if you don't bail out, we will end up being like the liquidationists of the 1929s and lead to a Great Depression. And I think that
Totally correct. But I think that the impulse now is the opposite, which is, “liquefy, liquefy, liquefy.” So even in 2008, if you needed to intervene to prevent a financial crisis, they can argue that was justifiable. But after that, look at what happened: the Fed continued to do quantitative easing, to lift asset prices, thinking that high asset prices will be good for economic growth, whereas it only increased inequality because the owners of asset prices tend to be the rich, and put many aspirants of the property market, or buyers of the property market, it put those things out of reach.
And then also the fact that you had, in terms of the last year, when you had the Silicon Valley Bank problems, you intervened again to try and prevent it. And again, the fear is raised that if you don't intervene, you'll have the Great Depression. So I think we need to find a balance here, which is that you've gone from “liquidate, liquidate, liquidate” to now “liquefy, liquefy, liquefy.” And I think that we need to get back to some balance and, yeah, there is a role for government, like in a terrible crisis, but you can't have a government always be there to intervene at this slightest hint of trouble. That is what causes problems.
Populist policy (14:38)
If we indeed live in a populist political age, I would assume that I would see more intervention, and more bailouts, and more pain-prevention, because that's what populism is. Populism is saying that you've been taken advantage of by the elites, it's not your fault, and if it's not your fault, you shouldn't feel any pain, therefore we will intervene to help you. Government will help you. So wouldn’t you expect in the near future just to have more of this kind of thing: more pain-management public policy?
I think that you're right that this trend can continue, because the default path of politicians is to keep doing what they've been doing unless there's a crisis which forces them to reverse course and forces, I'd say, American society then to wake up and say, “This is not working.” But to say that this is a populist thing, I'm not sure, because this is the strategy which has been pursued for the last few years, if not decades, which is what I've argued in the book. So if you're going to keep pursuing the same strategy, then you should not expect different outcomes. If the strategy of so-called populism of more intervention was working, then why should so many people be upset with what's going on?
I've been in the last few days talking about this book. I'm surprised at how receptive people have been to the idea that it's not the government's business to be bail out private sector companies. It's not the government's business to be instituting 3,000 new regulations a year. The average number of new regulations that the government would put into a year has gone up dramatically over the last couple of decades. Now, again, it is not the job of the government to be putting in so many regulations because the more regulations you put in, the harder it is for small and midsize businesses to thrive, and it creates a barrier to entry, because the cost of regulation has gone up significantly, it can be borne only by big businesses, the average business finds it more difficult. And also the lobbyists, then, are able to rewrite regulations in their favor, and they're hired typically by the Big Tech firms or the people with all the money.
I think a populist agenda would favor deregulation. A populist agenda would say that you can't have the Fed run monetary policy which keeps on inflating asset prices, particularly property prices, and makes it harder and harder for Americans to afford a home. And similarly, a populist agenda should be against bailouts, because bailouts typically help the entrenched powerful companies rather than the average person. And as I said, it's time to try something different also because what's been tried so far is clearly not working in terms of making Americans happy.
Catalyzing change (17:32)
Let me ask you this, then: What do you see, then, that makes you think we're going to do something differently?
I don't see anything as yet in terms of that, just because, as I said, the politicians will keep doing the same thing unless there's a crisis. What forces countries to change course, including the countries which are held by liberals as these great paragons of economic virtue like Sweden or the other Nordic countries?
The bond market? Is that the action-forcing entity here? Would it be the bond market?
Eventually, yes. I think that that's it, which that only when the markets stop financing these incredible deficits do you begin to see a course-correction. But what my book tries and does here is, as I said, the first step to a cure is to at least diagnose the problem correctly. So far, the diagnosis that the Biden administration offered was, “Oh, the government needs to double down because we have had this era of small government for the last 30 to 40 years. We need to double down on what the government's been doing to correct all these problems.” But the problem is this, that we never had an era of small government! So to say that now we need to reverse that era is just wrong. As I demonstrate in the book, capitalism has been about just expanding the government over the last 30 to 40, at least the capitalism practiced by the Western societies in the last few decades.
So I would say that you're right that the bond market, eventually, is obviously the ultimate disciplinarian. It did that to Greece, it did that to Sweden, and other countries before that. But we don't seem to be at that point yet, but we could be, because the next time you have an economic downturn in the US, the budget deficit will widen to nine, 10 percent of GDP. As revenues fall, more and automatic spending kicks in, and, after that's done, the ability for interest rates to come down, which typically do to help a country go through a recession, I think will be quite limited in countries such as the United States, because your government spending would've been so high already by then.
If I were to put together a short-but-sweet economic agenda to deal with some of these, I might say, “Boy, you're worried about debt and deficits?” I would say, “What we need is a value-added tax and we need to at least lower the projected pace of spending on social security and Medicare.” So Medicare reform, social security reform, a value-added tax—and if I tried running on that agenda, I think they would run me out of town.
Since we don't have politicians generally calling for that, isn't the revealed preference of American voters for more of the same: “Don't tax me, but also don't take away my benefits,” and off we go. If there was this deep swell for any change that you think might be there, based on maybe your conversations with folks, wouldn't it be already reflected by our politicians? But I don't see that.
Yeah, you're absolutely correct that politicians don't live in a vacuum, they react to a societal chain. But my point is that no one seems to be doing . . . I just pointed out to you that outside of this so-called “tough medicine,” which you just highlighted, there are other things which can be done in terms of drawing the line on bailouts, drawing the line on instituting new regulations. Even Trump came to par, he spoke about the fact that he's going to withdraw two regulations for every one that he puts into place. In the end, though, he ended up putting out as many regulations per year as the Obama administration had done, so there was no change in that.
So I think this tough medicine, which is hard to administer, and societies only bear tough medicine once they see an apparent crisis, not something more insidious, like what I'm describing in the book, maybe yes, but there are things, as I said, whether it's got to do with the conduct of monetary policy, or it's got to do with regulation, or the culture of bailouts, I think these things can change, and the American voter will be much more receptive to it.
What's the option? That the politicians keep doing what they're doing today, the person in office will keep being unpopular and keep losing elections. As I pointed out in the book, as well, and in one of my FT [Financial Times] columns, that if you look at it, the number of politicians across the Western world who are getting reelected is declining. Typically, if you were in office, you had the platform, that if you stood for reelection, you'd get reelected. In the last few years, being an incumbent has become a disadvantage rather than an advantage. So clearly the people are unhappy, and if the politicians are going to keep doing the same thing, they should be more prepared to lose elections once in office.
Well, I think maybe they should be prepared to lose elections. Maybe they all should pray for an AI-driven productivity boom, that'll cover up a lot of bad policy.
Yes, we got that briefly in the late ’90s and early 2000s when you had a big tech productivity boom, which took place with the internet, but that has faded over time, and as I argue in the book that you can keep doing all this great technological stuff, but if you're going to undermine the fabric of capitalism, the creative-destructive fabric of capitalism with much more government intervention and more statist policies, then you should be prepared also to see a decline in productivity, and that will offset the gains coming from technology like AI.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
. But is there a consensus as to why Atomic Age 1.0 came to an end? Obviously we still get a lot of energy from nuclear, but not what people had imagined 40 years ago.
There are a variety of reasons. We did build a lot at one point, and we were building 10 plants a year, pretty extensive builds out there. We did then have Three Mile Island in the late ’70s, and then we got costs started going up, and schedules started increasing on the builds, and we ended up not having a lot of energy growth, in fact, we went for a long period where we weren't having a lot of energy growth, and we had a lot of other energy sources, natural gas, coal, and all. We had a lot of other energy sources out there as well. So yeah, we became pretty stagnated around 20 percent of the electricity. But now, like you say, yeah, there's been a big change in what we think the needs are for nuclear going forward, for a variety of reasons.
My background is journalism, and as a journalist I’ve written, I know, multiple stories in my life about a Nuclear Renaissance. So I'm wondering why this time looks to be different. You suggested in your previous answer that there might be some reasons. What are those reasons that we may be entering a new age where we will see an expansion in the nuclear sector?
I do think we will see that expansion, and, in fact, I think we have to see that expansion, and it's because of a lot of the positive attributes of nuclear right now. Obviously there's a lot of focus on trying to get more clean energy out there, and nuclear is a large base load source of clean energy. And it's not just CO2 emission, but it doesn't emit particulates and all, as well, so it's good air, good quality of life. So it has those key attributes. But there are other clean energy sources as well: renewables, hydro, and all that. But I think the recognition that, if you are going to go toward decarbonization, you need still base load electricity too. You need base load electricity to help intermittent sources like renewables to be able to expand more as well. So nuclear is very good at enabling decarbonization, not just by adding clean electricity to the grid, but enabling you to expand out other renewables like wind and solar and all, as well.
Additionally, nuclear is very reliable. Of the energy sources, it has the highest capacity factor of any of the energy sources. In the United States, we run 93 percent of the time, so the existing fleet that we have out there of 94 plants, they're producing a 100 percent of the power 93 percent of the time, which dwarfs what any other energy source does out there as well.
Nuclear is safe. At times people are concerned about safety, but, in reality, it's actually one of the safest energy sources out there and continues to demonstrate that.
It's resilient for different weather-related events. It can still produce electricity out there as well. It also has a lot of energy security. And as we've learned, unfortunately, from Russia's unprovoked and unjustified invasion of Ukraine, we recognize energy security is national security, so nuclear really does help us on that national security front. It provides an energy source that we can largely on-source from us and our allies. We’ve got assured fuel supplies, and provides that long-term power. You can put fuel in it and it can last for two years or so.
And I guess one other thing I'll add out there as well, is it's a job creator. Of the different energy sources, the amount of jobs associated with nuclear are some of the highest on the amount of electricity produced. And when you actually start building nuclear, like we saw in Vogtle in Georgia where they were building the two plants, it creates huge amounts of jobs. In fact, I heard a stat recently that 35,000 union workers were trained as part of the construction of the Vogtle power plant, so it's a good job creator in all, as well. And again, the power density is great, it doesn't take up a lot of space, and with the advanced technologies that we've developed in the United States, you've continued to increase in the safety, you can have plants of a variety of different sizes that can be easily deployed to, say, retiring coal plants. It just has a lot of flexibility that it hasn't had in the past, but also it's that key recognition of its clean energy attributes, but its energy security attributes as well.
Major concerns (7:37)
I did not major in nuclear science, I majored in history and political science, but I remember I took a class as an undergraduate at Northwestern University on the nuclear fuel cycle, and I remember to this day that my professor — of course, this was obviously a while ago, and I think what most of the students knew about nuclear energy was probably Three Mile Island — and I remember to this day distinctly the professor saying, “If they wanted to build a nuclear reactor in my backyard, I would be totally fine with it.” He had zero fear on the safety issue. Now when you give that rap that you just gave me about the wonders of nuclear energy before regular people, what is their response? Do they worry about the nuclear waste? Do they worry about safety? Are they immediately sold, or what are the concerns that typically get raised to you?
You brought them up. I mean, safety is brought up because you do see these high profile accidents like Three Mile Island, Chernobyl, Fukushima, which were accidents. They weren't good things that you want to have happen, but the industry's also a very learning industry. The improvements that come out of those events have just made the industry even safer and safer. And again, it's still safer than most any other electricity-producing industry out there as well.
Waste does get brought up. We have not implemented a final disposal solution for the spent fuel from our reactors, but we have safely stored and managed the spent fuel over the last six decades, and the amount of fuel that's generated, I think the stat that gets tossed around, you could fit it all in a Walmart parking lot. This is not a lot of material because it's a high energy-density fuel. It's not a lot of material, and again, we safely manage that and store that. We have countries now that are moving forward with geological repositories, which we need to be doing in the United States. In fact, just last week, I went and visited the repository that hopefully will be operating next year in Finland for disposing of their spent fuel. We can do that, it’s not a technical issue, so we can safely manage the spent fuel.
The other issue that always comes up is still cost. We do have to demonstrate now that we can build these plants safely, and efficiently, and at a reasonable cost. On the Vogtle plant there were cost overruns and schedule overruns, but between Vogtle Unit 3 and Unit 4, there was about a 30 percent reduction in costs between those plants, so we are starting to get to where we can be deploying nth-of-a-kind cost plants out there as well. And hopefully with some of the small modular reactor designs and all that are going to rely more on modular construction, we can even get to nth-of-a-kind cost even quicker. It still takes some pushing and understanding to make sure that people do understand the advancements that have been made on nuclear technology, that it's not our parents' nuclear technology, there's a new round of technology out there.
Out of practice (11:04)
You raised two good points there. The cost issue, and that's a great stat about the Vogtle plant and the reduction between the two reactors. Is it your sense that the fact that we haven't been consistently building reactors and learning from the previous build, and having trained people who've worked on multiple reactors, that each one has become like this bespoke mega project? It’s my sense, and it seems logical, at least to me, that that has been a cost driver, that we haven't been able to churn these out like 10 a year, every year, decade after decade, because clearly, if that was the case, I don't see how we don't learn how to build them better, faster, and more efficiently. But that's not what we've been doing, obviously.
That's right. It's not. Even when I say with Vogtle, you had to stand back up the whole supply chain, you had to retrain the workforce, so there was a lot of learning in that process, even though, too, we did recognize on that plant you need to have designs very well finalized and standardized as well. One of the problems we realized from the buildout of the 90-something plants that we have now is no two plants were ever that similar. Everyone wanted to make a tweak in their plant, so we never got to where we had standardized designs.
So I think now that we're getting that trained workforce, getting the supply chain up there, and our vendors are really saying, “We're doing standardized plants. If someone else wants to make a tweak on this plant, they have to go somewhere else,” that people are going to go with standardized designs so we can really replicate these and get that cost benefit from it. The challenges that you brought up, we have to overcome, and I think we're set up now to be able to overcome that. I appreciate all the effort that went into building Units 3 and 4 at Vogtle. We've got enough benefit from that learning there and hopefully build very soon here.
There's a world where we have tripled our nuclear generating capacity, as Secretary Granholm said. Can that be a world where we get all our nuclear power from light water nuclear reactors, or must there be different kinds of reactors? You mentioned the small modular reactors, and I've interviewed startups doing microreactors, I don't know, maybe they'll be used to power data centers, but can that world of greatly increased nuclear generation, even with improvements in light water reactors, must there be different kinds of reactors?
I wouldn't say “must.” I think there will be. I think we will have that variability. I think we will still have large plants being built. I think maybe five years ago you wouldn't hear that people were talking about building gigawatt-sizes plants again. I think we'll have the gigawatt-size plants, we'll have the small modular reactors that are water-cooled, but I think we will get some of those advanced reactors out there: the Generation IV reactors, the sodium-cooled fast reactors that have the capacity to be able to burn waste better and also increase the sustainability of the amount of fuel they use. I think you’ll also have the high-temperature gas reactors that are helium-cooled, that use TRISO fuel. You'll have those because we need to not only decarbonize the electricity sector, we've got to decarbonize the industrial sector. That's much more challenging, and the high temperatures that can be provided from those reactors will help us in that decarbonization process.
So I think we will have a mixture out there. There are cases where the Gen IV systems are going to be better than the gigawatt-sized plants for the needs that are out there, but large power plants are going to be needed as well. Especially, like you say, you bring up the data centers, the amount of growth that we're hearing for electricity right now, I think again, we'll see gigawatt-sized plants will be needed to be able to meet that growth.
Yeah, I tell you, nothing frustrates me more than reading about what AI could perhaps do for our economy and then having people say, “Well, but we know we can't do it because we can't supply the power” or “We can't supply enough clean power,” I mean, well then it’d be sure great to have more nuclear energy. And I wonder, as you sort of tick off some of the potential advances and new kinds of reactors, maybe I look backward too much, but I can't help but wonder what nuclear reactors would be like today, where we would be today, maybe we would already have fusion reactors had we proceeded with this kind of momentum every decade since 1980. It drives me crazy, and you're a nuclear engineer, that must drive you crazy.
It does, I've been doing this . . . my first job in the nuclear industry was almost 40 years ago when I was still in college, and there have definitely been ups and downs in funding. In fact, there were some periods where there was almost zero research and development dollars spent in the government on nuclear energy. Luckily, though, the thing that we have is, under the four presidential administrations, there's been a real steady climb in the recognition of the importance of nuclear, and the funding to support it. So I'm happy that we have had this period that goes back to the early 2000s that's been really steady growth in recognition of nuclear. If we would've not had some of those laws in the late ’80s and ’90s, yeah, we could probably be further ahead, especially on some of the advanced technologies. Because yes, some of those advanced technologies started on research that was back in the ’50s, ’60s and ’70s: the sodium-cooled fast reactor, the molten salt reactor, all of those were based on R&D that we did back in the early days, as well.
Next-generation policy (17:38)
Which leads me to this question: You work for the government. I work for a public policy think tank, so of course I'm going to think about: Given where we are today, what government needs to do going forward, both on the R&D front and on the regulatory front, are we doing enough basic research for whatever the next, or the next next generation of nuclear is, and do we now have the kind regulatory framework we need for that next generation of reactors?
I'll go to the research one first—and I should note, my background is, I'm an R&D person, I came out of the national labs, so of course we always need more research and development. But that said, we have been blessed by funding from Congress and the administration that there's a significant amount of money for research and development in the United States. And I'll say that's good, because the one thing I will note, I do believe innovation in the US, as far as the nuclear technology, we are the best. The technologies that we're developing and our vendors are deploying it, really, it is the cutting edge technology, so it's good we have that R&D, and it's important, as you know, we need to continue to have it to move forward on that next generation of technologies and continue to make improvements on the technologies out there. So I think we have a good research base.
There's some infrastructure that we still need if we start deploying, say, when we mentioned that sodium-cooled fast reactor, we don't have a testing capability for that type of system. We shut down our last testing system on a fast reactor in 1994. We would probably need some additional infrastructure. But again, we have a pretty good base. And I'll say that also on the regulatory side. We do have a pretty good base as well. The Nuclear Regulatory Commission is obviously focused on light water reactors throughout its history, but they've actually been doing a good job at being able to work with some of the developers.
We have three entities out there that are working on Generation IV reactors. TerraPower did submit their construction authorization to the Nuclear Regulatory Commission, and they've accepted it, so they're working well with them, even though they have a water-based system. Hopefully X-energy, who's doing a high-temperature gas reactor, working with the government and all, as well, will be moving forward, as well. And we've had a third that's working in the molten salt space, a molten salt-cooled reactor that has already received a construction permit to go forward on a prototype reactor, a Kairos company.
I'm sure there's got to be reforms still on the Nuclear Regulatory Commission and make sure that we are timely and responding to license applications, but they are moving in the right direction. There's been a lot of interface with various laws, whether it's the NEICA (Nuclear Energy Innovation Capabilities Act), or NEIMA (Nuclear Energy Innovation and Modernization Act), two bills that were passed a little while back looking at reforming. And I think there still needs to be improvements and still need to be increase in the resource and capacity of the Nuclear Regulatory Commission, but they're heading in the right direction.
We have a good regulator, and that's one of the things that helps us make sure we feel that we can deploy this technology safely here, but also helps us in exporting our technology, where we can say, “Our technology has been licensed by the Nuclear Regulatory Commission,” which has such a high view externally in other countries, that helps us. So I want them to continue to be that safe regulator, but again, they are continuing to work to improve and streamline the process. Hopefully we get toward where we're standardizing, that we don't have to have a lot of interface and we don't — that'll come to the utilities, too — we don't make changes once we've got something approved, so we hopefully can speed up the process from the utility side, and all is well.
Human capital (21:48)
Are we going to turn out enough nuclear engineers? I imagine that, for a while, that probably seemed like a hard sell to someone who had an interest in science and engineering, to be in this industry versus some others. Probably a little easier sell; are we going to have enough people going into that to build all these reactors?
We are going to need to continue to increase it. We’re already seeing the uptick, though, in that area. I'll note: Our office, the office of Nuclear Energy, we've really — going back to the 2010 timeframe — really recognized that we needed to do more in that area, so we actually started investing almost 20 percent of our R&D budget to the universities to hopefully foster that next generation. And in fact, this year we just hit the mark where we've now spent $1 billion since the start of those programs on the universities to make sure we're doing R&D there and getting that next generation of folks out there. It’s something that we've got to continue to focus on to make sure that we do. Because yeah, if we triple, it's going to need a lot more nuclear engineers. But I also note, the thing I'm concerned about also is making sure we have the right trades and all, as well. If we're building these plants, making sure you have the welders, the pipe fitters, and all, that's going to be a big challenge, as well, especially if we're going to start building, say, 10 plants a year. That's a lot of people out there.
Fusion forecast (23:12)
I’m excited about the prospects for nuclear fusion, and I've talked to people at startups, and it has probably looked as promising as it ever has. How promising is it? How should I think about it as being part of our energy solution going forward, given where we're at? In fact, there are no commercial nuclear fusion reactors right now. Obviously people at startups give a lot of optimistic forecasts. How should I even think about that as being a partial solution in the coming decades? How do you look at it, at least?
I think it can be part of the solution in the coming decades. I think some of the changes that's taken place, especially over the last two years where there is more of a change to focus on, not fusion as a science program, but fusion as deployment, as an energy producer, you look at it as an applied energy. I think that's an important change that's occurred over the last two years, and the fusion programs within the Department of Energy are much more focused to that. It's similar to what's happened somewhat with fission. Fission, about 15 years ago, it was government-driven, and you pull along industry, until about 15 years ago you started having industry investing a lot of money and pulling along the government. You're now starting to see that happen in fusion, where people are doing a lot of a private investment, they're pulling along the government, and the government's working to see, how can we use the resources of the government to enable it?
So I think it will happen. I don't think fusion is going to be producing electricity to the grid this decade, but I think the vision that's been put forth by the government is their bold, decade-old vision to have a fusion pilot facility sometime within the decade. I think that is feasible. So maybe before the 2050s you can start having fusion generating some of our electricity. I'm a fission person at my heart, but I think fusion is, we're getting much more focused on moving it forward as an electricity source, and that'll help it be able to be deployed sometime here in our lifetime.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
But the second part of your question alluded to not the potential, but actually the actual impact. And if there's one takeaway from that exercise from the paper that we put together and from my conversation with so many academics and policymakers around the world, is that this is not just a watershed moment, it's not just a pivotal moment in human history, it's a fragile moment as well. This story can easily be a story of missed opportunity. I think that we so easily take for granted the fact that, yes, we will of course develop AI and deploy it and apply it very successfully. And it's so easy to get caught in the moment, particularly as the nation that advanced the science, I think somewhere in the back of the minds of all of us, there's that presumption that we will be the global leaders in deployment of AI. I am actually worried about that. To ensure that we are, it’s a tumultuous, fragile, and careful process that we’ve got to be really thoughtful about, with a lot of deliberate action about what we do, how do we proceed, and how do we ensure that we are indeed the ones that capitalize on the potential?
It's remarkable how quickly the narrative around American tech has shifted. Not long ago, Silicon Valley faced criticism for focusing on social media rather than groundbreaking innovations like the Apollo program or cancer cures. Now, they've unveiled generative AI, potentially the most significant technology of our era.
Regarding fragility, it's worth considering why AI might need special handling. Unlike the seamless diffusion of technologies like the internal combustion engine or electricity, AI seems more akin to nuclear power - a technology that was stifled by regulation. Do we need a proactive agenda to prevent AI's potential from being similarly constrained?
That's a great question, Jim. I'm so tempted to go back to your first part of the question about the importance of digital technologies, and economists get a really bad rap, but try to be a librarian these days. We tend to overlook the tremendous importance of information as a driver of economic growth in our economies. And even if you look just at small businesses and the tremendous opportunities that digital technologies have provided them. To think that a mom and shop store today can actually run a marketing campaign, analyze its customer base on large databases, export products to far markets, those are things that used to be the exclusive domain of just a few large companies that today are actually available broadly and widely through digital technologies. And maybe it's the fault of economists that we are not shouting off the tops of mountains frequently enough about the tremendous power of information and digital technologies and the accumulation of knowledge as a driver of economic growth.
The application of knowledge and intelligence — that seems to me to be pretty important.
I cannot agree more! And I think, to a great degree, it explains some of the tremendous optimism around AI as a technology that really reduces the barriers to interact with technology and democratizes its use in a way that we haven't seen before.
Risks and benefits (7:46)
We quickly shifted from marveling at AI's potential to fixating on its risks — existential threats, job losses, and disinformation. But let's step back for a moment. Can you elaborate on why you see this as an exciting technology with significant benefits? It seems many people aren't fully aware of its upside potential.
That's a great question, this is really the reason why we at Google, too, we paused for a minute and kind of wanted to think about this. We're at a sector where enthusiasm is in no short supply, so what does it actually mean when we say that this is a pivotal moment in human history? What does it mean for economists? I think it really boils down to this question of: Is AI an ordinary technology, or is it really a general purpose technology? That is the term of art that economists use, and I think it's actually important to pause for a minute and think about that, because it's critical. A general purpose technology is not just pervasive in use, it is a technology that enables productivity-enhancing applications to be applied across all segments and entire economies in ways that are not just advancing and accelerating economic growth, but are also expanding the frontier of innovation and technology. It's a source of ongoing and continual innovations.
And if you think about it for a minute, if you think about the prior general purpose technologies that we've had, if it's electricity, if it's personal computers, or it's the steam engine, their impact was tremendous. And at the time that they were launched, I think nobody had the perfect vision of where . . . we of course knew where we started, in the very same way that we do today about AI, but it's really difficult, if not impossible, to know where we will end. The compounding nature of these technologies is immense, particularly when you're looking at a general use technology and multi-domain technology that can lead to applications on such a broad basis. I don't think that today we can envision what new occupations, new applications, new sectors will emerge as a result of AI. And I think the fact that it's not an ordinary technology, but rather a general purpose technology, that is important, that does imply that we're probably looking at the most profound economic transformation in our generation. That is huge.
It's relatively straightforward to assess AI's ability to replicate current human tasks. But predicting the new possibilities it might unlock, like accelerating scientific discovery, is far more challenging. These potential upsides are difficult to quantify or model economically.
While we can more easily grasp potential downsides like job automation (which isn't necessarily negative), the upsides are less tangible. They depend on entrepreneurs creating new businesses and scientists leveraging AI for breakthroughs. This makes it harder to definitively argue that the benefits will outweigh any drawbacks.
Oh my God, Jim, I cannot agree more, and I think that there's two issues, and you have written about this just recently, and I think that there's really two issues that come up, at least in my mind, as a reaction to some of the studies that really focus on the measurement. We're trying to really drill this question of, “What will be the productivity gain from AI over the next five or 10 years?” I don't want to dismiss that question —
And can you give it to me within three decimal points, right?
Exactly! But we're doing such a huge disservice as economists when we focus on that. I think it really pertains to two reasons that you brought up. The first one relates to measurement. We are really looking, these studies are primarily based on occupational exposure of existing work streams. Little do we know today about what new work streams, occupations, tasks, creativity, or human endeavors will actually be triggered by this new technology. In a way, we're really just looking under a flashlight rather than thinking about the broader issue, the broader economic benefits that will emerge, kind of like the unknown unknowns that we know today about this technology.
Just to put it in perspective, think about the printing press that led to a scientific revolution. Think about the steam engine that led to an industrial revolution, to an electronic circuit that led to the digital age. We are at that point with AI today, and to think that we're looking at SOC, standard occupational codes, to look at the future impact on productivity, I think minimizes the value of our profession.
The other point that you touched on, which I think is so incredibly important: We're missing the point. It's really not about the third decimal point of our estimates. It's about the fact that we can reshape technology. Rather than measuring its benefit, let's actually make sure that we can capitalize on the potential. That is far more important than anything else. And at some point, we'll go back to your other question about fragility, but there are genuine barriers that we need to address collectively as a society. And if we are not going to do it, other countries will, right? And I think economists have a role in that conversation. I think that is the critical issue that we need to focus on.
Barriers to a boom (14:27)
We have a technology that seems, right now, it's fast evolving, but it seems pretty darn important. It's hard to believe that we've only really been having the specific conversation about generative AI for maybe a year and a half or so publicly. So what are the barriers? If this turns out not to be an important technology that's widely diffused throughout the American economy, what went wrong? What are the barriers that concern you?
I think there are three main categories that we focus on. First and foremost, you need digital infrastructure. I think it's a misconception, and I think we will learn that very quickly over the next few years, that AI or digital infrastructure is limited to broadband. It is increasingly becoming more so about access to data, large data centers, and compute power. And I think not just the US, but many other countries, will realize, or are in the process of understanding very soon, that those are the type of investments that one needs to make in order to deploy the technology. That's one category.
Another one is the regulatory environment and the legal standards. You know Jim, and this is something you've of course written about a lot, I don't think that no single country deliberately chooses to fall behind, and I think that we often fail to recognize the long-term impact and unintended consequences of regulations. We, of course, have a duty to protect, and there are areas that raise concern, but we have to balance that duty to protect with the desire to capitalize on the potential, to foster innovation, and to make sure, at the end of the day, that we emerge as the global leaders of this technology, that we lead its deployment. I think that the legal ecosystem is incredibly important in that respect and an important current dimension of competition between countries and the future over the deployment of AI.
And the third one is our workforce readiness. We need a workforce transition strategy. Let me pause here for a minute. If our workforce is not ready for an AI transition, our employers and our companies would find it very difficult to actually implement and adopt AI. It's simple as that. And if our companies do not adopt AI applications or technologies, we will quickly find out that we will fall behind. If you're looking at the history of our labor markets, we have been not just resilient, consistently resilient in our institutions and labor market operations, but we've also been highly effective at transitioning individuals from low-productivity to high-productivity occupations.
We used to be a primarily agrarian economy in the 19th century. We transitioned successfully to manufacturing, which at some point was about 27 percent of our workforce, now it's below 10 percent. From there, we switched on to services. We absorbed women into the workforce in an effective way. We have highly effective labor markets, which is a competitive advantage when we're thinking about global competition.
At the very same time, it's not without a cost. And in a lot of ways, I do think that it can be a double-edged sword, because the competitiveness of our labor markets also implies, at least factually, that the relationships that we have between employer and employee tend to be less permanent than they are in other economies, and that implies that employers have less of an incentive to actually invest in employees. That may put us at a relative disadvantage compared to other countries that have longer relationships between employers and employees and can afford for actually employers to participate, to take part, whether it's through apprenticeships or training programs, in making sure that their workforce is ready for an AI transition. That, Jim, worries me. I think it's more than just making sure that individuals that may lose their job get reinstated in the workforce, it's really an economic strategic objective for us. Unless we take care of our workforce, we will find it exceedingly difficult to implement AI on an economy-wide basis.
Investing in employees (19:16)
While Washington isn't dictating data center construction, companies are investing heavily in this infrastructure. Shouldn't the same logic apply to workforce development? If understanding and working with AI technology is crucial for business survival, there's a strong private incentive to invest in employees' skills. This holds true even considering the unique structure of the American labor market compared to, say, Europe's.
Let me pause for a minute, take it back one step so that we can think about why is investment in worker training and vocational programs are so difficult? Why are they so challenging, and why do they perhaps create externalities, the way that we just discussed, more broadly? Why are they ultimately a strategic concern for the US economy? So look, these programs are exceptionally difficult to get right in an ordinary course of business. And we at Google have invested a tremendous amount of resources on these programs, which are not a core product for us. They're not even a monetizable product for us. And we're not the only ones. A lot of other tech companies have done the same, to be honest.
Now, what are the challenges with these programs? First and foremost, they have to be relevant and they have to provide education, skills, programs that are actually relevant, that keep up to date with the advancements in technology. That is something that's really difficult to do. You have to make sure that employers are actually buying in. We may have the best program, but unless it enables the individuals who graduated from the program to signal to employers that these are individuals with high qualifications because they went through a program, let's say at Google, the program is simply not going to work.
And then the third thing, think about it from the employee perspective: For an individual, it's not about giving somebody a pamphlet, “Hey, let's participate in this great program.” It's really about, can you take time off, at a tremendous opportunity cost of time with your family, career, and work, to invest in a serious program that would end up in an outcome which actually lands you at a better career, more stable job, that is better paid. Those are tremendously difficult in the ordinary course of business, let alone when we're going through a transition where we don't even know today how tasks and occupations will evolve. Now, as I mentioned, private tech firms, because of also market expertise and access to occupational data that is far better, in a lot of ways, than what the government data that we have on occupations are perfectly positioned to carry out those programs. The question is whether they can be actually carried out independently, unilaterally, without the collaboration of federal agencies, whether it’s local governments, state or federal, without the participation of employers, colleges and other institutions —
It sounds to me like employers will have to be part of this.
For sure. Jim, maybe let me just mention one thing that we don't want to do. We've been in this movie before. Following NAFTA (North American Free Trade Agreement), we had the trade adjustment programs where we invested a lot of money in rescaling and training employees, and the results were very minimal, at best. So I do think that this is the type of a grand challenge, if you will, that no single actor can really solve independently. And I know that we're naturally, as economists, we're naturally hesitant about government intervention, but what better role for a government can you think of other than identifying a market failure that is of strategic importance to the US economy, and in a thoughtful way, collaborating with other relevant constituents to come up with solutions that are effective. Scaling those programs to a national level is going to be a real challenge. And I do think that there's a role for governments to actually lead, in collaboration with other constituents, those efforts.
And of course you are aware of the sort of deep skepticism among people about these programs.
Yes.
So obviously we talk about innovation, technological innovation, we also need program innovation, education innovation here.
Jim, I’ve got to be perfectly honest here — and this is just my individual experience — as economists, I would be lying if I wouldn't say that I share that skepticism and concern. At the very same time, I think that we need to consider the other ramifications and what is truly an issue. We are at a certain disadvantage because of the lower incentives that our businesses have, our employers have, to invest in employees, and we see it. Apprenticeship programs is one example that is working phenomenally well in other places, but not in the US. So I do worry about that. In the paper, we didn't offer any descriptive solutions, but we really highlight the challenge here: How do we find market-based, thoughtful solutions to scaling and vocational programs so that our workforce can be ready for an AI transition?
Human-complimenting AI (25:29)
I'm skeptical of the idea that we can guide AI development through policy to ensure it complements rather than just automates human work. It's unclear what policy levers could effectively achieve this — tweaking the tax code seems unlikely to produce specific AI outcomes.
But where we can make a difference is in human capital development. If we want AI that complements human skills and enables new business creation, we need to ensure people understand this technology. Currently, many don't, given its novelty. Focusing on education and skill development seems a more practical approach to shaping AI's impact.
You know, Jim, it’s really interesting, so Chris Pissarides, the Nobel Prize winner from London School of Economics, has a phenomenal paper about this question. He comes up with a very interesting finding that countries that actually invest in right regulatory environments and legal standards, that have the right infrastructure, that have the right environment to foster innovation, ultimately witness less concerns about substitution because the technology that's being advanced tends to be more complimentary, and as an economist, that makes a lot of sense to me.
Let me pause for a minute and explain why: There is a genuine concern about whether AI is being deployed or used to substitute labor. And I think if you think about the Turing Trap that Eric Brynjolfsson has written about, this notion that you can come up with the most myopic, plug-and-play, cheapest AI application, put that in some individual function in your business, and replace existing work streams that are being done by humans, that is a genuine concern, particularly for firms that are looking for the highest rate of return, at the lowest cost, without really investing and transforming their business.
As an economist, I can understand why that happens, but keep in mind that when that happens, those businesses are actually really failing to leverage and capitalize on the full potential of the technology. They're really going for the plug-and-play, cheapest applications. That's not good for labor because it leads to substitution. But certainly that's not good to the business, itself. In a competitive market—and I think one thing that we need to stress is the importance of competition in our markets—you'd anticipate that firms that actually go through the effort to invest, to reform, to transform their businesses, to reinvent themselves, are the ones that will prevail.
And I would think that would be a very powerful lesson for other businesses if that is indeed the case, right?
Exactly. And the question is, how can we promote this broader, more meaningful, more valuable application and adoption of AI? And I think it goes back to the fundamentals: You need the AI infrastructure, the right legal institutions and regulatory standards, and ultimately a workforce that is ready to transform. And I think once you put those together, I do believe (and I’m deliberately saying belief, because I don't know that we can really study this explicitly) that that will lead to more complementarity and augmentation, and less substitution.
Historical precedent suggests that extreme job loss scenarios, like robots taking all jobs, are unlikely. While AI will undoubtedly cause disruption, do you believe it will follow the pattern of past technologies? That is, replacing some tasks, enhancing others, and creating entirely new job categories. Given the policies we've discussed, are you confident that this balanced outcome is achievable with AI, or do you have doubts?
Oh my God, that's a tough question, Jim. You saved it for last.
I'll add an addendum, I'll add a qualifier: within the next 20 years. I don't know what it'll look like in 100 years from now, but within our immediate lifetimes as workers, you and me.
Let me address it in this way. Yes, absolutely, I am, I want to say, cautiously optimistic, because if we learned one thing from the last century, a period of time that reflects the most advanced technological progress in human history, is that we didn't witness an increase in unemployment and we didn't witness a decline in labor participation. That leads me to be optimistic about the future of AI as well. Having said that — I think you alluded to this — history doesn't always repeat itself, and we've never faced a technology that can automate such a wide range of human tasks and activities. So that should be concerning for us. We also should mention that, even if AI does not lead to mass unemployment or to net loss of jobs, there will be significant occupational and sectoral shifts if we get this right, which I am optimistic about us doing so. So that will be something that we will need to consider as well. So I would say, optimistic: absolutely. Cautiously optimistic: that's probably more correct.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Pethokoukis: You obviously love skyscrapers, you're fascinated by them. You wrote a whole book on them. So I want to just start the very basic question: Why do skyscrapers fascinate you, and the people who aren't fascinated by them, what are they missing?
Barr: Great questions. Well, I grew up on Long Island, and so I was always really fascinated with Manhattan. I grew up in the ’70s, and so New York back then was a very dark, mysterious place for a youngster. So when I grew up, actually when I was in college, I started hanging out in the city. So to me, the skyline of Manhattan and New York City, they're just two sides of the same coin. I really developed an interest in tall buildings through my interest and fascination with Manhattan’s and New York City's history.
So when I came to Rutgers Newark, I just started doing research on tall buildings, especially in New York City: what was driving the heights of these buildings; there's all these interesting height cycles over the last 150 years. So I wrote my first book on the Manhattan skyline, that was calledBuilding the Skyline, and then after that I thought, let's see what's happening around the rest of the world. So to me, the tall building is an interesting thing because it's part and parcel with urbanization, and I just personally don't think you could have one without the other.
I think some people might think that skyscrapers are, at least for rich countries, that they're kind of a 20th-century thing that we did as we were growing, and cities were getting bigger, and skyscrapers are a part of that, but now they're for other parts of the world, parts of the world which are still urbanizing, which are still getting richer. Are skyscrapers are still a thing for America?
The short answer is yes, but, given how dense cities are, tall buildings are just being added a lot more slowly. In New York, the population's kind of slowly growing, and so tall buildings are either replacing old buildings that are wearing out, or there's always this push by big global corporations to be in the newest and latest tall building. And obviously there's this international demand from people abroad to have an apartment — or national demand — global demand to have some kind of residential presence in New York. But the thing is, people in other countries: cities, planners, residents in other countries, they look to New York, they look to Chicago, and I think, for many of them, they see New York as something they want to emulate, and New York is, on just about almost any metric, it's probably the top global city. And so I think cities today, especially in China, and Asia more broadly, they're trying to kind of replicate that, what you might call “the Manhattan magic,” and I don't really think people in this country realize how much tall building construction is going on in other cities around the world. People in this country are a little bit more cynical about the role of the tall building in urban growth and in housing affordability and stuff like this, but other cities are basically going gangbusters, is a way to put it.
Is that driven by fundamental economic forces? Is it kind of a “national greatness” kinds of signaling projects? Are there fundamental reasons, not just to build skyscrapers, but to build very, very tall skyscrapers?
“All of the above” is the answer. Fundamentally, if there's many, many people who want to be working, living, playing in the center, the only way to accommodate the demand to be in the center is to make more land in the center, so the skyscraper, at its heart, is what I would say is “land in the sky.” You just go vertical because there's constraints on how much land there is in the center.
Having said that, definitely the skyscraper is seen as a kind of way to advertise, a way to increase confidence in the place, and so you boost foreign direct investment. Observatories are huge money makers, there’s a big tourism component. A lot of critics will say, “Oh, it's all about spectacle and ego.” But really, for the book, and just more broadly my research, when you drill down on the economics of these super tall buildings, not all of them are profitable or profit-maximizing, but they all have a strong economic rationale.
Now, I just also want to say, China has its own thing going on, which sort of compounds the skyscraper construction-building there because of their unique governance structure and land ownership structure, but China is building tall buildings because, at the end of the day, there's a kind of, what I call, a “tall building bling.” There's just something that says, “This city is growing, this city is drawing population.” So we build a tall building and we boost confidence in the city. And it works, really.
The pictures don't have to be too old, if you look at a picture of Shanghai, it looks a lot different not too long ago. It's almost as if a whole other city just kind of fell from the sky, a city of skyscrapers, and where there were once goats or something grazing, there's now a bunch of massive skyscrapers.
Yeah, absolutely, and there's a few reasons for this. One is, I think Chinese residents more broadly see tall building as a natural way to live. I've talked to many Chinese residents, whether it's Shanghai or other cities, and to them, to own an apartment in the sky is like the greatest thing. It's their equivalent of the single family home in the United States. Living in the clouds is something many people aspire to. The other aspect of it is, Shanghai, and the Pudong neighborhood in Shanghai, was chosen basically to become a financial hub. Basically, the leaders were looking at Hong Kong and they thought it was a, to quote, I forgot the author, but to quote him in the book, the Shanghai officials and the National Party officials saw Hong Kong as that frustratingly free city, and so they wanted to create a kind of a financial hub in Shanghai. And so the Shanghai Tower, for example, is part of that plan to really draw people's attention to Shanghai, itself. So it was part of a master plan.
The end of the skyscraper (9:00)
I certainly remember that, after 9/11, I heard about “the end of the skyscraper,” and then during the pandemic, I heard about “the end of the city.” Now I'm guessing that cities will continue to exist and we're going to continue to build tall buildings.
Absolutely. What 9/11 did was just make sure that we make our building safer with fire protection measures. In many Asian countries, every 20 floors, let's say, are mechanical floors, so you have the electric equipment, and the heating, and the cooling, and water tanks. They can also surround these in concrete, and so if something's on fire, if a floor is on fire, they can go to this hermetically sealed floor, a refuge floor, and stay there and be protected. And the elevator cores, they're made of concrete, and so you wouldn't have something like what happened on 9/11. So it didn't really impact the demand; 9/11 didn't impact the demand for the tall building, it just made us make tall buildings safer. And of course the downside is if you want to go into an office building, you have to have a swipe and you have to have an entry, so the negative of 9/11 was more about heightened security and increasing protections in a way that engenders a little bit more mistrust of us. But the demand didn't go away.
Same thing with Covid. For big cities like New York and San Francisco, I'm sure the empty-office problem is going to dissipate. It'll take a while. This may be an overly broad statement, but the truth is, our present and future is in cities. The funny thing about the internet and social media and all that, it was supposed to allow us to suburbanize more, or run away from these big, overcrowded cities, but the truth is, social media and internet technology has just made cities even more important. So, as long as cities are growing, there'll be a demand for tall buildings
Of the tallest, I don't know, half-dozen buildings, have you been to all of them?
That's a good question. I've been to the Shanghai Tower, which was the second-tallest building in the world, now is the third-tallest. The one that replaced it, I think it's [Merdeka 118] in Kuala Lumpur, I believe. I didn't go to that one yet because that just opened up recently. I've been to the Burj Khalifa, which is the world's tallest building. I'd have to look at the list. I've been to the Sears Tower, Empire State Building . . . Anyway, so I've been to a handful of them. I can't say I've been to every single one of the super-tall buildings in the world
And in any of those super-tall buildings, can you open a window? Why can't you open the windows in these skyscrapers?
Well, the wind forces are just tremendous! The biggest problem engineering tall buildings is making sure that the building doesn't sway so that people feel it. The really fascinating thing about engineering tall buildings is this question: How do you allow the building to sway enough so that you don't have to — you don’t want to over-engineer a building so that you make it perfectly stiff because that's just completely uneconomic to do that, but you want to make sure the building sways just enough so if you're sitting there reading a newspaper or drinking a cup of coffee on the top floor, you don't feel it. And so the wind forces high, a thousand feet in the air, are just so tremendous. I think if you open the window, everything would just would just blow away.
I was thinking about some of those very, super-skinny residential buildings, which I guess seem to be becoming more popular, and do those people really feel the motion?
From what I can tell, the short answer is no. There's one lawsuit in Manhattan, in particular, where the engineering wasn't exactly perfectly right, but I think that represents the exception that proves the rule. The building is safe, that's not the problem, it's just that, when you're dealing with these super-skinny buildings — these are kind of a new kind of breed of super-tall buildings, so sometimes the engineering isn't perfectly right, so they will figure out ways to kind of fix those problems. The problems are solvable, but sometimes if you don't get it 100 percent right, people complain, and obviously there's lawsuits and you have to go back in and tweak the engineering. But these things are selling for 70, 80, 100 million dollars for a penthouse on the 90th floor, so people still value them, and if motion sickness was a problem, they'd be worthless.
Pillars of commerce (14:05)
In the book, you run through a number of myths: tall buildings being only for the rich, that they drive up housing prices in cities, again, that you mentioned a little bit earlier, that they're somehow bad economic deals. All these myths all tend to be very negative.
I'm not going to rename your book, but I could call it “Cathedrals in the Sky,” I mean, I think these are beautiful buildings that say a lot about human aspiration and to create a sense of awe. Boy, but some people just do not see it that way.
I think there's a few strands; I've been thinking about this. There's a kind of a NIMBY strand, and sort of a NIMBY/gentrification strand. So people in the middle income, let's say, they see their housing prices going up, their rent going up, and then they see these billionaire condos, and so they, in my opinion, or based on my research, there's a confusion of correlation and causation. So the most visible manifestation in people's minds of gentrification and affordability problems are the super-slim buildings, but New York City has something like 3.6 million housing units, and if you look at the outlying areas of Queens and Staten Island, they're just covered in one- and two-family homes. Those neighborhoods have added barely any housing. So all of the housing — I'm exaggerating here when I say the word “all,” but the vast majority of new housing units happens in the center where either the zoning is more permissive, or old industrial sites come online and things like this, so people don't realize that the problem of housing affordability is citywide, it just looks naturally to be in their neighborhood where high rises are going up.
Then there's another strand, which I would say is kind of the “Jane Jacobs strand” / the anti-public-housing strand. Jane Jacobs has some great points in her book,The Death and Life of Great American Citiesabout walkability, about eyes on the street. She wasn't a big fan of tall buildings, and this has kind of given rise to this whole movement of “human scale,” where five-story Greenwich Village buildings, or 10-story Parisian mansard-roof-type buildings are perfect, and any other deviation from that is somehow destroying the city. So there's that part of it, that people see tall buildings as somehow destroying the feel or the perfect fabric of the city. And lastly, obviously, some of the failures with the public housing policy has made people convinced that it's unhealthy to live in these tall buildings. I think that gets at what you're inquiring about. I think there's those different strains.
I wonder if part of it stems for a confusion about what are cities for, and I mean cities are, for a large part, are where people to come together for jobs and to conduct commerce. And if you think of them that way, then certain things make sense; but if you think of them as, I don't know, some sort of urban retreat, where it's kind of like a garden or . . . I don't know, but it's a very different view, and perhaps it is not just about bike paths, but it's about what facilitates people to connect.
Without jobs, without a labor market, there's no purpose to have a city. Maybe in the 18th or 19th century, you can create a city for the king or the empire, as are many examples, or the Vatican or something like this; so you can have these sort of political capital cities, or even Washington DC, but, fundamentally, 99 percent of the world’s cities are places where people go to work, and so, if you don't allow the labor market to function properly, which means having a functioning housing market, then all these ideas about “the good city” and “the perfectly crafted city,” they kind of are irrelevant.
So you have to start with: what makes a city grow, what makes people productive, and then how do we accommodate that? To the extent that we can improve design, all the better. There's always a million ways to make things better for people design-wise. I think bike lanes are great, and I think pedestrian-friendly cities are better than car-centric cities, but you can't start with designing the city first and then seeing what happens. You have to start with “let's make an attractive place to live and work” first, and then work on the design feature second.
The sky’s the limit (18:36)
How tall are these buildings going to get?
Okay, well, the next world's tallest building is going to be one kilometer: The Jeddah Tower, which had started, I think back in 2013 or 2015, had been stalled, there was some sort of political turmoil in Saudi Arabia, and they've just restarted this Jeddah Tower in the city of Jeddah. And so when that's completed, that's going to be one kilometer. There were some plans floated to have a two-kilometer building in Riyadh. I don't think anyone really thinks that's going to happen.
How long does it take to get up to your office in a two-kilometer building?
Well, that's the thing. They're coming up with new ways to get people up there faster. The old conventional steel cables could maybe go 500 meters or something like that, which is maybe 80 floors or something. Maybe if you had a really good cable, you can get people to 80 floors and then they'd have to switch. Now they have these composite . . . it’s KONE UltraRope, which could go 1000 meters, which could go basically one kilometer continuously. So if you can get people from the ground floor to wherever their destination is within a minute, that's kind of like the golden rule here. People are not willing to wait more than a minute once they get in the elevator. The trick really is the ear pressure, and that's probably the hardest part because you're going up so quickly, the air pressure changes, so you have to figure out ways to make sure the cabin remains pressurized, and then there's the air pressure up on the highest floors. So that, I would argue, is the fundamental issue that's going to be coming next on the horizon is how to efficiently pressurize the highest floors. Let's say you're a mile high; if you're a mile high on the top floor, that's the equivalent of going from New York to Denver in a minute, or two minutes. So you have to figure out a way how to pressurize the entire building so it has a constant air pressure.
If I were to look at the skyline of major American cities 50 years from now, would you expect them to be radically different, futuristic looking, maybe not two-kilometer buildings, but a lot of very, very tall buildings? Or is it again, if they're not growing, if population isn't growing, then that won’t happen?
People are always asking me what I think about the doom loops and all that. Pick New York as one end of the spectrum: It's always going to be adding new buildings, that's just in its DNA, and so you're going to have this kind of collage of different building styles. But other cities, smaller cities, maybe where people are moving now because working from home, they'll add a few tall buildings here, they'll have mini-skylines. Then the other cities, like a St. Louis, that's just going to have to kind of figure out a plan for growth. So I don't see the world as a kind of Jetsons-type world.
I mean, you never know what's going to happen with the technology. There’s one company, TK or Thyssenkrupp elevators, they're working on Maglev elevators, and this can actually be a game changer because you have these shafts, so the Maglev elevator cars, they can go up or down or they can go horizontal. Part of the goal with that is that everybody has their own — if it's an apartment building, they have their own elevator car, it takes them up to their apartment, it becomes the door. So that could be a real game changer . . . And then you could run these things horizontally. So if you have these Maglev elevators, you can not only run them horizontally within the building, but, in principle, if you could work out property rights or whatever, you could connect these things across buildings. But at the end of the day, it's really about preferences and a kind of cultural perception of the tall building, and I just don't see us in the United States us having a dramatic, country-wide rethinking of where we live. There's always going to be this desire for the single-family home in the suburbs. Now maybe that'll diminish to some degree, but as long as people see their own little house as their own little castle . . .
Unlike China, where there seems to be a great desire to live in these kinds of buildings.
Manhattan extension (23:04)
Have you had any takers about your proposal to make Manhattan bigger?
No.
You would extend it by about 2000 acres and maybe build some tall buildings on that, I don't know.
The idea would be to create a new mini-Manhattan extending Manhattan into New York Harbor. Just briefly, the idea was both to add more housing and add more land, and to protect lower Manhattan against sea level rises and so forth. I proposed this in a New York Times op-ed piece, and, naturally, I would say the majority of commenters and people had this sort of kneejerk reaction against it.
I had a kneejerk reaction for it! I loved it!
You are part of a small, select core of appreciators, let's say. Having said that, in the 21st century it’s just not something I think most people are willing to wrap their heads around. So
To me, that's an idea with the future, and I think you should not be dejected that it was not initially well-received. I think that kind of idea might actually have some legs.
Trends and styles (24:23)
Finally, let me ask you, whether it's because of computers or new materials, would we expect skyscrapers in the future to look any differently? I think some people would love to go back to the 1930s style. They love that style of skyscraper, and they don't like the glass-and-steel, very rectangular skyscraper; they want it to look like Gotham City or something.
Actually, if you look in Manhattan, in Brooklyn there's one, I think they're calling it something like the “Dark Knight Tower” or the “Gotham Tower.” It's in Brooklyn and it has this almost art deco sort of —
It slipped my mind, I was thinking art deco, yes.
And there's a high rise apartment near Columbia University, which uses the same color masonry as the surrounding buildings. I think it's the Union Theological Seminary, which sold some of the land to build a high rise. It sort of blends in. So I guess the question is really architecturally speaking, and it's sort of hard to say. I think maybe there'll be some neo-historical buildings coming up here or there, but there's two things: One is that people like glass windows. People love to have light and views, and so that's really just pushing the glass buildings. I think developers like glass too, because it's easy to work with, and architects — if you're a developer and you want a super-tall building, you usually go to a handful of architects and you have some kind of design competition, and, chances are, you're going to get something that looks full of glass and has some funky geometry to it.
But they seem more twisty than they used to, so they're not just perfect rectangles.
Right, so you're creating a lot of illusion. The interesting thing is, at the end of the day, you can only have certain internal shapes because you need functional spaces, so you have to have illusion with the twisting and these sort of Jenga towers, and a lot of that is due to massive improvements in computer technology; so the rendering software has dramatically improved, the engineering know-how, the engineering technology improved, you can send your designs right to the manufacturer where they can then use the computer programs to design exactly the shapes and sizes.
So it's the learning curve of every building that you do adds to the knowledge of how to do something a little bit different, or some version of something before, and also just massive computer power. I think there'll be a lot more of these sort of funky architectural shape. How they hold up, only time tells. In the ’80s there was this massive postmodern boom with all kinds of pastiche-type buildings with all kinds of references to old buildings, and funky buildings, and some of those haven't held up as well.
Frankly, I'm from Chicago, and I know exactly what you're talking about. Also being from Chicago, I appreciate you calling that building the Sears Tower rather than what other name they try to put on it. Last question: Do you have a favorite skyscraper?
I’m from New York and I like the Empire State Building, and it's not just because architecturally a classic building, but it speaks to New York as a city of strivers. And the more research I did into the Empire State Building, the more I appreciate the sheer guts of these guys who built this building. And the thing is, when it was completed in 1931, Great Depression was really starting to kick into high gear, and so the building was unrented, and it kind of gave this whole mythology about how these guys didn't know what they were doing, but when you crunch the numbers, they knew exactly what they were doing. They knew what the landscape looked like for New York, and the costs, and the revenues. Nobody saw the Great Depression coming, and so to say that the Great Depression showed how foolish these were, I just think it's a bad standard to hold them to. And if you look at the revenues and them building value over its 90-whatever, 93-year history, it's been a money-maker for almost a century. After the Great Depression, it recovered and has become an icon and a moneymaker, so what's not to love about that?
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Pethokoukis: The Dutch futurist, Frederik Polak famously put it that any culture without a positive vision of the future has no future. It's a light paraphrase. And I kind of think that's where we are right now, that despite the title of your book, I feel like right now people can only imagine dystopia. Is that what you think? Do I have that wrong?
Bostrom: It's easier to imagine dystopia. I think we are all familiar with a bunch of dystopian works of fiction. The average person could rattle offBrave New World,1984,The Handmaid's Tale. Most people couldn't probably name a single utopian work, and even the attempts that have been made, if you look closely at them, you probably wouldn't actually want to live there. It is an interesting fact that it seems easier for us to imagine ways in which things could be worse than ways in which things could be better. Maybe some culture that doesn't have a positive vision has no future but, then again, cultures that have had positive visions also often have ended in tears. A lot of the times utopian blueprints have been used as excuses for imposing coercively some highly destructive vision on society. So you could argue either way whether it is actually beneficial for societies to have a super clear, long-term vision that they are staring towards.
I think if we were to ask people to give a dystopian vision, we would get probably some very picturesque, highly detailed visions from having sort of marinated in science fiction for decades. But then if you asked people about utopia, I wonder if all their visions would be almost alike: Kind of this clean, green world, with maybe some tall skyscrapers or something, and people generally getting along. I think it'd be a fairly bland, unimaginative vision.
That would be the idea of “all happy families are alike, but each unhappy family is unhappy in its own unique way.” I think it's easy enough to enable ways in which the world could be slightly better than it is. So imagine a world exactly like the one we have, except minus childhood leukemia. So everybody would agree that definitely seems better. The problem is if you start to add these improvements and you stack on enough of them, then eventually you face a much more philosophically challenging proposition, which is, if you remove all the difficulties and all the shadows of human life, all forms of suffering and inconvenience, and all injustice and everything, then you risk ending up in this rather bland future where there is no challenge, no purpose, no meaning for us humans, and it then almost becomes utopian again, but in a different way. Maybe all our basic needs are catered to, but there seems to be then some other part missing that is important for humans to have flourishing lives.
A utopian thought experiment (5:16)
Is your book a forecast or is it a thought experiment?
It's much more a thought experiment. As it happens, I think there is a non-trivial chance we will actually end up in this condition, I call it a “solved world,” particularly with the impending transition to the machine intelligence era, which I think will be accompanied by significant risks, including existential risk. In my previous book,Superintelligence, which came out in 2014, focused on what could go wrong when we are developing machine super intelligence, but if things go right—and this could unfold within the lifetime of a lot of us who are alive on this planet today—if things go right, they could go very right, and, in particular, all kinds of problems that could be solved with better technology could be solved in this future where you have superintelligent AIs doing the technological development. And we might then actually confront the situation where these questions we can now explore as a thought experiment would become pressing practical questions where we would actually have to make decisions on what kinds of lives we want to live, what kind of future we want to create for ourselves if all these instrumental limitations were removed that currently constrain the choices set that we face.
I imagine the book would seem almost purely a thought experiment before November 2022 when ChatGPT was rolled out by OpenAI, and now, to some people, it seems like these are questions certainly worth pondering. You talked about the impending machine superintelligence—how impending do you think, and what is your confidence level? Certainly we have technologists all over the map speaking about the likelihood of reaching that maybe through large language models, other people think they can't quite get us there, so how much work is “impending” doing in that sentence?
I don't think we are in a position any longer to rule out even extremely short timelines. We can't be super confident that we might not have an intelligence explosion next year. It could take longer, it could take several years, it could take a decade or longer. We have to think in terms of smeared out probability distributions here, but we don't really know what capabilities will be unlocked as you scale up even the current architectures one more order of magnitude like GPT-5-level or GPT-6-level. It might be that, just as the previous steps from GPT-2 to GPT-3 and 3 to 4 sort of unlocked almost qualitatively new capabilities, the same might hold as we keep going up this ladder of just scaling up the current architectures, and so we are now in a condition where it could happen at any time, basically. It doesn't mean it will happen very soon, but we can't be confident that it won't.
I do think it is slightly easier for people maybe now, even just with looking at the current AI systems, we have to take these questions seriously, and I think it will become a lot easier as the penny starts to drop that we're about to see this big transition to the machine intelligence era. The previous book, Superintelligence, back in 2014 when that was published—and it was in the works for six years prior—at that time, what was completely outside the Overton window was even the idea that one day we would have machine superintelligence, and, in particular, the idea that there would then be an alignment problem, a technical difficulty of steering these superintelligent intellects so that they would actually do what we want. It was completely neglected by academia. People thought, that’s just science fiction or idle futurism. There were maybe a handful of people on the internet who were starting to think about that. In the intervening 10 years, that has changed, and so now all the frontier AI labs have research teams specifically trying to work on scalable methods for AI alignment, and it's much more widely recognized over the last couple of years that this will be a transformative thing. You have statements coming out from leading policy makers from the White House, the UK had this global summit on AI, and so this alignment problem and the risks related to AI have sort of entered the Overton window, and I think some of these other issues as to what the world will look like if we succeed, similarly, will have to come inside the Overton window, and probably will do so over the next few years.
So we have an Overton window, we have this technological advance with machine intelligence. Are you as confident about one of the other pillars of your thought experiment, which is an equally, what might seem science-futuristic advance in our ability to edit ourselves, to modify ourselves and our brains and our emotions. That seems to hand-in-hand with the thought experiment.
I think once we develop machine superintelligence, then we will soon thereafter have tremendous advances in other technological areas as well because we would then not be restricted to humans trying to develop new technologies with our biological brains. But this research and development would be done by superintelligences on digital timescales rather than biological timescales. So the transition to superintelligence would, I think, mean a kind of telescoping of the future.
So there are all these technologies we can see are, in principle, possible. They don't violate the law of physics. In the fullness of time, probably human civilization would reach them if we had 10,000 years to work on it, all these science fiction like space colonies, or cures for aging, or perfect virtual reality uploading into computers, we could see how we might eventually . . . They're unrealistic given the current state of technology, but there's no (in principle) barriers, so we could imagine developing those if we had thousands of years to work on them. But all those technologies might become available quite soon after you have superintelligence doing the research and development. So I think we will then start to approximate the condition of technological maturity, like a condition where we have already developed most of those general purpose technologies that are physically possible, and for which there exists some in principally feasible pathway from where we are now to developing them.
The plausibility of a solved world (12:53)
I know one criticism of the book is, with this notion of a “solved world” or technological maturity, that the combinatorial nature of ideas would allow for almost an unlimited number of new possibilities, so in no way could we reach maturity or a technologically solved state of things. Is that a valid criticism?
Well, it is a hypothesis you could entertain that there is an infinite number of ever-higher levels of technological capability such that you'd never be able to reach or even approximate any maximum. I think it's more likely that there will eventually be diminishing returns. You will eventually have figured out the best way to do most of the general things that need doing: communicating information, processing information, processing raw materials, creating various physical structures, et cetera, et cetera. That happens to be my best guess, but in any case, you could bracket that, we could at least establish lower bounds on the kinds of technological capabilities that an advanced civilization with superintelligence would be able to develop, and we can list out a number of those technologies. Maybe it would be able to do more than that, but at least it would be able to do various things that we can already sort of see and outline how you could do, it's just we can't quite put all the pieces together and carry it out yet.
And the book lists a bunch of these affordances that a technologically mature civilization would at least have, even if maybe there would be further things we haven't even dreamt of yet. And already that set of technological capabilities would be enough to radically transform the human condition, and indeed to present us with some of these basic philosophical challenges of how to live well in this world where we wouldn't only have a huge amount of control over the external reality, we wouldn't only be able to automate human labor across almost all domains, but we would also, as you alluded to earlier, have unprecedented levels of control over ourselves or our biological organism and our minds using various forms of bio technologies or newer technologies.
In this kind of scenario, is the purpose of our machines to solve our problems, or, not give us problems, but give us challenges, give us things to do?
It then comes down to questions about value. If we had all of these capabilities to achieve various types of worlds, which one would we actually want? And I think there are layers to this onion, different levels of depth at which one can approach and think about this problem. At the outermost layer you have the idea that, well, we will have increased automation as a result of advances in AI and robotics, and so there will be some humans who become unemployed as a result. At the most superficial layer of analysis, you would then think, “Well some jobs become unnecessary, so you need to maybe retrain workers to move to other areas where there is continued demand for human labor. Maybe they need some support whilst they're retraining and stuff like that.”
So then you take it a step further, like you peel off another layer of the onion and you realize that, well, if AI truly succeeds, if you have artificial general intelligence, then it's really not just some areas of human economic contribution that gets affected, but all areas, with a few exceptions that we can return to. But AIs could do everything that we can do, and do it better, and cheaper, and more efficiently. And you could say that the goal of AI is full unemployment. The goal is not just to automate a few particular tasks, but to develop a technology that allows us to automate all tasks. That's kind of what AI has always been about; it's not succeeded yet, but that's the goal, and we are seemingly moving closer to that. And so, with the asterisk here that there are a few exceptions that we can zoom in on, you would then get a kind of post-work condition where there would be no need for human labor at all.
My baseline—I think this is a reasonable baseline—is that the history of technology is a history of both automating things, but then creating new things for us to do. So I think if you ask just about any economist, they will say that that should be our guide for the future: that this exact same technology will think of new things for people to do, that we, at least up to this point, have shown infinite creativity in creating new things to do, and whether you want to call those “work,” there's certainly things for us to do, so boredom should not be an issue.
So there's a further question of whether there is anything for us to do, but if we just look at the work part first, are there ways for humans to engage in economically productive labor? And, so far, what has been the case is that various specific tasks have been automated, and so instead of having people digging ditches using their muscles, we can have bulldozers digging ditches, and you could have one guy driving the bulldozer and do the work of 50 people with a shovel or something. And so human labor is kind of just moving out of the areas where you can automate it and into other areas where we haven't yet been able to automate it. But if AIs are able to do all the things that we can do, then that would be no further place, it would look like, at least at first sight, for human workers to move into. The exceptions to this, I think, are cases were the consumer cares not just about the product, but about how the product
They want that human element.
You could have consumers with just a raw preference that a particular task was performed by humans or a particular product—just as now sometimes consumers play a little premium sometimes if a little gadget was produced by a politically favored group, or maybe handcrafted by indigenous people, we may pay more for it than if the same object was made in a sweatshop in Indonesia or something. Even if the actual physical object itself is equally good in both cases, we might care about the causal process that brought it into existence. So to the extent that consumers have those kinds of preferences, there could remain ineliminable demand for human labor, even at technological maturity. You could think of possible examples: Maybe we just prefer to watch human athletes compete, even if robots could run faster or box harder. Maybe you want a human priest to officiate at your wedding, even if the robot could say the same words with the same intonations and the same gestures, et cetera. So there could be niches of that sort, where there would remain demand for human labor no matter how advanced our technology.
Bostrom asks how creatures very much like him might want to live for eons if they had total peace, vast wealth, and full eternal control of extremely competent AI that could do everything better than they. He . . . tries to list as many sensible possibilities as possible . . .
But I found it . . . hard to be motivated by his key question. In the future of creatures vastly more capable than us I'm far more interested in what those better creatures would do than what a creature like me now might do there. And I find the idea of creatures like me being rich, at peace, and in full control of such a world quite unlikely.
Is the question he would prefer you answer unanswerable, therefore you cannot answer that question, so the only question you can answer is what people like us would be like?
No, I think there are several different questions, each of which, I think, is interesting. In some of my other work, I do, in fact, investigate what other creatures, non-human creatures, digital minds we might be building, for example, AIs of different types, what they might want and how one might think of what would be required for the future to go well for these new types of being that we might be introducing. I think that's an extremely important question as well, particularly from a moral point of view. It might be, in the future, most inhabitants of the future will be digital minds or AIs of different kinds. Some might be at scales far larger than us human beings.
In this book, though, I think the question I'm primarily interested in is: What if we are interested in it from our own perspective, what is the best possible future we could hope for for ourselves, given the values that we actually have? And I think that could be practically relevant in various ways. There could, for example, arise situations where we have to make trade-offs between delaying the transition to AI with maybe the risk going up or down, depending on how long we take for it. And then, in the meantime, people like us dying, just as a result of aging and disease and all kinds of things that currently result in people.
So what are the different risk tradeoffs we are willing to take? And that might depend, in part, on how much better we think our lives could be if this goes well. If the best we could hope for was just continuing our current lives for a bit longer, that might be a different choice situation than if there was actually on the table something that would be super desirable from our current point of view, then we might be willing to take bigger risks to our current lives if there was at least some chance of achieving this much better life. And I think those questions, from a prudential point of view, we can only try to answer if we have some conception of how good the potential outcome would be for us. But I agree with him that both of these questions are important.
It also seems to me that, initially, there was a lot of conversation after the rollout of ChatGPT about existential risk, we were talking about an AI pause, and I feel like the pendulum has swung completely to the other side, that, whether it's due to people not wanting to miss out on all the good stuff that AI could create, or worried about Chinese AI beating American AI, that the default mode that we're in right now is full speed ahead, and if there are problems we'll just have to fix them on the fly, but we're just not going to have any substantial way to regulate this technology, other than, perhaps, the most superficial of guardrails. I feel like that's where we're at now; at least, that's what I feel like in Washington right now.
Yeah, I think that has been the default mode of AI development since its inception, and still is today, predominantly. The difficulties are actually to get the machines to do more, rather than how to limit what they're allowed to do. That is still the main thrust. I do think, though, that the first derivative of this is towards increased support for various kinds of regulations and restrictions, and even a growing number of people calling for an “AI pause” or wanting to stop AI development altogether. This used to be basically a completely fringe . . . there were no real serious efforts to push in this direction for almost all decades of AI up until maybe two years ago or so. And since then there has been an increasingly vocal, still minority, but a set of people who are trying hard to push for increased regulation, and for slowing down, and for raising the alarm of AI developments. And I think it remains an open question how this will unfold over the coming years.
I have a complex view on this, what would actually be desirable here. On the one hand, I do think there are these significant risks, including existential risks, that will accompany a transition. When we develop superintelligent machines, it's not just one cool more gadget, right? It's the most important thing ever happening in human history, and they will be to us as we are to chimpanzees or something—potentially a very powerful force, and things could go wrong there. So I do agree with the C.
So I've been told over the past two years!
And to the point where some people think of me as a kind of doomsayer or anti-AI, but that's not the full picture. I think, ultimately, it would be a catastrophe if superintelligence was never developed, and that we should develop this, ideally carefully, and it might be desirable if, at a critical point, just when we figure out how to make machines superintelligent, whoever is doing this, whether it's some private lab, or some government Manhattan Project, whoever it is, has the ability to go a little bit slow in that, maybe to pause for six months or, rather than immediately cranking all the knobs up to 11, maybe do it incrementally, see what happens, make sure the safety mechanisms work. I think that might be more ideal than a situation where you have, say, 15 different labs all racing together first, and whoever takes any extra precautions just immediately fall behind and become irrelevant. I think that would seem . . .
I feel like where we're at right now—I may have answered this differently 18 months ago—but I feel like where we're at right now is that second scenario. At least here in the United States, and maybe I'm too Washington-centric, but I feel we're at the “crank it up to 11,” realistically, phase.
Well, we have seen the first-ever real AI regulations coming on board. It's something rather than nothing, and so you could easily imagine, if pressure continues to build, there will be more demand for this, and then, if you have some actual adverse event, like some bad thing happening, then who knows? There are other technologies that have been stymied because of . . . like human cloning, for example, or nuclear energy in many countries. So it's not unprecedented that society could convince itself that it's bad. So far, historically, all these technology bans and relinquishments have probably been temporary because there have been other societies making other choices, and eventually, just like each generation is, to some extent, like a new role of the die, and eventually you get . . .
But it might be that we already have, in particular with AI technologies that, if fully deployed, could allow a society in a few years to lock itself in to some sort of permanent orthodoxy. If you imagine deploying even current AI systems fully to censor dissenting information—if you had some huge stigmatization of AI where it becomes just taboo to say anything positive about AI, and then very efficient ways of enforcing that orthodoxy by shadow banning people who dissent from it, or canceling them, or you could imagine surveilling anybody not to do any research on AI like that, the technology to sort of freeze in a temporary social consensus might be emerging. And so if 10 years from now there were a strong global consensus of some of these issues, then we can't rule out that that would become literally permanent. My probably optimal level of government oversight and regulation would be more than we currently have, but I do worry a little bit about it not increasing to the optimal point and then stopping there, but once the avalanche starts rolling, it could overshoot the target and result in a problem. To be clear, I still think that's unlikely, but I think it's more likely than it was two years ago.
In 2050, do you feel like we'll be on the road to deep utopia or deep dystopia?
I hope the former, I think both are still in the cards for what we know. There are big forces at play here. We've never had machine intelligence transition before. We don't have the kind of social or economic predictive science that really allows us to say what will happen to political dynamics as we change these fundamental parameters of the human condition. We don't yet have a fully reliable solution to the problem of scalable alignment. I think we are entering uncharted territories here, and both extremely good and extremely bad outcomes are possible, and we are a bit in the dark as to how all of this will unfold.
Faster, Please! is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Discover shows related to Faster, Please! — The Podcast, based on actual content similarities. Explore podcasts with similar topics, themes, and formats, backed by real data.
