Jordan McGillis: Welcome back to the 10 Blocks podcast. I’m Jordan McGillis, economics editor of City Journal. Joining me on the show today is Jim Pethokoukis. Jim is a senior fellow and the DeWitt Wallace Chair at the American Enterprise Institute.
He’s a regular television presence on CNBC, and he writes the Faster, Please! newsletter on Substack. Most pertinently for our conversation today, Jim is the author of a recently released book, The Conservative Futurist: How to Create the Sci-Fi World We Were Promised. Jim, thanks for joining me.
James Pethokoukis: Jordan, thank you so much for having me on.
Jordan McGillis: Now Jim, I’m tempted to ask you when my hover pack will be out for delivery, but instead I’m going to go to the fundamentals of technological and economic progress, GPT. For you listeners, I don’t mean ChatGPT, which stands for generative pre-trained transformer, I’m talking about general-purpose technologies. Jim, what is a GPT, and why do they matter?
James Pethokoukis: Yeah, a general-purpose technology is a technology that has a significant economic impact across a broad range of the economy, throughout many sectors. Sort of a classic case of a GPT, to go back a little bit, would be the steam engine, would be electrification, the internal combustion engine. The computer, especially when combined with the internet, is a general-purpose technology.
And these are really important, because when you look back through history, economic growth and productivity growth, the output per worker is fairly steady, and then you get a general-purpose technology or if you’re lucky, a cluster of them, and those can really accelerate growth. And it’s a hope of my book at least that we may be seeing a cluster of important technologies right now.
Jordan McGillis: You referenced the first industrial evolution, the steam engine. Do you think that that phenomenon of industry taking off and then becoming globalized is essential for something to be a GPT? Or was there anything prior to that like metallurgy or anything that predates industrialization that would also be a GPT?
James Pethokoukis: Yeah, that’s an interesting question. I mean, one might call the printing press—it’s a technology that enables other technologies. I’m not sure if it meets the classic definition or not, but certainly that was a very important technology. There were a lot of important technologies before the industrial revolution. The compass was a very important technology, the printing press. Various kinds of metallurgy were very important, but what you saw starting with the first part of the industrial revolution, then really the second part where you really saw electricity and the combustion engines and the chemical industry and the early communications industry, telegraphs, you saw lots of very important technologies, especially a technology that can enable another technology, I think is also ideally a key component that the computer also makes other technologies work better. It can be applied throughout the economy, so it’s not just a computer on every desk, but how those computers might be used. And again, they’re super important, and when you can get one, you should feel lucky.
Jordan McGillis: Can you give our listeners a sense of those different waves of industrial revolution—the first and second?
James Pethokoukis: Well, when people talk about the industrial revolution, the first part, a lot of it was a steam engine, and it was improvements in textile manufacturing. Early on you saw the beginnings of locomotion and the railroad, but that initial period did not really create the sort of rapid growth we associate with the industrial revolution. One, it often takes time, takes time for these technologies to be figured out, to be used efficiently by business. That lag is common throughout all these technologies. Again, sort of the classic case is that in the second wave, where you saw again chemicals, electricity, and the classic example is electrification, where even though people saw, business owners saw the advantage of electrifying a factory floor over steam engines, it took a long time before those factories change, before they changed from being big, tall buildings to being in much flatter buildings to train new workers or those new workers to leave and new workers to come on.
And even though the advantages and basic concepts of electrification were known in the late 19th century, it really wasn’t until the 2020s, where you saw a huge impact on productivity growth. So that was also the case with computers. Right before we saw the big tech boom of the 1990s, people were writing articles about where is the productivity and the economic impact of computers? They’ve been around since the 1960s. Where is it? Well, it took a while, took a while for companies to understand how to use them and to be connected to the internet. And then finally in the 1990s, we really saw a massive productivity boom that was long awaited, and some had lost hope would ever happen.
Jordan McGillis: As we sit here talking in the early age of artificial intelligence, I’m thinking about how this new technology can potentially remake the industrial world in the way that electricity was able to. Do you think that AI is a general-purpose technology that can accelerate productive capacity within industrial sectors?
James Pethokoukis: I think that’s the hope, and I think there’s good reason for hope. I mean, already we’re seeing it. It’s not early days for AI or machine learning, it may be early days for this latest advance of what they’ll call generative AI and these large language models, ChatGPT was suddenly released in November, 2022, but already you’re seeing how it can affect different industries—law, customer service, healthcare. We start to see these early examples of it being adopted. Then it starts to look like a technology that could be used anywhere where there’s sort of knowledge work going on. And then if it can be used, I think it’s helpful if it can sort of make what customer service representatives that can make them more efficient and can help the low performing ones do better. That’s all really important.
But if it also helps researchers or technologists become more efficient, where the technology acts as a super research assistant where it helps them go through past research more efficiently, maybe connect the dots with something, which, in my own use of ChatGPT, has helped connect the dots and help them come up with plausible theories for why something might happen. Go through the whole stock of human knowledge far more efficiently than we currently can, and again, makes them more efficient, then that is massive. Because one supposed cause of sort of the slowdown as measured by economic statistics and scientific productivity has been the notion that we have to kind of climb higher up on the tree of knowledge and we need more people climbing. But if we can make each researcher far more productive, whether they’re in drug research or material science or advanced aeronautics, then we have really, I think then you have really accomplished something that cannot be captured even in these early economic estimates about how much AI might affect productivity growth and then more broadly economic growth. So it’s pretty exciting. It’s what I’ve been waiting for my whole life.
Jordan McGillis: Are you able to call to mind any examples from your own experience interacting with GPT where it did help you connect the dots on research?
James Pethokoukis: Yeah, one thing I’ll do is I’ll try to come up with the pros and cons of a particular policy and over and over I’ll come up with five and then there’ll be one or two that ChatGPT will come up with and I’ll think, oh, okay, maybe given another day, I might’ve come up with those or maybe I wouldn’t have. But those are actually good points. Sometimes the points aren’t so good, but they make me more productive, and I think more efficient in building on my current knowledge base where I have enough knowledge to say, okay, that’s a good idea, but that’s not a very good idea. So it allows me to build on what I already know, and maybe I can do in an afternoon what it would’ve taken me a couple or three days to do. And it’s pretty helpful when coming up with images for my newsletter, because we were only relied on my sketching to come up with those images, it wouldn’t be very good. And that’s a big help too.
Jordan McGillis: It’s also helping my cocktail creations at home.
James Pethokoukis: Jordan, don’t accidentally, don’t create some sort of toxic virus that you thought was a cocktail. Be careful.
Jordan McGillis: No lab leaks will occur from the McGillis kitchen. Total factor productivity. Jim, what does that mean? And what are the trends we’ve seen over the last 100 or even 200 years?
James Pethokoukis: So you can make people more productive by giving them a machine, giving them more machines. You can make them more productive by training them better, giving them better education. But when you look at all those kinds of things, there’s something left in the economic equations, there’s a gap or what they call a residual. There’s something else to make the numbers all meet and round off nicely. So what is that residual? And for lack of a better interpretation, that residual seems to be innovation. It seems to be, often it’s called technological progress, but it’s really any sort of innovation, whether it’s a new technology or just a new way of doing things, that seems to be pretty important in making a society more productive. Like half of that productivity comes from this mystery residual, which is called total factor productivity.
And you could add all the machines, do all the trainings, but that will, especially for an advanced economy like the United States where we exist on the technological frontier, push that frontier forward. You need to become more innovative and do things a different way. And that factor, that TFP is something I think a lot about. It’s something policymakers should think about, and it’s fundamentally going to be the difference going forward, whether we are at a 1.5 percent GDP growth economy, which is like half of what it’s been since World War II, or whether we can try to grow as fast in the future as we have in the past, or if everything cuts right even faster. And there’s huge ramifications for that.
Jordan McGillis: In terms of rapid accelerations of growth, certainly it’s the case that we need to be pushing the frontier. Do you think that there is still more productivity and growth to be had in a country like the United States through better education of our existing citizenry?
James Pethokoukis: I think that’s absolutely the case. There are some statistics, and I point in my book, if we were just to sort of score as well as like, you have these international tests, if we just go up the next group of students in countries like Canada, if we could score as well as those students, I mean you might be talking about another half a percent or a full percentage point in economic growth, which is a huge number. And anytime you get big numbers in economics, you should immediately question them. But fine. Now what if it’s half that or a quarter of that? When you have an economy that’s only growing, supposedly only able to grow without inflation at 2 percent, to be able to jump like a quarter of a percentage point or half a percentage point is a massive improvement.
And what I think is important, and what I try to get at in the book is there’s lots of these—AI will be fantastic, but there’s also lots of these other places where there are these sort of inefficiencies where if we could pick up a quarter point here, a 10th of a point here, they really do all add up. And that will be the difference. Those percentages, the little, tiny increments year after year, decade after decade is the difference between us let’s say in 50 years having a $60 trillion economy or having a $160 trillion economy, I could probably think of a lot of things to do with an extra $100 trillion.
Jordan McGillis: As could I. I want to ask one more question on this education point. Would something like adopting the math curriculum from a country like Singapore be useful, or is it just an entirely different cultural context? What are some of these things we can do at the margin to get more people working in productivity-advancing fields?
James Pethokoukis: Well, I think that’s sort of two different questions. Can we teach mathematics and science better? And is that a curriculum issue or is it a teacher quality issue? I think a lot of it might be a teacher quality issue, which is hard to fix because we are a big country, and you have to persuade people with lots of other options who are talented to do something else in a profession that is not certainly seen as elite, or high-profile, or high status as some others. So, while it would be really great if there were people who right now, maybe they’re considering becoming investment bankers, maybe become teachers, that would be great. But I think ultimately there’s going to be a lot more promise at integrating AI to help poor-performing teachers or middle-performing teachers do better, because you’re already seeing that in some other fields, where they’re trying to implement AI where it really helps the lower-performers improve.
So that’s one thing. The other thing is what is the status in this country of people more broadly who are sort of creative, who are in science and technology, who are entrepreneurs? Would it help if those kinds of people were held up as representing occupations and fields worth doing? He’s become sort of controversial in politics, which I don’t much care about, is somebody like Elon Musk, who has been an entrepreneur in the technology field. And I just don’t think it helps when you begin to look at people’s achievements through a political lens, and you begin to say things like, if you don’t like Elon Musk’s politics and you say, well yeah, so what? So, he’s built a rocket, big deal. Other places build rockets. That is diminishing, like, human achievement of people in fields that we need smart people in. It’s not easy to build a rocket company.
Other people have been trying to build reusable rockets and haven’t been doing a very good job and to sort of take it from nothing to what SpaceX is, is something that should be lauded even if you don’t much care for Elon Musk’s politics. To begin to diminish entrepreneurs, especially in technological businesses, as somehow easy or not important, to me that has to filter down into schools and what kids want to do. You should want kids to want to start their own SpaceX or Tesla, or pick whatever tech company that should be. We should feel enthusiastic and not say, well no, it really isn’t because I don’t like the politics of that industry, or they’re just going to be oligarchs, or we don’t like billionaires, then we won’t get as many of those, and we won’t get the benefits they produce for our society.
Jordan McGillis: As you talk about in the book, there was this golden period of futurism where young people were thinking in that way, we can point to the 1950s and the ‘60s as that golden period. And then as you described, we’ve entered what you call, I think it’s a great term, the great downshift, both economically and in many cultural ways, that are reflected in the ways you’ve described. Can you talk about the great downshift and why that’s the term that you’ve selected?
James Pethokoukis: So statistically, there was a noticeable downshift in U.S. productivity growth in 1973, so 50 years ago. And that is sort of the statistical hub of this book, where something changed. Where all the sort of early—primitive as they were—forecasts, especially in the 1960s of how productive the U.S. economy would be, how fast the economy would grow, those all proved to be badly optimistic. And of course those would be driven by continual advances across a number of technological fields, but that downshift, that lack of growth is one reason why all those crazy Jetsons-like sci-fi dreams didn’t happen. And at the beginning, it was not clear that that’s what was happening. We had, there was an oil shock, people were like, oh, okay, there’s an oil shock, and economies go up and down. But pretty soon we’ll be back to like it was in the 1960s.
But it didn’t happen. And certainly, by the late 1970s and 1980, you had the government and the Office of Management and Budget and Congressional Budget Office and so forth trying to figure out what happened to that productivity tech boom of the 1950s and ‘60s. Is it just oil? Is it something else? And they’re still trying to fully figure it out. I offer some kind of a roundup of the best theories we have and some of these things were sort of external, such as all the productivity gains of the inventions of old, we sort of have gotten the productivity, everything that could be electrified was electrified, and there’s combustion engines everywhere. So that’s part of it. But I think what we did, our own decisions played a key role, which the good part is if maybe if we make different decisions today, we can get back to that period of very rapid growth that we saw in those immediate post-war decades and then in the late nineties as well.
Jordan McGillis: What are a couple of decisions that you think would spark a renaissance of a kind?
James Pethokoukis: To me, the two things that really stand out was that, and again the book is called the Conservative Futurist, but it’s really a book for anybody across the political spectrum who thinks that humans can create the tools and we have the will and the agency and enough smarts and wisdom to solve problems. So I think two things which might have appeal across the spectrum are one, after Project Apollo, there was project nothing. And I think had we continued to spend a similar share of the economy on research and development, that that would’ve played an important role in helping keep economic growth high across the economy, productivity growth, advancing technologies that we’re only seeing really today, maybe we could have had earlier. Instead of a nuclear fusion breakthrough in December, 2022, maybe it would’ve happened in December of 1992 or what we’re seeing in AI, maybe that would’ve happened 30 years ago.
All these kinds of things. So I think not spending on something that most people think government should do, which is kind of the blue-sky or earlier-stage research at minimum, I think that decline is a pretty important bad decision. And the other decision is to create environmental regulations with little concern about how they might ultimately impact our ability to innovate and build in the real world. Heaven help you if you want to build a high-speed rail or transmission line or highway extension or subway line or geothermal well or nuclear reactor or small nuclear reactor or someday commercial fusional reactors. It is just far too hard to do that in this country in a timely and fiscally responsible manner. And that was not considered and we’ve had 50 years to fix it. Maybe now since people are realizing, well, you want a green revolution. Well good luck building a transformer.
Forget about installing a wind turbine. Good luck getting the factory to build the wind turbines built in a timely manner. Like, to meet the Biden’s goals for wind power, we need to build like 50 of these factories across America. Right now we have one unless we want to get them done by the year 2050, I don’t think that’s the plan, we need to have a regulatory revolution in this country if you want an energy revolution. So those are two, to me, those are two obvious bad decisions that we’ve made that we have yet to fix.
Jordan McGillis: Okay, last question I’ve got for you. This pertains to the future of conservatism. As you describe in your book, your outlook is along the same lines of George Will’s, that conservatism in the American tradition is custodianship of the classical liberal outlook of markets and limited government and the pursuit of happiness. In some ways, however, at this time, that is a position that has fallen from ascendancy in the broader right-of-center U.S. political constellation. How do you see this battle on the right shaping up over the next five to 10 years?
James Pethokoukis: I hope if we get even with suboptimal policy—we’ve talked somewhat about AI, but there’s lots of things going on in biotech and energy and space that, in a better growth environment some of these sort of populist impulses will be less powerful. And we begin to think about what we really value, and what I really value is that sort of inheritance of personal and political and economic freedom, which is an inheritance to be used well today and then passed along to future generations. Because I think one thing and sort of the subsections here is that Donald Trump, but I think one thing Donald Trump said, which really struck me as right is when he called America a developing country. If you look at the IMF and World Bank and things, the U.S. is a developed country just like Germany and Great Britain and France are, but in a way, we’re still developing, and we’re all poor. America is poor compared to where it will be, and hopefully will be, 75 years from now or 100 years from now.
So this isn’t the end of the game. There are lots of problems to fix, diseases to cure, creating a country and a human civilization that won’t get wiped off the map because of a stray comet or the next pandemic. We can be richer, healthier, have a more resilient world and more opportunity. Not just if you’re lucky enough to live in the United States or the West, but everywhere. So if any of that, and I hope there’s enough happening right now with technology and what it can bring us and the beginnings of which I think we only saw with powerful vaccines, when we start seeing some more CRISPR cures maybe, I’m hoping that there’ll be a tailwind for this new kind of thinking that can be seen on the left and the right. We’ll have our different flavors, and maybe some different policy prescriptions, but if you believe that we can solve big problems, I’ll sit down with you left or right, and I think we probably hammer out some pretty good ideas.
Jordan McGillis: I can sincerely say that that is a hope I share. Jim, where can our listeners keep up to date with your work?
James Pethokoukis: My Substack, Faster, Please! I would love everybody to subscribe. Of course, my book, The Conservative Futurist: How to Create the Sci-Fi World We Were Promised, still available everywhere, audio, Kindle or a hard copy. And of course, you can also find me at the AEI website and on the AEIdeas blog. So all those places, come visit, and I hope you see something you like.
Jordan McGillis: I second that. I encourage you all to buy The Conservative Futurist and to subscribe. Jim, thank you very much for joining me today.
James Pethokoukis: Jordan, outstanding. Thank you.
Jordan McGillis: As always, you can follow City Journal on the website formerly known as Twitter @CityJournal and on Instagram @CityJournal_MI. And of course, if you enjoyed listening today, please like, rate and subscribe. Thank you.