AI Won’t Decide the Future of Work—We Will (with David Autor)

The rising inequality and growing political instability that we see today are the direct result of decades of bad economic theory. The last five decades of trickle-down economics haven't worked. But what's the alternative? Middle-out economics is the answer. Because the middle class is the source of growth, not its consequence. That's right. This is Pitchfork Economics with Nick Hanauer, a podcast about how to build the economy from the middle out. Welcome to the show. Hey, Pitchfork listeners, I'm Freddie, producer at Pitchfork Economics. Every new wave of technology comes with the same promise. Progress will make everyone better off. But executives have been bragging on earnings calls about using AI to cut jobs and shrink payrolls, as if their massive campaign of layoffs is proof the future has arrived. This week, we're revisiting how AI could help rebuild the middle class, a conversation with MIT labor economist David Autor, who argues the real danger isn't artificial intelligence itself, but an economy that uses innovation to replace workers instead of expanding what they can do. We've been here before. Technology moves fast. Workers are told to adapt, and somehow most of the gains end up in the bank accounts of the wealthy. But technology doesn't decide who benefits. The rules of the economy do. And this conversation explains why that matters so much right now. So we're recording this about a month after the election, Nick, and Obviously, there's a lot of things for me to feel really anxious about. And just one more thing. I just need to ask you, are you planning to replace me with AI? If I could, I would, buddy. If I could, I would. Spoken like a venture capitalist. I think we're a few years away, though. I'll tell you what. You know, I don't know that AI can ever figure out how to do what I do because I've never figured out how to do what I do. So I don't know what it would learn from. Yeah. But it's a great way to lead into today's conversation because, you know, we're talking about the potential future of AI. Yeah. And we get to talk to David Autor, who is one of the country's most highly respected economists. Labor economists. Labor economists. But economists more generally. He's an economist at the Massachusetts Institute of Technology. He studies technological change and globalization and how it affects workers principally. But he's a very consequential guy and has out a really interesting new paper arguing that AI, rather than decimating the middle class, may actually help them. And it's an interesting article. I would say there are a lot of assumptions baked into the argument. Really? Economists relying on assumptions? Never. I'm not sure I agree with. But, you know, I think that AI is here. It's here to stay. And we're going to have to wrestle with what to do with it. And it's certainly a good thing that people like David are thinking about it and talking about it. So with that, let's jump into the conversation with David and find out what he thinks. My name is David Otter. I'm the Daniel and Gail Rubenstein Professor of Economics at MIT. but basically I'm a labor economist and I work on things that affect jobs, skills, mobility, employment, income, and inequality. That's awesome. So I guess our main question is how long before you were replaced by AI? Is that where we're going, Nick? Yeah, exactly. I would be delighted, man. I could really use a break. So you've got this fascinating new article out where you make this contrarian argument that AI, rather than being bad for the middle class, may in fact be good for it. And before we get into that, just help us understand why you think that AI is not just part of the tech industry hype cycle. Yeah, no, I think it is not. I mean, let me be clear. There's lots of hype. There's going to be lots of companies that fail. There'll be a bubble, et cetera. But in the long run, and the medium run, the technology would be extremely consequential because it unlocks capabilities that we just didn't have before. AI can do things. So just to make the contrast, traditional computing followed rules and procedures. It could do what we could program it to do and nothing more. That was incredibly useful, but it never had new ideas. It didn't improvise. It didn't solve a new problem. It didn't think of a better way of doing things. We had to explicitly understand how to do something to write a program so the computer could follow those steps to do it. And AI learns a lot like we do, not the same, which is it learns inductively. It learns experientially. It learns tacitly, right? Most of what we know, we don't know formally. We don't know formally how the structure of language works, how to make a joke, how to have a hypothesis. We don't even know formally how to recognize a child in a photograph or ride a bicycle. These are all things that we understand tacitly. And for that reason, we couldn't really teach computers to do them because we didn't know how to write the program. And AI can learn, as we do, from unstructured information, can recognize patterns that we're not told to look for. It can produce new ideas that we didn't program into it. And that makes it applicable to a whole host of problems that we couldn't previously apply computing to, whether that's in discovering the structure of proteins or whether it's in recognizing things in medical scans or whether it's helping people write and come up with ideas or make inferences from data. So it's really quite broadly applicable, though I will say we don't really know what it's for yet. I mean, we're still in the kind of the early age of experimenting and trying to figure out what is it good for and what is it not good for. And a lot of what it's not good for is what it's being used for at present. Because, you know, the first thing you're doing in new technology is you say, well, what does it allow me to do better, cheaper, faster than what I already, that I was already doing? But there's two problems with that. One is AI can do many things that you couldn't do with the old technology. So you're kind of missing half the picture. And the other is it's not good at a lot of things that old technologies are good for. It can't reliably do math or keep facts straight, right? Think of that relative to traditional computers. And so there's a lot of instinct to let's use it for this calculation. Let's use it to write a law brief. Let's use it to write a letter of recommendation. And it's going to do those things badly. It's going to confabulate. It's going to not make a calculation incorrectly. Yeah, I've actually been shocked by how bad the analytics are sometimes. Yeah. Like, you know, you ask it a question and then you go look for the data and you're like, dude, this is not close to the real data. And they're like, and it's like, sorry, I did. Yeah. And even if it was right and you told it was wrong, it would still say sorry because it wouldn't know. No, you can challenge it to rock, paper, scissors, which is a game, a simultaneous move game. And then it will say rock and you can say paper and then they'll go, oh, you won. And I guess you're lucky, right? So there are many ways you can convince yourself it lacks an understanding of the world, but that doesn't mean it's not incredibly powerful for many things that we need to do. Because many things we are looking at data and recognizing patterns and making predictions. And that's a lot of what cognition is about. That doesn't mean it's better than us, but it can help us to do those things. Even when we're good at it, we're not consistent, we're inattentive, we get tired. So not good at math, but could be good for, I don't know, enslaving humanity. Sure, if that's what you're into. Well, I look at some of the billionaires who are funding AI, and I worry that that is, in fact, what they're into. Like if Elon Musk could create Skynet, he would just to prove the point. Yeah. I actually agree. I agree with this point that a lot of AI research is driven by the idea of how do we replicate human capabilities in machines? And the truth is we have lots of human capabilities, and I don't think we're scarce. We're running short of those. We want additional capabilities that we don't have. Right. Flight was not a human capability. You know, penicillin was not a human capability. Telecommunications was not a human capability. What makes technologies transformative, actually, is not that they automate the stuff we already do, but they enable us to do things we couldn't previously do. And that's where AI has so much potential. That's a really good point. So you make this broad argument that AI may, in fact, be good for working in middle class people because it will allow, if I may summarize your argument, it will allow people with less skills to apply themselves in areas which require a lot of skill and therefore be more productive and ideally compensated better. You also make the point, I think, that's very important in your piece, that the world is undergoing a demographic transition in which we are very rapidly going to run out of people. And the idea that we're going to run out of jobs is something you do not subscribe to, which I'm in violent agreement with. I think that as long as humanity faces problems, we're going to have jobs. But what I want to understand from you is why you are optimistic that these workers who are just more productive will share in the benefits of that productivity. Like, how do you see this unfolding? Sure. So let me, there are a lot of laws in there. Let me respond to a few of them. Yeah. So we are definitely running into it, just on the demographic point, because we have low birth rates, our populations are aging. So we have large and growing retired populations and actually numerically smaller number of people in prime working years to support them And that scarcity and that in the industrialized world and we seen that really intensively since the pandemic Now, we should recognize that's artificial, right? There's a lot of people in the rest of the world who would like to come to these countries and work. There's not really true labor scarcity worldwide. There's labor scarcity in rich countries, but rich countries don't seem that interested in bringing those people in. Okay, but just to be clear, if you go out 25 or 30 years, you are seeing a massive decrease. You go out 100 years, you're going to see a massive decrease. And across the world. Yeah. I mean, the problem is it's not – there's nothing wrong with having a smaller world population. It's the transition between sizes that creates this demographic imbalance. It's always those pesky transitions, isn't it? Yeah, exactly. So about allowing what AI can enable. So I like to – rather than the word skill, which sounds kind of one-dimensional, right? People have more education, have more skill. I like to use the word expertise. So expertise could be the expertise to code an app, but it could be to remodel a kitchen, right, to land a plane, you know, to fix a washing machine, to, you know, do accounting. And expertise is really what makes labor valuable in the rich world. It's the capability to do valuable things. Not everyone knows how to do them, requires specialized knowledge. And our rich world is dominated by experts. and they really are kind of a choke point on everything in education, in healthcare, in law, in design. More and more of our national income goes to highly educated people who are at the tops of their fields and making these high stakes decisions. And we're really beholden to their expertise. I mean, I can point the finger myself. I'm an academic, so I'm a paid expert. And 50 years ago, we had a lot of people in the middle doing office work, clerical, admin, and also a lot of production and operative work. And a lot of that work was automated by computing, right? Because it followed well-understood rules and procedures that could be encoded in software and done by machines. And so that has moved a lot of people who aren't part of this elite, the 40% of people have college degrees, for example, four-year college degrees, into services, food service, cleaning, security, entertainment, recreation, socially valuable work, but it's poorly paid because it's non-expert. It doesn't require specialized training or skills. So even some of those valuable jobs, you know, being a crossing guard or a daycare teacher, right, these are life and death activities. They're poorly paid substantially, not entirely, but substantially because most people of able-bodied adults can do that work without specialized training or certification. So we've kind of created this hollowed out middle. And what we would like is a world in which more people who are not at the, you know, the hyper elite educated could do more valuable work in software development, in, you know, medical, medical technical work, you know, administering care doesn't have to be, you know, doing surgery or in creating legal documents or in, you know, designing, doing kitchen redesign and the engineering around that. I think what AI is complementary to is the judgment and knowledge to do those things without having as much of the specific technical formal frontier knowledge. Right. So I like the example I give in my articles. You know, let's say I wanted to change the, you know, the breaker box in my house. Sorry, the fuse box in my hundred year old house to a breaker box. Right. Old technology, new technology. You know, I could go to YouTube and there's hundreds of videos telling me how to do that. Right. But imagine I have no electrical skills. I've never, you know, worn a pair of insulated gloves or used a wire stripper. I could watch one of those videos and I will surely electrocute myself, set my house on fire, cannot work out well. And probably produce another really great YouTube video in the process. Something that will go viral on TikTok. Exactly. And if, of course, if I'm a master electrician, I'm not going to watch the video. What's the point? But if I actually, you know, if I have some basic knowledge, if I know, you know, I've done electrical projects, I have an ohmmeter, I know some basic stuff. I could, with this tool, I could extend my expertise further, right? So my judgment and knowledge is a complement to the tool, not a substitute. And so the good scenario is where we use AI as a tool that enables people to take on important problems, expert problems, without being at the frontier. And what I mean the frontier is I'm not imagining a world where everyone can do everything. That's not even a good world, actually, from a later point of view. But a world where someone who's a registered nurse could do more care tasks. or someone who's a junior law associate could do more legal work or someone who is, you know, not didn't do a four-year degree in CS, computer science from MIT, could do good coding. And AI is very supplementary or, you know, it supplements judgment, but it requires something we've seen from tons of research now in AI. And there's lots and lots of case studies. AI is really good when you have enough judgment to know if it's doing well and it's dangerous if you don't. And so it requires domain expertise and training, but it doesn't necessarily require, if you have that, you can apply that tool to do things. I can code in a computer programming language I don't know, because I know enough to know when it's producing the right output, so I can use this tool. What makes you confident that workers will actually reap the monetary benefits of using these tools rather than the owners of the intellectual property? Because I can see a dystopian future of just massive rent-seeking from a handful of AI giants. Right. So this is a good question. So I'm speaking of a possibility, not a certainty. I think there's a case where we can do this. Whether we will is not clear. And I think there's two issues, right, that both of which you've now raised. One is, of course, how are the tools being designed? Are they being designed to basically replace people or to, you know, leverage their expertise? So that matters, right? We have a lot of choice about how we use it. They're incredibly flexible, right? There's a million ways to use them. And so that's one question. And then the next one is, well, is the market structure such that they get the returns to that productivity or not? And it depends, in my opinion, substantially to whether that expertise remains scarce and valuable, right? Like no one thinks that, you know, people with bachelor's degrees, certainly PhDs are underpaid relative to what they do, right? Why they have scarce expertise. And, you know, and because of that, there's a lot of competition for their services. A good example would be, you know, registered nurses, right? For your degree and they pay, they make a very decent income in the United States. So I think having expertise remains scarce. I actually think a lot of adjunct professors would disagree with you. You're right. You're absolutely right. But that's again about competition, right? It's about market structure. And the concern you're raising, which is a valid one, is maybe this will all be monopolized by a few companies. And I think that's a risk that we have to be really vigilant about. I don't want to take it as a given that this all goes well. The case you're making, of course, is plausible or possible. And what I've said a million times is that we cannot fear technology. What we have to fear is an economic system like neoliberalism that privatizes all the benefits of technology and socializes all of the disadvantages of technology, where a few people at the top get all the benefits and everybody else gets screwed. And I guess, given history, take truck drivers. At one time, one person could move very little in terms of goods from one place in the country to another place in the country, because they would either have to carry it on their back or use a donkey or something like that. And now a big truck can move a lot of goods from one place to another. But truck drivers are very poorly paid generally, even though they are creating a huge amount of value relative to the alternative. That's a consequence of the fact that truck drivers have no power. We deregulated the trucking industry. In fact, truck drivers earned a middle class wage before deregulation. Yeah. Likewise, administrative assistant, right? I mean, before computers, That job was X productivity. With a modern computer, you can obviously accomplish a heck of a lot more, do 10 times as much work, really, relative to what you could with a typewriter. But those folks don't earn more today. They earn less than they used to. Their wages have compressed down over the last 50 years. By the way, this is true of people with bachelor degrees, too. in general, people with bachelor degrees, their wages have either plateaued or decreased. This is the story of the last 50 years, is that the only people who really benefited from economic growth were the people in the top one or 2%. I'm not agreeing with that, by the way. Okay. Top 10%. No. Basically, people who have a college degree on up, which is 40%, have seen real income growth, but there's been a lot of stagnation. But they have not maintained their same share of GDP. OK. So this is actually, this is a point I want to make. Yeah. So six out of every $10 in the US economy, approximately, is first paid to workers. And it's fallen somewhat, although it's still pretty high. What's actually remarkable, and I think many people find this surprising, is you might think in the richest countries, the countries that have the most capital, the most technology, the most infrastructure, that labor's share of economic activity would be relatively small. But in fact, it's larger, right? So the world average is about 50%. In parts of North Africa, it's 45%. One of the amazing things actually about the kind of industrialized countries is even as we've gotten all this technology, we've made our labor more valuable, not just in terms of we pay higher wages, a larger share of all that kind of activity has gone to workers. Now, we should be concerned about that declining, but it's still actually relatively high here. And I would argue in the case of truckers, this is part of the reason wages haven't risen with productivity. you can make the same argument about cashiers or many activities is because the labor is not scarce, right? The people who can do it or the capability to do it is abundant. And that's why I really focus on this notion of what makes expertise valuable. And the problem that we have is that people who do not have, or a problem that we have is that the majority of people who do not have a college degree have been pushed downward into non work clerical work production work That required real skills and knowledge and training that made it valuable over long periods of time And in a world in which people are doing work that many, many others can do, it's very difficult for that to pay well. Now, I don't want to make the case that it can't pay well. Obviously, it pays more in some countries than others, and that's a function of the way they structure their economies. Well, power. Yeah, sure. I mean, some people who work at McDonald's make $20 an hour. Some people who work at McDonald's doing exactly the same job make $7.25 an hour. Yeah. That is not a reflection of rarity. That's a function of we somehow delivered more power to the people who were earning $20 than the people who earned $7.25. I mean, I'd say, you know, so in almost all countries, the sort of the ranking of earnings across occupations activities is very similar, but the disparities very enormously. Right. Right. So in Scandinavia or even in Canada, wages are substantially higher for people doing that same work. And that is not an economic inevitability. That's a choice of power, as you say. How does the social system organize? So again, going back to these two questions, one is can we augment the value of expertise? Can we enable more people to compete in these high-paid activities? It won't necessarily be good for doctors or lawyers or software coders, But if more people who are not the elite were able to do that work, it would be better for them. And it would also bring down the prices of goods and services. So it would improve the quality of jobs for some. It would improve the availability of a lot of expensive stuff for others. But then there is the question that was just asked, well, what's to prevent firms from monopolizing that? And it depends on healthy labor market competition. It absolutely does. That means the way we structure regulations, the ability, you know, worker voice and institutions to protect that, labor standards and social norms about what we view as tolerable. Yeah. So I don't think it's either or. I think we need to be productive. Ultimately, how much we can distribute depends on how much we produce. And we want to use these tools well, but we don't want to use them in a way that just makes human labor unnecessary or superfluous. We want to use it in a way that makes more people's judgment and knowledge and expertise more applicable across valuable things. So earlier when we were talking about the impending demographic, I don't want to use the term crisis, collapse, decline, whatever. It certainly doesn't have to be a crisis. You emphasize it's the transition that's the difficulty. Not the smaller global population, but the transition to it. As a student of history and technology, broadly say that new technologies in the long run have proven beneficial. Like we have a much higher living standard now than we did 200 years ago. And we're all better off for that transition. But in the moment, that transition immiserated people. It was very difficult. And we all live in the moment. So the people who are going to live through the AI revolution, a lot of them risk being displaced and demiserated. What are the policies that we need to focus on to get through the transition without, I don't know, undermining democracy? Just to put a fine point on it. Yeah. Yeah. So first, let me agree with you that even if the long run outcome is a good one, the transitions are often really costly. I mean, like the transition into the Industrial Revolution, working class incomes did not rise in Britain for the first 60 years of the Industrial Revolution. Right. And so people were miserated. And even when they started to rise, there's still there's lots of people who are displaced. So people on the upside are usually not the ones on the downside. Right. Like Pittsburgh has reconstituted itself after losing steel and coal, right? But the people in Pittsburgh who are doing well now are people in education, people in healthcare. They're not former steel workers or former coal workers, right? So even if you think it's all going to work out great, that doesn't mean every person is going to benefit. But then there's the even harder problem. Like let's say it's all going to be great for people in the long run, but not in the first 20 years. I think that's a very, very serious challenge. And so part of it depends on how fast it changes, right? And this is a very important part. Like if you told me 20 years from now, all trucks will be self-driving, I'd say, well, we can manage that. A 20-year transition, people will retire, people won't enter. If you told me that's going to happen on New Year's Day, I would say that's a crisis. So part of it is how fast it happens. But then the other, of course, is the support systems that we have to enable people to thrive and rebound. And that's just something where the U.S. falls uniquely short among developed countries. We spend one-tenth as much per capita or per GDP as does, for example, Denmark. on retraining people and getting them back in the workforce. We used to believe that we had a national job security program. It was called Get a New Job. And we've learned that that just isn't such a good program. And that's why I worry the most about the U.S. I think the U.S. does so little to buffer against the damaging forces of inequality and the scars they leave that it actually inhibits economic mobility, inhibits opportunity. It's not just cruel, it's also inefficient. We have a thriving meritocracy from the median on up. And then for everybody else, we underinvest in their skills and their security and their health care and their safety and their education. I think that's the danger that we run. And survey evidence says that US workers are more nervous about technology than people in other countries. And you say, why? We're creating this technology. Why shouldn't we love it? And the answer, of course, is what you said earlier, because they recognize the economic system. It's not they're afraid of the technology. They're afraid of the cost they'll personally pay. You know, the science fiction writer Ted Chiang said, you know, fear of technology is not fear of the machines. It's fear of capitalism. It's the fear of I don't win when we get more productive because someone else, something else takes my job. That's what I said. Yeah. It's not the technology. It's the neoliberalism. That's right. That kills you. Exactly. So I think we are an amazing, we are at a transitional period. There is incredible, once you simultaneously be fearful and optimistic in the sense that there's a lot to be worried about. And the things outside the labor market are arguably more worrisome than even what's going to happen in the labor market. But there's also enormous opportunity. And there's no question that we will enhance the rate of technological progress. We will be able to do better things with addressing some of our most challenging problems. And if we do it well, we will enable more people to do valuable work. But it's going to be, that doesn't mean, even if that's good over a while, it doesn't mean it's good for everyone by any stretch of the matter. Are you suggesting that labor markets do not inevitably return to a Pareto optimal equilibrium? That we can just trust the invisible hand? I mean, you're an economist. I don't know what you teach in your classes. I think labor markets, well-functioning labor markets, are central to democracy. You cannot have a well-functioning democratic system where most people don't perceive themselves to be stakeholders and aren't perceived to be productive contributors. A society where everybody is seen as a ward of the state is not politically stable because the people who perceive themselves as the income generators will say, well, why are we supporting all these people? So the best case scenario is a country like Kuwait where you have no taxation and no representation, but at least the dictatorship is relatively benevolent. But we have a real stake in maintaining labor's value, making sure that work pays well and that people have economic security that is tied to it. Work is not just about the displeasure of going to a job, right? It's about community. It's about identity. It's about esteem. For many people, it's also a source of meaning and fulfillment. I think... By the way, I think one of the things that we can be most optimistic about is that a greater and greater proportion of humans on planet Earth every year have jobs from which they can derive meaning. This has been the best 30 years in the history of humanity in terms of elevating people out of poverty. You know, from the perspective of the West, we say, well, it's been kind of mid, you know, we haven't grown that much. But billions of people have been lifted out of poverty actually by China's rise and the prosperity is created in sub-Saharan Africa, in Central and South America. So yes, it absolutely is the case. And, you know, these are, you know, China's, it's not a democracy. It is a market economy. And that has helped. Yeah, for sure. So just a couple more questions, David. If you were in charge of the world, we have a benevolent dictator question. You know, political constraints aside, what would you do with respect to AI to lead to the promised land that you described? You know, a couple of things. First of all, I would, the first thing I would do just because it's easy is I would regulate the use of intellectual property. I actually think there's enormous IP theft occurring in industrial scale. Intellectual property law was not meant to protect people against machines that would read everything and regurgitate it. So I think people should be paid for their creations. And that's a first order problem, but more, you know, kind of at a deeper level, I think we should be investing in these good use cases. We should say, how do we make healthcare a place where more people can use expertise, where it's cheaper and more available? How do we make education more effective and accessible? And actually, AI is a way, a tool that we can use to make AI, sorry, to make learning more like simulation, right? More like a flight simulator, less like a chalkboard. And so I would be investing for good use cases. And then I would build lots of labor market supports that are not really around AI. They're just around insecurity. We have trade protection programs in the United States for people whose jobs are displaced by trade, but we don't have technology displacement protection programs. Why not? If I lose my job, I lose my job. I don't care if it's because I lost it because of a robot or I lost it because of a competitor in another country I just need economic security We have models of countries that do this much more effectively than we do Now it hard We can just become Denmark But we shouldn think that it impossible to do because we know it is possible to do And certainly even during the Obama administration they stood up what was called a wage insurance program for people displaced from trade. And essentially, it said, if you take a new job at a lower wage and you've been displaced by trade, we'll make up some of the difference for a while. And that got people back into the labor market faster. It didn't help them recover earnings per se, didn't get them back to where they were, but it made that transition at least more successful. And long-term unemployment is devastating for people, not just economically, but psychologically. And one final question, why do you do this work? It just seems the most natural thing in the world to care about work and how people make their living. I just think it's actually, I think the labor market is the most central social institution, more than the church, more than the government, more than the bowling alley that's talked about by people who study social capital. And it's constantly changing, right? It's at the center of these tectonic forces of globalization, of trade, of technology. And I see how it affects people. I worked for several years in nonprofits. I did education for poor kids and adults. I did volunteer work in South Africa. And I saw how the world was changing around them and how difficult it was for the people to adapt and how the institutions were not in place to enable them to thrive. And so I feel like the first thing we need to do is understand so we can respond appropriately. And I do hope that my work and as my colleagues and many other people in my field have helped us better understand both how computerization has changed a set of jobs and the skills we need, how labor market institutions shape opportunity, shape wages, shape allocation and fairness, and how globalization has been such a mixed bag. Something that economists were really rah-rah about, my work, and that was my co-authors, is it really suggested that it does lead to economic growth, but the direct cost to individuals are enormous. And this will be the same with AI in some ways. It will create growth, but there will be huge displacement, and the beneficiaries will be pretty diffuse, and the people who are on the losing side will be very concentrated. They'll know why they lost their job, and we should invest in them and try to, both for their sake, because it's morally just, because it's the right thing to do. And collectively, it's really in our interest to have a society where people feel that their opportunity is protected and preserved and invested in. And, you know, we would want them to do the same for us. And we need that for, you know, political function and social harmony and for human well-being. And democracy. Democracy. Yeah. Weird. Social cohesion is efficient. Well, David, thank you so much for being with us. It's obviously a fascinating and timely topic. I do hope you're right. You know, so I hope I'm right. You know, I want to say, as I say in the end of the article, you know, I'm painting, I think, an intellectually and morally coherent idea. It's not really a prediction. It's not a prediction. It's a possibility. It's saying, like, this is something we could attain and we should try. Yeah. It's not out of our reach. No. It's actually more within our reach than it has been. I am 100% in agreement with that. The future you describe is absolutely possible. It will just take a remarkable amount of political and economic leadership to get us there. Yeah. And my guess, Nick, it's not what your buddies in venture capital are hoping for when they invest in these A. That is not what Marc Andreessen is thinking. No. No. Even not knowing him personally, I'm sure that that is correct. Anyway, thank you so much for being with us. It's been fantastic. Thank you very much. Great show and appreciate the opportunity to speak with you. I think it's important, Nick, to reiterate what David said, that he is not predicting the future. Yes. He's not predicting this utopian future. He's just painting a portrait of a possible future in which AI benefits everybody. Yeah. And I think that it is absolutely true that that future is possible. Right. I just think it's not probable. Well, not probable under our current institutions. Yeah, yeah, yeah. Yeah. I think that, you know, we have had tens of thousands of years of technological advancement. It is certainly true that over time that technological advancement has benefited people broadly. but as we spoke about in the chat that it's those pesky transitions that usually are pretty terrible right our lives are so much better for the millions of people who were mutilated in the mills of the 19th century and enslaved in the cotton plantations such nice fabric Right. Because of all that suffering, the modern world is made possible, the genocides, all that. We're better off for it. And I think the important thing is one could imagine that we could have gotten to this future in less awful ways. Yeah, that's right. I really, truly do believe that that is true, that there is a way to go towards the future ambitiously without the horrific downsides that these transitions generally generate. And, you know, I do think it's all about recognizing the downsides. But I guess if I have to quibble with David, it's just that, you know, I think that there's this view among economists generally that talent and ability is connected to benefits. Right. And that, you know, the most expert people, you know, are most highly rewarded. And, you know, in some abstract sense, some of that may be true. But, you know, this narrow ability, for example, to trade, you know, stocks on Wall Street or, you know, some other obscure financial thing, which produces an income of $100 million a year against somebody who teaches high school calculus. it's very hard to make the argument that the person trading financial derivatives or something like that is more expert than the person who is teaching high school calculus. Or more meritorious. I mean, David did make this comment that we basically have a meritocracy from the median up, and I wouldn't, you know, we've talked about this before. We don't believe it is a meritocracy. I think that the real danger going forward is to pretend that it's not about power. Right. And that the market will handle this, that the Mark Andresens and Elon Musks and Peter Thiel's, I mean, you look at their comments and it's like, they actually think we need the immiseration. Yes. I mean, Musk talking about we need a big economic collapse. There's got to be a lot of pain and suffering coming up. He thinks that's a good thing. And maybe that's because he looks at history and thinking all those people who suffered in his family's emerald mines created a much better world in the future. But there's this mean-spirited ideology among the super rich, which – No, you're Larry Summers. It's not all of you. No. Or Larry Summers thinking that the only way out of inflation is to put 7% of Americans or 8% of Americans out of work. That's the only solution. And it turned out there was another path. Yeah, there was. Well, there were a lot of other paths. But making the little people suffer seems like, to a lot of people, the only path, which is, I think, the problem. And so I think the – I don't see a problem in the technology itself other than the possibility of Terminator style like the AI becomes self-aware and says, oh my god, humanity does what? Yeah. And decides to enslave or destroy us. I guess that's a possible future too. Yeah. But to me, it's ownership. It always comes down to ownership. Are we going to allow the owners of the IP, of the intellectual property, of the AI to extract all of the benefits for themselves or most of the benefits for themselves? Or are we going to regulate it in a way that benefits, that broadly benefits society? And if we do the latter, things quite possibly will be okay, the vision that David put out here. If we allow Elon Musk and Mark Andresen and Peter Thiel and so forth to determine how it is used and who will benefit, it's not going to be such a good outcome. That transition is going to be very, very bad. Yeah. Well, it's a great conversation. And we will provide a link in the show notes to David's piece. AI could actually help rebuild the middle class. And for links to everything we just mentioned, plus transcripts and more, visit our website, pitchforkeconomics.com. As always, from our team at Civic Ventures, thanks for listening. See you next week.