Russ Roberts

Tyler Cowen on Inequality, the Future, and Average is Over

EconTalk Episode with Tyler Cowen
Hosted by Russ Roberts
PRINT
David Epstein on the Sports Ge... Oster on Pregnancy, Causation,...

Tyler Cowen of George Mason University and blogger at Marginal Revolution talks with EconTalk host Russ Roberts about his latest book, Average is Over. Cowen takes a provocative look at how the growing power of artificial intelligence embodied in machines and technologies might change labor markets and the standard of living. He tries to predict which people and which skills will be complementary to smart machines and which people and which skills will struggle.

Size:27.5 MB
Right-click or Option-click, and select "Save Link/Target As MP3.

Readings and Links related to this podcast

Podcast Readings
HIDE READINGS
About this week's guest: About ideas and people mentioned in this podcast:

Highlights

Time
Podcast Highlights
HIDE HIGHLIGHTS
0:33Intro. [Recording date: September 16, 2013.] Russ: Our topic for today is your new book, Average is Over. It is, for me, your most thought-provoking book. It outlines a rather daring, and daunting, possible future for America, and the world--a mix of great optimism with, it seems to me, some pessimism. So let's start with the title--"Average is Over." What are you saying with that title? Guest: This is a book about the increase in income inequality in America, and it's trying to trace through that trend and see what it looks like 10, 20 years out: what are the advantages, what are the disadvantages, and what's driving it. So, "average is over" is getting at exactly that increase in inequality of income. Russ: You are really suggesting that there is going to be a bi-modal distribution, though. You are making a much stronger suggestion than just income is going to get less equal. Guest: Well, I think over time a lot of people in the middle class, in fact most of them, think we'll either grow into being quite wealthy, or they will have more stagnant income in the lower-middle class range. So, I think what we thought of as middle class America in, say, the 1960s, will be viewed as a historical anomaly. Lots of people will benefit greatly from this change, but not quite everyone. Russ: So, I think you know I'm somewhat of a skeptic about the inequality data--that I think it's distorted by changes in family structure, immigration, and other factors. And that boils down ultimately to some statistical discussions we don't have to have. But if you are going to push your viewpoint that this is a real trend, an important trend, what do you see as the causal, underlying causal mechanism for this growing inequality? And in particular, the stagnation of the middle and the bottom? So I don't have any trouble with the idea that the highest incomes are going to get higher than they were before relative to the median or the mean--because of globalization, technology, a lot of reasons. But you are saying something much stronger than that. So why do you see that happening? Guest: Well let me first point out that even if you are skeptical about the macro data, there's a lot of micro or sector-specific evidence that inequality is going way up. So, the Harry Potter books, or Da Vinci Code, they earn much, much more than best sellers of the past. But if you look at average advances for authors, those are in fact way down. If you look at the economics profession, at the top end, people like Paul Krugman, Greg Mankiw, again earn much, much more than their predecessors; but there are more adjuncts being hired than tenure track professors. So I think you can look at the economy sector by sector and in most sectors you see some version of this. Russ: Okay, I'm going to challenge that. I don't find that compelling at all. I'll tell you why and see what your response is. So it's true--this is my point about the technology and globalization--it's the argument that Sherwin Rosen and Ed Lazear made in their "Economics of Superstars" paper, that the highest people at the top of their profession, whether it's sports or the arts are going to make a lot more than they used to because of that phenomenon. But your second point is the one I dispute; you said, well, the average advance is going down. I'm not surprised at that. And that's because a lot more people are writing books. So, a lot more people are coming into the distribution, pulling down the average. But the people who would have been there 20 years ago could be making more than they did before. So I think that's a very distorted measure. Guest: Well, I would say, as I often have, that I'm a happiness optimist but a revenue pessimist. So if you look at the total amount of revenue going into authors' advances, again, if you take away that top part of the market, I think that's definitely down. I think you see the same thing in the music sector. But just to turn to the main question I think you wanted to ask me: Two of the driving forces, one of them is simply the Internet and automation and the general rise of a new class of machines connected to computers and smart software. I think those extend the reach of the most productive workers and I would include smart phones and cell phones and better communications in that. So there are many more global markets and the people who are truly talented can sell to more people and sell to them with better marketing. That accounts for some of the increase on the upper end. But at the same time, automation is replacing some jobs. But we don't have the same kind of flexible economy that we've had in many times past. And I think on net you'll see labor force participation rates are falling, so fewer people are working. And thus we are getting a lot of stagnation with the other income classes. The other driving force behind this I think is globalization: that there are pressures which equalize wage rates of labor across countries. So there have been fantastic gains in India, China, as well as many other places, as you know. Much of this is positive-sum, but I think a lot of it is actually zero-sum. And there's a kind of arbitrage applied to the price of labor, as there would be to, say, the price of apples. Russ: So when the price of labor goes down worldwide but up in the places that were low, so-- Guest: That's right. So global inequality is way down; but within virtually every country inequality is very much up. If you look at the United Kingdom or Germany you'll find most of the growth there has been in lower-wage jobs. And again, the typical or median wage of the last 10-12 years has slightly been going down.
6:33Russ: So, you paint a very interesting, really incredibly interesting vision of how smart machines might change our lives. And we are going to talk about the implications of that. But I think for most people they've only seen the tip of the iceberg. They might use a GPS (Global Positioning System) device in their car or on their smartphone. They might ask Siri on their iPhone for directions or a recommendation for a restaurant or a movie. And that's fun. It's kind of cool. It's not life-changing. But you see some rather extraordinary things coming very soon, when smart devices get married to big data or other types of information technology. Talk about some of the things you see coming. Guest: Driverless cars are coming fairly soon. They basically work now. We need to set up a legal and regulatory environment; that will take a while. But the technology works. This will ease commutes; it will mean cars can drive much faster, spread themselves out more; it will reshape a lot of economic activity. And of course there will be driverless trucks, other kinds of driverless vehicles. This of course will eliminate a lot of jobs, create some others. But it will really be a big change in our lives. The artificial intelligence program Watson is now being applied to medical diagnosis. In less than 5 years and maybe less than 2, this will be online. I think rapidly the price, whatever it is, will fall. And there will be very high-quality medical diagnosis available to a lot of people, no matter where they are located. Even if they are not very well off in terms of income. In general, a lot of the service sector jobs are now being automated; a lot of what legal assistants used to do. It's possible now that you have artificial intelligence programs which can grade essay exams, which is what you and I have spent a chunk of our lives doing. So I think there's been a series of related breakthroughs in smart software which are really going to change the job market. Russ: We have more to say about this; you have a lot to say about it and we'll continue to get into it. But some people--Robin Hanson on EconTalk in the past and Ray Kurzweil--have a very different vision than yours of what this growth in artificial intelligence, machine intelligence, whatever you want to call it, how it might affect our lives. So, quickly lay out their vision and why you dispute it. Guest: Well, the Ray Kurzweil view is within 50 or so years we'll all become computer uploads; our brains will be uploaded into computers. I just find that very, very hard to believe. I think it's a new kind of secular religion. Robin believes that, too. But he also believes that something akin to robots will basically only lower wages, because robots will compete with humans. I would stress much more that humans can always complement robots. I'm not saying every human will be good at this. That's a big part of the problem. But a large number of humans will work very effectively with robots and become far more productive, and this will be one of the driving forces behind that inequality. Russ: We're not just going to be used as an energy source for them, as they suck us dry or whatever. You're not that pessimistic. Guest: No. Again, I think 'pessimistic' and 'optimistic'--they are words everyone in this debate is applying too quickly. I think in this debate the radical thing to do is to write a book which isn't trying to be too normative; just try to think through what will things be like. And by keeping the evaluation at much more of a distance I think we'll actually get further with the analysis. Russ: Yeah. I actually enjoyed that a great deal. It does lull one into thinking you are just describing--it's just a history book that happened to fall into our hands today. So there's a little risk there. But it's very provocative. Guest: I've had people confront me in outraged fashion: How can you accept all of this? But look: as a writer the point is to try to figure it out as best you can, and at the end of the book if one wants to say, let's not go that route, well that's worth a discussion. But the notion that at every intermediate point you have to inject your emotional outrage I think has become one of the worst features of this whole debate. Russ: It cuts across other debates, too, doesn't it? Guest: Absolutely.
11:01Russ: So let's talk about what you've learned as a chess fan. And you write at some length. At first I was rather taken aback by this, but I grew to find it quite fascinating. You write at some length about the role of machines in chess tournaments, and particularly in freestyle. Talk about that and why it's a nice potential template for future human interaction. Guest: Freestyle is a form of chess where a human teams up with a computer. So, if you play human-and-computer against computer, for the most part human-and-computer, if it's a practiced human, will beat the computer. Even though computers per se are much stronger than humans at chess, it's the team that's stronger than either one. And I think this is a good metaphor for a lot of what our job market future will look like. So there's a big chunk of the book that looks rather closely at freestyle chess and tries to see what we can learn from it. Russ: The thing I found most provocative about that is that the best freestyle teams do not necessarily have the best human players. In fact that could be something of a handicap. Guest: That's right. The really good human players are too tempted to override the computer and substitute in their own judgment. The best freestyle teams, they are quite epistemically modest, the human or humans involved. And what they are really good at is asking questions. So they'll run two or three different computer programs and then just check on where do those programs disagree. And then they'll probe more on those points. And that's what the humans do well that the computers, at least not yet, aren't able to copy. So it's knowing what questions to ask that has become the important human skill in this freestyle endeavor. Russ: So, applying that to the medical diagnosis example you gave earlier, it suggests I don't want the guy or the woman who had the best grades in medical school or the most arrogant--which is often in today's world, can be, the best doctor. I might want the most modest doctor, or not the most modest, but someone who is willing to let the diagnosis provided by the machine be the "right" one. Guest: That's right. So, wisdom and modesty will become much greater epistemic virtues in the future scheme. I think that's overall a good thing. We should revere those qualities more. And we will have to, looking forward. Russ: You also have a lot to say about conscientiousness and advantages that women might have over men. Explain. Guest: Well, one thing we are going to get very good at in the future--you see it now--is just measuring quality. So, whether it's doctors, lawyers, economics professors, there's always a randomized control trial now; there are always numerical ratings. Everything has a Yelp rating or an Amazon rating or something. And we all know these are highly imperfect but basically they are still better than not being informed at all. So it's like in the future there's a credit score for everything. So people who test well young I think will have a lot more invested in them early in their lives, early in their careers; and they'll have a head start. And another way to think of this is, I think, within 5 years the world's best education will be available online and it will be free. Arguably that's already the case. But the question is: Who is there to learn from this? It's the people who are disciplined and conscientious, which is still distinct from just raw intelligence. Now, if you ask the question if you compare men to women on average which group is less conscientious, I think you have to hand that one to the men. At least the lower tail of the distribution. So I think we already see in higher education and many other areas women doing better. And not just better because there is less prejudice. They are just outright doing better and out-competing the men. And I think that trend will be magnified by this increase value for conscientiousness. Russ: Explain again why conscientiousness is going to be more important then than it is now? Guest: There will be many more free resources; and there's already a lot. So the person who is just disciplined enough to sit down and, say, listen to EconTalk podcasts or read blogs or go through Kahn Academy or whatever it is they ought to be doing, that will be there for them free. So, what you'd call the 'shadow value' of conscientiousness in economic terms will be much, much higher. And you see that--students in India, they take Coursera classes and they are brilliant; but they are also really determined to work hard. And they are outperforming in general a lot of these top Stanford students. And that's being measured and picked up. And those people I think will do very well in this new water[?], and that's again a case of conscientiousness paying off. Russ: I guess it's not just the advantage of being disciplined to sit in front of the computer on your own time, but it's also the fact that possibly it will be harder to get somebody to be drilling you and pushing you face to face. Guest: That's right, especially for people with lower incomes. But the free things will be there and the people who are maybe brilliant in some strange way but need to be in the middle of a lot motivation and pushing and peer effects, those people may be worse off. Russ: You have a description in the book of something I thought of as Siri on steroids or Google Glass run wild. You are in a negotiation and your phone vibrates or your glasses send you a signal that says the person is lying and encourages you to kiss your date because temperatures changed and something sensitive that you are wearing or carrying. What I found fascinating about that--I'm a little skeptical about its accuracy but that's okay; some of it's definitely coming. I really was interested in your point about the cultural challenge that we face with these kind of devices. Guest: I think we'll record a lot more than what we are doing; we'll measure a lot more. And again, when your smart phone tells you to kiss your date or whether someone is lying, the point is not that this is perfectly accurate. It just has to be a little bit better than you would do on your own. So, you'll walk into a Whole Foods and your smartphone will tell you: Here are the 5 things on sale. It will tell you, talking to your refrigerator: These are the 3 items to throw out and don't really finish eating. It will give you a lot of advice that will one way or another communicate to you the content of reviews. So it's like going around through life with all of recorded intelligence available at your fingertips. If you are willing to listen. Russ: But you suggest that the cultural transformation might be a little bit challenging for some of us. We might not be so willing to listen. Guest: I think we are all a bit reluctant to become so subservient to computers and to software programming. And those people who are not, those people who can listen yet without losing their sense of self, without losing their inner propulsion, I think that will be a highly propitious mix of character qualities to have. But I think a lot of smart people--again, they'll try to do too much themselves, like some of the lesser freestyle players, and they'll end up getting in a lot of trouble. They'll be out-competed by the wiser and more modest individuals. Russ: Now, most of this sounds pretty great. A lot of what you are talking about is going to make life more fun. Maybe more silly. Maybe more gadget-centric. But a lot of it is just going to be pleasant. Guest: Like GPS. It's mostly an advantage. It's not perfect. Russ: Right. Occasionally it takes you down a one-way street. Or takes you on a detour that was not necessary but somehow it thinks it is. Or roads have changed. All those things are going to happen, and as you point out, there will be a little bit of an arms race. The person who wants to lie, maybe wearing a bag of ice on them somewhere to keep their temperature down. Or there will be ways that their phone will chill them. Whatever. We don't know what will actually happen. There may be all kinds of interesting market forces at work in these equilibria. We don't know how it will actually pan out. But most of these things are really great. As you go by the movie theater it says: You'll like this one. Or whatever the things that are there. Guest: There's something oppressive about it. Because all the mistakes you make get recorded; they get measured. And we're not used to that. We're used to a lot of things being overlooked. Like I give the example of a guy who is on a date with a woman. And like he's a nice guy, but he just smiles a bit too long at the waitress, and her smartphone tells her that afterwards. Russ: The date, not the waitress. Guest: So the whole thing's against people and we're going to have to back out of that much more. I don't think that will be so easy. Russ: I agree with that. We see it happening now. Guest: One wrong Tweet and you get in a lot of trouble.
19:54Russ: Yep. You make a fool of yourself; you lose your job. But the point I'm trying to make is those things are all--I think it's an incredibly interesting time to be alive and I think the health applications are going to be mind-boggling. And the driverless car might save 35,000 lives, potentially, if it's done well, and allow us to get places faster while saving lives. Those are just wonderful, glorious things. The average-is-over part of the book, though, is not so attractive. Who are the people who are going to lose from this technology? These advances in technology? Who are the losers and why do you think they are numerous? Guest: Say that right now you are someone who is a truck driver, which often is a reasonably well-paying job. You don't get rich doing it, but you are not lower middle class if you have a good truck-driving job. And over time all of these vehicles become self-driving. We'll see more and more workers basically crowded into service sectors. There will be more people trying to do things like be yoga instructors or be personal servants. And as those kinds of jobs grow, I don't see the corresponding wage growth. I think there will be a fair amount of stagnation, basically in service sectors for people who are not extremely well educated at working with technical material or computers. Russ: Well, I guess the issue would be, you'd think the demand for a really good yoga instructor would go up, from these very wealthy people who would have all these great productivity enhancements. So I think the question is: Will you not even need a yoga instructor because your phone will tell you, or your Google Glass will tell you what posture to be in and show you a 3-D hologram of it in front of you. So it's even better than a yoga instructor. In fact, as you said, it will tell you when you did it wrong. You'll be talking to the cameras in the room and it will give you advice on how to do it better the next time. Which the yoga instructor may struggle to do in a room of 20 students. If that's the future, where the yoga instructor can't make a living, I agree with you. You suggest at one point that really rich people like to be fawned over and complimented; and people who are good at that will be able to make a living, for sure. But that's very depressing. Most of the tasks that really wealthy people used to pay for, they don't pay for any more. They don't have to have a cook because they can do take-out. Take-out is cheap now. They don't have to have a dish washer--a person who washes dishes--because they have a dishwasher. We all are rich that way. That's great. That's been the glorious transformation of the middle class, not only of the upper class. But you are suggesting that that's going to come to an end. Guest: Well, I think there will be a lot more jobs in service and restaurant sectors, but except at the very top, I don't think those jobs will pay so well. There is a lot of competition to do them. Indirectly there's competition from overseas, which will only get stronger. So I don't see significantly rising wages in those sectors. And I see more workers flowing into them because of these other jobs' disappearing. So that's the downside, in my view. But let me just note, I think there's some much longer term, which I'm not trying to forecast in the book, but say, 50-70 years out, where you get enough neat, new stuff that's almost like scarcity isn't there any more. So I think that the very long-term picture is quite utopian. But you know, go back and look at the Industrial Revolution. Let's say that starts in the 1770s or 1780s. In England real wages are not going up at a high pace until the 1840s. So there's there 50-70 years of transition, where even free market economists like David Ricardo are writing about the machinery question and then worried about this. Russ: But it turned out okay. Guest: Well, I guess after 50-70 years. And I think 50-70 years out, our own scenario will be much more utopia than dystopia. But, say, 20-30 years out, I think is much more mixed. Russ: And that's where Robin Hanson disagrees with you, I guess. He argues that the 50-70 year scenario is even worse. Guest: Well, he doesn't think it's bad. He thinks we're all computer uploads earning subsistence, but there's a lot of us that are just fine. I'm not sure how to adjudicate that ethically. But it's fair to say most people don't like the idea when they hear it, right? Russ: No, they don't. Guest: But Robin thinks that's great. Russ: I didn't realize that. Guest: He does. Russ: Oh, okay. Guest: That is utopian for him. Russ: Maybe that's what he'll get. Maybe it will be a choice in the world, but maybe the rest of us will be able to do something different.
24:44Russ: One area you do see a lot of potential for is coaching and marketing. Explain first what you mean by marketing. You don't mean what I think what everybody else might mean by it. So, talk about what you mean by marketing and coaching and why they are both likely to be very useful. Guest: Coaching and marketing, they are both forms of motivation, and one thing I think computers are not good at is motivating us. So you can write a program to send yourself text messages and that does a bit, but what really motivates us is what inspires us, human role models, examples, stories, narratives--things that are quite powerful emotionally. Teachers we met when we were young. And I think those will be big growth sectors. And in terms of marketing, imagine, say, that 20% of America is millionaires or even richer. There will just be that much more competition for their attention, because attention is still scarce. But they'll have a lot more money. So in competing for their attention, basically in different ways we'll all be doing more marketing. And I see marketing as really the single biggest growth sector of the future, viewed in these somewhat unusual terms. Russ: Of course, we're all marketing these days in all kinds of unusual ways already. Because attention is scarce. Guest: That's right. And that will become more scarce, because the number of hours in a day will not really go up. Russ: And the two eyes problem--you only have two. Awkward. You can't see behind you. Guest: You'll get that extra hour in your driverless car. So in that sense we'll augment the supply of attention somewhat. But I think the supply of goods and services will outrace that effect. Russ: I agree with that. Let's talk about games. You have some interesting things to say about games and the application toward education. Guest: Well, if you ask in what sphere of modern life has education really succeeded, I think it's gaming. These games are incredibly complex. When I look at them I feel they are too complicated for me; I could never learn them. But the people who are interested--the game itself teaches people how to play. It's all done by software and very often online. And it seems it really works. They teach you in steps. They make it hard enough to be interesting but easy enough that you feel you are making progress. And in my view the big educational breakthrough has already come with games. It's just a question of how do we apply that to educating everyone else to do other things. And we're far behind on that. Russ: What are some of the ways that might be applied? Guest: Imagine an intro to economics textbook but structured more like a game, where you move on to different levels and maybe you capture pieces or you acquire points. And there's a competitive angle. And I'm not saying everyone has to do it that way. But I think there's really a big chunk of people who get interested in games, who would otherwise, say, never be interested in medieval history or battles or whatever else, but the games get the interested. And I think that's a big frontier for education, where we've actually solved the problem. We don't quite even know we've done it yet. Now we just need to apply what we've already learned. Russ: And do you think that's coming? You've alluded to a number of successful, moderately successful--it's early--online educational activities: Coursera, the Kahn Academy; and EconTalk. Do you think EconTalk as a game would do better? Guest: I'm not sure how we would make EconTalk into a game. You could make the testing of it into a game. But if you think of big intro courses in subjects like English, economics, where a lot of people take them, I don't see why it isn't profit-maximizing to have a gamified version of those classes. Which may be, you know, a third of students who choose to opt for it; and they do it as a game. And I think they would learn the material very well with a lot of immediate feedback. And enjoy it a lot more. So, yes, I think we will do that, pretty soon. Russ: A lot of people would say that undergraduate education is a game: the grading system and how you pick your classes and how you pick your professors is a game. But you are talking about something a lot more--gamey. Guest: Yeah; that's gaming it to be easy or to be a kind of fraud. This would be an actual true game. As chess is a game, or World of Warcraft is a game. And you actually play it in your spare time, which has now become your study time. And that's what keeps you interested, is the competitive and gaming aspect. Russ: Although I agree with you that you learn something about medieval life from some of these games, warfare, violence, strategy and so on, it's not obvious to me that you could turn your intro Literature class or Econ class into a game. I think that's part of the reason it hasn't happened; it's not the only reason; obviously there's a lot of other entrenched forces. And you sound very Smithian in the book about your view of the education industry, just as he did in the Wealth of Nations. He has some not-so-positive things to say about the vested interests of professors. But I suspect that's going to be challenging. Guest: Yeah; a top-level game together, private companies, might spend $200 million on it, which is really a lot. Now, are companies right now ready to spend $200 million creating a gamified version of, say, an Introduction to Calculus class? I think until they know it will be accredited and accepted, they are reluctant. But I do see the gains from trade there, and I think there will be multiple ways of learning all of these topics, which will compete with each other. Probably no one will win out; different people learn according to different styles. But I think the main styles will be turned into reality. There will be like Kahn Academy style, there will be gamified style. There will just be narrative chat style; there will be listen-to-podcast style. And so on. Russ: Yeah, I agree with that.
30:52Russ: Talk about the future of economics teaching generally. And the profession. You have a section on that. Talk about what you think is coming. Guest: As you know, in the United States, what--70-80% of students go to state universities of some kind. And really most of those institutions feel squeezed by budget. So the idea that each of these schools will support a full line of tenure track faculty I think is coming to an end. I think the top schools will--the top schools will do more; they'll fund more superstars. But again, there will be this thinning out. I think in some ways this will damage research. It will be harder to be a successful researcher in the middle of the sector. I think there will be demand for professors to teach much more. A professor will work in conjunction with online programs, and the people who are the best teachers will not necessarily be those who know the most economics but will be more like tutors. Like, what can you do to motivate your students to spend time with those online programs? And you see this already in Hong Kong and in South Korea. There's a very active private tutoring sector, and the best tutors can earn millions of dollars a year selling their services. And again, they are not Nobel Prize winners. They are geniuses at motivation. Russ: That's a huge part, I think, of high school education certainly is the ability to motivate adolescents to stay focused. I think that's a huge part of why face-to-face in high school is still necessary to some degree. And only the most dedicated and conscientious students are going to be able to take online classes in their entirety. And there also are going to be superstars who present the material, who do the online classes. Those people are going to make a lot of money--great teachers/tutors are going to make a lot of money. The beautiful thing, to me, is that there won't be as many of them and the average quality will be much higher. Than what we have now. Guest: I agree with that. Russ: But at one point you talk about the role of empiricism in economics. I don't have the exact quote at my fingertips, but you talk something about the typical good economics paper of 2013--you see economics potentially moving, and you see it has already moved, into a very different mix of theory and empiricism. Describe it. Guest: In my view there aren't really any new theories left in economics. I don't think we'll discover it 20 years or 5 years down the road. I think we have most of the core theoretical ideas. We can argue about their relative importance. But great papers today are mostly: What data set did you find? They are not from mining or re-mining the available data sets. That's already been done to death. It's people working in areas like development economics finding or creating their own data and then doing something creative with it. But this I think will continue. I think also finding data sets is something machines cannot do very readily. So that will remain the province of humans for the foreseeable future. And we'll have a completely empirical profession. And I see already in young graduate students--a lot of them, they just don't learn standard old-line, like Chicago school, UCLA microeconomics. It's a kind of dying art. Russ: I agree. The question is what is replacing it. Which is--I'll quote an unnamed professor who told me that when he studies the labor market he doesn't have any pre-conceived theories; he just listens to the data. He concludes whatever the data concludes. So, it's theory-free, but 'it's better,' because it's just what the numbers say. And I find that an untenable vision. Because the numbers are very loud and cacophonous and they say whatever we want them to say. When I think of the best economics paper of 2013, you are going to say it uses a new data set. I'm not sure we are making a lot of progress. You are much more optimistic than I am about, say, macroeconomic stabilization policy and our ability to access, using data, better policies in the future rather than using better theories. You think the data are going to get us there. You want to try to defend that? Guest: Well, I share a lot of your worries. I think we've already moved to around what I call 'common sense theory' is strongly undervalued. What's valued is intricate theory, which I think is often of quite low marginal value. And the notion that the empiricists will just run wild, that worries me, as it worries you. But that also said, I think there's something called 'big data' coming. I think we'll learn a fair amount from it. I'd like to see what we learned filtered more by common sense theory than what we are likely to do. I think macroeconomics, we've actually learned a lot of the key insights. It's just people are never that happy with the stabilization program because there are big losers in recessions and depressions. But I think a lot of countries are responding pretty well to their crises. In part the problem is it's just not always possible to do better. Russ: Well, I agree with that.
36:16Russ: I do feel you give insufficient shrift--which is a word I've never used without 'short' in front of it--short shrift to the knowledge problem that Hayek identified. To the causal density that Jim Manzi talked about on EconTalk. For example you talk about: I'll go pick a lawyer, I'll be hiring a lawyer and my smartphone or device will tell me what grades this person earned in law school, the quality of their law school, their success rate in trial. But those things don't tell us very much. Those aren't very helpful. It's like when hospitals publish their success rates--sometimes against their will. We all understand that if they have difficult cases then their success rate is not a good measure of their quality. You think we're going to be able to parse all that out and independently control for all those factors? Guest: Well, to what extent? Keep in mind that every client and every case will have its own metric of quality. So, if you lose a lot of cases taking on poor people accused of being murderers when the gun was found on the site with their fingerprints, those cases get graded too. And your marks are adjusted for that. But again, you don't have to believe these grades will be extremely reliable, any more than credit scores are. You just have to think they are better than not really knowing anything. So, the grade will be supplemented with personal recommendations. We already use personal recommendations. And I think the grades will be viewed and are viewed, like Yelp reviews, Amazon reviews sell books, as being informative. And I think down the path of this world where everything is graded. But no, I don't think it's all going to be that accurate. That's one of my worries, actually. Russ: What's Google doing to our brains? Guest: With Google it's easier to forget things, as long as you know how to find them. I find, in the case of myself, that's how I operate. I remember fewer facts, but I've really honed my ability to define the right keywords. Russ: Do you think that's a real phenomenon? Guest: You mean affecting a lot of people? Russ: Yeah. It's true for you. I still know Ty Cobb's batting average. 367. Guest: 367. Russ: I'm stuck with that. I can't get rid of it. I'd love to--I'd really like to forget it; I'd like to make room for something more useful. When I was 17 or 14 I thought it was extremely important. And I'm stuck with it now. So, do I really have that kind of control, to forget stuff? Guest: You and I have this privileged position of spanning two different human cultures--pre-Internet and post-Internet. And that's actually fantastic, and I think we'll go down in history for a kind of richness of our vision, because we've lived in both worlds. But I think looking forward there will be many fewer people who have it stuck in their minds as the two of us do that Ty Cobb hit 367. Russ: You don't think my kids are going to remember it? Guest: I think they will remember fewer things. They may remember that one. But they'll be much better at finding information than either of us. Russ: One thing you didn't talk about in the book is books. I think about this a lot. I went the other day and I gave away, I don't know, maybe 200 books to the Friends of the local library, the organization that's a fundraiser for the library. I felt better doing that than throwing them out. But I have too many books. Like you, I get lots of books, still. People send me books. Increasingly they are digital. But people still send me physical books. And I oscillate between thinking: Why do I have any physical books, and between thinking: Oh, I miss them so much already. And I have hundreds and hundreds still. Maybe a few thousand all together. I've culled my collection very modestly. But I'm in this in-between era that you are talking about. I've got this still love of the physical and I have this love of the Kindle. And my iPad. So, what do you think is going to happen there? Do you think books are going to die out? Physical books? Guest: I don't think they'll die. And even vinyl LPs (Long Play records) have not died. They are making somewhat of a comeback. But I do think there will be a big generational shift away from the physical product. There will be a general decline of stuff, including the desire to own a home of your own. And we see this in the data--lower rates of household formation, lower rates of owning stuff. And I think we will be viewed as this group of transitional people who were lucky enough to acquire a love of both worlds. Russ: Time out. Now, I have fewer books, maybe. Actually I don't. It's just that the rate of growth is slowing. I probably have more books than I had a year ago. But I have a lot more devices. I have an iPhone, an iPad, a MacBook Air, and a MacBook Pro. I'm not particularly proud of that. And I've got some leftover ones from the past, just like the books I've finished. These devices are sitting around; they are being used by my kids; they are sometimes just decorating my basement; I have a Mac Classic down there from 1984 that I can't bear to throw away. You think people are getting rid of stuff? They are getting rid of books; I don't think they are getting rid of stuff. Guest: Well, what we see in collectibles markets is that overall the prices of collectibles have fallen after the advent of e-Bay. That to me suggests a net desire to get rid of stuff, because to get people to hold the same stock--you are reselling them at lower prices. Russ: Yep. Guest: And I think your children will not own as many books as you do. And Yana will not read as many physical books as I do--not close to it. And she reads a lot. Russ: Well, my daughter is a bit of a rebel. She still eagerly acquires physical books, and I sometimes have to carry them out to the car in her suitcase. It's awkward. But I think you're right. I think the next generation, and next two, are going to be very, very different in how they--they obviously are already--consume their information. Guest: I went to Best Buy one day ago and I wanted to buy a CD player. I assumed they would have it. They don't sell CD players to put into a stereo system. I was shocked. Call me naive. Russ: And so you had to find it online somewhere? Guest: Correct. Russ: It's still there. Like you said. You can still get it. And it's reasonable. It's not like buying a new typewriter. I don't even know if you can buy a new one. You probably can. But there are lots of used ones. You can find them. And so you are able to still buy one. But they are not extinct. They are just scarce.
43:21Guest: But the way we will use physical space in our cities is really going to change. There will be much more about entertainment and looking good. And I think you see a lot of scientists already saying Manhattan. Or San Francisco. Russ: As opposed to what? You said 'looking good,' 'entertaining.' Hasn't that always been the point of a house? Guest: Well, cities. Cities used to have more manufacturing. The 20th century city also is a big home for poor people. And I think 50 years from now we'll look back on that and be shocked at the notion that so many poor people got to live in Manhattan, Los Angeles, wherever we are talking about. Because in the future they will be in other places with lower rents. Russ: Why don't you talk about that? It's an interesting idea; I'm a little skeptical of it. Talk about your view on the world's most livable cities and where you think population distribution is going in the world of Average is Over. Guest: What I see in the data is that wealthy people want to live in really nice cities. That they are capable of gentrifying them. A city like Washington, D.C. has gotten much nicer. And that pace if anything seems to be accelerating. People of lower income are moving out; they are going to the Carolinas or wherever. They are doing fine; it gets a lot of them out of high-crime areas. It seems to me to be a better outcome for virtually everyone. The cities are much more expensive, even though real estate as a whole is kind of a flat-line investment, you look at quality coastal property and we are passing the previous peaks of right before when the housing bubble passed. So we are not producing any more of that land; you and I would probably agree we should deregulate it; you and I would probably agree we are not about to do that. We are not even going to deregulate Fairfax land supply. So that stuff gets more expensive, wealthier, gentrified; and poorer people will move more to the South, the Midwest, Texas, Oklahoma. And I think there will be a much more geographic segregation in the United States. And elsewhere. Russ: Talk about the Texas phenomenon and the relationship between size of government, government services, cost of living, and mobility. Because you have some interesting things to say. Guest: Well I think one way to get a glimpse into the future is just to ask: Where is it that people want to move today? Now I'm not saying those movers are representative of all Americans. But still you see a net tendency. And the state attracting the most migrants is Texas. Three of the five fastest-growing cities in the United States are in Texas. And you might ask: What does Texas have? Does it have great weather? great wealth? Probably not. Crime rates are mixed, but they are not great. Public services are probably below average. But they have a fair number of jobs and really cheap housing. And that suggests to me what people really, really like is cash in their pocket which they are in control of spending. And I know a paternalist may disagree with that. I'm not trying to judge this normatively. But for the future I predict we'll see more of that: people moving to get more cash in their pockets. Russ: Which goes against, I think, the way we think about migration. We think about moving to Manhattan, Washington, D.C., San Francisco for either the work climate or air--the physical climate--or the beach. Most of these--I said the work climate--that was a bad example. But a lot of non-monetary intangible benefits. You are saying that a lot of it is going to be monetary. That's going to drive where people want to live. Because of the cost of housing. Guest: Of course. And the elderly migration is the next big story. And the average old person--you know, their net wealth in the United States when you take away the house they own is well below $100,000. So they will want to live in cheaper areas, for the most part. And as long as these cheaper areas have an okay hospital, I think you are going to see a great deal of migration to them by the elderly. Russ: Well, you are correct to point out the importance of that because there are going to be a lot more elderly people. Guest: And budgets will be squeezed. The Federal budget will be squeezed; private budgets. Russ: But they are willing to pay a premium to be where it's warm. They tend not to move to Buffalo, NY and Duluth, MN. Obviously there's a tradeoff. They tend now to move in large numbers to expensive places, not cheap ones. Guest: But a lot of America's warmest real estate is also low cost. Places like Oklahoma are pretty warm; places like Alabama. Parts of the Carolinas. And those areas are seeing considerable influxes. So not everyone can move to Newport Beach. Of course, maybe everyone wants to. Russ: The entry price there is different. And of course, as you point out in the book, one of the reasons Texas is growing quickly is it--certainly Houston doesn't have the zoning restrictions other places have. That means that the real estate doesn't get bid up as much when people try to move there. Guest: I think we'll just see a lot more effective economization. So, wages and measured wealth will become less accurate as real indicators of wellbeing. We'll just see a lot more variations in terms of the costs people face and the benefits they derive from the Internet. So think of it as a kind of ethical view similar to a lot of religions, where we've said for centuries: Well, money isn't the main thing; it's the values you have and how you treat people and how you take advantage of what you've got. Very corny stuff. But it's true, and I think it's going to become a lot more true. Russ: Why isn't it more true already? I agree with you that these are universally important things, that money is overrated, things are overrated. And yet you are suggesting--it's like when people talk about the budget issues at the national level. And we'll get into those in a minute. People say, oh, states are running out of money, or the Federal government is running out of money. They spend a lot more than they used to. So to suggest that there's some crisis coming that's going to force people out into Oklahoma in greater numbers that used to go to California and Florida--I don't see why. What's going to precipitate that? Guest: Well, your general question: as you know, long-run elasticities are always greater than short-run elasticities. So, in the longer run there's always more adjustment and adjustment matters more. But I think also the world now is changing more rapidly again. So there are a lot of changes that have come along. A lot of them cued or indicated by the financial crisis. And we're just starting to adjust to those. I call it sometimes the Great Reset. But we're not close to being done with those adjustments. Now with state budgets, that's all about Medicaid. So, health care costs more; they spend more. But I think other stuff is getting crowded out, like higher education. Russ: Sorry to press you on this on short notice, but what do you think has happened to total spending on education at the state level in the last 20 years? You think it's down? You think it's crowded out by Medicaid? Guest: It really depends what number you take, not over the last 20 years but over the last 5 years, a lot of systems have seen very large cuts. And those will continue. Russ: Well, I think that's recession-driven. I'm not sure that's permanent. Guest: In my view it's mostly permanent. Not in every state. And the U. of North Dakota is probably going to get a whole lot better. But overall, you look at the U. of California system, in relative terms that's been declining relative to say budgets at private schools. State aid to that has been declining in absolute terms. Also in Virginia there have been and will be cutbacks in terms of percentage of the state university's budget which comes from the state. So the old average of over 30% will probably end up within 5 years as being 10-15%. Russ: I'm just raising the possibility that it's not Medicaid so far that's pushing those numbers out, but rather expanding access actually for many, many larger numbers of people. Guest: Oh, sure. That, too.
51:35Russ: Let's turn to the politics of the future and this greater inequality you have. You don't think the revolution is coming. Guest: I don't. I know we'll be much older, and old people are less likely to rebel. But they love to vote for law-and-order candidates. Russ: Yeah, I love that. That's a great insight. It's undeniably true. Guest: Inequality is way up, at least according to the numbers I look at; and crime is way down. And an awful lot of people have a hard time accounting for both of those things being true. Russ: They expect people to be on the barricades. Guest: That's right. What people resent are the people who are close to them--their colleagues, people they went to high school with, in-laws. For the most part, Americans do not resent the wealthy. Russ: Academics do, I think. And you suggest so as well. Explain why. Guest: I'm not sure I know why, but it's as if there are two worldviews competing. One worldview is: the richest people deserve more status. And the other worldview is: the most academically smart people deserve more status. And there's a clash there. So that is kind of a local rivalry. But for ordinary Americans like the home of envy is Facebook. Russ: Explain. Guest: That's where envy blossoms. You look at the people you grew up with, the people you know, you see how well they are doing; if they are doing better than you, you feel bad. That's what I observe. You're not worried about the Titans of Silicon Valley--you know, he earned another billion, I hate that guy. I don't see much of that. Russ: I like your point about people who worry about inequality as being a threat to the safety of the country: that if the 1% keep oppressing the 99%, violence is inevitable. You point out why not, part of it the age factor and part of it as you say, people don't seem to be so concerned about the faceless 1% rather than the 1% near them, wherever they might be. But you also suggested that maybe the people who are claiming this really aren't so worried about it. If they were, they'd have a different focus. They'd be trying to discourage people from worrying about it, and they don't. They go the other way. Guest: Exactly. It's as if they are egging them on. Like, Worry about this, worry about this. Russ: Yeah. Well, I think that is part of it. I think part of the reason we think we have an inequality problem is we spend a lot of time telling people we have one. So they start to think it's true. And as you point out, I think people look at the people around them, and in fact--they like rich people who aren't near them. They love looking at, reading about Bill Gates's house and the lifestyles of the rich and famous. They don't find--it doesn't enrage them. They find it entertaining. Guest: Sure. I think we have a quite genuine jobs problem, and to some extent an income problem; and those can get mislabeled as inequality problems. Because it sounds like a similar phenomenon. But conceptually they are quite distinct. Russ: Yeah. I think it's important to keep them separate.
54:40Russ: Let's talk about the politics in particular. So, you are also not so worried about massive redistribution from rich to poor that might be spawned by this bi-modal distribution. Guest: Well, imagine the future where say 15% of Americans are, like, millionaires. First, virtually all of those people are going to vote. But also, they'll be very socially influential and politically influential. And I think basically they'll build a coalition with a lot of the elderly, also preserving benefits for the elderly, and arguably underinvesting in the poor and the young by a number of criteria. And I'm not saying this is a good coalition, but I think it's our most likely political future. And you see in the world today: Medicare is quite safe, and Medicaid is under a lot of fire. So this strikes me as the direction we're headed in. Russ: Well, you also point out that a lot of poor people are not exactly politically radical. Guest: That's right. A lot of them, when they do vote-- Russ: How do you explain that? Guest: Some of it is common sense wisdom, that they don't think radicalism will solve their problems. And of course, that's usually correct. But I think some of it also is just disengagement. That to be a political radical, you need a whole conceptual apparatus, whether it's true or not. And some poorer people have it, but a lot of poorer people just have a more common sense view of the world. Again, often for the better. And that means they are not going to end up as radicals. They embrace traditional values. Russ: Well, they are trying to put food on the table, which tends to make it harder to indulge a vast philosophical apparatus. But the idea that they are also more conservative socially and other--at least so far in American history--is an interesting idea. Guest: Yes. Russ: It bothers people on the left a great deal. They view it as a form of, I guess, self-destructive behavior. Guest: Sure. But the progressive left is largely in the United States an upper middle class phenomenon. Russ: You claim that's true of Occupy Wall Street as well. Is that correct? Guest: Yes. I think the Occupy Wall Street movement has not had much resonance and it will have even less resonance if the new slogan has to be: We are the 85%. Russ: Let's close with the topic you don't really talk about; and you wanted to deflect a little bit in the opening of our discussion, which is optimism-pessimism. You are right now in the top 15%--you are right now in the comfortable-working-with machines, hasn't been displaced by machines. We both understand that the future may be very different for our children. The advantages that we've given them or tried to give them may not pay off in this new world. We maybe invested for them, gave them things that we thought would pay off in a 2013 world; but in a 2023 world might not be there. Overall, for yourself and for the future generally, how would you describe the future--the near future that we are talking about? With optimism? I mean, you are saying there's not going to be a revolution, not going to be violence in the streets. I think that's a good thing. Maybe other people would disagree. But I think that's a good thing. But it's not a very cheerful vision to me of 15% of the people really rich and the other 85%, they've got entertaining phones and they are working for the 15% keeping them entertained or flattered. Where do you see--say I ask it a different way--the human enterprise going in the near future? Guest: Well, I think it will have elements of both utopia and dystopia. We'll have a lot more wealth; there will be a lot of new and exciting technological developments, which won't just be metallic machines--they'll be technologies which help us realize what it means to be human or to feel emotions or to connect with each other. More people will be liberated from unhappy jobs. But I also see for a lot of people higher levels of risk, a lot of economic precariousness, a lot of people being underemployed in say part time or temporary service sector jobs that they just keep on stringing along. And that worries me. I think that's also going to be part of the future.

COMMENTS (26 to date)
Martin Brock writes:

Cowen's analysis assumes static standards of forcible propriety, but I doubt that these standards can remain unchanged.

Both Roberts and Cowen assume unchanging standards, particularly intellectual property standards, in the future. J. K. Rowling is incredibly wealthy, compared with authors in the past, not only because her potential market is larger and more rapidly growing thanks to technological innovation. She is relatively wealthy compared with the past, because standards of propriety channel the value of these technological changes to her and not so much to others involved in the development and employment of these technologies and their consumers.

This outcome is not inevitable and need not persist. Recent statutory changes in these standards, like TRIPS and other international conventions and the Millennium Copyright Act in the U.S., seem to move these standards in an unsustainable direction or to fix them at an unsustainable point, and many economists have addressed the likely effects and possible alternatives.

For example, an alternative copyright could protect an author's monopoly only until a standard financial remuneration is reached, possibly dependent upon the author's accounting for the cost of producing a work, rather than granting the monopoly for a specified period of time, Also, the scope of copyright could contract rather than expanding, so that Rowling has no right to exclude Tanya Grotter stories for example.

It's not obvious to me why states should enforce broader intellectual property rights. States don't care what's obvious to me, of course, but statutory standards do change. Speculating about the future in 1800, while assuming the persistence of slavery, led to many erroneous conclusions.

[Typo fixed: name of guest changed to Cowen--Econlib Ed.]

Scott Small writes:

The assumptions of Cowen in this discussion are untenable. When considering technology and society the problem of prediction is too complex.

One problem is that Cowen does not adequately account for how dynamic these systems are. Not only do the data problems lack tractability, it is not clear what relevance the data will have over time. It is unclear what feedback loops or how the chains of causality my flip or change at different thresholds. Is it not plausible that if technology reduces medical costs that there would be less pressure on the elderly to find retirement in lower cost geographies?

A second problem is that Cowen assumes "big data" actually has the potential to enhance the productivity of the larger system. It is shocking how little evidence there is for the effectiveness of "big data" schemes. Most of the use case testing touted by vendors, users, and academics is limited to short term and limited scope effects. Most of the examples of success such as climate models, Google, and econometrics have not proven to necessarily produce net productive outcomes for the relevant domains. Is it not plausible that Google is simply redistributing the allocation of knowledge and not really increasing it in a productive way?

A third problem is that Cowen states that using some data to improve a decision is better than none. How do you know if the data will really improve the outcome? It is all based on the underlying theory of how the data might impact the outcome.

DougT writes:

Yes, Cowen repeatedly states that "Some information is better than no information." That was not the case in the old Soviet Union, where messengers were literally shot for bringing (accurate) bad news. The result: no bad news. The problem: inaccurate data. The consequence: poor decisions.

Bad information is worse than no information. If my cousin has a recommendation of a restaurant or a lawyer, I trust her a lot more than I trust a Yelp-troll who might be paid to provide a positive review. You could say that Yelp is here, deal with it. And we all have to. But until there are reporting standards, the rule will continue to be "caveat emptor."

John Hawkins writes:

Highly disagree that we have a full grasp of all the theoretical problems in macroeconomics. In-cred-i-ble amount to be learned and incorporated from network theory and much to be learned still about the disequilibrating effects of money (we don't understand money or debt a lick)

Macro is just now taking a turn from using physics as its model to using biology as the glasses through which it sees the world (as far as being more institutional, feedback effects, interconnectivity). I am very excited about the theoretical ideas that will come from this.

Kevin Smith writes:

Wonderful discussion. I welcome our...integrated software overlords.

I can only speak about one part of this discussion with expertise and that is the medicine part. It will be awhile before software can replace a doctor, but it will happen. It will not be long at all before all doctors should be pre-screening their patients with some software.

Software demonstrations of accurate diagnosis assume that you a priori know what the patient has and test against it. These often assume classical presentations and here software would be a big help in increasing the potential differential a doctor should consider. For example I see an output like this would be a huge leap forward:

46 yo single Hispanic female with indigestion after dinner: pre-test probability of diagnosis - GERD 84%, ulcer 14%, cancer 2%. The following tests would be confirmatory: 1 month trial of PPI (cost $150), endoscopy (cost $475), endoscopy with biopsy (cost $750). The post-test probability of successful PPI course PPV: 95%; failure of PPI course NPV of 40%, etc.

The physician would then be able to discuss costs and benefits intelligently and make decisions about how to proceed. It will take longer to get good at software understanding symptoms and longer still at signs and physical exam but I expect those will all happen in my lifetime. Modern standards management is pretty easy - choices and relative trade offs is much harder and I suspect most people would like some human facilitator to help with making those choices (just because software says to take HTN medication does not mean best choice given relative other real world trade-offs - I cannot imagine a family meeting to discuss end of life care with sons, daughter, and wife crying being led by an iPad anytime soon). Unassisted robot procedures are much farther off...

It is an exciting time to live. I see self driving cars as perhaps the biggest and soonest change most people will notice. IF your car can drop the kids off at school - and they can sit and study - do you care how far you live from the charter school or the good school anymore? Does the concept of school districts matter to you anymore? If it doesn't than the current school district driving housing paradigm disappears and many interesting things will change. And if I can work while I drive do I care if my commute is an hour anymore? And think of all the benefits of a no stress commute in a big city. General well being will go through the roof! And if I can watch netflix on my way to work instead of working is the demand for digital infrastructure going to increase. Anyway - I think self-driving cars are going to change the world in ways we cannot even imagine. That will be an unbelievably big change and if that means I go from upper middle class doctor to lower middle class computer facilitator doctor it may very well be worth it living in a world with so much less stress and so many more possibilities. Life is more than income.

Henry Makansi writes:

I enjoyed the discussion. Cowen like many authors seeking to propagate a novel meme, paints a tad simplistic and extreme picture. The point that resonated with me was conscientiousness becoming an increasingly important psychological trait in the future. Being the father of two sons and someone who loves to learn, my sense this is as much a case of seeking out free resources such as econtalk, edge.org and ted as it is forgoing temptation with respect to the plethora of an instantly gratifying addictive but vacuous entertainment on offer in the form of Breaking Bad, online poker, internet porn, Angry birds and the horrors of Grand theft auto. I think where we have gotten to is that even the below median income proportion of the population have access to smart phones and round the clock entertainment. At the level of the brain it is developing the prefrontal cortex to override the instant dopamine surges of the reptilian brain.

The ability to deploy useful strategies to withstand temptation has always been important in helping one get on the right side of the bimodal distribution (just imagine if Keynes had access to an online day trading account with Schwab, or Davinci was tempted by a mini iPad) going forward I suspect it's going to increasingly be a case of shutting out tempting but mindless entertainment and picking up a book (bounded not electronic) instead.

Speed writes:

Tyler Cowen said, " Driverless cars are coming fairly soon. They basically work now. We need to set up a legal and regulatory environment; that will take a while. But the technology works."

The Wright brothers' first flight was in 1903 and an optimist might have said shortly after that "airplanes basically work now." The first aircraft autopilot was flown in 1912 and we're just now (100 years later) getting "driverless" aircraft in the form of un-manned drones or UAVs. Don't expect to soon see many travelers buying tickets on driverless flights -- even to Las Vegas.

With respect to changes in employment ...
US agriculture has gone from employing 70%-80% of US workers to 2%-3% during a time when incomes have risen. I don't see why any other equally large change in the productive side of our economy should be any more unpleasant. Tyler Cowen is completely dismissing the human ability to invent new things.

Predicting the future is hard.

David McGrogan writes:

Fascinating stuff.

Four thoughts:

1. There was a huge elephant in the room when the discussion turned to higher education, which is this: going to university is not about learning for most students. It is about growing up, moving away from home, having fun, meeting new and interesting people, and developing into an adult. Admittedly I work at a British university where going to a university far from home is the norm, so this may not be as true in other countries, but here I don't see a future in which university students mostly work online, because going to university is so culturally important.

2. I'm an academic at a law school (in the UK law is an undergraduate discipline like economics) and I think law, and certain other subjects, are resistant to change to a greater degree than, say, maths. A big element of being a law student is learning presentation skills, client liaison skills, and so forth - things that can't really be learned online. And, going back to point 1, going to law school is even more about making contacts and getting a training contract by pressing the flesh, so to speak, at events with local law firms, mooting competitions, etc.

3. Politics is real. No, I don't think there will be a revolution either, but nor do I think technological change will sweep away the old order so quickly. Public sector unions are powerful. The urge towards protectionism in the face of globalisation is powerful. Well organised, influential professions with vested interests are powerful. Lawyers are one of these; there are many others.

4. Tyler has to be applauded for being so measured, brave, and insightful. However, predicting the future is hard. The only certain thing is that he won't be correct about much.

Brit writes:

@Speed: your comparison of driverless cars to the Wright brothers is kind of silly. Driverless cars are coming - and it's a lot closer to happening than from the Wright brothers to present-day.

There are driverless cars that have accumulated over 100,000 miles on California highways. There are lots companies building them (Mercedes-Benz, General Motors, Google, Continental Automotive Systems, Autoliv Inc., Bosch, Nissan, Toyota, Audi), and three states have already passed laws allowing for autonomous cars. There's also an obvious market for them (replacing taxis, replacing truck drivers, giving drunk and older people a safe transportation system). At this point, it's about getting the cost low enough that people would favor driverless-vehicles over driver-based vehicles (which would be less expansive).

Speed writes:

Tyler Cowan said, "The artificial intelligence program Watson is now being applied to medical diagnosis. In less than 5 years and maybe less than 2, this will be online. I think rapidly the price, whatever it is, will fall. And there will be very high-quality medical diagnosis available to a lot of people, no matter where they are located. Even if they are not very well off in terms of income."

IBM says ...
"Watson then mines the patient data to find relevant facts about family history, current medications and other existing conditions. It combines this information with current findings from tests and instruments and then examines all available data sources to form hypotheses and test them. Watson can incorporate treatment guidelines, electronic medical record data, doctor's and nurse's notes, research, clinical studies, journal articles, and patient information into the data available for analysis."

"Watson will then provide a list of potential diagnoses along with a score that indicates the level of confidence for each hypothesis."
http://www-03.ibm.com/innovation/us/watson/watson_in_healthcare.shtml

Watson will help to diagnose and suggest treatments but I still see a lot of human effort getting information from the patient into Watson and then treating the patient. Like most "breakthroughs" Watson will be evolutionary, not revolutionary.

Guy Crouchback writes:

This a wonderful podcast. I thought it was going to be along familiar "rich-getting-richer-others-left-behind" lines. There is a bit of that - although Cowen's idea of rising income gaps is subtler than the 1% vs 99% - but it's about so much more than that.

Trying to predict the future is notoriously tricky. But when done smartly, as is the case here, it is compelling. Cowen reminds me of Tocqueville: he identifies a long-term trend driving change (in Cowen's case, it is a technological trend - ever-smarter machines - rather than a social one) and he looks at the implications in every conceivable field: economics, culture, politics, etc.

Unlike Tocqueville, of course, Cowen foresees rising inequality rather than equality. But the systematic approach and balanced tone are similar. And both have the same, fundamental concept of revolution as a slow, relentless process.

Cowen sees some good things and bad things about the way technology is changing our world. Like Tocqueville, he is saying: we'd better get used to it and make the most of this dubious deal. I hope Cowen does not turn out to be as good a prophet as his 19th-century predecessor, but he too has a talent for telling a fascinating story.

David McGrogan writes:

To clarify my comment above, it should have read "going to university is not *just* about learning for most students".

Speed writes:

@Brit: When the technology is mature, the cost is reasonable, the infrastructure is in place, regulations are updated and people are convinced, then we will start (start!) seeing semi-autonomous (not driverless) vehicles on the road. Compare this monumental task to the growth of cell phones from expensive, unreliable, automobile mounted talk-only devices in 1940 to what we have today (Wikipedia).

David writes:

This statement by Cowen - "This is a book about the increase in income inequality in America, and it's trying to trace through that trend and see what it looks like 10, 20 years out: what are the advantages, what are the disadvantages, and what's driving it." - reminded me of an excellent Atlantic Monthly article from about 20 years ago.

That article began by saying something like, "Whenever someone begins an analysis with the phrase, 'if present trends continue,' you can ignore the rest of the analysis. Present trends never continue."

So, listening to the rest of the podcast through this conceptual filter, I found it entertaining but insubstantial. Cowen may or may not be right. We have no way of knowing. However, if the present trend of present trends not continuing continues, we can be fairly certain he's completely overlooked the disruptive events that will render his analysis moot.

Doug Tree writes:

Like some of the previously posted comments, I too have been considering the interesting social implications of driverless cars. Consider how the disabled or elderly might benefit. In our current world, being mobile means a huge change in lifestyle. For the disabled and elderly, a loss in mobility has enormous social and economic impacts. I also think driverless cars might actually decrease car ownership. Why own a car to simply commute, when taxi services will become drastically cheaper if no humans are required to drive them? Just text my local taxi service and it shows up and takes me somewhere for the price of gas, capital depreciation and a small rate of return... With this change in price, maybe municipal busing and other local public transportation will become obsolete.

don rudolph writes:

1. If there is a revolution I think it will be instigated by intellectuals rather than the poor. We know revolutions can happen under conditions of inequality because they have in the past. What profound change has taken place in society that would prevent revolution? I felt Cowen started to drift into wishful thinking. If he stated these are the reasons that inequality might not spawn revolution, he would be on safer ground.

2. The idea of a society of poor fawning over the rich too put food on the table sounds pretty dystopian to me. I didn't hear any sense of outrage at that possible future.

Dan Hanson writes:

Three words for Tyler Cowen: Remember the Segway.

The Segway was supposed to change the very structure of our cities. It was going to be a revolution in transportation. Whole communities were going to be designed around it. And this wasn't some futuristic device - it was already in production when these predictions were made.

So what happened? To this day it's hard to pinpoint exactly how the Segway failed. It did exactly what it claimed it could do, but it just didn't find traction with the public. Too geeky, too expensive, too limited in the places it could go.

Then there were the regulatory problems. Dean Kamen thought certifying the Segway for sidewalk use was a no-brainer, since it didn't really go any faster than a fast walking speed. But there's a difference between running into a soft person at 5mph and running into a steel contraption with a metal plate at ankle-breaking height. Before long, the Segway was banned from many urban places, and that was the end of it being anything but a niche device.

But the Segway could have succeeded. If it was something the public really wanted, the regulatory path could have been cleared. But for whatever reason, the device that many technologists thought was a revolution was met with a shrug by the public. Society is complex and unpredictable.

I suspect driverless cars will go the same way. A couple of high-profile crashes and suddenly there will be bans, and if people don't trust the technology 100%, they'll be unwilling to put their life in its hands. And so driverless cars will wind up being niche vehicles used in controlled environments.

But I could be completely wrong. The history of futurism is a history of serious predictions that were so wrong as to make for hilarious reading. Society and the economy are complex adaptive systems, and no one can predict them.

Even the best futurists in the past were wrong far more often than they were right, and the things they were right about were often obvious or trivial, or just random chance: make enough predictions and some are bound to be right. Then just advertise the heck out of the correct ones, and you too can be a futurist. The same trick works with fortune telling and stock advice.

Another thing I take issue with is the notion that computers are on the verge of replacing most jobs. Two problems with this: One is that jobs today are not the jobs of tomorrow. The other is that Tyler grossly over-estimates the ability of computers to replace generalized knowledge and creative thinking.

Computers excel at rote calculation, and they can do it so fast it can appear like they are 'thinking'. But they're not. Artificial intelligence has come a long way, but is still a very, very long way from being able to do anything close to a majority of tasks the human workforce engages in daily. And by "a very long way", I mean decades, or maybe never. Software does not improve at the rate of silicon. The AI we have today is made possible by using fairly simple algorithms and applying them to a lot of data. That's a far cry from having real judgment.

But I could be wrong. The future is unpredictable. That should be the standard disclaimer of every 'futurist'. It's a fun brain exercise to speculate on what might happen in the future, and it was an enjoyable podcast, but it's probably not much more accurate than throwing darts at a dartboard.

SaveyourSelf writes:

Cowen said: “If you ask in what sphere of modern life has education really succeeded, I think it's gaming.”

He's right! This is a REALLY big deal! But he is way ahead of the curve. These learning games are in their infancy.
For any non-gamers out there, I would like to introduce you to “Everquest.”
http://en.wikipedia.org/wiki/EverQuest

Everquest was the first 3-D massively multiplayer online role playing game. It was a virtual-reality simulation where thousands of people lived together in an enormous virtual world. It was horribly addictive.

Exploring the world was a big part of the game. The environment was big and varied and beautiful. Importantly, players were not allowed to explore new areas in the world until they had achieved a requisite level. “Experience-points” were used to advance levels. Experience-points were accumulated by completing tasks. The tasks varied in difficulty. The rewards for completing hard tasks were greater than the rewards for completing easy tasks.

A second important aspect of Everquest was social. Players could interact and compete with each other. They would go to incredible lengths to express their individuality by acquiring unique abilities, unique cloths, unique titles, unique names, etc. And, importantly, players could help each other.

A good education game functions exactly the same as the old Everquest. There is a competitive, virtual world where students can interact with each other and with the world. Students are not allowed to move to new areas in the world until they have mastered certain skills. Skills--be they math solutions or memorized facts--are LITERALLY the keys that unlock the virtual doors and allow students to move forward. Rewards for accomplishing educational objectives are weighted so that difficult tasks give more rewards than simple tasks. As students gain mastery of difficult tasks, those tasks become less rewarding. Eventually the tasks become trivial and provide no reward at all. In this way, students are encouraged ever forward through the subject matter and the world. Along the way students find ways to express their individuality, interact with others, and help each other.

There are many benefits to this artificial, conceptual, virtual learning environment. First, the difficulty is tailored to each individual student. This allows students to progress at their own pace. Second, the information provided in the game is uniform. Every student must master the foundation-material before progressing. No one can slip through the cracks. Fourth—and most importantly—is its speed. A computer simulation can give IMMIDIATE feedback (both rewards and punishments) in response to a student’s efforts.

The importance of the speed of feedback cannot be understated. "Behavior is most sensitive to stimulus changes that occur immediately after, or within a few seconds of, responses...'Events that are delayed more than a few seconds after the response do NOT directly increase its future frequency'" (Applied Behavior Analysis: Second edition, p34). Human teachers cannot provide feedback at anything close to the speed and consistency of computers. With learning, Milliseconds matter.

On the other hand, one of the great weaknesses of computers is their inability to innovate. They can only teach what is already known. And they have no capacity for understanding or rewarding creativity or experimentation.

Extrapolating on these different competitive-advantages, I postulate that in the future, teachers will design and program learning-games. Aids—not teachers—will manage the day to day interface issues between students and the software. The aids will also act as the real world interface between the outcomes in the game and outcomes in the real world [something akin to spending carnival-game-tickets to purchase real world prizes].

Advancing the boundaries of a subject (and therefore the game) will probably remain the slow, tedious, expensive, risky, trial and error purview of entrepreneurs and scientists.

Ken P writes:

I enjoyed this podcast. Good to see that Cowen is moving beyond stagnationism. I bought the book this weekend.

ziad abdelnour writes:

[Comment removed pending confirmation of email address and for irrelevance. Email the webmaster@econlib.org to request restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Sheldon Z. Goldberg, MS, MD writes:

I am a medical oncologist. I have integrated the internet in my practice since I started in 1988. I update my information for every patient using online textbooks (e.g. UpToDate), guidelines (e.g. NCCN), and databases ( National Library of Medicine, Web of Knowledge). I also try to investigate the sources of opinions and try to assemble several sources to support important decisions. I often do this in real time , while with the patient, while I record the history (vetted by the patient) and write a note (including a document for the patient) using (Dragon) voice recognition software.

I am trying to use the free style chess methodology in medicine, as you suggest in the podcast.

The elements of computer assisted patient care that are available reveal the correctable flaws.
The electronic medical record systems( EPIC, etc.) do not support the analysis of patient data over time. These systems were designed to provide Medicare and other payers with long documents, to increase billing. They do a poor job of communicating the patient's situation and do not even attempt to assists in diagnostic and therapeutic decisions.


The data upon which medical decisions are based are studies that are presented as aggregated, median data. The details are not accessible (to anyone but the investigators). Often these studies are done on groups that are heterogeneous Often, the heterogeneity is discovered years after the study, Meaningful information is buried, and, perhaps, some patients are buried because of the inaccessibility of the raw data. The idea of reanalyzing such data, in the light of the heterogeneity is considered "data dredging" : a (frequentist) statistical crime.

I imagine having access to the details and outcomes of thousands of cases that can be sorted by parameters to match the patient in front of me, with an electronic medical record that aids in the analysis and decisions. What an improvement that could be!


I would love to advance this cause.

I hope that you can help mel

Dan Hanson writes:

"I imagine having access to the details and outcomes of thousands of cases that can be sorted by parameters to match the patient in front of me, with an electronic medical record that aids in the analysis and decisions. What an improvement that could be!


I would love to advance this cause.

I hope that you can help me!"

This may not be as good an idea as you'd think. I work in human factors and machine interaction, and there was a paper written a few years ago about the impact of digitization on the medical field, and one of the problems uncovered was a tendency of doctors to keyword search their patient database and then cut-and-paste diagnoses from one patient to another when given an easy method for doing so.

The claim was that this was causing them to miss small differences between patients or to fall back on a simpler, pattern-matching mode of diagnosis rather than using harder, more time-consuming but ultimately more effective inductive reasoning and judgment.

Todd Kreider writes:

I was disappointed that Russ Roberts asked about Ray Kurzweil but Tyler Cowen doesn't seem to have read or heard much about what he has said. Instead he dismisses "brain uploading" in 50 years and blows it off as a religion.

Kurzweil thinks a)human intelligence will start to be enhanced by 2030 and b)everyone will have creative jobs where everyone will have to retrain often. I wish the second part was discussed instead of Cowen knocking down a straw man.

Cowen clearly does not understand that at least for now we are on an exponential computer power curve in this talk or any other I have heard. He is correct about the beginning of the Industrial Revolution and the time it took for wages to rise, but the GDP/capita was growing at about 0.5% in 1780 and 1% in the early 1800s. It has more than doubled at the end of the 20th century, so it isn't reasonable to assume one can just shift the same 50 to 70 year period in the 21st century. (And I guarantee that neither Cowen nor I have a clue if life will be more utopian in the year 2080.)

With respect to free style chess, it tells you something if the better humans lose. Free style chess will soon be over when computers smash humans one on one.

The same holds with medicine as someone above said. Software/robotics won't augment doctors/surgeons but eventually replace them. Personally, I think much sooner than many assume.
Sure, there will be humans in the link for a while but the salaries will likely plummet as the job dramatically changes. At some level of quality, the AMA can hold this back for a nanosecond... Cowen also assumes coming pills couldn't prevent most diseases in the first place.

Ken P writes:

Todd, I agree with your Kurzweil point. Cowen dismisses Kurzweil primarily due to the "mind upload" idea. Kurzweil has more to offer than that.

I read Cowen's book and it's very good and insightfult, but much of it does rest on the premise that human+machine > machine for the forseeable future. You don't have to be a "foom singularity" believer to give plausability to a machine > human+machine scenario occurring in the near future.

In the book, Cowen brings up the idea that machines will move beyond our understanding of science. I found that concept very interesting. I expect it would be like machines can approximate science in a way that is way more accurate than we can but is totally unintelligible to humans.

isomorphismes writes:

Russ. Nice point about new authors dragging down the apparent average and perhaps disguising the trend with the people who would have made the cut before. With professional economists is there something similar? More econ PhD's than ever before? (Maybe we can blame you! ;) ) If you come up with data on either of these points it would be great to read a blog post on it because I think in theory yours could be a crucial contention.


Driverless cars. Prediction: driverless automotive will be legalised first in Western U.S. states like AZ, SD, where inebriated drivers frequently kill themselves by veering off the road in zero traffic. Then American trucking companies, which: [a] spend a lot on drivers, [b] have the cash to invest in driver less technology, [c] carry goods across Western US States, will be the main early driver of demand for computer-driving technology. When accidents occur, courts will have to decide whose fault is what—and that will be an interesting and influential bit of law as the brake company countersues the driverless-software companies countersues the county that let the lighting go out etc.

Grading essays. This seems in-credible. Who would ever accept that the computer's low grade of their child's essay was correct?


Yoga instructors, whole foods, truck drivers, computer programmers. Guys, take a look at the dictionary of occupational titles. These sound like stimuli you two would be bombarded with in N Virginia or California but they are not representative stories.


Lag from industrial revolution to wage rises. Great point Tyler.


Gamification. All of this depends upon the crucial assumption that economics, history, and maths classes improve labour productivity—not by giving the winners passkey to the jobs in which they're most productive, but by intrinsically making them do some tradeable, productive task better.


No new economic theories. Just you wait...


"We will be viewed as a lucky transitional people" I'm struggling to think of an historical example of a "lucky people who spanned two worlds". Pre/post WW1? Pre/post French Revolution? I think most people are much too ignorant of the past to venerate or judge the people of previous times except with uninformed, listless, presentist stereotypes. Why should we think our descendants will view us any differently than most of us view our forebears? (which is mainly to forget about them)


Texas, Alabama, Oklahoma. Good reasoning Tyler. Also good points about the cold Russ. Possibly the best part of the interview.


Coalition of millionaires, old. Another great idea Tyler.

Ari Tai writes:

Seems similar to the migration of farm labor from the country to the cities in the early 1900s. A wrenching disruption but one arguably enabled everything good that happened after (in terms of an additional 90% of the country's intellects being freed for more productive use, of which 5-10% are the innovative and entrepreneurial).

Be interesting to see the correlation between giant agribusiness and not only hunger but quality of life. (S) Korea has almost finishing moving the last third of its rural population to its cities (as another major national rebuild of infrastructure nearing completion). China is doing it by force. India? (Japan?) (France?). Arguably the U.S. is doing it for the poorest of the Mexican farmers (why Mexico isn't Venezuela.. yet).

Sounds like we're in for another generational shift - where those that adapt (perhaps already investing one night a week in their own education) will get the good jobs and those that don't will get warehoused.

Comments for this podcast episode have been closed
Return to top