WEBVTT 00:00:06.950 --> 00:00:09.910 We're speaking today with Eric Brynjolfsson, a professor here 00:00:09.910 --> 00:00:11.950 at MIT Sloan School of Management, 00:00:11.950 --> 00:00:15.520 the director of the MIT Initiative on the Digital 00:00:15.520 --> 00:00:18.520 Economy, and co-author of a bestselling book, 00:00:18.520 --> 00:00:21.510 The Second Machine Age, a book that really 00:00:21.510 --> 00:00:23.270 has gotten the conversation going 00:00:23.270 --> 00:00:27.290 about how important today's digital economy is, 00:00:27.290 --> 00:00:29.480 how the technologies that are coming along 00:00:29.480 --> 00:00:32.350 are going to have an enormous effect on the future of work. 00:00:32.350 --> 00:00:34.260 So Eric, thank you for joining us. 00:00:34.260 --> 00:00:35.260 It's a real pleasure. 00:00:35.260 --> 00:00:35.770 Good. 00:00:35.770 --> 00:00:36.810 Let's get started. 00:00:36.810 --> 00:00:39.880 Why don't you tell us why you think 00:00:39.880 --> 00:00:43.780 the current technological wave of innovation 00:00:43.780 --> 00:00:47.530 is as important as the first Industrial Revolution, 00:00:47.530 --> 00:00:50.035 with the steam engine and all the things that went with it. 00:00:50.035 --> 00:00:51.410 Well, you know, technologies make 00:00:51.410 --> 00:00:53.576 a huge difference in the living standards of people. 00:00:53.576 --> 00:00:56.010 For centuries, living standards were essentially stagnant 00:00:56.010 --> 00:00:58.870 until James Watt and others developed a better steam 00:00:58.870 --> 00:01:01.340 engine, and that ignited the Industrial Revolution. 00:01:01.340 --> 00:01:03.330 And ever since then, living standards 00:01:03.330 --> 00:01:07.710 have been growing about 2% per year, which adds up to a lot. 00:01:07.710 --> 00:01:09.810 What those technologies really did 00:01:09.810 --> 00:01:12.650 was they augmented and automated a lot 00:01:12.650 --> 00:01:17.280 of muscle power from humans and animals to machines. 00:01:17.280 --> 00:01:19.320 What we're seeing now is that machines 00:01:19.320 --> 00:01:21.400 are beginning to be able to do the same thing 00:01:21.400 --> 00:01:23.370 for our brains and our minds. 00:01:23.370 --> 00:01:26.240 We have computers and software and big data 00:01:26.240 --> 00:01:29.300 that are learning how to help us make decisions, 00:01:29.300 --> 00:01:31.490 how to extend our mental capacity. 00:01:31.490 --> 00:01:36.260 And that's at least as profound a difference as what the steam 00:01:36.260 --> 00:01:38.230 engine and, later, the internal combustion 00:01:38.230 --> 00:01:40.650 engine and other technologies did for our muscles. 00:01:40.650 --> 00:01:46.040 Well, it's very clear that technological changes like this 00:01:46.040 --> 00:01:49.650 are very good for the economy, but they also affect work. 00:01:49.650 --> 00:01:53.150 Tell us a little bit about how technologies affect 00:01:53.150 --> 00:01:55.970 work in what ways-- in positive ways, in negative ways-- 00:01:55.970 --> 00:01:57.521 and then we can go on from there. 00:01:57.521 --> 00:01:58.020 Sure. 00:01:58.020 --> 00:01:59.510 Well, the big headline is the first one 00:01:59.510 --> 00:02:01.500 you mentioned, that the pie is getting bigger. 00:02:01.500 --> 00:02:03.830 And so we have to keep sight of that, 00:02:03.830 --> 00:02:06.040 that basically these are technologies that create 00:02:06.040 --> 00:02:08.090 more abundance, more wealth. 00:02:08.090 --> 00:02:10.990 And there's more wealth now than there ever has been in history, 00:02:10.990 --> 00:02:12.690 and global poverty is going down. 00:02:12.690 --> 00:02:13.960 So that's the good news. 00:02:13.960 --> 00:02:16.400 But the reality is that there's no economic law that 00:02:16.400 --> 00:02:19.030 says that everyone's going to benefit evenly from this. 00:02:19.030 --> 00:02:22.210 It's possible for some people to be made worse off. 00:02:22.210 --> 00:02:24.219 In fact, there's no economic reason 00:02:24.219 --> 00:02:26.010 that you couldn't have a majority of people 00:02:26.010 --> 00:02:29.654 be made worse off, even as other people get much, much better. 00:02:29.654 --> 00:02:31.070 Through most of history, there was 00:02:31.070 --> 00:02:34.380 a rising tide that lifted most boats, almost all boats. 00:02:34.380 --> 00:02:37.240 But in the past 20 years or so, there's 00:02:37.240 --> 00:02:40.310 been what we call a decoupling, where productivity 00:02:40.310 --> 00:02:43.700 has continued to grow, GDP per capita has continued to grow, 00:02:43.700 --> 00:02:47.160 but the median income, the 50th percentile and those below it, 00:02:47.160 --> 00:02:50.320 have seen their incomes stagnate or even fall. 00:02:50.320 --> 00:02:53.060 And that reflects the fact that these technologies 00:02:53.060 --> 00:02:55.770 don't affect everybody evenly. 00:02:55.770 --> 00:02:57.300 So let's play this out. 00:02:57.300 --> 00:03:00.560 How do you see these technologies affecting work 00:03:00.560 --> 00:03:02.100 in the future? 00:03:02.100 --> 00:03:04.834 Well, the technologies continue to advance, 00:03:04.834 --> 00:03:07.000 and they continue to affect different kinds of jobs. 00:03:07.000 --> 00:03:09.680 But what really matters is how we go about using them. 00:03:09.680 --> 00:03:12.580 What we've seen so far is that many of the technologies 00:03:12.580 --> 00:03:14.910 have been used to automate a lot of routine 00:03:14.910 --> 00:03:19.410 information-processing work, things like bank tellers, 00:03:19.410 --> 00:03:22.160 bookkeepers, a lot of middle managers. 00:03:22.160 --> 00:03:25.070 In fact, about 60% of Americans primarily 00:03:25.070 --> 00:03:27.850 do information-processing work, so that's a big category. 00:03:27.850 --> 00:03:29.992 And machines are very good at automating 00:03:29.992 --> 00:03:31.450 a lot of these, especially the more 00:03:31.450 --> 00:03:34.090 repetitive, routine information-processing tasks. 00:03:34.090 --> 00:03:37.632 It turns out that the people who do those jobs are typically 00:03:37.632 --> 00:03:39.340 in the middle of the income distribution, 00:03:39.340 --> 00:03:41.600 and what they've seen is that as machines get better 00:03:41.600 --> 00:03:44.010 at those tasks, there's less demand for humans 00:03:44.010 --> 00:03:45.290 to do the identical tasks. 00:03:45.290 --> 00:03:47.579 Or if they do still have jobs, they 00:03:47.579 --> 00:03:49.620 have to compete with machines that are doing them 00:03:49.620 --> 00:03:50.840 more and more cheaply. 00:03:50.840 --> 00:03:53.160 And that his pushed down wages. 00:03:53.160 --> 00:03:54.980 It's led to fewer job opportunities 00:03:54.980 --> 00:03:56.750 in those categories. 00:03:56.750 --> 00:03:59.810 At the same time, there's been growth in job opportunities 00:03:59.810 --> 00:04:02.170 at the other ends of the spectrum, what 00:04:02.170 --> 00:04:06.010 David Autor and other people call job polarization. 00:04:06.010 --> 00:04:09.210 At the high end, you get creative people 00:04:09.210 --> 00:04:11.294 or entrepreneurs are making, in some cases, 00:04:11.294 --> 00:04:12.460 more money than ever before. 00:04:12.460 --> 00:04:13.990 It's probably the best time in history. 00:04:13.990 --> 00:04:15.650 If you have some kind of special talent 00:04:15.650 --> 00:04:17.470 that could be replicated digitally, 00:04:17.470 --> 00:04:19.640 you might end up being a billionaire. 00:04:19.640 --> 00:04:22.670 On the other hand, the other end of the spectrum 00:04:22.670 --> 00:04:24.930 has also seen some growth, and those 00:04:24.930 --> 00:04:29.150 are jobs that require a lot of dexterity or physical tasks 00:04:29.150 --> 00:04:31.300 or interpersonal skills, things that machines 00:04:31.300 --> 00:04:32.320 aren't very good at. 00:04:32.320 --> 00:04:34.720 The thing is, often those jobs don't pay very well. 00:04:34.720 --> 00:04:36.770 And just to be specific, those are the kinds 00:04:36.770 --> 00:04:40.860 of jobs like a waiter or a gardener 00:04:40.860 --> 00:04:43.080 or a janitor, those kinds of tasks. 00:04:43.080 --> 00:04:46.140 So what is your advice, then, to workers 00:04:46.140 --> 00:04:48.940 facing this world of changing technology? 00:04:48.940 --> 00:04:51.150 What should they do in order to be 00:04:51.150 --> 00:04:53.530 able to prosper in this world? 00:04:53.530 --> 00:04:55.750 Well, what you'd like to try to do 00:04:55.750 --> 00:04:59.180 is strengthen those areas where humans have 00:04:59.180 --> 00:05:00.306 an advantage over machines. 00:05:00.306 --> 00:05:02.596 You don't really want to be competing against machines. 00:05:02.596 --> 00:05:04.580 Ideally, you'd like to do things were machines 00:05:04.580 --> 00:05:06.120 can leverage your talents. 00:05:06.120 --> 00:05:07.944 Data scientists are more in demand 00:05:07.944 --> 00:05:09.360 than ever before, because machines 00:05:09.360 --> 00:05:11.250 have created an abundance of data, 00:05:11.250 --> 00:05:12.580 but they can't analyze it. 00:05:12.580 --> 00:05:15.380 They can't even ask the right questions themselves. 00:05:15.380 --> 00:05:16.870 You need humans to do that. 00:05:16.870 --> 00:05:18.990 And so if you have those kinds of skills, 00:05:18.990 --> 00:05:20.390 you're more valuable than ever. 00:05:20.390 --> 00:05:22.650 And Silicon Valley, and now companies 00:05:22.650 --> 00:05:24.550 throughout the country and the world, 00:05:24.550 --> 00:05:27.830 are greatly demanding those kinds of workers. 00:05:27.830 --> 00:05:29.960 And so there's a set of categories around that. 00:05:29.960 --> 00:05:32.501 I mentioned earlier, if you have some kind of special talents 00:05:32.501 --> 00:05:35.180 or creativity, machines aren't very creative. 00:05:35.180 --> 00:05:39.790 People who can sing well or who can invent new businesses, 00:05:39.790 --> 00:05:44.470 as entrepreneurs, many kinds of scientists, artists, 00:05:44.470 --> 00:05:46.940 if you're a great writer or write books, 00:05:46.940 --> 00:05:48.990 those are things that are probably 00:05:48.990 --> 00:05:50.870 more valuable now than before. 00:05:50.870 --> 00:05:53.670 Another big category that's gotten to be more valuable is-- 00:05:53.670 --> 00:05:55.480 I touched on it earlier-- is some of these interpersonal 00:05:55.480 --> 00:05:56.120 skills-- 00:05:56.120 --> 00:05:58.310 nurturing and caring. 00:05:58.310 --> 00:06:01.750 I don't think it would have been very motivating 00:06:01.750 --> 00:06:04.930 if, at the half time of the football games last weekend, 00:06:04.930 --> 00:06:06.760 the coach had been a robot coach. 00:06:06.760 --> 00:06:08.830 You need a human to do that. 00:06:08.830 --> 00:06:11.770 And that means that leadership, team work, 00:06:11.770 --> 00:06:12.860 those kinds of things. 00:06:12.860 --> 00:06:15.500 We still connect with other people a lot better, 00:06:15.500 --> 00:06:17.030 and we will for some time. 00:06:17.030 --> 00:06:19.780 And I think all those skills, from the creative ones 00:06:19.780 --> 00:06:22.650 to the interpersonal, the teamwork, the nurturing skills, 00:06:22.650 --> 00:06:23.890 they can all be developed. 00:06:23.890 --> 00:06:26.090 And I would encourage people to work harder 00:06:26.090 --> 00:06:27.990 at developing those kinds of skills. 00:06:27.990 --> 00:06:31.110 Become a better salesperson, negotiator, nurturing, 00:06:31.110 --> 00:06:34.180 caring for people, nursing. 00:06:34.180 --> 00:06:36.310 Those are all things that I expect to grow, 00:06:36.310 --> 00:06:37.940 even as machines take care of more 00:06:37.940 --> 00:06:41.210 and more of the routine, repetitive kinds of tasks. 00:06:41.210 --> 00:06:45.030 So that's great advice to the workforce. 00:06:45.030 --> 00:06:46.910 Let's talk a bit about the people who are 00:06:46.910 --> 00:06:48.820 designing these technologies. 00:06:48.820 --> 00:06:54.420 How can we encourage more technological design 00:06:54.420 --> 00:06:57.940 that complements work, that utilizes these skills, that 00:06:57.940 --> 00:07:00.220 thinks about ways to enhance them 00:07:00.220 --> 00:07:04.910 or to work with them to drive productivity 00:07:04.910 --> 00:07:06.246 and economic performance? 00:07:06.246 --> 00:07:07.870 Well, I'm glad you asked that question, 00:07:07.870 --> 00:07:09.286 because most people don't even get 00:07:09.286 --> 00:07:10.706 as far as asking that question. 00:07:10.706 --> 00:07:12.330 I think there's a widespread assumption 00:07:12.330 --> 00:07:14.150 out there that technology just happens, 00:07:14.150 --> 00:07:16.600 and there's nothing you can do to shape the path of it. 00:07:16.600 --> 00:07:19.300 The reality is that you can encourage people 00:07:19.300 --> 00:07:23.580 to have technologies that create more inclusive innovation, that 00:07:23.580 --> 00:07:25.800 helps people more broadly, or that 00:07:25.800 --> 00:07:28.940 mostly focuses on substituting for people, for that matter. 00:07:28.940 --> 00:07:31.670 And we, as a society and as individuals, 00:07:31.670 --> 00:07:33.490 can shape the direction of technology 00:07:33.490 --> 00:07:36.160 to quite a significant extent. 00:07:36.160 --> 00:07:39.810 One of the things we can do is change around our tax policy 00:07:39.810 --> 00:07:42.130 and our government policies. 00:07:42.130 --> 00:07:44.250 One of the oldest rules of economics 00:07:44.250 --> 00:07:47.350 is if you want less of something, you tax it. 00:07:47.350 --> 00:07:49.350 If you want more of something, you subsidize it. 00:07:49.350 --> 00:07:52.650 Right now, ironically, we are taxing work. 00:07:52.650 --> 00:07:57.400 If two entrepreneurs come up with a billion-dollar idea, 00:07:57.400 --> 00:07:59.510 and one of them involves employing lots of people 00:07:59.510 --> 00:08:01.630 and one of them doesn't, our current tax system 00:08:01.630 --> 00:08:04.620 will put more taxes on the one that employs more people, 00:08:04.620 --> 00:08:06.510 and that's probably not we want to do. 00:08:06.510 --> 00:08:09.919 Conversely, we don't tax pollution, carbon, congestion, 00:08:09.919 --> 00:08:11.460 things that we'd like to see less of. 00:08:11.460 --> 00:08:12.918 So we could rejigger the tax system 00:08:12.918 --> 00:08:15.130 to tax the things that we don't want more, 00:08:15.130 --> 00:08:22.460 and then we'll have less of them, and lower the taxes, 00:08:22.460 --> 00:08:24.410 or even subsidize, work. 00:08:24.410 --> 00:08:26.510 And I think that would guide entrepreneurs 00:08:26.510 --> 00:08:28.120 to be more creative about inventing 00:08:28.120 --> 00:08:31.650 things that have less pollution and more widespread labor. 00:08:31.650 --> 00:08:35.210 The Earned Income Tax Credit is a good example, but too small. 00:08:35.210 --> 00:08:36.816 Could be expanded at that. 00:08:36.816 --> 00:08:38.190 Another kind of thing we could do 00:08:38.190 --> 00:08:42.030 is just recognize reward and motivate people. 00:08:42.030 --> 00:08:45.030 One of the things that I've learned over the past few years 00:08:45.030 --> 00:08:47.330 is how much people can be motivated 00:08:47.330 --> 00:08:49.490 by having a prize and a goal in front of them. 00:08:49.490 --> 00:08:51.660 I've been watching the DARPA Robotics 00:08:51.660 --> 00:08:53.930 Challenge, the DARPA Grand Challenge that 00:08:53.930 --> 00:08:55.640 led to the driverless car. 00:08:55.640 --> 00:08:57.620 And I see my colleagues in engineering, 00:08:57.620 --> 00:09:00.412 grad students, professors working nights and weekends 00:09:00.412 --> 00:09:02.870 to solve this challenge that has been put in front of them. 00:09:02.870 --> 00:09:05.170 And many times, they do come around 00:09:05.170 --> 00:09:07.550 to making huge strides in that. 00:09:07.550 --> 00:09:09.920 Why don't we try to reward and motivate 00:09:09.920 --> 00:09:13.090 business leaders and economists to reinvent 00:09:13.090 --> 00:09:15.150 the businesses and economy the same way 00:09:15.150 --> 00:09:17.820 that technologists have been reinventing the technology? 00:09:17.820 --> 00:09:20.140 And in fact, here at MIT, we're launching something 00:09:20.140 --> 00:09:23.140 called the Inclusive Innovation Competition, which 00:09:23.140 --> 00:09:25.690 is designed specifically to recognize those business 00:09:25.690 --> 00:09:29.060 models that use technology to create broad, 00:09:29.060 --> 00:09:30.190 shared prosperity. 00:09:30.190 --> 00:09:34.370 We call it shared prosperity for the many, not just the few. 00:09:34.370 --> 00:09:36.070 It's just getting launched, but we 00:09:36.070 --> 00:09:39.720 think this is going to be a good way to recognize and highlight 00:09:39.720 --> 00:09:40.830 people who have done that. 00:09:40.830 --> 00:09:42.280 I think that's a great idea, and I 00:09:42.280 --> 00:09:45.130 hope that it creates a lot of attention, 00:09:45.130 --> 00:09:49.380 and that we get lots of people really thinking along 00:09:49.380 --> 00:09:55.630 these ways, because I think it's the genius of people-- 00:09:55.630 --> 00:09:59.560 using technologies to address important problems that 00:09:59.560 --> 00:10:01.940 will help make progress. 00:10:01.940 --> 00:10:03.550 Yeah, and we all have choices. 00:10:03.550 --> 00:10:05.560 We can shape the future in that direction. 00:10:05.560 --> 00:10:06.990 And why don't we encourage people 00:10:06.990 --> 00:10:09.370 who are doing it in a way that does create this shared 00:10:09.370 --> 00:10:10.960 prosperity? 00:10:10.960 --> 00:10:13.640 So let me ask this final question, because it's 00:10:13.640 --> 00:10:15.260 obviously on everybody's mind. 00:10:15.260 --> 00:10:18.560 And that is, from time to time in history, 00:10:18.560 --> 00:10:21.160 as you know better than I, people have worried 00:10:21.160 --> 00:10:23.490 that technology is going to replace work, 00:10:23.490 --> 00:10:24.870 it's going to be the end of work, 00:10:24.870 --> 00:10:26.500 we're not going to have enough jobs. 00:10:26.500 --> 00:10:30.010 And that's an issue on people's mind today. 00:10:30.010 --> 00:10:32.140 How do you think about that problem, given 00:10:32.140 --> 00:10:34.720 all of the innovation that's going on 00:10:34.720 --> 00:10:36.620 and that could come down the road? 00:10:36.620 --> 00:10:39.760 Well, down the road, I could certainly 00:10:39.760 --> 00:10:42.611 imagine technology creating kind of a Star Trek economy. 00:10:42.611 --> 00:10:44.610 And I don't think that's necessarily a bad thing 00:10:44.610 --> 00:10:47.440 if technology's able to take care of all of our basic needs 00:10:47.440 --> 00:10:48.000 for us. 00:10:48.000 --> 00:10:50.090 But the more immediate issue is not 00:10:50.090 --> 00:10:51.640 that technology is going to eliminate 00:10:51.640 --> 00:10:54.330 all the jobs, but the types of jobs that are affected. 00:10:54.330 --> 00:10:58.090 The reality is that technology's always been destroying jobs. 00:10:58.090 --> 00:10:59.450 It's always been creating jobs. 00:10:59.450 --> 00:11:01.770 But most importantly, it's been changing the mix. 00:11:01.770 --> 00:11:04.450 And what we're seeing recently, as I touched on earlier, 00:11:04.450 --> 00:11:07.504 is that mix is changing in a way that a lot of people 00:11:07.504 --> 00:11:09.420 are being made worse off, and other people are 00:11:09.420 --> 00:11:11.000 being made much better off. 00:11:11.000 --> 00:11:14.180 And we want to try to shape the direction of the technology. 00:11:14.180 --> 00:11:16.980 We want to also give people the skills and education 00:11:16.980 --> 00:11:18.790 so they can adapt to the new kinds of jobs 00:11:18.790 --> 00:11:19.710 that are available. 00:11:19.710 --> 00:11:21.560 And if we do that, I think we're going 00:11:21.560 --> 00:11:24.474 to be able to have that kind of world of shared prosperity. 00:11:24.474 --> 00:11:26.390 The technology will march on, and we shouldn't 00:11:26.390 --> 00:11:27.770 try to stop the technology. 00:11:27.770 --> 00:11:29.400 That's how we get that bigger pie, 00:11:29.400 --> 00:11:30.820 that growth that I mentioned. 00:11:30.820 --> 00:11:32.400 But we can direct it. 00:11:32.400 --> 00:11:34.590 In the earlier industrial revolutions, 00:11:34.590 --> 00:11:37.710 when the steam engine and other technologies came along, 00:11:37.710 --> 00:11:39.640 we adapted our institutions. 00:11:39.640 --> 00:11:41.760 We invented mass public education, 00:11:41.760 --> 00:11:43.690 first at the primary level, and then later 00:11:43.690 --> 00:11:44.940 at the secondary level. 00:11:44.940 --> 00:11:48.220 We invented social security and a whole bunch of other policies 00:11:48.220 --> 00:11:51.240 that helped smooth that transition 00:11:51.240 --> 00:11:53.200 to a new kind of work force and cushioned 00:11:53.200 --> 00:11:55.580 the people who otherwise would have been left behind. 00:11:55.580 --> 00:11:57.580 We're going to have to reinvent education again. 00:11:57.580 --> 00:11:58.996 And this time, we're going to have 00:11:58.996 --> 00:12:01.490 to do our best to shape the technology that 00:12:01.490 --> 00:12:03.590 is aligned with our values. 00:12:03.590 --> 00:12:06.120 Ultimately, it's not what the technology does to us. 00:12:06.120 --> 00:12:07.840 Technology is a tool. 00:12:07.840 --> 00:12:10.940 It's always been a tool, whether it's a hammer or an enterprise 00:12:10.940 --> 00:12:12.380 resource planning system. 00:12:12.380 --> 00:12:15.770 We have more powerful tools now than we ever had before, 00:12:15.770 --> 00:12:18.470 and that means we have more power to shape the future 00:12:18.470 --> 00:12:19.690 than we ever did before. 00:12:19.690 --> 00:12:22.300 But it starts with understanding that we have that power, 00:12:22.300 --> 00:12:25.181 and aligning our actions with our values. 00:12:25.181 --> 00:12:27.430 Well, Eric, I think that's a really important message, 00:12:27.430 --> 00:12:31.530 and one that I'm delighted that we can deliver in this class 00:12:31.530 --> 00:12:33.760 and in other settings where people can really 00:12:33.760 --> 00:12:34.510 make a difference. 00:12:34.510 --> 00:12:35.886 So thanks for joining us today. 00:12:35.886 --> 00:12:37.010 It's a tremendous pleasure. 00:12:37.010 --> 00:12:39.010 Thanks a lot, Tom.