1 00:00:00,360 --> 00:00:03,280 Speaker 1: Bloomberg's David Weston spoke with Larry Summers, the former US 2 00:00:03,360 --> 00:00:06,920 Speaker 1: Treasury Secretary, and Steve Ratner, the CEO of Willet Advisors, 3 00:00:07,280 --> 00:00:09,720 Speaker 1: to discuss the changes that will all experience with AI, 4 00:00:09,960 --> 00:00:11,799 Speaker 1: as well as some of the consequences too. 5 00:00:13,440 --> 00:00:15,640 Speaker 2: I think there are two things that have to be 6 00:00:15,800 --> 00:00:22,240 Speaker 2: added to the AI conversation. The first is that technological 7 00:00:22,400 --> 00:00:29,920 Speaker 2: change really does have some major distributional and opportunity consequences. 8 00:00:30,440 --> 00:00:34,920 Speaker 2: When I first heard, was first exposed to these arguments 9 00:00:35,320 --> 00:00:40,600 Speaker 2: as an undergraduate student at MIT in the early nineteen seventies, 10 00:00:41,159 --> 00:00:45,880 Speaker 2: there were people explaining about how terrible automation was going 11 00:00:45,960 --> 00:00:49,360 Speaker 2: to be, and then there were the smart people for 12 00:00:49,479 --> 00:00:55,080 Speaker 2: me at that time, personified by Nobel Prize winner Robert Solo, 13 00:00:55,120 --> 00:00:59,920 Speaker 2: explaining that ultimately things evolved and new jobs were created 14 00:01:00,560 --> 00:01:05,360 Speaker 2: and this was progress and it's all right. Just one thing. 15 00:01:06,240 --> 00:01:11,880 Speaker 2: In the nineteen sixties, ninety six percent of American American 16 00:01:12,040 --> 00:01:16,760 Speaker 2: men twenty five to fifty four were working and only 17 00:01:16,800 --> 00:01:22,440 Speaker 2: four percent were not. Today it's more like fourteen percent 18 00:01:23,040 --> 00:01:25,000 Speaker 2: who are not working. 19 00:01:25,160 --> 00:01:26,560 Speaker 3: Are there things we should be doing right now that 20 00:01:26,600 --> 00:01:29,200 Speaker 3: we fail to do with automation and with globalization? To 21 00:01:29,360 --> 00:01:33,920 Speaker 3: think about those distributional effects potentially AI to make sure 22 00:01:34,080 --> 00:01:36,759 Speaker 3: that we bring more people along with the progress. 23 00:01:36,959 --> 00:01:39,960 Speaker 4: Absolutely. Look, I don't think actually Larry and I really disagree. 24 00:01:40,040 --> 00:01:43,160 Speaker 4: I understand the problem he's talking about. It actually relates 25 00:01:43,200 --> 00:01:47,319 Speaker 4: to automation a lot, and also to trade. And where 26 00:01:47,360 --> 00:01:49,600 Speaker 4: I think, frankly, the economists and Larry may jump down 27 00:01:49,680 --> 00:01:51,640 Speaker 4: my throat for this, but I think the economists got 28 00:01:51,640 --> 00:01:53,920 Speaker 4: it wrong on trade, which is similar in a lot 29 00:01:53,920 --> 00:01:57,120 Speaker 4: of respects to automation or other technological improvements in terms 30 00:01:57,120 --> 00:02:00,480 Speaker 4: of its impact, is that trade had huge macroeconomic benefits 31 00:02:00,480 --> 00:02:04,160 Speaker 4: for the country. We missed the marcroeconomic impacts those workers 32 00:02:04,440 --> 00:02:06,960 Speaker 4: in Flint or Detroit or in Ohio. Some of their 33 00:02:07,040 --> 00:02:09,320 Speaker 4: jobs were and I actually just read something the other day, 34 00:02:09,360 --> 00:02:12,280 Speaker 4: you know, rough justice. Maybe half their jobs were lost 35 00:02:12,280 --> 00:02:14,919 Speaker 4: to automation, the other half were lost to trade. And 36 00:02:15,200 --> 00:02:17,839 Speaker 4: didn't we had this little trade adjustment assistance program which 37 00:02:17,880 --> 00:02:20,480 Speaker 4: basically did nothing. And we have not really done a 38 00:02:20,520 --> 00:02:23,399 Speaker 4: great job as a society both in getting the benefits 39 00:02:23,480 --> 00:02:26,480 Speaker 4: of technology into the hands of everybody. There's been this 40 00:02:27,200 --> 00:02:29,920 Speaker 4: lack of wage growth commensurate with the productivity growth over 41 00:02:29,960 --> 00:02:32,679 Speaker 4: a fairly long period of time. Now as well as 42 00:02:32,840 --> 00:02:35,080 Speaker 4: individuals and finding them things to do where they can 43 00:02:35,120 --> 00:02:36,480 Speaker 4: be more productive and happier. 44 00:02:37,040 --> 00:02:39,760 Speaker 3: Lurie, what about it? Did the economists, and yes, the 45 00:02:39,760 --> 00:02:42,280 Speaker 3: policymaker in Washington, it sort of let us all down 46 00:02:42,520 --> 00:02:44,440 Speaker 3: with respective both automation and trade. 47 00:02:44,560 --> 00:02:49,440 Speaker 2: We should have done more to cushion the various changes 48 00:02:49,520 --> 00:02:53,079 Speaker 2: associated with trade. I agree with that, I'm not sure 49 00:02:53,080 --> 00:02:57,600 Speaker 2: I agree with Steve's quantification, and I think that a 50 00:02:57,720 --> 00:03:02,040 Speaker 2: full calculus on trade to recognize a large number of 51 00:03:02,080 --> 00:03:06,280 Speaker 2: benefits in terms of jobs created and in terms of 52 00:03:06,320 --> 00:03:09,720 Speaker 2: real wages enhanced. But that brings me to the other 53 00:03:09,760 --> 00:03:14,720 Speaker 2: point I wanted to make about AI, and I don't 54 00:03:14,960 --> 00:03:19,120 Speaker 2: know for sure about this, but if my suspicion is right, 55 00:03:19,720 --> 00:03:24,440 Speaker 2: it's very big. Most of the technological changes we've had 56 00:03:24,520 --> 00:03:32,639 Speaker 2: before came for working people doing relatively routine things. They 57 00:03:32,680 --> 00:03:38,920 Speaker 2: were automatic ways of picking cotton that came from agricultural workers. 58 00:03:39,440 --> 00:03:46,320 Speaker 2: They were things that replaced typists or telephone operators. As 59 00:03:46,360 --> 00:03:51,520 Speaker 2: Steve mentioned, I have a suspicion that AI is coming 60 00:03:51,600 --> 00:03:55,160 Speaker 2: for the cognitive class. And part of the reason you're 61 00:03:55,240 --> 00:04:01,520 Speaker 2: seeing such hysteria now is that it's the people who 62 00:04:01,560 --> 00:04:07,480 Speaker 2: write articles and their friends, the people like the three 63 00:04:07,560 --> 00:04:13,320 Speaker 2: of us, who are more at risk from AI competition 64 00:04:14,040 --> 00:04:17,440 Speaker 2: than has been the case with most of the technological 65 00:04:17,480 --> 00:04:23,760 Speaker 2: innovations in the past. I would say that there's a 66 00:04:23,839 --> 00:04:28,560 Speaker 2: substantial chance the AI is going to be much more 67 00:04:28,560 --> 00:04:32,239 Speaker 2: of a threat to IQ than it is to EQ. 68 00:04:32,760 --> 00:04:38,279 Speaker 2: It will be a very long time before AI will 69 00:04:38,400 --> 00:04:43,120 Speaker 2: replace many of the kinds of direct physical work. Think 70 00:04:43,160 --> 00:04:48,800 Speaker 2: of working in a garden, for example. So I have 71 00:04:48,880 --> 00:04:55,040 Speaker 2: a suspicion that the distributional consequences of AI for the 72 00:04:55,080 --> 00:04:59,640 Speaker 2: bosses versus the boss may be very different than the 73 00:04:59,800 --> 00:05:05,080 Speaker 2: just retributional consequences of many of the other technological revolutions, 74 00:05:05,120 --> 00:05:08,920 Speaker 2: and that affects how bosses are going to think about 75 00:05:08,960 --> 00:05:13,000 Speaker 2: it in profound ways. They're going to be much more scared, 76 00:05:13,560 --> 00:05:17,240 Speaker 2: and on the other side, may be more benign from 77 00:05:17,960 --> 00:05:20,279 Speaker 2: the point of view of some of those who've been 78 00:05:20,320 --> 00:05:21,760 Speaker 2: traditionally left behind. 79 00:05:22,720 --> 00:05:24,680 Speaker 3: So Larry, I agree. I agree with that. 80 00:05:25,200 --> 00:05:27,840 Speaker 4: I think it is the cognitive classes as you call them, 81 00:05:27,839 --> 00:05:29,839 Speaker 4: who are most at risk. I might make a discis 82 00:05:29,839 --> 00:05:31,520 Speaker 4: I'm not sure. I would think about it as bosses 83 00:05:31,560 --> 00:05:33,760 Speaker 4: and boss and I'll use it, But I will use 84 00:05:33,760 --> 00:05:36,360 Speaker 4: this historical analogy to give us a little bit of hope. 85 00:05:36,480 --> 00:05:39,760 Speaker 4: When I started on Wall Street as a young investment banker, 86 00:05:39,880 --> 00:05:42,080 Speaker 4: I had nothing. I had an early HB twelve C 87 00:05:42,279 --> 00:05:46,080 Speaker 4: calculator in my hand. We had no we had no Excel, 88 00:05:46,200 --> 00:05:48,400 Speaker 4: we had no computers to speak of, we had know nothing. 89 00:05:48,480 --> 00:05:50,960 Speaker 4: All of our spreadsheets were done by hand. They took 90 00:05:50,960 --> 00:05:53,320 Speaker 4: a really long time. They had to then be typed up. 91 00:05:53,320 --> 00:05:55,760 Speaker 4: We'll put the typists aside, and then if I wanted 92 00:05:55,800 --> 00:05:57,640 Speaker 4: to make a change, I had to start all over again. 93 00:05:57,920 --> 00:05:59,520 Speaker 4: And now that can all be done with the click 94 00:05:59,560 --> 00:06:02,400 Speaker 4: of a mouse with an Excel program by anybody with 95 00:06:02,440 --> 00:06:05,880 Speaker 4: a small personal computer. And yet the number of people 96 00:06:05,920 --> 00:06:08,159 Speaker 4: doing what I did forty years ago when I started 97 00:06:08,200 --> 00:06:12,000 Speaker 4: on Wall Street has multiplied since then, and so it 98 00:06:12,120 --> 00:06:16,720 Speaker 4: became a productivity enhancing tool, not a jobucting destructive tool. 99 00:06:17,040 --> 00:06:19,760 Speaker 4: I'm perfectly prepared to believe that this may come out 100 00:06:19,839 --> 00:06:21,920 Speaker 4: a different way. All I'm saying is I don't think 101 00:06:21,960 --> 00:06:25,040 Speaker 4: we know yet, and I think history is probably still 102 00:06:25,080 --> 00:06:26,880 Speaker 4: on the side that we will find our way through 103 00:06:26,920 --> 00:06:28,080 Speaker 4: this in a positive way.