WEBVTT - Larry Summers Talks Retail and AI

0:00:02.440 --> 0:00:08.280
<v Speaker 1>Bloomberg Audio Studios, Podcasts, radio News. We start with the

0:00:08.360 --> 0:00:10.760
<v Speaker 1>US economy and some signs that the consumer may be

0:00:10.920 --> 0:00:13.920
<v Speaker 1>pulling back. Has seen this weekend the retail sales numbers,

0:00:14.120 --> 0:00:16.400
<v Speaker 1>and we welcome down our very special contributor, Larry Summers

0:00:16.440 --> 0:00:18.919
<v Speaker 1>of Harvard. So we had the retail sales numbers which

0:00:18.920 --> 0:00:21.440
<v Speaker 1>came in lower and expected by the housing starts were

0:00:21.480 --> 0:00:24.759
<v Speaker 1>really down substantially. Are we seeing a slowing economy?

0:00:25.520 --> 0:00:28.040
<v Speaker 2>We may be, and it's certainly not at the pace

0:00:28.160 --> 0:00:31.920
<v Speaker 2>that it once was. I think it's a real question

0:00:32.080 --> 0:00:37.040
<v Speaker 2>whether we're really seeing a profound slowing or month to

0:00:37.159 --> 0:00:41.479
<v Speaker 2>month fluctuations. My guess is this is still in the

0:00:41.520 --> 0:00:46.080
<v Speaker 2>world of month to month fluctuations with an underlying picture

0:00:46.600 --> 0:00:52.480
<v Speaker 2>around continuing growth. But you can't be sure, and I'd

0:00:52.520 --> 0:00:56.120
<v Speaker 2>certainly agree that the data has been more on the

0:00:56.240 --> 0:00:59.360
<v Speaker 2>slow side for the last month or two than on

0:00:59.440 --> 0:01:00.280
<v Speaker 2>the rapid.

0:01:00.600 --> 0:01:03.120
<v Speaker 1>So last week we spoke. We talked about the possible

0:01:03.160 --> 0:01:06.200
<v Speaker 1>economic policies of a Donald Trump if he were returned

0:01:06.400 --> 0:01:08.119
<v Speaker 1>to the White House, because he had floated the idea

0:01:08.160 --> 0:01:11.640
<v Speaker 1>of maybe having tariffs replaced some or all of income taxes.

0:01:11.760 --> 0:01:14.479
<v Speaker 1>You were not very enthusiastic about that. Since then, he's

0:01:14.480 --> 0:01:18.080
<v Speaker 1>responded to you very specifically on a podcast called all

0:01:18.200 --> 0:01:20.959
<v Speaker 1>In where they asked him about well, how he responded

0:01:21.000 --> 0:01:23.959
<v Speaker 1>to your thoughts on tariffs, and he said he respected you.

0:01:24.080 --> 0:01:26.840
<v Speaker 1>He gave you nice plaut it said, he respects you. You

0:01:26.600 --> 0:01:28.920
<v Speaker 1>speak your mind, is I think what he said. But

0:01:28.920 --> 0:01:31.080
<v Speaker 1>at the same time, he really likes tariffs, he said,

0:01:31.120 --> 0:01:34.600
<v Speaker 1>because it shows the power of a country, both economically

0:01:34.680 --> 0:01:38.280
<v Speaker 1>and politically. What do you make of his endorsement of

0:01:38.319 --> 0:01:41.040
<v Speaker 1>tariffs as a major tool of policy.

0:01:41.520 --> 0:01:47.640
<v Speaker 2>I don't see the evidence for his belief. First of all,

0:01:48.000 --> 0:01:51.920
<v Speaker 2>the tariffs he proposes are going to be levied against Canada,

0:01:52.000 --> 0:01:56.680
<v Speaker 2>They're going to be levied against our traditional European allies.

0:01:57.320 --> 0:02:01.400
<v Speaker 2>If they are a tool of power and timidation, it

0:02:01.440 --> 0:02:04.280
<v Speaker 2>seems to me they should be used much more selectively

0:02:04.880 --> 0:02:10.359
<v Speaker 2>than he has proposed. Second, when you launch attacks, then

0:02:10.520 --> 0:02:15.200
<v Speaker 2>others respond and the whole thing can spiral. The classic

0:02:15.320 --> 0:02:19.200
<v Speaker 2>example of a major tariff policy in American history was

0:02:19.320 --> 0:02:25.359
<v Speaker 2>Smooth Hawley, and it contributed to making the depression great.

0:02:25.760 --> 0:02:29.480
<v Speaker 2>As I look around the world at countries who have

0:02:29.720 --> 0:02:35.440
<v Speaker 2>seen tariffs as the center of their economic strategy. To

0:02:35.560 --> 0:02:40.400
<v Speaker 2>make a nationalist point vis a vis other countries. The

0:02:40.600 --> 0:02:45.480
<v Speaker 2>examples look like places like Argentina and a number of

0:02:45.520 --> 0:02:51.120
<v Speaker 2>places in Latin America where it has not been so successful.

0:02:51.520 --> 0:02:58.920
<v Speaker 2>So I'd like to see what the case is. But

0:02:59.440 --> 0:03:05.480
<v Speaker 2>this is a case swhere. It's about as universal among

0:03:05.800 --> 0:03:13.000
<v Speaker 2>economists who studied these things that you shouldn't pursue systematic

0:03:13.160 --> 0:03:19.720
<v Speaker 2>across the board tariffs for long periods of time.

0:03:20.280 --> 0:03:22.760
<v Speaker 1>Let's talk about change, because there's an awful lot of

0:03:22.760 --> 0:03:25.119
<v Speaker 1>things that are changing around us right now. Whether it's

0:03:25.120 --> 0:03:29.440
<v Speaker 1>climate that you've talked about at fair amount, geopolitics, global economy,

0:03:29.480 --> 0:03:31.400
<v Speaker 1>there are a lot of changes going at the same time.

0:03:31.760 --> 0:03:34.920
<v Speaker 1>Are we going through a particularly tumultuous time of change

0:03:34.960 --> 0:03:37.640
<v Speaker 1>right now? Do you think around us? Or has it

0:03:37.680 --> 0:03:38.160
<v Speaker 1>always been?

0:03:38.200 --> 0:03:43.600
<v Speaker 2>Thus, here's the sense I have, David, I was fortunate

0:03:43.720 --> 0:03:50.560
<v Speaker 2>enough to welcome baby granddaughter, Francis Joanne into the world

0:03:50.640 --> 0:03:55.680
<v Speaker 2>when my daughter had my first granddaughter, about ten days ago.

0:03:56.120 --> 0:03:59.320
<v Speaker 2>So I've been thinking about her life and it made

0:03:59.360 --> 0:04:04.480
<v Speaker 2>me think about my grandmother. My grandmother lived from nineteen

0:04:04.600 --> 0:04:09.720
<v Speaker 2>hundred to nineteen seventy four. She saw indoor plumbing come,

0:04:10.080 --> 0:04:15.160
<v Speaker 2>she saw electricity come, she saw a first telephone, then radio,

0:04:15.440 --> 0:04:23.599
<v Speaker 2>then TV movies come. She saw air conditioning, She saw antibiotics,

0:04:23.680 --> 0:04:28.839
<v Speaker 2>which made childhood death a rarity. She saw the ability

0:04:28.880 --> 0:04:32.880
<v Speaker 2>to no longer ride in horses, but instead to be

0:04:32.920 --> 0:04:38.440
<v Speaker 2>able to fly across the country in five hours. My

0:04:38.640 --> 0:04:44.240
<v Speaker 2>life almost now as long as hers was yours. We've

0:04:44.279 --> 0:04:48.400
<v Speaker 2>seen a lot of history. We've seen a lot of change. Yes,

0:04:48.600 --> 0:04:53.560
<v Speaker 2>we've seen computers. We've seen the cell phone. We've seen yes,

0:04:53.800 --> 0:05:01.560
<v Speaker 2>the Bloomberg terminal. We've seen more modern financial markets. But

0:05:01.680 --> 0:05:05.119
<v Speaker 2>I think you'd have to agree that we've saw much

0:05:05.279 --> 0:05:11.640
<v Speaker 2>much less change than my grandmother's generation did. I have

0:05:11.680 --> 0:05:16.919
<v Speaker 2>a suspicion that my granddaughter is going to witness history

0:05:17.800 --> 0:05:24.960
<v Speaker 2>like my grandmother did. And most importantly, I think we're

0:05:25.000 --> 0:05:29.640
<v Speaker 2>going to see a step change with what happens in

0:05:30.440 --> 0:05:36.360
<v Speaker 2>artificial intelligence. As I've said before, the wheel was awfully fundamental,

0:05:36.920 --> 0:05:39.880
<v Speaker 2>but once you have the wheel, you don't automatically get

0:05:39.880 --> 0:05:44.800
<v Speaker 2>more and better wheels. Same thing with electricity. But artificial

0:05:44.880 --> 0:05:50.920
<v Speaker 2>intelligence has the capacity to make better artificial intelligence, and

0:05:50.960 --> 0:05:55.880
<v Speaker 2>that puts in a kind of upward exponential ratchet that

0:05:56.320 --> 0:06:01.200
<v Speaker 2>isn't a feature of any other technological change. So my

0:06:01.360 --> 0:06:07.480
<v Speaker 2>daughter's gonna witness seismic change, and the granddaughter is gonna

0:06:07.520 --> 0:06:10.760
<v Speaker 2>witness seismic change, and the issue is going to be

0:06:11.480 --> 0:06:17.560
<v Speaker 2>can we manage it so we avoid the catastrophes that

0:06:17.880 --> 0:06:24.440
<v Speaker 2>were also part of my grandmother's interval on this earth.

0:06:25.080 --> 0:06:27.240
<v Speaker 1>I just want to pick up on the artificial intelligence point,

0:06:27.279 --> 0:06:31.920
<v Speaker 1>because you brought artificial generative artificial to us here at

0:06:31.960 --> 0:06:34.159
<v Speaker 1>Wall Street week right about the time I think chat

0:06:34.200 --> 0:06:36.120
<v Speaker 1>Gepeat came out. Now, since then you've got on the

0:06:36.120 --> 0:06:37.960
<v Speaker 1>board of Open AI, so I want to ask what

0:06:38.279 --> 0:06:41.320
<v Speaker 1>secret things in open eye? But from what you understand

0:06:41.440 --> 0:06:45.640
<v Speaker 1>now about generative AI, has your view changed about how

0:06:45.680 --> 0:06:47.800
<v Speaker 1>big it could be and also about what the possible

0:06:47.839 --> 0:06:50.400
<v Speaker 1>dangers are because we know now more than we did then.

0:06:51.200 --> 0:06:58.479
<v Speaker 2>Look, I think, David, this can be transcendent in its

0:06:58.520 --> 0:07:04.000
<v Speaker 2>importance for the reason and that that I just describe.

0:07:04.640 --> 0:07:10.280
<v Speaker 2>It has this recursive aspect where it's able to improve itself,

0:07:11.040 --> 0:07:17.480
<v Speaker 2>and that differentiates it from other technologies. When we'll get

0:07:17.960 --> 0:07:22.160
<v Speaker 2>where very hard to know how much will there be

0:07:23.320 --> 0:07:32.040
<v Speaker 2>last mile problems that will stop some uses. That's another

0:07:32.320 --> 0:07:36.800
<v Speaker 2>very important question where I don't think anybody can give

0:07:37.360 --> 0:07:45.080
<v Speaker 2>a completely confident answer. God knows as companies as a society.

0:07:46.120 --> 0:07:51.480
<v Speaker 2>We cannot leave war to generals, and we cannot leave

0:07:52.120 --> 0:07:59.600
<v Speaker 2>AI only to AI developers. That's why it's absolutely essential

0:08:00.280 --> 0:08:07.600
<v Speaker 2>that public authorities take a strong role here to make

0:08:07.640 --> 0:08:12.600
<v Speaker 2>sure this technology is used for good. But equally, any

0:08:12.680 --> 0:08:16.800
<v Speaker 2>effort to stop this or to just slow it down

0:08:17.280 --> 0:08:22.000
<v Speaker 2>for the sake of slowing it down without also thinking

0:08:22.040 --> 0:08:26.800
<v Speaker 2>about its positive development, would be to see the field

0:08:27.400 --> 0:08:31.120
<v Speaker 2>to the irresponsible, would be to seed the field to

0:08:31.880 --> 0:08:36.240
<v Speaker 2>potential adversaries of the United States, would be to seed

0:08:36.360 --> 0:08:42.560
<v Speaker 2>the field to those whose vision of AI might be

0:08:42.679 --> 0:08:47.240
<v Speaker 2>as a tool of totalitarianism rather than as a tool

0:08:47.880 --> 0:08:55.880
<v Speaker 2>of human emancipation. That's why there's a phrase that I

0:08:56.200 --> 0:09:03.240
<v Speaker 2>like that the that open eyes I often uses responsible

0:09:03.400 --> 0:09:09.080
<v Speaker 2>iterative deployment. That is, you don't you proceed in stages

0:09:09.800 --> 0:09:14.640
<v Speaker 2>and you're very focused on doing it responsibly. Now, that's

0:09:14.760 --> 0:09:21.680
<v Speaker 2>easier said than done and requires enormous thought, but I

0:09:21.720 --> 0:09:25.880
<v Speaker 2>don't think we have any other viable alternative.

0:09:26.200 --> 0:09:28.000
<v Speaker 1>Larry, thank you so very much for being with us

0:09:28.000 --> 0:09:30.880
<v Speaker 1>again this week. That's our special contributor, Larry Summers of

0:09:31.000 --> 0:09:31.440
<v Speaker 1>Harvard