1 00:00:02,520 --> 00:00:07,200 Speaker 1: Bloomberg Audio Studios, podcasts, radio news. 2 00:00:08,000 --> 00:00:10,520 Speaker 2: Good morning everyone. I'd hope that we'd start by talking 3 00:00:10,520 --> 00:00:13,200 Speaker 2: about productivity and some of the data that you've just said, well, 4 00:00:13,240 --> 00:00:18,720 Speaker 2: no one really knows. And the question with AI and 5 00:00:18,760 --> 00:00:22,239 Speaker 2: the US economy is what has happened thus far? Right? 6 00:00:22,760 --> 00:00:26,920 Speaker 2: And so I'll hit you with some of the official 7 00:00:27,000 --> 00:00:28,760 Speaker 2: data that I've been tracking, and you can tell me 8 00:00:29,200 --> 00:00:31,720 Speaker 2: whether the utility of it or not. Right, which is 9 00:00:31,760 --> 00:00:34,280 Speaker 2: really in US productivity? I always go with the Bureau 10 00:00:34,320 --> 00:00:38,280 Speaker 2: of Labor Statistics measure of output per hour x non 11 00:00:38,360 --> 00:00:41,199 Speaker 2: farm business sector. Right, And if you look at the 12 00:00:41,240 --> 00:00:43,640 Speaker 2: data over fifty years, that chart was really interesting, the 13 00:00:43,680 --> 00:00:47,680 Speaker 2: side by side of electricity and AI over fifty years, 14 00:00:47,720 --> 00:00:50,080 Speaker 2: the average quarterly reading is about one point nine percent 15 00:00:50,320 --> 00:00:54,520 Speaker 2: annual rate on productivity. But something's happened in the last 16 00:00:54,560 --> 00:00:57,160 Speaker 2: ten quarters where it's higher, Yes, you know, just under 17 00:00:57,240 --> 00:01:00,360 Speaker 2: three percent two point seven percent. Do we really know 18 00:01:01,000 --> 00:01:02,920 Speaker 2: what that is? And is it AI? 19 00:01:03,440 --> 00:01:04,120 Speaker 3: We don't know. 20 00:01:04,200 --> 00:01:06,400 Speaker 1: I mean, that's the part that makes it hard is 21 00:01:06,800 --> 00:01:10,360 Speaker 1: in productivity numbers, especially when they're happening in what I 22 00:01:10,360 --> 00:01:13,480 Speaker 1: would think of as real time, it's very challenging to 23 00:01:13,600 --> 00:01:16,320 Speaker 1: assess or draw it back to exactly what the factors 24 00:01:16,360 --> 00:01:18,280 Speaker 1: are that have shaped it. You know, in fact, people 25 00:01:18,319 --> 00:01:20,320 Speaker 1: still don't agree on what happened in the nineties all 26 00:01:20,319 --> 00:01:22,080 Speaker 1: the time if you look at research. So it's just 27 00:01:22,120 --> 00:01:25,040 Speaker 1: something to keep in mind. So then what you do 28 00:01:25,280 --> 00:01:28,600 Speaker 1: would any good economist or person, any industry person would do, 29 00:01:28,680 --> 00:01:29,600 Speaker 1: is they'd say, well. 30 00:01:29,520 --> 00:01:31,679 Speaker 3: What am I seeing? What am I seeing? 31 00:01:31,680 --> 00:01:34,399 Speaker 1: And so right now, while we can't find it in 32 00:01:34,440 --> 00:01:39,280 Speaker 1: the macro studies, it would do very sophisticated empirical econometrics 33 00:01:39,319 --> 00:01:41,720 Speaker 1: and ask the questions how much of this is AI. 34 00:01:42,520 --> 00:01:45,480 Speaker 1: We still can see that there's something going on there. 35 00:01:45,840 --> 00:01:49,520 Speaker 1: The question is is it happening? The question is how 36 00:01:49,520 --> 00:01:53,520 Speaker 1: long will it persist? And so clearly something's happening in 37 00:01:53,560 --> 00:01:56,480 Speaker 1: the economy. But if you make a series to go 38 00:01:56,560 --> 00:01:59,160 Speaker 1: back to your question about productivity, if you make a 39 00:01:59,280 --> 00:02:02,560 Speaker 1: series of one time adjustments, so say you automate a 40 00:02:03,040 --> 00:02:05,800 Speaker 1: production line or you use AI to help in loan 41 00:02:05,840 --> 00:02:10,080 Speaker 1: application process, you save money once. You don't save money forever. 42 00:02:10,639 --> 00:02:12,280 Speaker 1: I mean, you keep saving that money, but you don't 43 00:02:12,280 --> 00:02:14,520 Speaker 1: get growth out of that. You don't get productivity growth. 44 00:02:14,520 --> 00:02:17,880 Speaker 1: You get one time adjustments to the level of productivity of. 45 00:02:17,840 --> 00:02:19,560 Speaker 3: Your employees or your process. 46 00:02:19,960 --> 00:02:24,639 Speaker 1: So what we're looking for is a technology to give 47 00:02:24,720 --> 00:02:29,880 Speaker 1: us consistently good changes in productivity so that all industries 48 00:02:30,040 --> 00:02:33,200 Speaker 1: at scale get better, industries figure out new ways to 49 00:02:33,280 --> 00:02:37,000 Speaker 1: generate revenue, new ways to do product design, new ideas 50 00:02:37,040 --> 00:02:39,600 Speaker 1: to come and shape the economy. That's the thing that 51 00:02:39,639 --> 00:02:43,519 Speaker 1: has a sustained productivity growth part. So it's undeniable productivity 52 00:02:43,560 --> 00:02:46,960 Speaker 1: growth has gone up. What's not as clear is how 53 00:02:46,960 --> 00:02:48,280 Speaker 1: long will that last. 54 00:02:48,639 --> 00:02:53,480 Speaker 2: Broadly, people want to see and understand how AI impacts 55 00:02:53,520 --> 00:02:57,600 Speaker 2: workforce and more recently maybe inflation. So if we go 56 00:02:57,680 --> 00:03:01,640 Speaker 2: back to the nineties and what green Span saw in 57 00:03:01,760 --> 00:03:06,400 Speaker 2: productivity gains contributing to economic growth, there was a consideration 58 00:03:06,480 --> 00:03:09,079 Speaker 2: around both of those things. Absolutely, you said that it's 59 00:03:09,120 --> 00:03:11,760 Speaker 2: not the playbook to go back to what happened nineties 60 00:03:11,760 --> 00:03:14,320 Speaker 2: and apply today, But what do you see in those things? 61 00:03:14,400 --> 00:03:18,880 Speaker 2: Is it possible that AI is driving productivity games resulting 62 00:03:18,880 --> 00:03:22,560 Speaker 2: in economic growth, but without the inflation it is. 63 00:03:22,520 --> 00:03:25,240 Speaker 1: Absolutely possible and something we have to interrogate. I mean, 64 00:03:25,320 --> 00:03:29,960 Speaker 1: right now, as you know to well, inflation still above 65 00:03:29,960 --> 00:03:34,040 Speaker 1: our target are two percent target, and price level has 66 00:03:34,080 --> 00:03:36,400 Speaker 1: been high for much higher for a long time, and 67 00:03:36,440 --> 00:03:40,240 Speaker 1: people are feeling stretched by the high inflation that they see. 68 00:03:40,440 --> 00:03:42,960 Speaker 1: And now oftentimes people say, well, now AI is going 69 00:03:43,040 --> 00:03:45,839 Speaker 1: to take hurt the labor market, and so now I'm 70 00:03:45,920 --> 00:03:50,120 Speaker 1: in double doom, as people say. But I think ultimately 71 00:03:50,880 --> 00:03:54,880 Speaker 1: the way you think many people think about AI is 72 00:03:54,920 --> 00:04:00,560 Speaker 1: the investment part of any technology can actually boost demand 73 00:04:00,680 --> 00:04:05,800 Speaker 1: for good services in people and can then raise the 74 00:04:05,800 --> 00:04:09,400 Speaker 1: pressure on inflation. But then the productivity part comes and 75 00:04:09,440 --> 00:04:11,760 Speaker 1: that that's a disinflationary part. 76 00:04:12,080 --> 00:04:12,520 Speaker 3: You can see. 77 00:04:12,560 --> 00:04:14,680 Speaker 1: This is all about the timing, and so what we 78 00:04:14,880 --> 00:04:19,359 Speaker 1: end up investigating is not just the models but asking questions. 79 00:04:19,440 --> 00:04:24,000 Speaker 1: Are the buildout of data centers raising prices for construction workers, 80 00:04:24,320 --> 00:04:28,680 Speaker 1: are the buildout of data centers raising prices for metals 81 00:04:28,720 --> 00:04:31,160 Speaker 1: and other things that go into them? The raw materials 82 00:04:31,520 --> 00:04:34,560 Speaker 1: are the productivity gains. And then on the other side 83 00:04:34,600 --> 00:04:37,839 Speaker 1: of that, are the productivity gains already affecting the cost 84 00:04:37,880 --> 00:04:40,720 Speaker 1: structure of firms? Do they see that and even if 85 00:04:40,720 --> 00:04:44,400 Speaker 1: a series of one off adjustments can actually change the 86 00:04:44,440 --> 00:04:47,479 Speaker 1: cost structure, And if you look at profit margins when 87 00:04:47,600 --> 00:04:50,680 Speaker 1: prices haven't been raising as rapidly as they once were, 88 00:04:51,000 --> 00:04:52,919 Speaker 1: and firms are saying they don't have as much power 89 00:04:52,960 --> 00:04:55,320 Speaker 1: to pass through, you would think that they're doing something 90 00:04:55,360 --> 00:04:58,240 Speaker 1: to help margin protection. And so I think this is 91 00:04:58,520 --> 00:05:00,840 Speaker 1: there's something going on here. Whether we wanted to link 92 00:05:00,880 --> 00:05:03,160 Speaker 1: it all back to AI and then use that as 93 00:05:03,200 --> 00:05:06,880 Speaker 1: a demonstrated proof that we're in a transformative state, I 94 00:05:06,920 --> 00:05:11,080 Speaker 1: think that's a little bit too far, but certainly something's happening. 95 00:05:11,400 --> 00:05:15,000 Speaker 1: And thinking about the productivity growth is exactly what you 96 00:05:15,080 --> 00:05:17,440 Speaker 1: know we did back in the nineteen nineties. We saw 97 00:05:17,600 --> 00:05:21,440 Speaker 1: evidence firms were being more productive. We were interrogating how 98 00:05:21,480 --> 00:05:25,560 Speaker 1: long that would last. And interestingly, the nineteen nineties when 99 00:05:25,560 --> 00:05:27,800 Speaker 1: I said it was the Roaring nineties that followed. It 100 00:05:27,880 --> 00:05:30,800 Speaker 1: was good growth, but it was also a good labor market, 101 00:05:31,080 --> 00:05:34,280 Speaker 1: a really strong labor market, and so those two things 102 00:05:34,320 --> 00:05:37,680 Speaker 1: went together. Because ultimately we had this conversation in the 103 00:05:37,760 --> 00:05:40,599 Speaker 1: roundtable and one of the participants made a great point, 104 00:05:40,640 --> 00:05:41,760 Speaker 1: it's true economics. 105 00:05:41,960 --> 00:05:43,239 Speaker 3: This is how economics works. 106 00:05:43,320 --> 00:05:47,640 Speaker 1: Is if an employee using AI gets much more productive, 107 00:05:48,240 --> 00:05:49,320 Speaker 1: you hire more of them. 108 00:05:49,200 --> 00:05:50,599 Speaker 3: Right, not fewer of them. 109 00:05:50,680 --> 00:05:54,320 Speaker 1: So you know, the economy grows faster, the product development 110 00:05:54,360 --> 00:05:57,560 Speaker 1: goes goes faster, and demand gets stronger. 111 00:05:57,600 --> 00:05:59,440 Speaker 2: I'm going to jump ahead to data center I'd been 112 00:05:59,480 --> 00:06:03,359 Speaker 2: saving it, but it's highly relevant to San Jose the 113 00:06:03,400 --> 00:06:06,800 Speaker 2: build out of data center. Very recently, the CEO of 114 00:06:06,839 --> 00:06:09,600 Speaker 2: PG and E, Patty Poppy, Game on the program and 115 00:06:09,880 --> 00:06:13,719 Speaker 2: made the argument that it's possible that the data center 116 00:06:13,760 --> 00:06:16,719 Speaker 2: build out within PG and e's jurisdiction actually brings down 117 00:06:16,880 --> 00:06:21,040 Speaker 2: wholesale electricity prices because the hyperscalas take on the capital 118 00:06:21,080 --> 00:06:25,120 Speaker 2: burden and they are buyers and aggregate of electricity. But 119 00:06:25,320 --> 00:06:28,560 Speaker 2: many people, you know, your constituents in the twelfth district 120 00:06:28,839 --> 00:06:32,320 Speaker 2: will find it hard to see that argument playing out. 121 00:06:32,520 --> 00:06:34,960 Speaker 1: Well, I think we have to separate what we're talking 122 00:06:35,000 --> 00:06:37,279 Speaker 1: about into now, next, later. 123 00:06:37,760 --> 00:06:38,880 Speaker 3: So let's think about now. 124 00:06:39,279 --> 00:06:43,720 Speaker 1: Right now, we have more demand than we have supply 125 00:06:44,160 --> 00:06:47,000 Speaker 1: for energy for electricity. Right if you talk to you, 126 00:06:47,120 --> 00:06:50,960 Speaker 1: we regularly have CEO round tables with the big power 127 00:06:50,960 --> 00:06:54,240 Speaker 1: companies across the twelfth district. You can look throughout the nation. 128 00:06:54,800 --> 00:06:58,840 Speaker 1: Demand for power is higher than the supply of power, 129 00:06:59,200 --> 00:07:02,080 Speaker 1: and there's a lot of reasons why supply is falling behind. 130 00:07:02,440 --> 00:07:05,000 Speaker 1: One is demand's just gone up rapidly, but another part 131 00:07:05,080 --> 00:07:07,320 Speaker 1: is that they've got an aging infrastructure. They have to 132 00:07:07,360 --> 00:07:11,000 Speaker 1: get those things built out. It's a highly regulated industry, 133 00:07:11,120 --> 00:07:13,720 Speaker 1: so the infrastructure doesn't just come on like a light switch. 134 00:07:14,120 --> 00:07:16,960 Speaker 1: Then you have there are supply chain issues that made 135 00:07:16,960 --> 00:07:19,840 Speaker 1: it hard to get the transformers and other things. So 136 00:07:19,920 --> 00:07:23,920 Speaker 1: all of this just adds to the imbalance of demand 137 00:07:24,000 --> 00:07:27,880 Speaker 1: versus supply. But the remedy for that isn't to take 138 00:07:27,920 --> 00:07:30,120 Speaker 1: away demand, it's to increase supply. 139 00:07:30,520 --> 00:07:31,440 Speaker 3: So when. 140 00:07:33,640 --> 00:07:36,040 Speaker 1: They talk about any CEO of a power company says, 141 00:07:36,080 --> 00:07:39,720 Speaker 1: we can solve this problem by adding more supply. Absolutely, 142 00:07:39,800 --> 00:07:43,640 Speaker 1: but that's a next and later. And so what you said, 143 00:07:43,640 --> 00:07:47,160 Speaker 1: my constituents, what consumers and businesses are saying is I'm 144 00:07:47,160 --> 00:07:49,440 Speaker 1: worried my electricity prices are going to rise, and they've 145 00:07:49,440 --> 00:07:54,040 Speaker 1: already been going up. And the CEOs of power companies 146 00:07:54,040 --> 00:07:56,440 Speaker 1: are saying, but if we just keep building, that will 147 00:07:56,440 --> 00:07:58,120 Speaker 1: go down, and both are true. 148 00:07:57,960 --> 00:07:59,880 Speaker 2: Go down as far as it will be disinflation. 149 00:08:00,000 --> 00:08:02,640 Speaker 1: And you know, it's hard to say energy could be 150 00:08:02,680 --> 00:08:06,360 Speaker 1: disinflationary if we get to a point where supply is 151 00:08:06,400 --> 00:08:08,640 Speaker 1: greater than demand. Right now, I'm just looking for a 152 00:08:08,680 --> 00:08:11,080 Speaker 1: supply to equal demand, and that would be a big 153 00:08:11,120 --> 00:08:13,960 Speaker 1: benefit to consumers because it would mean that we wouldn't 154 00:08:14,040 --> 00:08:17,320 Speaker 1: keep seeing inflationary pressure coming out of the energy sector. 155 00:08:17,680 --> 00:08:19,160 Speaker 2: The other thing I wanted to ask you through the 156 00:08:19,200 --> 00:08:22,120 Speaker 2: lens of constituents of the twelfth districts is one reason 157 00:08:22,640 --> 00:08:25,800 Speaker 2: you might focus on productivity is there is a direct 158 00:08:25,840 --> 00:08:29,120 Speaker 2: read through to GDP growth and other data sets that 159 00:08:29,160 --> 00:08:32,280 Speaker 2: you can look at. But the anxiety in the real 160 00:08:32,320 --> 00:08:36,120 Speaker 2: world is, well, a job, an AI talk can make 161 00:08:36,160 --> 00:08:39,640 Speaker 2: me more productive or it can displace me altogether. Where 162 00:08:39,640 --> 00:08:41,880 Speaker 2: do you see that tension in the economy right now? 163 00:08:42,040 --> 00:08:42,800 Speaker 3: So one of the. 164 00:08:42,720 --> 00:08:46,240 Speaker 1: Things that is true is that the labor market has slowed, 165 00:08:46,840 --> 00:08:49,839 Speaker 1: but it slowed for a whole variety of reasons. And 166 00:08:50,000 --> 00:08:52,200 Speaker 1: much like when you said, well, productivity is risen, Mary, 167 00:08:52,320 --> 00:08:55,240 Speaker 1: so shouldn't we isn't that AI? I think we always 168 00:08:55,240 --> 00:08:58,800 Speaker 1: want to be a little humble about the correlations we 169 00:08:58,880 --> 00:09:00,480 Speaker 1: see and ascribing. 170 00:09:00,120 --> 00:09:01,080 Speaker 3: Causality to them. 171 00:09:01,160 --> 00:09:04,800 Speaker 1: So I wanted to temper your enthusiasm for thinking all 172 00:09:04,840 --> 00:09:07,080 Speaker 1: the productivity growth is AI might be, but it could 173 00:09:07,120 --> 00:09:10,480 Speaker 1: just be general cost management in a slowing economy or 174 00:09:10,520 --> 00:09:13,840 Speaker 1: a slowing you know, or to manage tariff costs, etc. 175 00:09:14,400 --> 00:09:17,560 Speaker 3: So on the labor market. The labor market is slowing. 176 00:09:17,920 --> 00:09:21,160 Speaker 1: It's slowing in industries that are directly telling us that 177 00:09:21,160 --> 00:09:23,480 Speaker 1: they're using AI and it's slowing in industries that aren't. 178 00:09:23,559 --> 00:09:25,480 Speaker 1: So it's one of the things that I just want 179 00:09:25,520 --> 00:09:28,720 Speaker 1: to be cautious. So what I talk to, We talked 180 00:09:28,720 --> 00:09:31,640 Speaker 1: to workers, We talk to you know, communities all the time. 181 00:09:32,240 --> 00:09:36,760 Speaker 1: What's true is in technologies is a really interesting thing. 182 00:09:36,880 --> 00:09:41,280 Speaker 1: No technology ever reduces net employment, not in the history 183 00:09:41,320 --> 00:09:45,480 Speaker 1: of technologies, but it does change what that employment looks like. 184 00:09:46,000 --> 00:09:49,920 Speaker 1: And so there's a period of replacement right now. It's 185 00:09:50,000 --> 00:09:54,199 Speaker 1: replacement of tasks. So if your job has certain tasks 186 00:09:54,200 --> 00:09:56,840 Speaker 1: in it, AI can do those for you. And the 187 00:09:56,880 --> 00:10:01,439 Speaker 1: next part is augmentation, so every place augment and create. 188 00:10:02,280 --> 00:10:06,839 Speaker 1: What's interesting about AI is that, unlike say electricity, when 189 00:10:06,880 --> 00:10:10,280 Speaker 1: the candle lighters or the lamplighters or the candle makers 190 00:10:10,800 --> 00:10:15,320 Speaker 1: got displaced before the US completely became electrified, you know 191 00:10:15,760 --> 00:10:18,719 Speaker 1: that this is going more quickly if you go to 192 00:10:18,840 --> 00:10:21,920 Speaker 1: a firm. I was on a panel at the Reagan 193 00:10:22,800 --> 00:10:29,200 Speaker 1: National Library Economic Forum with Patrick Collison from Sprint. I'm Sprints, right, Gosh, Stripe, Sorry, 194 00:10:29,400 --> 00:10:33,960 Speaker 1: he's going to kill me, Stripe, don't tell him, okay, 195 00:10:34,440 --> 00:10:39,320 Speaker 1: but from Stripe, and and interestingly he said, I am 196 00:10:39,760 --> 00:10:43,199 Speaker 1: hiring more coders that I'm laying off, but I am 197 00:10:43,320 --> 00:10:47,120 Speaker 1: laying off coders whose technology skills didn't advance or they 198 00:10:47,160 --> 00:10:48,080 Speaker 1: weren't the right workers. 199 00:10:48,120 --> 00:10:49,440 Speaker 3: And you're seeing this right. 200 00:10:49,320 --> 00:10:54,240 Speaker 1: You're seeing you know, businesses reskill them their cells to 201 00:10:54,640 --> 00:10:57,359 Speaker 1: meet the AI moment, and that's going to cause worker anxiety. 202 00:10:57,400 --> 00:11:00,200 Speaker 1: And right now, worker anxiety is high. People were a 203 00:11:00,240 --> 00:11:04,080 Speaker 1: low firing, low hiring environment. That's already going to make 204 00:11:04,080 --> 00:11:07,120 Speaker 1: people feel vulnerable. If you haven't found a job and 205 00:11:07,120 --> 00:11:10,080 Speaker 1: you're newly minted out of college, you just think I 206 00:11:10,120 --> 00:11:12,160 Speaker 1: was supposed to get a job before I graduated. Now 207 00:11:12,160 --> 00:11:15,839 Speaker 1: I still don't have one. That's very worriesome to people. 208 00:11:16,160 --> 00:11:18,880 Speaker 1: And then if you're thinking, well, I might lose my job, 209 00:11:19,000 --> 00:11:20,360 Speaker 1: but I don't know how long it will take to 210 00:11:20,360 --> 00:11:22,600 Speaker 1: get another one, then you're worried about that. So I 211 00:11:22,640 --> 00:11:25,720 Speaker 1: think it's natural for the sentiment to feel nervous. But 212 00:11:25,840 --> 00:11:29,040 Speaker 1: it's not the same things. AI is taking all the jobs, 213 00:11:29,040 --> 00:11:34,200 Speaker 1: because what we're really seeing is AIS is replacing tasks, 214 00:11:34,679 --> 00:11:37,160 Speaker 1: augmenting workers. When we talk to firms, most of those 215 00:11:37,200 --> 00:11:39,319 Speaker 1: firms are saying I'm augmenting my workforce. 216 00:11:39,600 --> 00:11:40,880 Speaker 3: If you're in big. 217 00:11:40,679 --> 00:11:44,000 Speaker 1: Manufacturing firms, they don't have enough workers that do skilled labor, 218 00:11:44,240 --> 00:11:47,040 Speaker 1: and so they're looking to augment their workforce, and then 219 00:11:47,160 --> 00:11:50,480 Speaker 1: we're also seeing jobs created. It's interesting I gave this talk. 220 00:11:50,800 --> 00:11:53,080 Speaker 1: I gave a talk on this in twenty twenty three, 221 00:11:53,559 --> 00:11:56,760 Speaker 1: and I used prompt engineers as the jobs they were creating. 222 00:11:56,840 --> 00:11:58,880 Speaker 1: But those jobs are now being replaced. 223 00:12:00,360 --> 00:12:00,600 Speaker 3: But a. 224 00:12:02,679 --> 00:12:05,439 Speaker 2: Or it's a change, it's. 225 00:12:04,840 --> 00:12:07,720 Speaker 1: A warning, you could think of it that way. Or 226 00:12:07,840 --> 00:12:11,080 Speaker 1: it's an indicator. So let's take the warning. The warning 227 00:12:11,240 --> 00:12:15,560 Speaker 1: is you can't keep up. I would say, let's use 228 00:12:15,600 --> 00:12:18,599 Speaker 1: it as an indicator. It's an indicator that technology is 229 00:12:18,640 --> 00:12:24,200 Speaker 1: evolving really fast and workforces need to skill endurable skills, 230 00:12:24,480 --> 00:12:25,560 Speaker 1: and durable skills. 231 00:12:25,640 --> 00:12:27,280 Speaker 3: Are be AI ready? 232 00:12:27,559 --> 00:12:30,360 Speaker 1: Be able to use AI to lift yourself in the 233 00:12:30,480 --> 00:12:33,880 Speaker 1: educational space. You know, use the use the technologies that 234 00:12:33,920 --> 00:12:36,320 Speaker 1: are out there to build your skills up, because you 235 00:12:36,320 --> 00:12:40,679 Speaker 1: can learn a lot fast if you train yourself to 236 00:12:40,840 --> 00:12:41,439 Speaker 1: look at AI. 237 00:12:41,679 --> 00:12:44,160 Speaker 3: Say give me a lesson on how. 238 00:12:44,040 --> 00:12:47,240 Speaker 1: To write I've been thinking about this, how to write 239 00:12:47,280 --> 00:12:50,720 Speaker 1: a smart contract from end to end? What sort of 240 00:12:50,800 --> 00:12:52,360 Speaker 1: software would I need? How would I do it? What 241 00:12:52,400 --> 00:12:54,320 Speaker 1: would the code look like? How would I test the code? 242 00:12:54,360 --> 00:12:56,600 Speaker 1: How would I know it's right? Before I execute on 243 00:12:56,640 --> 00:12:59,720 Speaker 1: this smart contract, and so you can do these things 244 00:13:00,160 --> 00:13:03,000 Speaker 1: in an evening and then it's just about being able 245 00:13:03,040 --> 00:13:05,720 Speaker 1: to do that. So I think that's the message for workers, 246 00:13:06,120 --> 00:13:08,480 Speaker 1: and I would have taught my young self this same thing. 247 00:13:08,679 --> 00:13:14,600 Speaker 1: Is if you put off technology because you're afraid of it, 248 00:13:15,320 --> 00:13:19,319 Speaker 1: then you won't be in the first place of trying 249 00:13:19,360 --> 00:13:22,520 Speaker 1: to use the technology to further your own abilities. 250 00:13:22,600 --> 00:13:25,240 Speaker 2: Can we extend that to the FED? Now bear with 251 00:13:25,280 --> 00:13:27,920 Speaker 2: me on that one. Okay, all right, you talked about 252 00:13:28,000 --> 00:13:34,120 Speaker 2: disaggregated data but also improved measurement, citing Greenspan in that sense. 253 00:13:34,120 --> 00:13:39,959 Speaker 2: If AI is so good, can it process larger sets 254 00:13:40,000 --> 00:13:45,000 Speaker 2: of data and make more accurate economic forecasts than traditional 255 00:13:45,000 --> 00:13:45,640 Speaker 2: FED models? 256 00:13:45,679 --> 00:13:50,600 Speaker 1: Can? We know we don't right now use AI in 257 00:13:50,600 --> 00:13:55,120 Speaker 1: our monetary policy work, but we do use it in 258 00:13:55,200 --> 00:13:58,360 Speaker 1: research as researchers. If you go to any academic institution, 259 00:13:58,800 --> 00:14:03,280 Speaker 1: you're going to see researchers using AI to see what 260 00:14:03,360 --> 00:14:05,440 Speaker 1: they can do better on coding and other things, but 261 00:14:05,480 --> 00:14:07,319 Speaker 1: also data analytics. 262 00:14:07,480 --> 00:14:10,719 Speaker 3: What do you see The place. 263 00:14:10,480 --> 00:14:14,760 Speaker 1: We are there is AI doesn't give you answers to problems. 264 00:14:14,800 --> 00:14:17,920 Speaker 1: It helps you get to the discovery perspective. So if 265 00:14:17,960 --> 00:14:20,840 Speaker 1: I use AI as a researcher to look at a 266 00:14:20,880 --> 00:14:24,160 Speaker 1: bunch of data. I still have to test my hypothesis. 267 00:14:24,160 --> 00:14:26,080 Speaker 1: I have to go in with a hypothesis. What am 268 00:14:26,120 --> 00:14:28,680 Speaker 1: I trying to answer? So that's the human person. And 269 00:14:28,720 --> 00:14:32,480 Speaker 1: so that's why it's not particularly well tooled right now 270 00:14:32,760 --> 00:14:39,840 Speaker 1: to replace our forecasters and our thinkers, our scholars who. 271 00:14:39,800 --> 00:14:41,920 Speaker 2: Comproduce a more accurate neutral rates. 272 00:14:41,960 --> 00:14:45,080 Speaker 1: For example, Well, you're still going to get estimates that 273 00:14:45,120 --> 00:14:47,920 Speaker 1: are between eleven and negative three on the neutral rate 274 00:14:47,960 --> 00:14:50,960 Speaker 1: of interest, So I'm not kidding. Models can the models 275 00:14:51,040 --> 00:14:54,360 Speaker 1: that we have can produce an estimate from negative three 276 00:14:54,720 --> 00:14:58,320 Speaker 1: to positive eleven, right, And so there. 277 00:14:58,320 --> 00:14:59,240 Speaker 3: What does that tell you? 278 00:14:59,640 --> 00:15:02,440 Speaker 1: The the neutral rate of interest is not a truth 279 00:15:02,520 --> 00:15:03,320 Speaker 1: with a capital T. 280 00:15:03,840 --> 00:15:04,680 Speaker 3: It's an estimate. 281 00:15:04,720 --> 00:15:07,720 Speaker 1: It's a theoretical construct to help us understand how to 282 00:15:07,760 --> 00:15:11,920 Speaker 1: benchmark policy. But you can't use it as a threshold 283 00:15:11,960 --> 00:15:15,560 Speaker 1: that you can do surgical adjustments around. No one calibrates 284 00:15:15,840 --> 00:15:20,040 Speaker 1: monetary policy surgically with a neutral rate of interest estimate 285 00:15:20,160 --> 00:15:23,360 Speaker 1: for those reasons. So we are using AI though at 286 00:15:23,360 --> 00:15:25,480 Speaker 1: the FED, and many people may be surprised about that. 287 00:15:25,520 --> 00:15:26,640 Speaker 1: Would you like to learn about that? 288 00:15:26,960 --> 00:15:27,840 Speaker 2: Yes? Please help? 289 00:15:28,920 --> 00:15:32,160 Speaker 1: So I know many might think I work in an 290 00:15:32,160 --> 00:15:35,800 Speaker 1: institution that waits for elect we're still getting electricity. You 291 00:15:35,840 --> 00:15:39,520 Speaker 1: might think that, but no, we actually are not the 292 00:15:39,560 --> 00:15:43,480 Speaker 1: earliest adopters because remember we're fiduciary stewarts of public funds 293 00:15:43,560 --> 00:15:46,480 Speaker 1: but also as fiduciary stewards of public trust, and so 294 00:15:46,520 --> 00:15:48,720 Speaker 1: we really have to make sure that we're working in 295 00:15:49,120 --> 00:15:53,560 Speaker 1: the most risk free and risk managed environment we possibly can't. 296 00:15:54,520 --> 00:15:57,800 Speaker 1: But we have been at this since really in twenty 297 00:15:57,840 --> 00:16:00,120 Speaker 1: twenty three. So the first thing that we did it 298 00:16:00,160 --> 00:16:02,200 Speaker 1: as a system, and I'll really speak about the twelve 299 00:16:02,320 --> 00:16:04,920 Speaker 1: at a reserve banks that are across the country. We 300 00:16:05,000 --> 00:16:08,040 Speaker 1: worked as a system to say, well, we need to 301 00:16:08,080 --> 00:16:13,520 Speaker 1: make sure our employees, our teams are ready to understand AI. 302 00:16:13,800 --> 00:16:15,240 Speaker 3: So what do we need to do. 303 00:16:15,320 --> 00:16:19,760 Speaker 1: We need to have lessons, work playing, you know, work gatherings, 304 00:16:19,760 --> 00:16:23,520 Speaker 1: et cetera, get people familiar and get people focused in 305 00:16:23,640 --> 00:16:25,800 Speaker 1: areas that we can practice with. So we built a 306 00:16:25,840 --> 00:16:30,080 Speaker 1: practice environment that was completely ring fenced and not in production. Right, 307 00:16:30,120 --> 00:16:32,480 Speaker 1: it's just a practice environment trying it out, and of 308 00:16:32,520 --> 00:16:35,000 Speaker 1: course we got what most businesses got. The other businesses 309 00:16:35,040 --> 00:16:36,280 Speaker 1: did exactly the same thing. 310 00:16:36,480 --> 00:16:38,239 Speaker 3: And what do you get? You get the early adopters. 311 00:16:38,400 --> 00:16:40,200 Speaker 1: But the good news about our early adopters as are 312 00:16:40,200 --> 00:16:43,440 Speaker 1: often ambassadors. So then we're holding like tech cafes and 313 00:16:43,480 --> 00:16:45,280 Speaker 1: things to help other people learn that. 314 00:16:45,400 --> 00:16:46,080 Speaker 3: So that was then. 315 00:16:46,400 --> 00:16:48,880 Speaker 1: So then in twenty twenty four and twenty five we 316 00:16:48,960 --> 00:16:52,200 Speaker 1: really made a full court press push March Madness is 317 00:16:52,200 --> 00:16:52,640 Speaker 1: coming up. 318 00:16:53,720 --> 00:16:53,920 Speaker 3: You know. 319 00:16:53,960 --> 00:16:57,280 Speaker 1: We really went hard at making sure that people had 320 00:16:57,320 --> 00:16:59,840 Speaker 1: not just the if they're interested, do it, but that 321 00:17:00,040 --> 00:17:01,400 Speaker 1: this is something that we really want. 322 00:17:01,320 --> 00:17:01,640 Speaker 3: You to learn. 323 00:17:01,720 --> 00:17:03,640 Speaker 2: This is the operations within the system. 324 00:17:03,680 --> 00:17:05,639 Speaker 1: This is the operations. I should have said that before. 325 00:17:05,920 --> 00:17:08,000 Speaker 1: I'm sorry. Let'll tell you that if you were at 326 00:17:08,000 --> 00:17:10,720 Speaker 1: a reserve bank, and again little known facts. These are 327 00:17:10,720 --> 00:17:14,160 Speaker 1: like facts that people don't know about the FED if 328 00:17:14,160 --> 00:17:16,840 Speaker 1: you go to a reserve bank or any of our operations. 329 00:17:17,080 --> 00:17:20,080 Speaker 1: Most of the people who work with us are operations people. 330 00:17:20,320 --> 00:17:26,159 Speaker 1: We process cash, We do all the electronic payment system backbones, 331 00:17:26,280 --> 00:17:28,119 Speaker 1: make sure they operate on time. If you're in the 332 00:17:28,119 --> 00:17:31,159 Speaker 1: financial sector, you know FED wire or acch FED Now 333 00:17:31,400 --> 00:17:32,840 Speaker 1: all of that is operated by. 334 00:17:32,760 --> 00:17:35,760 Speaker 3: Our operations teams. We also support Vice Chair. 335 00:17:35,720 --> 00:17:39,960 Speaker 1: Bowman in supervision of banks and all of those things. 336 00:17:40,800 --> 00:17:42,600 Speaker 1: And then we have all our support people who help 337 00:17:42,640 --> 00:17:45,560 Speaker 1: make sure that that occurs. All of that can be 338 00:17:45,640 --> 00:17:48,360 Speaker 1: easily if you can do AI and you can use it, 339 00:17:48,560 --> 00:17:51,840 Speaker 1: you can think of opportunities. So the next thing we did, 340 00:17:51,960 --> 00:17:54,720 Speaker 1: get our workforce ready is number one. For next thing 341 00:17:54,760 --> 00:17:57,080 Speaker 1: you do is and this is again like all the 342 00:17:57,119 --> 00:17:59,800 Speaker 1: businesses we talk to, you see what your vendors are 343 00:17:59,800 --> 00:18:02,720 Speaker 1: our offering, right if you're you have a technology vendor, 344 00:18:02,760 --> 00:18:05,320 Speaker 1: an accounting vendor, an HR vendor, and then you just 345 00:18:05,359 --> 00:18:07,560 Speaker 1: turn the service on. But if you turn the service 346 00:18:07,600 --> 00:18:10,160 Speaker 1: on before your team is ready, you don't really get 347 00:18:10,160 --> 00:18:12,560 Speaker 1: an ROI out of it. So again, fiduciary stewarts of 348 00:18:12,600 --> 00:18:15,360 Speaker 1: public funds, we have to make sure so we're all 349 00:18:15,480 --> 00:18:19,440 Speaker 1: instilled in this ring fenced proprietary environment because the public 350 00:18:19,480 --> 00:18:22,520 Speaker 1: has to know that we're not introducing risk to this. 351 00:18:23,040 --> 00:18:25,440 Speaker 3: And then of course the place we're. 352 00:18:25,280 --> 00:18:28,320 Speaker 1: Working now is where many many people are working, which 353 00:18:28,400 --> 00:18:31,760 Speaker 1: is if you have a lot of technology workers, then 354 00:18:31,880 --> 00:18:36,760 Speaker 1: the coding assist is just so important and we're one 355 00:18:36,760 --> 00:18:42,280 Speaker 1: of the things back to workforce. It doesn't create a 356 00:18:42,400 --> 00:18:45,199 Speaker 1: massive change in who you have working for you. What 357 00:18:45,240 --> 00:18:47,879 Speaker 1: it means is they can do their work faster, better 358 00:18:48,160 --> 00:18:50,520 Speaker 1: and more effectively. And if you think of the three 359 00:18:50,800 --> 00:18:55,840 Speaker 1: timples of the FED, we need to be efficient, effective. 360 00:18:55,480 --> 00:18:56,280 Speaker 3: And resilient. 361 00:18:56,480 --> 00:18:59,439 Speaker 1: So it also builds in that resilience for us because 362 00:18:59,440 --> 00:19:03,560 Speaker 1: we have you know, quality assurance and unit testing. All 363 00:19:03,600 --> 00:19:05,240 Speaker 1: the things that can slow you down if you're not 364 00:19:05,359 --> 00:19:09,840 Speaker 1: right or interrupture your ability to serve. Those things can 365 00:19:09,880 --> 00:19:13,080 Speaker 1: all be assisted with AI in a really positive way. So, 366 00:19:13,600 --> 00:19:18,960 Speaker 1: again not monetary policy, but definitely like all other companies 367 00:19:19,240 --> 00:19:22,000 Speaker 1: who are working on this space, making sure we're not 368 00:19:22,119 --> 00:19:25,840 Speaker 1: behind and delivering good value for I mean, the shareholders 369 00:19:25,920 --> 00:19:29,000 Speaker 1: of the FED are the American people, and we owe 370 00:19:29,040 --> 00:19:32,919 Speaker 1: them the effort to make sure we're modernizing ourselves and 371 00:19:33,000 --> 00:19:35,080 Speaker 1: keeping up with the things that can help us do 372 00:19:35,119 --> 00:19:36,920 Speaker 1: our work faster, better, and more effective. 373 00:19:37,080 --> 00:19:40,480 Speaker 2: I do have two questions relating to AI and monetary 374 00:19:40,480 --> 00:19:42,520 Speaker 2: policy quickly, and I know we want to get some 375 00:19:42,600 --> 00:19:44,720 Speaker 2: audience questions as well, some conscious there are students in 376 00:19:44,720 --> 00:19:47,480 Speaker 2: the room who will go out into the workforce. I 377 00:19:47,560 --> 00:19:50,280 Speaker 2: think that the main thing reflecting back on the nineties 378 00:19:50,880 --> 00:19:55,320 Speaker 2: is that there are anticipated impacts yes from AI on 379 00:19:55,359 --> 00:20:01,760 Speaker 2: the economy and PCEE is the preferred gauge inflation running 380 00:20:02,200 --> 00:20:06,080 Speaker 2: higher beyond two percent. How do you manage that. You know, 381 00:20:06,119 --> 00:20:10,880 Speaker 2: many would argue that those anticipated AI driven productivity gains 382 00:20:11,160 --> 00:20:15,639 Speaker 2: would justify lower rates, but they are that anticipated, yeah, and. 383 00:20:15,640 --> 00:20:18,480 Speaker 1: I think that, you know, really, it's important to recognize 384 00:20:18,640 --> 00:20:21,439 Speaker 1: that monetary policy is a forward looking business, but it's 385 00:20:21,480 --> 00:20:24,240 Speaker 1: also an evidence based business. And so there will be 386 00:20:24,320 --> 00:20:28,439 Speaker 1: a point in time when we'll have enough confidence that 387 00:20:28,480 --> 00:20:33,240 Speaker 1: the anticipated effects are materializing. And where would you look 388 00:20:33,280 --> 00:20:35,800 Speaker 1: for that? You look for that in what's happening with 389 00:20:35,840 --> 00:20:39,280 Speaker 1: price pressures? Not just aggregate inflation, but if you disaggregate 390 00:20:39,320 --> 00:20:41,840 Speaker 1: it and you ask yourself a question, are the AI 391 00:20:42,240 --> 00:20:46,720 Speaker 1: using sectors just doing less pass through of input costs 392 00:20:46,880 --> 00:20:49,720 Speaker 1: into prices? Well, maybe that tells you something. So that's 393 00:20:49,720 --> 00:20:51,480 Speaker 1: where the research really becomes important. 394 00:20:51,560 --> 00:20:52,360 Speaker 3: You ask questions. 395 00:20:52,359 --> 00:20:56,880 Speaker 1: But you also do research where you can disaggregate firms, 396 00:20:56,880 --> 00:20:59,159 Speaker 1: you can disaggregate prices, and you can ask where do 397 00:20:59,160 --> 00:21:00,800 Speaker 1: we see price pressure. 398 00:21:00,680 --> 00:21:02,719 Speaker 3: And how do we think they will evolve. 399 00:21:03,440 --> 00:21:06,200 Speaker 1: That's so that's important, and you can't wait because remember 400 00:21:06,240 --> 00:21:08,960 Speaker 1: Monterrey policy as a twelve to eighteen month lag So 401 00:21:09,080 --> 00:21:13,480 Speaker 1: right now we're modestly restrictive, slightly restrictive depending on who 402 00:21:13,520 --> 00:21:15,480 Speaker 1: you talk to. If you have a neutral rate of 403 00:21:15,520 --> 00:21:18,439 Speaker 1: around three percent interest. Remember that's the one with the 404 00:21:18,440 --> 00:21:20,840 Speaker 1: big range. But if you have a neutral rate of interest, 405 00:21:21,240 --> 00:21:23,920 Speaker 1: think this is around three percent. We have some ways 406 00:21:23,960 --> 00:21:26,639 Speaker 1: to go seventy five basis points roughly before we get 407 00:21:26,680 --> 00:21:29,359 Speaker 1: to that level. But we need to get inflation down 408 00:21:29,480 --> 00:21:32,080 Speaker 1: and we need to make sure that it's on a 409 00:21:32,080 --> 00:21:36,280 Speaker 1: good path. I'm certainly looking at AI and productivity growth 410 00:21:36,320 --> 00:21:39,880 Speaker 1: as one mechanism that continues to help us bring inflation 411 00:21:39,960 --> 00:21:42,880 Speaker 1: leve but we also have restrictive policy and other factors 412 00:21:42,880 --> 00:21:43,960 Speaker 1: that are all bringing in. 413 00:21:43,920 --> 00:21:47,600 Speaker 2: How are you thinking about the labor market now, particularly 414 00:21:47,640 --> 00:21:51,679 Speaker 2: post January jobs, which showed essentially the most hiring in 415 00:21:51,720 --> 00:21:54,080 Speaker 2: more than a year. It was an interesting data point. 416 00:21:54,320 --> 00:21:57,000 Speaker 1: Well, you know, one of the things that I'll offer here, 417 00:21:57,040 --> 00:21:59,680 Speaker 1: and it's something probably most of us stone. 418 00:21:59,720 --> 00:22:01,120 Speaker 3: I mean, you don't look at. 419 00:22:01,080 --> 00:22:03,760 Speaker 1: I'd look at it a lot, but is that a 420 00:22:03,800 --> 00:22:06,760 Speaker 1: lot of the job growth in our nation right now 421 00:22:07,000 --> 00:22:10,959 Speaker 1: is located in health care and education. And while it's 422 00:22:11,000 --> 00:22:13,680 Speaker 1: not bad to have jobs growing in health care and education, 423 00:22:13,760 --> 00:22:16,200 Speaker 1: if you look at the rest of the economy, there 424 00:22:16,200 --> 00:22:18,640 Speaker 1: hasn't really been any job growth, and in fact, there's 425 00:22:18,680 --> 00:22:21,639 Speaker 1: been job decline, you know, negative job growth. 426 00:22:21,680 --> 00:22:23,360 Speaker 3: Basically job losses. 427 00:22:23,760 --> 00:22:27,639 Speaker 1: And so that just makes me put an underscore on 428 00:22:27,680 --> 00:22:30,080 Speaker 1: this idea that the labor market has a no hiring, 429 00:22:30,359 --> 00:22:33,840 Speaker 1: no firing that's already making you a little vulnerable to 430 00:22:34,400 --> 00:22:36,520 Speaker 1: a negative shock pushing you below. 431 00:22:36,960 --> 00:22:38,920 Speaker 3: But also if all your jobs. 432 00:22:38,600 --> 00:22:41,320 Speaker 1: Are in health care and education, think of all those 433 00:22:41,359 --> 00:22:45,119 Speaker 1: workers trained for other sectors who are not and are 434 00:22:45,160 --> 00:22:46,520 Speaker 1: not getting opportunities. 435 00:22:46,840 --> 00:22:47,760 Speaker 3: And I think that's. 436 00:22:47,600 --> 00:22:49,720 Speaker 1: Where, you know, we have more work to do to 437 00:22:49,760 --> 00:22:54,080 Speaker 1: make sure that there's no vulnerability doesn't turn into fragility. 438 00:22:54,320 --> 00:22:57,560 Speaker 1: But that's less about AI and more about the diversified 439 00:22:57,600 --> 00:23:01,359 Speaker 1: growth in the economy. And if companies are able to 440 00:23:01,600 --> 00:23:06,880 Speaker 1: really see positive output growth as uncertainty for the positive 441 00:23:06,920 --> 00:23:11,000 Speaker 1: demand growth as the uncertainty decreases, then I think, you know, 442 00:23:11,119 --> 00:23:13,320 Speaker 1: that's a possibility that would be a positive boost for 443 00:23:13,359 --> 00:23:15,520 Speaker 1: the economy. So then it's about should we look at 444 00:23:15,520 --> 00:23:18,680 Speaker 1: a positive boost for the economy as an inflationary event 445 00:23:19,000 --> 00:23:21,000 Speaker 1: or should we think that a positive boost for the 446 00:23:21,040 --> 00:23:24,800 Speaker 1: economy comes with AI and doesn't actually induce inflation. 447 00:23:25,000 --> 00:23:27,119 Speaker 2: So diversity in the economy is where I want to 448 00:23:27,200 --> 00:23:29,399 Speaker 2: end it before we take audience questions. One of the 449 00:23:29,760 --> 00:23:34,520 Speaker 2: features on the show regularly is compensation in the field 450 00:23:34,560 --> 00:23:39,399 Speaker 2: of AI, stock based compensation, competitive salaries, the newly minted 451 00:23:39,400 --> 00:23:42,480 Speaker 2: millionaires in the field, who you know buying property in 452 00:23:42,480 --> 00:23:45,840 Speaker 2: San Francisco but within the twelfth district. One of the 453 00:23:45,840 --> 00:23:47,960 Speaker 2: things I always reflect on is if I drive from 454 00:23:47,960 --> 00:23:50,680 Speaker 2: the Bay Area down to Socou on the five or 455 00:23:50,720 --> 00:23:54,040 Speaker 2: the one to one, it's the agricultural sector this state 456 00:23:54,119 --> 00:23:58,800 Speaker 2: in particular. But you could expand that to the other 457 00:23:59,160 --> 00:24:03,760 Speaker 2: regions of the district. There's a big sort of contrast there. 458 00:24:04,600 --> 00:24:06,840 Speaker 2: Could you reflect on both, you know, what you see 459 00:24:06,840 --> 00:24:09,680 Speaker 2: at the high end of the tech sector and what 460 00:24:09,720 --> 00:24:13,640 Speaker 2: you do or do not see in agriculture. Feeling benefit 461 00:24:13,680 --> 00:24:15,040 Speaker 2: from AI, well. 462 00:24:14,880 --> 00:24:16,000 Speaker 3: You know, it's interesting. 463 00:24:16,040 --> 00:24:19,399 Speaker 1: So we have, as I said, we have roundtables, and 464 00:24:19,440 --> 00:24:21,639 Speaker 1: I have one this morning, but I have them with 465 00:24:21,680 --> 00:24:23,920 Speaker 1: all kinds of industries. I like to do them by industry. 466 00:24:24,040 --> 00:24:26,960 Speaker 1: So we had an agricultural roundtable. How are you using 467 00:24:27,000 --> 00:24:31,160 Speaker 1: AI surprisingly for ahead of where you'd think, right, they're 468 00:24:31,240 --> 00:24:34,600 Speaker 1: using it to do everything from you know, think about 469 00:24:34,600 --> 00:24:38,359 Speaker 1: idea generation. How do you get better crops, more weather resistant, 470 00:24:38,520 --> 00:24:42,919 Speaker 1: drought resistant, fire resistant, you know, there's all smoke resistant. 471 00:24:43,520 --> 00:24:46,560 Speaker 1: AI can help there because it can help generate ideas. 472 00:24:46,880 --> 00:24:49,560 Speaker 1: Another thing they're doing is using AI to think about 473 00:24:49,720 --> 00:24:51,159 Speaker 1: what's the right planting season? 474 00:24:51,320 --> 00:24:52,680 Speaker 3: Right, how do I forecast weather? 475 00:24:52,800 --> 00:24:55,960 Speaker 1: It's predictive, it's predictive and so, and then of course 476 00:24:56,080 --> 00:24:59,240 Speaker 1: using it in their plants and processing to help augment 477 00:24:59,320 --> 00:25:04,520 Speaker 1: their technology along the production line. So AI is something. 478 00:25:05,040 --> 00:25:10,520 Speaker 1: This is why I think it's more pervasive than many understand, 479 00:25:11,000 --> 00:25:17,320 Speaker 1: is that we've had travel and entertainment, We've had consumer retail, 480 00:25:17,440 --> 00:25:24,159 Speaker 1: we've had builders, commercial developers, agricultural, you name it. Everybody's 481 00:25:24,240 --> 00:25:28,080 Speaker 1: trying to see how this can make their business work better. 482 00:25:28,359 --> 00:25:32,280 Speaker 1: And the question is when we finished this part which 483 00:25:32,280 --> 00:25:34,359 Speaker 1: I think we've been in of using it for cost 484 00:25:34,440 --> 00:25:38,200 Speaker 1: management and just getting your budgets right, is it going 485 00:25:38,240 --> 00:25:41,920 Speaker 1: to start to change into revenue generation etc. We're seeing 486 00:25:41,920 --> 00:25:44,280 Speaker 1: the seeds of that, using it for product development, etc. 487 00:25:44,880 --> 00:25:48,399 Speaker 1: But that's the uncertainty around this is when does it 488 00:25:48,480 --> 00:25:53,840 Speaker 1: move from something that's just in the development stages and 489 00:25:53,920 --> 00:25:57,560 Speaker 1: with electricity, the wealthy urban areas had it and the 490 00:25:57,640 --> 00:26:02,159 Speaker 1: rural areas didn't. In in this could this go faster? 491 00:26:02,400 --> 00:26:05,280 Speaker 1: Is the diffusion of AI and its use cases faster? 492 00:26:05,680 --> 00:26:08,119 Speaker 1: And we had a great discussion at this roundtable this morning, 493 00:26:08,119 --> 00:26:10,239 Speaker 1: and the tail to share and I'm not sharing all 494 00:26:10,280 --> 00:26:13,760 Speaker 1: of our points, so it is still Chathamhouse rules, guys. 495 00:26:13,840 --> 00:26:18,160 Speaker 1: But seriously, the learning is you know, there's a lot 496 00:26:18,200 --> 00:26:22,760 Speaker 1: of perspectives out there that say that AI could be 497 00:26:22,800 --> 00:26:26,000 Speaker 1: an equalizing force, and I think we need to interrogate 498 00:26:26,119 --> 00:26:30,080 Speaker 1: is an an equalizing force? As vigorously as we interrogate, 499 00:26:30,320 --> 00:26:33,800 Speaker 1: could it be driving further inequality. I don't think we 500 00:26:33,880 --> 00:26:36,960 Speaker 1: know the answer to that, and I under I'll end 501 00:26:37,000 --> 00:26:37,280 Speaker 1: with this. 502 00:26:37,680 --> 00:26:39,359 Speaker 3: In the end, the decision is going to be ours. 503 00:26:39,560 --> 00:26:40,840 Speaker 3: It's not gonna be the technologies. 504 00:26:40,880 --> 00:26:45,920 Speaker 1: The technologies don't you know, kind of inherently decide. 505 00:26:45,960 --> 00:26:46,760 Speaker 3: We decide. 506 00:26:47,080 --> 00:26:48,800 Speaker 2: We're going to take a couple of quick questions from 507 00:26:48,800 --> 00:26:51,199 Speaker 2: the audience. But while we find the mic, oh, we 508 00:26:51,240 --> 00:26:54,160 Speaker 2: have some in advance. I know that in the room. 509 00:26:54,200 --> 00:26:57,080 Speaker 2: We're here at San Jose State University, which. 510 00:26:56,920 --> 00:26:58,000 Speaker 3: I'm very excited to be at. 511 00:26:58,080 --> 00:27:01,040 Speaker 2: You know, there are there are those that will soon 512 00:27:01,080 --> 00:27:03,600 Speaker 2: be going into the workforce here. One of them is 513 00:27:03,640 --> 00:27:08,119 Speaker 2: student questioned tough one, what advice do you have for 514 00:27:08,240 --> 00:27:12,520 Speaker 2: new economists, especially those with a desire to enter public service. 515 00:27:12,840 --> 00:27:15,600 Speaker 2: We got in a little bit about how the fed 516 00:27:15,680 --> 00:27:18,439 Speaker 2: and fed a reserve Bank of San Francisco is or 517 00:27:18,440 --> 00:27:20,880 Speaker 2: isn't using AI but reflect on that. 518 00:27:21,359 --> 00:27:24,680 Speaker 1: Sure absolutely, So first of all, I will just say 519 00:27:24,720 --> 00:27:30,160 Speaker 1: thank you. You're an economist and you're going into public service. Fantastic. Seriously, 520 00:27:30,160 --> 00:27:34,240 Speaker 1: we need people like yourselves who are interested in doing this. 521 00:27:34,240 --> 00:27:38,040 Speaker 1: This is a very fantastic career. I would call it 522 00:27:38,080 --> 00:27:40,480 Speaker 1: a vocation to be in public service and serving on 523 00:27:40,560 --> 00:27:42,639 Speaker 1: the types of things that are the federal reserves and 524 00:27:42,680 --> 00:27:44,400 Speaker 1: other public institutions missions. 525 00:27:44,960 --> 00:27:46,359 Speaker 3: So that's important. 526 00:27:46,640 --> 00:27:49,960 Speaker 1: The important thing about public service that I think is 527 00:27:50,000 --> 00:27:54,040 Speaker 1: overlooked is one of the biggest skills you have to 528 00:27:54,040 --> 00:27:57,000 Speaker 1: have as an economist is being a detective. And a 529 00:27:57,040 --> 00:28:01,600 Speaker 1: detective never gets satisfied by looking at one thing. You 530 00:28:01,680 --> 00:28:05,960 Speaker 1: test your theories, you dig deeper. You're never really satisfied. 531 00:28:06,160 --> 00:28:08,240 Speaker 1: You know, people ask me, Mary, why are you constantly 532 00:28:08,280 --> 00:28:10,840 Speaker 1: curious and never really satisfied with the answers? And I said, 533 00:28:10,840 --> 00:28:14,879 Speaker 1: because you basically, the minute you get confident, you lose. 534 00:28:15,400 --> 00:28:18,119 Speaker 1: You want to be confident in the moment and humble 535 00:28:18,240 --> 00:28:19,880 Speaker 1: enough to ask again, is this right? 536 00:28:19,960 --> 00:28:21,720 Speaker 3: And why would it be wrong? And how do you 537 00:28:21,800 --> 00:28:22,119 Speaker 3: do that? 538 00:28:22,200 --> 00:28:24,679 Speaker 1: So that's an important thing I see that you know, 539 00:28:24,720 --> 00:28:29,720 Speaker 1: AI is a technology, it's not a miracle, and so 540 00:28:30,280 --> 00:28:32,879 Speaker 1: it's about how you find a way to relate to 541 00:28:32,920 --> 00:28:36,320 Speaker 1: AI that makes you better, a better detective if you're 542 00:28:36,320 --> 00:28:38,920 Speaker 1: an economist, a better public servant if you choose to 543 00:28:38,960 --> 00:28:39,840 Speaker 1: work in that field. 544 00:28:40,040 --> 00:28:41,200 Speaker 3: And that's how I use it. 545 00:28:41,240 --> 00:28:44,680 Speaker 1: I'm always trying to make myself better at serving those 546 00:28:44,760 --> 00:28:48,480 Speaker 1: who I've got the responsibility to serve, and doing that 547 00:28:48,520 --> 00:28:51,640 Speaker 1: with a technology or with just being out in the 548 00:28:51,680 --> 00:28:53,040 Speaker 1: factory floor and learning how. 549 00:28:52,960 --> 00:28:55,200 Speaker 3: Businesses are doing it. That's the magic there. 550 00:28:55,280 --> 00:28:59,280 Speaker 1: So you don't get yourself monoligned into only one skill. 551 00:28:59,520 --> 00:29:02,520 Speaker 1: It's really about having the detective range of skills and 552 00:29:02,560 --> 00:29:06,520 Speaker 1: recognizing those skills have to change to meet a changing environment, 553 00:29:06,560 --> 00:29:09,240 Speaker 1: but to also meet the moment. The skills I developed 554 00:29:09,480 --> 00:29:12,000 Speaker 1: in the mid nineties, I've certainly had to change and 555 00:29:12,040 --> 00:29:15,200 Speaker 1: augment those to be able to do my job today 556 00:29:15,440 --> 00:29:16,040 Speaker 1: present daily. 557 00:29:16,120 --> 00:29:19,600 Speaker 2: Quite a few of the other questions are on the 558 00:29:19,640 --> 00:29:22,000 Speaker 2: other side of the remit, which is regulation. You know, 559 00:29:22,040 --> 00:29:25,960 Speaker 2: in your speech you mentioned that financial services the financial 560 00:29:26,000 --> 00:29:29,520 Speaker 2: sector early adopters in many ways, and the question is 561 00:29:29,560 --> 00:29:34,760 Speaker 2: how do you balance regulation that ensure safety within the 562 00:29:34,760 --> 00:29:38,320 Speaker 2: financial system but also allows them to innovate, move faster. 563 00:29:38,560 --> 00:29:40,760 Speaker 1: So I do have to say because this is a 564 00:29:40,800 --> 00:29:43,200 Speaker 1: weird aspect of fetter reserve. 565 00:29:43,440 --> 00:29:44,360 Speaker 3: I don't know if it's weird. 566 00:29:44,440 --> 00:29:47,280 Speaker 1: I think it's right, but it's a unique aspect of 567 00:29:47,320 --> 00:29:50,920 Speaker 1: fetter reserves system. The Reserve Bank presidents don't do any 568 00:29:50,960 --> 00:29:54,400 Speaker 1: regular superviser and we don't even do any supervision. That's 569 00:29:54,480 --> 00:29:57,680 Speaker 1: all left with Vice Chairbowman and she at the border governors, 570 00:29:57,680 --> 00:29:59,880 Speaker 1: and the rules get made by the full border governors, 571 00:30:00,040 --> 00:30:02,600 Speaker 1: not the Reserve Bank presidents. That said, we can talk 572 00:30:02,640 --> 00:30:06,440 Speaker 1: about regulation more generally, not just in financial services. And 573 00:30:06,520 --> 00:30:08,760 Speaker 1: there's always attention. If you're an economist, you know this, 574 00:30:08,840 --> 00:30:11,800 Speaker 1: if your business you know this, right, there's always attension. 575 00:30:12,200 --> 00:30:16,640 Speaker 1: If you let fully unregulated innovation occur, you could do 576 00:30:16,840 --> 00:30:18,920 Speaker 1: customer and consumer harm. 577 00:30:19,240 --> 00:30:22,080 Speaker 3: If you do so much regulation. 578 00:30:21,680 --> 00:30:25,680 Speaker 1: That no innovation occurs, well then you will end in stasis. 579 00:30:26,040 --> 00:30:30,040 Speaker 1: And so somewhere in the middle is where the nation 580 00:30:30,320 --> 00:30:34,520 Speaker 1: has to go. Nations, and we have historically had a 581 00:30:34,680 --> 00:30:38,440 Speaker 1: very robust financial sector in the United States that's facilitated 582 00:30:38,520 --> 00:30:43,000 Speaker 1: a lot of intermediation and growth and sort of allowed 583 00:30:43,080 --> 00:30:45,480 Speaker 1: us to be the country that we've been in terms 584 00:30:45,560 --> 00:30:46,240 Speaker 1: of doing. 585 00:30:46,000 --> 00:30:47,960 Speaker 3: Things so we don't want that to stop. 586 00:30:48,200 --> 00:30:51,120 Speaker 1: But as new tools and technologies come out, it's not 587 00:30:51,240 --> 00:30:54,120 Speaker 1: about cutting them off. It's really about thinking about how 588 00:30:54,160 --> 00:30:58,240 Speaker 1: they can be done safely but still innovatively. And I 589 00:30:58,280 --> 00:31:01,840 Speaker 1: think that magic place is not something you get to 590 00:31:02,040 --> 00:31:06,160 Speaker 1: and then you're always there. It's constant recalibration, constantly asking 591 00:31:06,200 --> 00:31:09,040 Speaker 1: the question, you know, the bridle is too tight, or 592 00:31:09,080 --> 00:31:11,520 Speaker 1: the reins too tight, or are they too lose. It's 593 00:31:11,640 --> 00:31:13,880 Speaker 1: very much like monetary policy in that way. You know, 594 00:31:13,920 --> 00:31:15,640 Speaker 1: you don't get to a point and say great, we 595 00:31:15,720 --> 00:31:18,760 Speaker 1: want victory. You actually are always know if you've ever 596 00:31:18,840 --> 00:31:21,760 Speaker 1: ridden a horse, and if you haven't, I apologize, But 597 00:31:21,800 --> 00:31:25,480 Speaker 1: if you've ever ridden a horse, it's not my first vocation. 598 00:31:26,240 --> 00:31:29,120 Speaker 1: You know, if you pull too tight it stops on 599 00:31:29,160 --> 00:31:31,480 Speaker 1: a diamond, you're over the head, And if you let 600 00:31:31,520 --> 00:31:33,480 Speaker 1: go too much, it runs too fast and you're over 601 00:31:33,520 --> 00:31:37,040 Speaker 1: the back. So it's basically trying to manage the bridles 602 00:31:37,080 --> 00:31:40,480 Speaker 1: so that you get the innovation you want without exposing 603 00:31:40,800 --> 00:31:44,760 Speaker 1: consumers or other businesses or the society to harm. 604 00:31:45,200 --> 00:31:48,280 Speaker 2: Let me ask a final question and will end on 605 00:31:48,400 --> 00:31:51,920 Speaker 2: I guess a positive note. Oh good? Which data sets 606 00:31:52,320 --> 00:31:54,720 Speaker 2: and what you see in the real world, because you 607 00:31:54,760 --> 00:31:58,080 Speaker 2: still go out into the real world, gives you most 608 00:31:58,080 --> 00:32:01,880 Speaker 2: optimism about the impact that AI will have on the 609 00:32:02,000 --> 00:32:05,800 Speaker 2: US economy, and specifically that of the twelfth twelfth District. 610 00:32:05,920 --> 00:32:09,040 Speaker 1: So I will say that, you know, when I first 611 00:32:09,040 --> 00:32:12,160 Speaker 1: came to this job in nineteen ninety six, I am 612 00:32:12,160 --> 00:32:15,040 Speaker 1: going to work at the San Francisco feder Reserve. And 613 00:32:15,120 --> 00:32:18,680 Speaker 1: I had been to California a year before, down at 614 00:32:19,200 --> 00:32:23,120 Speaker 1: Southern California at the Rand Institute, and I remember going 615 00:32:23,160 --> 00:32:24,880 Speaker 1: to a conference there and we met a lot of 616 00:32:24,920 --> 00:32:28,959 Speaker 1: business people thinking about not AI, but something else. And 617 00:32:29,240 --> 00:32:32,840 Speaker 1: I came home and I told my wife, We've got 618 00:32:32,880 --> 00:32:36,000 Speaker 1: to go there. It's filled with entrepreneurs. It's filled with 619 00:32:36,080 --> 00:32:38,960 Speaker 1: people who have never heard the word no. They just 620 00:32:39,080 --> 00:32:42,719 Speaker 1: heard why not right. And what's interesting about the twelfth 621 00:32:42,720 --> 00:32:45,640 Speaker 1: District is all of those people don't live in California. 622 00:32:45,880 --> 00:32:48,160 Speaker 3: They live in Utah, they live in Idaho, they live 623 00:32:48,200 --> 00:32:49,640 Speaker 3: in Vegas. You know, they live. 624 00:32:49,520 --> 00:32:53,160 Speaker 1: In the entire inter mountain west west of the Rockies. 625 00:32:53,200 --> 00:32:55,800 Speaker 1: And I'm not saying anything about other places. They're all 626 00:32:55,880 --> 00:32:59,880 Speaker 1: very innovative too, But there is something here that gives 627 00:32:59,920 --> 00:33:03,640 Speaker 1: me optimism because it's not that people say, well, AI 628 00:33:03,760 --> 00:33:06,719 Speaker 1: is coming, and let me figure out how to you know, 629 00:33:07,400 --> 00:33:09,920 Speaker 1: not be eaten up by it. It's like AI is here, 630 00:33:10,040 --> 00:33:12,800 Speaker 1: let's figure out how to harness this tool to create 631 00:33:13,280 --> 00:33:17,280 Speaker 1: a better business, a better economy. And really the thing 632 00:33:17,320 --> 00:33:20,320 Speaker 1: that gets me jazzed and optimistic is people talk about 633 00:33:20,320 --> 00:33:23,440 Speaker 1: a better world. How do we make things better for people? 634 00:33:23,440 --> 00:33:26,360 Speaker 1: How do we change education so there's more quality, How 635 00:33:26,360 --> 00:33:30,440 Speaker 1: do we help the globe have more opportunity? How do 636 00:33:30,520 --> 00:33:33,200 Speaker 1: we harness what's sitting here in front of us with 637 00:33:33,200 --> 00:33:37,480 Speaker 1: all these people, into something powerful that changes lives and livelihoods. 638 00:33:37,520 --> 00:33:40,040 Speaker 3: So that's what gets me optimistic. And it's not this. 639 00:33:40,280 --> 00:33:42,760 Speaker 1: They're just not talking about what they might do. They 640 00:33:42,800 --> 00:33:45,400 Speaker 1: don't stay if I said now, next, later, there's not 641 00:33:45,440 --> 00:33:47,760 Speaker 1: too many conversations with people who live out here who 642 00:33:47,800 --> 00:33:50,680 Speaker 1: talk about way later. They're all talking about now and next. 643 00:33:50,960 --> 00:33:52,040 Speaker 1: That gets me excited. 644 00:33:52,600 --> 00:33:54,680 Speaker 2: With that, all that's left to do is to thank 645 00:33:54,720 --> 00:33:59,000 Speaker 2: the Silicon Valley Leadership Group, San Jose State University, our hosts, 646 00:33:59,040 --> 00:34:03,000 Speaker 2: and some Francisco fore the Reserve Bank President Mary C. Daily, 647 00:34:03,040 --> 00:34:03,480 Speaker 2: thank you very 648 00:34:03,560 --> 00:34:05,040 Speaker 3: Much, Thank you appreciate it.