1 00:00:02,520 --> 00:00:07,400 Speaker 1: Bloomberg Audio Studios, podcasts, radio news. 2 00:00:07,800 --> 00:00:12,040 Speaker 2: A master's and computer science from Stanford nineteen eighty eight. 3 00:00:12,320 --> 00:00:13,360 Speaker 3: An artificial intelligence. 4 00:00:13,440 --> 00:00:15,520 Speaker 2: I was a new partner at a law firm called 5 00:00:15,560 --> 00:00:18,239 Speaker 2: lower Color and Pickering in Washington during National trade? What 6 00:00:18,320 --> 00:00:20,600 Speaker 2: did you see in artificial intelligence? 7 00:00:20,600 --> 00:00:26,200 Speaker 1: As on It's honored to be with you all, I 8 00:00:26,239 --> 00:00:30,000 Speaker 1: wish I had been a behavioral economist. That was my 9 00:00:30,240 --> 00:00:33,960 Speaker 1: only real big regret in life. And so what did 10 00:00:33,960 --> 00:00:37,599 Speaker 1: I see in AI? Which is sort of related, what's 11 00:00:37,640 --> 00:00:42,360 Speaker 1: the nature of intelligence? I think I saw a dream 12 00:00:42,800 --> 00:00:48,720 Speaker 1: of understanding ourselves and our society a little better, and 13 00:00:48,760 --> 00:00:52,680 Speaker 1: that was sort of the primary motivator. I will say 14 00:00:52,760 --> 00:00:56,520 Speaker 1: that the AI that I studied was the equivalent of 15 00:00:56,680 --> 00:01:01,160 Speaker 1: the sun goes around the earth field. So we had 16 00:01:01,160 --> 00:01:04,679 Speaker 1: a fundamental theory in the nineteen eighties that was completely 17 00:01:04,720 --> 00:01:06,920 Speaker 1: false and didn't work, but we didn't know it at 18 00:01:06,920 --> 00:01:11,720 Speaker 1: the time. So me and many other people then fled 19 00:01:11,760 --> 00:01:15,960 Speaker 1: to all kinds of other fields. And it's exciting now 20 00:01:15,959 --> 00:01:20,319 Speaker 1: to see a new set of techniques really be transformative. 21 00:01:21,200 --> 00:01:27,360 Speaker 1: And to Mike's reference on the news this week, you know, 22 00:01:27,520 --> 00:01:33,160 Speaker 1: you know that I'm enamored with subscription models and now 23 00:01:33,920 --> 00:01:39,240 Speaker 1: I've become more aware about tender offers and I like 24 00:01:39,400 --> 00:01:45,200 Speaker 1: TV channels. So we're announcing today my tender offer for Bloomberg. 25 00:01:47,440 --> 00:01:51,800 Speaker 1: We hope you will consider it appropriately. Your board members 26 00:01:51,840 --> 00:01:56,840 Speaker 1: are willing to debate the transaction. But unfortunately, as you 27 00:01:56,840 --> 00:01:59,480 Speaker 1: would suspect as a board member, that's all I can 28 00:01:59,520 --> 00:02:01,600 Speaker 1: say on the news of the week. 29 00:02:01,720 --> 00:02:04,160 Speaker 3: So you're not going to answer my bosses question, who's going. 30 00:02:04,080 --> 00:02:06,000 Speaker 1: To I think it is all answered. 31 00:02:07,800 --> 00:02:10,359 Speaker 2: So looking back on AI as you saw it then 32 00:02:10,400 --> 00:02:12,919 Speaker 2: and whatever you saw then, and comparing it to now, 33 00:02:13,840 --> 00:02:16,640 Speaker 2: I hear a range of things being said. As extreme 34 00:02:16,760 --> 00:02:19,480 Speaker 2: as this is, as big an invention as the industrial 35 00:02:19,480 --> 00:02:23,440 Speaker 2: evolution sometimes say is the discovery of fire. Are we 36 00:02:23,600 --> 00:02:27,360 Speaker 2: overestimating or underestimating the effect generative A. 37 00:02:27,400 --> 00:02:32,480 Speaker 1: I will have us, Well, let's think back thirty forty 38 00:02:32,480 --> 00:02:35,240 Speaker 1: thousand years ago, and we're a bunch of Neanderthals and 39 00:02:35,240 --> 00:02:38,320 Speaker 1: we're in the entrance to our cave, and you know, 40 00:02:38,360 --> 00:02:42,080 Speaker 1: we've been the dominant human humanoid species for a couple 41 00:02:42,120 --> 00:02:45,560 Speaker 1: hundred thousand years. And the two of us look down 42 00:02:45,840 --> 00:02:48,919 Speaker 1: and we see Homo sapiens. We're like, look at those guys, 43 00:02:48,919 --> 00:02:54,959 Speaker 1: they're so skinny. Hairless, but they're pitiful. But those guys 44 00:02:54,960 --> 00:03:00,400 Speaker 1: were really intelligent, and ultimately those homelostapiens dominated the earth 45 00:03:00,880 --> 00:03:06,880 Speaker 1: and killed off us neandervals. So I would say intelligence 46 00:03:07,040 --> 00:03:11,880 Speaker 1: per species has been highly selected for, and that Homo 47 00:03:11,919 --> 00:03:16,000 Speaker 1: sapiens intelligence has allowed us to become the dominant species 48 00:03:16,040 --> 00:03:19,520 Speaker 1: on Earth because we use our intelligence to make tools 49 00:03:19,600 --> 00:03:23,320 Speaker 1: and do other things. So I think it's a lot 50 00:03:23,560 --> 00:03:27,839 Speaker 1: different than say, mechanizing muscle power. So if you think 51 00:03:27,880 --> 00:03:33,639 Speaker 1: of bulldozers and how did that transform you know, society, 52 00:03:34,120 --> 00:03:36,400 Speaker 1: it did a lot, but it did it over one 53 00:03:36,440 --> 00:03:41,200 Speaker 1: hundred years. And although muscles are important, they're not the 54 00:03:41,240 --> 00:03:44,680 Speaker 1: core human attribute. Thus the dislocation of the end with 55 00:03:44,760 --> 00:03:49,520 Speaker 1: the Neanderthals, which were much stronger than humans. So if 56 00:03:49,600 --> 00:03:54,920 Speaker 1: AI develops to actually be super intelligence, then it will 57 00:03:54,920 --> 00:03:58,360 Speaker 1: be a lot more profound, I think, than anything else, 58 00:03:58,560 --> 00:04:03,160 Speaker 1: and that we will have actually real species threats because 59 00:04:04,400 --> 00:04:07,480 Speaker 1: the AI will keep getting smarter and smarter and smarter 60 00:04:07,600 --> 00:04:11,720 Speaker 1: without limit. A natural selection works quite slowly in terms 61 00:04:11,760 --> 00:04:15,240 Speaker 1: of making humans more intelligent, so then we have to 62 00:04:15,280 --> 00:04:18,800 Speaker 1: sort of say, okay, how fast and how you know? 63 00:04:19,080 --> 00:04:21,640 Speaker 1: Will the computer is really thank Obviously when we all 64 00:04:21,720 --> 00:04:24,600 Speaker 1: use AI as consumers we can find. You know, it's 65 00:04:24,640 --> 00:04:28,839 Speaker 1: pretty miraculous in some cases, but in other cases it's 66 00:04:28,960 --> 00:04:33,800 Speaker 1: not very effective. But you know, we're understanding more and 67 00:04:33,839 --> 00:04:36,719 Speaker 1: more of the techniques. And again, one theory is it'll 68 00:04:36,760 --> 00:04:39,360 Speaker 1: be like Moore's law and just AI will get better 69 00:04:39,360 --> 00:04:43,599 Speaker 1: and better and better. But the other theory is it's 70 00:04:43,600 --> 00:04:46,680 Speaker 1: sort of like the War on cancer, where cancer developed 71 00:04:46,720 --> 00:04:50,000 Speaker 1: over a very long time. It has lots of different etiologies, 72 00:04:50,480 --> 00:04:52,560 Speaker 1: and we keep coming up with a solution for one 73 00:04:52,600 --> 00:04:57,040 Speaker 1: cancer but not another, and our overall progress and society 74 00:04:57,200 --> 00:05:01,080 Speaker 1: against cancer has been you know, pretty steady but flat. 75 00:05:01,120 --> 00:05:05,520 Speaker 1: It definitely not exponential, and so you know, it may 76 00:05:05,560 --> 00:05:08,880 Speaker 1: be that as AI gets better, it hits various walls 77 00:05:09,400 --> 00:05:12,320 Speaker 1: and that we've got some time to deal with it. 78 00:05:13,040 --> 00:05:15,520 Speaker 1: Or it may be that it stays on this exponential 79 00:05:16,240 --> 00:05:18,600 Speaker 1: So we're just going to have to watch. But I 80 00:05:18,640 --> 00:05:21,480 Speaker 1: think we need to be prepared for it to be 81 00:05:21,640 --> 00:05:25,480 Speaker 1: on the exponential, in which case we're going to have 82 00:05:25,760 --> 00:05:28,480 Speaker 1: a lot of societal stress over the next twenty years. 83 00:05:28,640 --> 00:05:31,560 Speaker 2: So your analogy is very helpful, but a little disturbing 84 00:05:31,720 --> 00:05:34,400 Speaker 2: because I don't see many Neanderthals around anymore. 85 00:05:34,839 --> 00:05:35,960 Speaker 3: Right, we have a lot of. 86 00:05:35,880 --> 00:05:39,880 Speaker 2: Homo saviens in your analogy, What can we do to 87 00:05:40,000 --> 00:05:42,279 Speaker 2: make sure that we survive, that that this thing we're 88 00:05:42,320 --> 00:05:46,520 Speaker 2: creating doesn't become so smart and perhaps I can have 89 00:05:46,560 --> 00:05:48,920 Speaker 2: our best interests so are well. 90 00:05:48,920 --> 00:05:53,440 Speaker 1: It's a big challenge. It's hard to slow down because 91 00:05:54,160 --> 00:05:57,000 Speaker 1: unlike chemical weapons, like we're kind of in a war 92 00:05:57,080 --> 00:06:00,360 Speaker 1: with a West in Russia and yet no one using 93 00:06:00,440 --> 00:06:03,760 Speaker 1: chemical weapons and no one's using nuclear weapons. So even 94 00:06:03,800 --> 00:06:06,480 Speaker 1: in this incredible state of war, we're able to put 95 00:06:06,520 --> 00:06:10,000 Speaker 1: some limits on what happens. The problem with AI is 96 00:06:10,040 --> 00:06:13,640 Speaker 1: it's very continuous. You know, your thermostat is an AI thermostat. 97 00:06:13,720 --> 00:06:17,040 Speaker 1: So there's no good way or even the major powers 98 00:06:17,080 --> 00:06:20,320 Speaker 1: to agree not to use AI because it's so integrated 99 00:06:20,360 --> 00:06:24,719 Speaker 1: through everything else that we do. So there's no real 100 00:06:24,800 --> 00:06:28,480 Speaker 1: practical scenario to take a break as human society, even 101 00:06:28,520 --> 00:06:30,320 Speaker 1: if we could get the politics to work on that. 102 00:06:31,080 --> 00:06:34,120 Speaker 1: So instead we're in the situation where we in America 103 00:06:34,360 --> 00:06:38,360 Speaker 1: better well damn win the race, and so everyone acknowledges that, 104 00:06:38,600 --> 00:06:43,599 Speaker 1: and so all the companies are going full out, both 105 00:06:43,680 --> 00:06:48,760 Speaker 1: for their own you know, competitive reasons, and so you know, 106 00:06:48,920 --> 00:06:52,520 Speaker 1: it is a race, hopefully to the top of what 107 00:06:52,760 --> 00:06:56,320 Speaker 1: can AI do for us? And there will be amazing 108 00:06:56,480 --> 00:07:00,680 Speaker 1: positive scenarios. You know, all kinds of medical yours, all 109 00:07:00,760 --> 00:07:05,640 Speaker 1: kinds of productivity. I certainly six years ago, AI was 110 00:07:05,720 --> 00:07:09,279 Speaker 1: very good at image of analysis. That was one of 111 00:07:09,279 --> 00:07:11,480 Speaker 1: the kind of workloads that it first got good at, 112 00:07:11,760 --> 00:07:15,760 Speaker 1: and in particular in radiology, it got significantly better than 113 00:07:15,800 --> 00:07:19,040 Speaker 1: the typical radiologists pretty quickly and has been better than 114 00:07:19,080 --> 00:07:23,080 Speaker 1: the typical radiologist for at least five years. And so 115 00:07:23,240 --> 00:07:25,880 Speaker 1: I would say, you know, weddings and with friends who 116 00:07:25,920 --> 00:07:28,680 Speaker 1: are radiologists, what's good that your kids are not going 117 00:07:28,720 --> 00:07:31,520 Speaker 1: in the business, because radiology is going to be the 118 00:07:31,560 --> 00:07:36,000 Speaker 1: first profession that's wrecked by AI. Good for humanity, but 119 00:07:36,200 --> 00:07:41,080 Speaker 1: not good for radiologists. Well, what happened is as radiology 120 00:07:41,120 --> 00:07:46,080 Speaker 1: got better because of AI analysis that helped radiologists, the 121 00:07:46,160 --> 00:07:49,800 Speaker 1: price of scans went down, and the number of scans 122 00:07:49,880 --> 00:07:52,240 Speaker 1: turned out to be hugely elastic. Now you walk in, 123 00:07:52,320 --> 00:07:55,160 Speaker 1: you've got a cough, boom, you get a scan, okay. 124 00:07:55,320 --> 00:07:57,840 Speaker 1: And so what's happened now is there's about thirty four 125 00:07:57,880 --> 00:08:01,360 Speaker 1: thousand radiologists in the US, which is a shortage. We 126 00:08:01,440 --> 00:08:06,240 Speaker 1: now have a labor shortage of radiologists okay, by about 127 00:08:06,280 --> 00:08:09,760 Speaker 1: five thousand. So it's an incredible story. Of of course, 128 00:08:09,840 --> 00:08:14,880 Speaker 1: elasticity of demand to improve lives, i e. More scans, 129 00:08:15,280 --> 00:08:17,600 Speaker 1: and so we may well see with AI that it 130 00:08:17,640 --> 00:08:22,120 Speaker 1: makes Wall Street analysts more productive and makes software engineers 131 00:08:22,200 --> 00:08:26,000 Speaker 1: more productive, and that in fact the elasticity and that 132 00:08:26,120 --> 00:08:28,800 Speaker 1: we grows and that we grow the economy. And when 133 00:08:28,840 --> 00:08:32,200 Speaker 1: people ask, you know, we're spending collectively as an industry 134 00:08:32,240 --> 00:08:34,640 Speaker 1: at half a trillion dollars on AI data center, how 135 00:08:34,720 --> 00:08:37,320 Speaker 1: is that ever going to get paid back? Well, you know, 136 00:08:37,600 --> 00:08:39,920 Speaker 1: add one or two points to GDP growth and it 137 00:08:39,960 --> 00:08:44,400 Speaker 1: gets paid back fast. So you know, I don't want 138 00:08:44,440 --> 00:08:47,319 Speaker 1: to like guarantee that it's always going to be like radiology. 139 00:08:48,040 --> 00:08:52,959 Speaker 1: I do think software engineering is particularly fascinating because all 140 00:08:52,960 --> 00:08:55,839 Speaker 1: the major AI companies are working on it, and so 141 00:08:56,120 --> 00:08:59,200 Speaker 1: it is the white collar canary in the coal mine. 142 00:09:00,280 --> 00:09:04,640 Speaker 1: If software engineering engineering jobs go down a lot over 143 00:09:04,679 --> 00:09:08,240 Speaker 1: the next five years, then that's probably going to happen 144 00:09:08,240 --> 00:09:11,640 Speaker 1: to law on architecture and many other things. If in fact, 145 00:09:11,720 --> 00:09:14,600 Speaker 1: because of the increased productivity, people are building a lot 146 00:09:14,640 --> 00:09:19,200 Speaker 1: more software, sort of the radiology example, then I think 147 00:09:19,240 --> 00:09:22,880 Speaker 1: in many professions we will see a big expansion and 148 00:09:22,920 --> 00:09:26,600 Speaker 1: we can be more confident of the high productivity. None 149 00:09:26,600 --> 00:09:29,320 Speaker 1: of that really answers the long term question that you asked, 150 00:09:29,320 --> 00:09:31,880 Speaker 1: which is how are we not the Neanderthals? So I 151 00:09:31,880 --> 00:09:37,160 Speaker 1: think reasonable chance high productivity rather than mass unemployment. Okay, 152 00:09:37,240 --> 00:09:40,120 Speaker 1: but then ultimately, what if they're smarter and smarter and 153 00:09:40,160 --> 00:09:42,640 Speaker 1: smarter than us. We're going to have to find ways, 154 00:09:42,679 --> 00:09:45,920 Speaker 1: and I don't know what they are, to both continue 155 00:09:45,960 --> 00:09:48,480 Speaker 1: to insist on alignment, and that's where you train the 156 00:09:48,559 --> 00:09:52,480 Speaker 1: AIS to care about human beings so they are aligned 157 00:09:52,480 --> 00:09:56,959 Speaker 1: with our values. Okay, but if somebody doesn't train, somebody 158 00:09:57,000 --> 00:09:59,600 Speaker 1: programs their AI to try to take over the world, 159 00:10:00,120 --> 00:10:02,440 Speaker 1: we're going to have to enlist the other AIS on 160 00:10:02,520 --> 00:10:06,160 Speaker 1: our defense to protect us. Okay. So you know there's 161 00:10:06,400 --> 00:10:10,839 Speaker 1: a number of scenarios out there, and probably for ten 162 00:10:10,920 --> 00:10:13,040 Speaker 1: twenty years, we're not going to know how serious the 163 00:10:13,080 --> 00:10:15,720 Speaker 1: threat is, but we will have tools. It's not just 164 00:10:15,760 --> 00:10:21,360 Speaker 1: that the AI biological species we have been selected for 165 00:10:21,520 --> 00:10:25,200 Speaker 1: dominance to try to grow our species. So AI is 166 00:10:25,240 --> 00:10:28,680 Speaker 1: not naturally trying to expand, not naturally it could be 167 00:10:28,720 --> 00:10:31,320 Speaker 1: programmed for that, but it can be also programmed to 168 00:10:31,400 --> 00:10:34,880 Speaker 1: keep humans on top. So it's not as scary as 169 00:10:34,920 --> 00:10:38,120 Speaker 1: a super powerful human, which we all kind of into it. 170 00:10:38,240 --> 00:10:41,680 Speaker 1: A super powerful human would be hard to hold back 171 00:10:41,720 --> 00:10:44,440 Speaker 1: from taking over the world. It's not as dire as that. 172 00:10:45,640 --> 00:10:49,239 Speaker 2: Insofar as we do make progress on trying to reshape 173 00:10:49,960 --> 00:10:52,600 Speaker 2: general of AI in a more positive direction. Does that 174 00:10:52,640 --> 00:10:55,240 Speaker 2: come from outside the industry or inside the industry? Does 175 00:10:55,280 --> 00:10:58,440 Speaker 2: that come from government regulation and agreement or does it 176 00:10:58,480 --> 00:11:00,800 Speaker 2: have to come from inside? Because as I talk to 177 00:11:00,880 --> 00:11:04,240 Speaker 2: many people in the business, they're much more focused on 178 00:11:04,280 --> 00:11:06,240 Speaker 2: the race you talk about than they are on the 179 00:11:06,240 --> 00:11:06,760 Speaker 2: safety part. 180 00:11:06,840 --> 00:11:08,360 Speaker 3: They don't want to get slowed down in that. 181 00:11:08,400 --> 00:11:10,120 Speaker 2: I mean, you're on the border anthropic, which I know 182 00:11:10,240 --> 00:11:11,920 Speaker 2: is trying to make a move in the other direction. 183 00:11:12,200 --> 00:11:14,160 Speaker 2: But can we do it from the outside or does 184 00:11:14,240 --> 00:11:17,080 Speaker 2: the industry itself somehow have to internalize the risk. 185 00:11:18,080 --> 00:11:20,800 Speaker 1: Well, I think lots of the industry is working on it. 186 00:11:20,800 --> 00:11:24,880 Speaker 1: So there's different sides of safety. So there's when you're 187 00:11:25,679 --> 00:11:31,080 Speaker 1: treating AI like a counselor and you know it, you 188 00:11:31,080 --> 00:11:33,320 Speaker 1: know helps you tie a news that's not a good thing. 189 00:11:34,400 --> 00:11:37,480 Speaker 1: And so those cases across the industry are getting you know, 190 00:11:37,559 --> 00:11:42,000 Speaker 1: more and more watched for it eliminated, So there's inevitable 191 00:11:42,040 --> 00:11:46,640 Speaker 1: safety bumps as any technology grows, so then there's more 192 00:11:46,720 --> 00:11:49,760 Speaker 1: macro safety, like none of the major AIS can be 193 00:11:49,840 --> 00:11:54,120 Speaker 1: used to design chemical weapons or biological weapons. But those 194 00:11:54,160 --> 00:11:58,800 Speaker 1: defenses in the AIS you know, aren't perfect, and we 195 00:11:58,880 --> 00:12:01,760 Speaker 1: have to constantly and best in them to prevent people 196 00:12:01,800 --> 00:12:05,840 Speaker 1: from using this super powerful technology plus some crisper to 197 00:12:05,920 --> 00:12:09,600 Speaker 1: do some really bad things. So there's active work across 198 00:12:09,640 --> 00:12:14,240 Speaker 1: the whole industry on those scenarios. Because from a commercial standpoint, 199 00:12:14,520 --> 00:12:17,320 Speaker 1: nobody wants their AI brand to be the brand that 200 00:12:17,400 --> 00:12:21,679 Speaker 1: develops some bad virus, so there is an incentive there 201 00:12:21,800 --> 00:12:26,120 Speaker 1: to protect against the negative cases. But the sort of 202 00:12:26,160 --> 00:12:28,720 Speaker 1: big long term case is that the AIS get so 203 00:12:28,840 --> 00:12:33,120 Speaker 1: smart that then we have to do things to make 204 00:12:33,120 --> 00:12:35,679 Speaker 1: sure that they are aligned with human values, that they 205 00:12:35,720 --> 00:12:39,480 Speaker 1: are programmed so that success is humans flourishing. 206 00:12:40,160 --> 00:12:42,319 Speaker 3: Give us your view, but we are likely to see 207 00:12:42,320 --> 00:12:46,320 Speaker 3: the biggest effect. You mentioned biomedicine. You mentioned computer software progners. 208 00:12:46,640 --> 00:12:48,000 Speaker 3: Let's talk turning the news. 209 00:12:47,800 --> 00:12:51,600 Speaker 2: Today with the deal between Disney and Open AI and 210 00:12:51,679 --> 00:12:53,520 Speaker 2: the entertainment space that you know well, I mean you 211 00:12:53,880 --> 00:12:57,160 Speaker 2: manage transition from my perspective, least from more traditional TV 212 00:12:57,440 --> 00:12:59,040 Speaker 2: cable broadcast. 213 00:12:58,840 --> 00:12:59,600 Speaker 3: Through to streaming. 214 00:13:00,080 --> 00:13:03,240 Speaker 2: What does the transition look like from streaming through the AI. 215 00:13:03,960 --> 00:13:07,360 Speaker 1: So just as an example, we can look at video 216 00:13:07,480 --> 00:13:11,560 Speaker 1: creation storytelling and sort of say, let's look at what 217 00:13:11,600 --> 00:13:14,520 Speaker 1: AI is going to do and the mechanics of generating 218 00:13:14,600 --> 00:13:18,960 Speaker 1: video and frankly, whether that's a news channel trying to 219 00:13:19,000 --> 00:13:24,000 Speaker 1: illustrate a concept or entertainers trying to do amazing special effects, 220 00:13:24,840 --> 00:13:28,600 Speaker 1: we're going to have higher quality special effects and that 221 00:13:28,679 --> 00:13:32,479 Speaker 1: will shift to be AI generated instead of manually generated 222 00:13:32,520 --> 00:13:36,560 Speaker 1: with sort of overseas visual effects companies. So there is 223 00:13:36,840 --> 00:13:40,320 Speaker 1: a shift there, But the core thing of storytelling is 224 00:13:40,440 --> 00:13:44,440 Speaker 1: very hard. How do you do long form character development, 225 00:13:44,600 --> 00:13:49,360 Speaker 1: creating tension, resolving tension. That's not a case that the 226 00:13:49,400 --> 00:13:53,960 Speaker 1: AI does well today. Eventually, ten twenty years, the AI 227 00:13:54,120 --> 00:13:57,640 Speaker 1: may win the Booker Prize, okay, and then if it does, 228 00:13:57,720 --> 00:14:00,720 Speaker 1: it's going to be able to do a movie script all. So, 229 00:14:00,760 --> 00:14:04,600 Speaker 1: but for now, the AI is really helping on sort 230 00:14:04,640 --> 00:14:07,760 Speaker 1: of industrial aspects of what we do, like an amazing 231 00:14:07,840 --> 00:14:12,040 Speaker 1: visual effects shot, and so it's you know, important, and 232 00:14:12,320 --> 00:14:15,160 Speaker 1: we want to stay on the front as Disney does, 233 00:14:16,520 --> 00:14:19,600 Speaker 1: but it's not doing the core thing, which is which 234 00:14:19,680 --> 00:14:22,160 Speaker 1: story is going to captivate human attention. 235 00:14:22,560 --> 00:14:25,160 Speaker 2: I'm not sure it says much about my job, about 236 00:14:25,160 --> 00:14:27,600 Speaker 2: whether I keep my job in the AI world, because 237 00:14:27,600 --> 00:14:29,400 Speaker 2: there's a lot of things that are done on television 238 00:14:29,480 --> 00:14:32,560 Speaker 2: with news and other things that frankly, are not creating 239 00:14:32,560 --> 00:14:33,520 Speaker 2: the next Lion King. 240 00:14:34,640 --> 00:14:38,400 Speaker 1: Well, I think interpreting the world in your case is 241 00:14:38,440 --> 00:14:42,240 Speaker 1: a broad general wisdom, and so I think you'll have 242 00:14:42,280 --> 00:14:47,440 Speaker 1: a relatively longstanding role in your success. 243 00:14:47,480 --> 00:14:48,360 Speaker 3: Given my age, I'm good. 244 00:14:48,600 --> 00:14:52,520 Speaker 1: Given your age, you're good. But you know, it is 245 00:14:52,840 --> 00:14:55,800 Speaker 1: a very big change. Again, you know, people want to 246 00:14:55,840 --> 00:14:59,320 Speaker 1: compare it to the Industrial revolution that happened over two 247 00:14:59,400 --> 00:15:02,880 Speaker 1: hundred years. This is going to happen over ten or twenty. 248 00:15:03,880 --> 00:15:08,280 Speaker 1: And it might be that what's caused political polarization in 249 00:15:08,280 --> 00:15:11,880 Speaker 1: the last decade or two is rate of change. So 250 00:15:12,040 --> 00:15:15,240 Speaker 1: think about you know, since NAFTA, you know, sort of 251 00:15:15,280 --> 00:15:21,800 Speaker 1: the rise of globalization, the redefining of marriage equality, immigration, 252 00:15:22,360 --> 00:15:25,960 Speaker 1: all kinds of change in society. And one view is 253 00:15:27,040 --> 00:15:31,120 Speaker 1: enough of our fellow Americans, it's just too much, and 254 00:15:31,200 --> 00:15:33,920 Speaker 1: it radicalizes them and they're willing to vote for things 255 00:15:34,000 --> 00:15:38,160 Speaker 1: they wouldn't normally vote for. And if that's fundamentally what's 256 00:15:38,200 --> 00:15:41,280 Speaker 1: going on in US society as well as you know, 257 00:15:41,440 --> 00:15:45,720 Speaker 1: Brexit and a couple other places. Then we're in for 258 00:15:45,760 --> 00:15:48,520 Speaker 1: a pretty big storm because the rate of change is 259 00:15:48,560 --> 00:15:51,520 Speaker 1: not going to slow down. The rate of change of society, 260 00:15:51,840 --> 00:15:55,080 Speaker 1: you know, partially or maybe largely driven by AI, is 261 00:15:55,120 --> 00:15:58,440 Speaker 1: going to be large. And so that may be that 262 00:15:59,360 --> 00:16:03,080 Speaker 1: you know, how everyone today is nostalgic for Reagan, you know, 263 00:16:04,200 --> 00:16:06,480 Speaker 1: there may be a day when we're nostalgic for Trump. 264 00:16:07,400 --> 00:16:13,560 Speaker 1: That our polarization has really continued to increase because the 265 00:16:13,680 --> 00:16:18,160 Speaker 1: rate of change has continued to increase. So that's why 266 00:16:18,400 --> 00:16:21,440 Speaker 1: I think it's so important that leaders like all of 267 00:16:21,440 --> 00:16:25,040 Speaker 1: you are thinking about how do we build bonds, how 268 00:16:25,040 --> 00:16:28,240 Speaker 1: do we have people of Americans care about each other 269 00:16:29,200 --> 00:16:33,640 Speaker 1: so that we can keep the society coherent and caring 270 00:16:34,400 --> 00:16:38,520 Speaker 1: despite the rapid amounts of change in generated by technology. 271 00:16:38,600 --> 00:16:40,480 Speaker 2: I want to come back to something you referred to 272 00:16:40,480 --> 00:16:42,360 Speaker 2: just explore a bit more, and that is how are 273 00:16:42,360 --> 00:16:44,600 Speaker 2: we going to pay for it? There is so much 274 00:16:44,640 --> 00:16:47,240 Speaker 2: money going into data center's investment right now, and to 275 00:16:47,280 --> 00:16:49,360 Speaker 2: something said, I think it's been supporting the markets because 276 00:16:49,400 --> 00:16:50,480 Speaker 2: there's been so much investment. 277 00:16:51,320 --> 00:16:51,920 Speaker 3: How do you get a. 278 00:16:51,920 --> 00:16:54,560 Speaker 2: Return on that investment without laying off an awful lot 279 00:16:54,560 --> 00:16:55,280 Speaker 2: of people. 280 00:16:55,960 --> 00:16:59,960 Speaker 1: With GDP growth, so you can either automate away the job. 281 00:17:00,320 --> 00:17:03,000 Speaker 1: That's one theory, and the other is that people will 282 00:17:03,040 --> 00:17:06,399 Speaker 1: produce more and I'm sure it'll be a mix, but 283 00:17:06,520 --> 00:17:10,639 Speaker 1: efficiency often generates more growth. I mean, we did the 284 00:17:10,720 --> 00:17:14,119 Speaker 1: radiology example, so it's sort of that at larger scale. 285 00:17:14,440 --> 00:17:17,520 Speaker 1: So the way we pay for it is more GDP growth. 286 00:17:17,800 --> 00:17:20,919 Speaker 2: You can do GDP growth with fewer people. And so 287 00:17:21,040 --> 00:17:22,760 Speaker 2: what happens to the people who are out of work 288 00:17:22,880 --> 00:17:25,440 Speaker 2: or have to take much more menial jobs than they 289 00:17:25,440 --> 00:17:27,359 Speaker 2: had before because it's the knowledge workers. 290 00:17:27,560 --> 00:17:30,480 Speaker 1: Well, I think it's a great question. Compared to globalization. 291 00:17:30,560 --> 00:17:33,040 Speaker 1: Probably this is a room full of globalists. We believe 292 00:17:33,080 --> 00:17:36,920 Speaker 1: in the benefits of trade, and yet despite that that 293 00:17:37,000 --> 00:17:40,000 Speaker 1: really did deliver in terms of the overall economy. There 294 00:17:40,080 --> 00:17:43,600 Speaker 1: was enough dislocation and enough of the countries that our 295 00:17:43,680 --> 00:17:48,080 Speaker 1: politics has been shifted. So I think there probably will 296 00:17:48,119 --> 00:17:51,960 Speaker 1: be things like that because of AI, But at least 297 00:17:52,000 --> 00:17:56,600 Speaker 1: it's not again, it's not a mass layoff scenario. Most likely, 298 00:17:56,680 --> 00:18:00,159 Speaker 1: it's much more likely that there's a growth response. We 299 00:18:00,280 --> 00:18:05,000 Speaker 1: also see significant GDP growth, you know, increased beyond what 300 00:18:05,040 --> 00:18:08,879 Speaker 1: we typically see because of the productivity of AI. 301 00:18:09,440 --> 00:18:12,640 Speaker 2: What I hear from you is AI has enormous potential, 302 00:18:13,080 --> 00:18:15,399 Speaker 2: a lot of it for good, some of for decidedly 303 00:18:15,560 --> 00:18:19,119 Speaker 2: not good. It's very complicated, it's coming very very fast. 304 00:18:19,800 --> 00:18:24,040 Speaker 2: Where will the leadership in directing AI come from? Are 305 00:18:24,080 --> 00:18:26,439 Speaker 2: there people who understand this, who get it, who have 306 00:18:26,480 --> 00:18:27,879 Speaker 2: the right values, who can help us go in the 307 00:18:27,960 --> 00:18:30,560 Speaker 2: right direction, particularly given how fast it's coming. 308 00:18:33,080 --> 00:18:35,480 Speaker 1: I would say that's an emerging area. I mean, I 309 00:18:35,520 --> 00:18:38,720 Speaker 1: think this was likely to be a big political issue 310 00:18:38,760 --> 00:18:43,960 Speaker 1: in the next couple cycles because many Americans will be 311 00:18:44,000 --> 00:18:47,760 Speaker 1: concerned about world changing too fast and too much, and 312 00:18:47,840 --> 00:18:51,199 Speaker 1: I think it's up to our politicians to sort of 313 00:18:51,359 --> 00:18:56,880 Speaker 1: understand that and channel it in some productive way. There's 314 00:18:56,960 --> 00:19:00,640 Speaker 1: leaders in the industry like Jeffrey Hinton that spends full 315 00:19:00,680 --> 00:19:03,600 Speaker 1: time now on these how do we keep humans at 316 00:19:03,600 --> 00:19:08,560 Speaker 1: the center of the system. So there definitely are emerging 317 00:19:08,640 --> 00:19:11,639 Speaker 1: leaders in that way. But again it's a new area. 318 00:19:11,960 --> 00:19:17,119 Speaker 1: Like you know, nuclear energy you know, or other DNA 319 00:19:17,280 --> 00:19:19,080 Speaker 1: for long time is very controversial. 320 00:19:19,880 --> 00:19:23,040 Speaker 2: Timing is everything, and it comes. The develops you talk 321 00:19:23,080 --> 00:19:25,360 Speaker 2: about come at a time when there's a decided rise 322 00:19:25,400 --> 00:19:27,520 Speaker 2: in populism in the United States as well as in 323 00:19:27,600 --> 00:19:30,840 Speaker 2: much of the Western world. Right now, the early things 324 00:19:30,920 --> 00:19:33,520 Speaker 2: I'm hearing this but on political from the people is 325 00:19:34,200 --> 00:19:37,760 Speaker 2: it increases our energy costs and it reduces our jobs. 326 00:19:38,200 --> 00:19:40,920 Speaker 2: We're against it, and some politicians are showing signs now 327 00:19:40,920 --> 00:19:43,600 Speaker 2: of trading on that and saying we should be against 328 00:19:43,640 --> 00:19:46,000 Speaker 2: Actually data centers, we should be against them. How do 329 00:19:46,040 --> 00:19:48,600 Speaker 2: you put together the rise of populism with the likely 330 00:19:48,640 --> 00:19:50,040 Speaker 2: effects of general AI. 331 00:19:51,119 --> 00:19:54,320 Speaker 1: Well, you know, the heart of democracy is we're going 332 00:19:54,400 --> 00:19:56,800 Speaker 1: to live in this country together and we're going to 333 00:19:56,840 --> 00:19:59,480 Speaker 1: have different views, and so you want a system that's 334 00:19:59,480 --> 00:20:03,479 Speaker 1: not civil war to work out those views. And I 335 00:20:03,480 --> 00:20:05,800 Speaker 1: think I think a lot of people will have those concerns. 336 00:20:05,800 --> 00:20:08,840 Speaker 1: They'll be those effects, and we have to take them 337 00:20:08,960 --> 00:20:12,920 Speaker 1: thoughtfully and seriously. And arguably during the era of globalization, 338 00:20:13,040 --> 00:20:16,840 Speaker 1: there was sort of lip service to retraining, but we 339 00:20:16,880 --> 00:20:19,919 Speaker 1: didn't really understand or take seriously the devastation of a 340 00:20:19,960 --> 00:20:23,480 Speaker 1: lot of communities. So I think, you know, the elites 341 00:20:23,480 --> 00:20:25,720 Speaker 1: in this room as an example of do a better 342 00:20:25,840 --> 00:20:28,400 Speaker 1: job at that, which is, you know, where is AI 343 00:20:28,560 --> 00:20:32,280 Speaker 1: really providing benefits to people's healthcare, to their day to 344 00:20:32,320 --> 00:20:34,600 Speaker 1: day lives, and if we can do that, then there's 345 00:20:34,680 --> 00:20:38,680 Speaker 1: more toleration of the dislocation, which is as you said, 346 00:20:38,920 --> 00:20:42,000 Speaker 1: cost of electricity, data centers, those things. 347 00:20:42,200 --> 00:20:43,320 Speaker 3: Are we over investing. 348 00:20:47,200 --> 00:20:51,200 Speaker 1: It's unlikely, but you know it's definitely possible. But again, 349 00:20:51,800 --> 00:20:54,520 Speaker 1: you know the Telkom boom in two thousand, I mean, 350 00:20:54,560 --> 00:20:57,879 Speaker 1: you know it's a slight over investment, you know, so 351 00:20:59,600 --> 00:21:01,679 Speaker 1: it could. But if you look at the in the 352 00:21:01,880 --> 00:21:04,399 Speaker 1: in the telecom one, everyone got so excited about the 353 00:21:04,400 --> 00:21:08,840 Speaker 1: Internet that they forward invested and were disappointed for a while, 354 00:21:08,920 --> 00:21:12,320 Speaker 1: but ultimately the Internet really did deliver. But if you 355 00:21:12,320 --> 00:21:15,320 Speaker 1: look at the mobile phone, it never had a bubble. 356 00:21:15,440 --> 00:21:17,280 Speaker 1: It came out and just grew and grew and grew 357 00:21:17,280 --> 00:21:19,960 Speaker 1: in impact. So you can get both things that live 358 00:21:20,080 --> 00:21:23,040 Speaker 1: up to the hype and things where the hype gets 359 00:21:23,080 --> 00:21:26,640 Speaker 1: bigger than the actual technology in the short term. And 360 00:21:26,720 --> 00:21:28,680 Speaker 1: you know, the market's a good way to work that out. 361 00:21:28,680 --> 00:21:31,399 Speaker 3: On AI, does it matter who gets their first? 362 00:21:32,040 --> 00:21:34,560 Speaker 2: I mean, is this like beta and VHS where the 363 00:21:34,560 --> 00:21:36,560 Speaker 2: point that gets their first gets to really set this up. 364 00:21:36,640 --> 00:21:38,560 Speaker 1: You're in countries, it matters. 365 00:21:38,240 --> 00:21:41,000 Speaker 3: A lot, So US versus China take that correct. 366 00:21:41,400 --> 00:21:45,119 Speaker 1: If you're talking companies you know, I care a lot 367 00:21:45,119 --> 00:21:47,280 Speaker 1: because I'm on the board of Entropic. But from a 368 00:21:47,320 --> 00:21:51,440 Speaker 1: society standpoint, it probably just matters that the technology is 369 00:21:51,520 --> 00:21:53,200 Speaker 1: developed and deployed in great ways. 370 00:21:53,800 --> 00:21:55,840 Speaker 3: But it does matter in the US versus China. 371 00:21:56,720 --> 00:22:00,280 Speaker 1: Yeah, absolutely. I mean, if another country doesn't have to 372 00:22:00,280 --> 00:22:04,320 Speaker 1: be China right gets ahead of us. There are significant 373 00:22:04,359 --> 00:22:07,680 Speaker 1: scenarios referred to in a class of the singularity where 374 00:22:07,680 --> 00:22:10,480 Speaker 1: the AI gets so good that it rights the next 375 00:22:10,520 --> 00:22:13,800 Speaker 1: DAI and that AI is much better, and then that 376 00:22:13,840 --> 00:22:16,520 Speaker 1: writes the next day I And so you know, being 377 00:22:16,680 --> 00:22:21,040 Speaker 1: first by even just a year gives you an enormous advantage. 378 00:22:21,080 --> 00:22:24,640 Speaker 1: So that's the exponential. But again that's a theory that's 379 00:22:24,680 --> 00:22:29,240 Speaker 1: not yet proven, but even its slight possibility means we 380 00:22:29,440 --> 00:22:31,720 Speaker 1: better don't get there first. And I think we're on 381 00:22:31,760 --> 00:22:32,359 Speaker 1: a good bathroom. 382 00:22:32,440 --> 00:22:33,240 Speaker 3: Well that's my question. 383 00:22:33,280 --> 00:22:35,880 Speaker 2: If that were your one goal in the United States, 384 00:22:36,200 --> 00:22:38,679 Speaker 2: what are the policies that get you there? What are 385 00:22:38,720 --> 00:22:40,760 Speaker 2: the things we need to do or avoid doing to 386 00:22:40,800 --> 00:22:42,440 Speaker 2: make sure we're as competitive as possible. 387 00:22:43,440 --> 00:22:44,960 Speaker 1: I think we're doing a pretty good job on that. 388 00:22:45,200 --> 00:22:48,760 Speaker 1: I mean, there's you would push the edge on light regulation, 389 00:22:49,000 --> 00:22:52,679 Speaker 1: make it easy to build data centers. You know, I 390 00:22:52,720 --> 00:22:58,960 Speaker 1: think the government's done open field running, so I can't 391 00:22:58,960 --> 00:23:00,480 Speaker 1: think I don't I don't think we need like a 392 00:23:00,520 --> 00:23:05,040 Speaker 1: government investment program or something like that. So there's a 393 00:23:05,080 --> 00:23:09,000 Speaker 1: lot of progress, and we are ahead of our competitors 394 00:23:09,040 --> 00:23:12,240 Speaker 1: as far as we can tell. And again, we don't 395 00:23:12,240 --> 00:23:15,439 Speaker 1: want to make it zero some because you'd like, just 396 00:23:15,480 --> 00:23:18,040 Speaker 1: like we did in globalization, we would like to bring 397 00:23:18,080 --> 00:23:21,080 Speaker 1: the rest of the world along with us so that 398 00:23:21,119 --> 00:23:24,640 Speaker 1: we've got a stable we have less likely chance of war. 399 00:23:25,720 --> 00:23:30,080 Speaker 1: So I mean, but that's happening. 400 00:23:30,680 --> 00:23:32,320 Speaker 3: Okay, So let's talk about what you really love. 401 00:23:32,359 --> 00:23:39,000 Speaker 2: This has been very entertaining snow powder and mountain resorts. 402 00:23:39,720 --> 00:23:41,159 Speaker 3: This is what you really love, you know. 403 00:23:42,119 --> 00:23:49,040 Speaker 1: I have a slight addiction to skiing, and then retiring 404 00:23:49,080 --> 00:23:52,359 Speaker 1: from Netflix, rather than do a sailboat or something like that, 405 00:23:52,680 --> 00:23:56,320 Speaker 1: I decided to take over a ski resort and try 406 00:23:56,359 --> 00:23:59,320 Speaker 1: to make another Yellowstone Club, but one that was sort 407 00:23:59,320 --> 00:24:02,639 Speaker 1: of more artists, Dick, and I'm a visitor else don 408 00:24:02,720 --> 00:24:06,000 Speaker 1: not a member. But it's a great place. But to 409 00:24:06,040 --> 00:24:08,119 Speaker 1: do something like that, yeah, So I've been learning the 410 00:24:08,119 --> 00:24:11,960 Speaker 1: real estate business. It's a real turnaround. It's a very 411 00:24:12,000 --> 00:24:15,240 Speaker 1: different thing for me. But when you've got overwhelming capital 412 00:24:15,320 --> 00:24:20,639 Speaker 1: relative to our project, things are pretty easy. So I 413 00:24:21,000 --> 00:24:24,640 Speaker 1: can't say it's it's it's gonna work out great for customers. 414 00:24:24,920 --> 00:24:27,160 Speaker 1: I'm not sure it'll be the best investment I ever made, 415 00:24:27,520 --> 00:24:29,920 Speaker 1: but I do love it. But you get to snowboard 416 00:24:29,920 --> 00:24:32,040 Speaker 1: a lot, but I get to snowboard all winter like 417 00:24:32,800 --> 00:24:34,920 Speaker 1: Powder Mountain. Utah's fantastic place. 418 00:24:35,080 --> 00:24:37,320 Speaker 3: Check it out. Thank you, thank you so much. 419 00:24:37,359 --> 00:24:39,720 Speaker 1: Really what an honor. Thank you all, thank you all.