WEBVTT - Big Tech’s Most Valuable Product Isn’t AI. It’s the Belief that AI is Inevitable - The Story 0:00:15.960 --> 0:00:19.040 Welcome to tech Stuff. I'm Os Voloshen. If you've been 0:00:19.040 --> 0:00:21.640 listening to this podcast, or if you spend any time 0:00:21.680 --> 0:00:23.880 on the Internet, you won't be able to miss the 0:00:23.880 --> 0:00:27.360 word in shitification. And our guest today is the man 0:00:27.440 --> 0:00:31.240 who coined it, Corey doctor Ol. He's a science fiction author, 0:00:31.440 --> 0:00:34.920 a technology activist, and a journalist. His new book is 0:00:34.960 --> 0:00:39.120 called The Reverse Centaur's Guide to Life After AI, How 0:00:39.120 --> 0:00:42.920 to Think about Artificial Intelligence before It's too late. Corey, 0:00:42.920 --> 0:00:43.839 Welcome to text Stuff. 0:00:44.159 --> 0:00:45.760 Thanks as it's a pleasure to be on. 0:00:46.600 --> 0:00:48.960 For anyone who did miss it. What is in ghentification? 0:00:49.320 --> 0:00:52.600 Well, I've worked for the Electronic Frontier Foundation, which is 0:00:52.840 --> 0:00:56.400 an NGO, for twenty five years on the question digital rights, 0:00:56.960 --> 0:00:59.360 and getting people to engage with that question is very 0:00:59.360 --> 0:01:03.000 hard because digital rights questions are abstract and technical and 0:01:03.080 --> 0:01:04.800 relate to things that are going to happen in the future, 0:01:04.880 --> 0:01:07.120 and for very good reasons, people mostly care about concrete 0:01:07.120 --> 0:01:12.320 things happening right now. And what I have done over 0:01:12.360 --> 0:01:14.600 the years has come up with different framing devices and 0:01:14.640 --> 0:01:17.840 similes and metaphors and narratives too. I'm a science fiction novelist, 0:01:18.080 --> 0:01:21.360 but it turned out that a dirty word initification was 0:01:21.400 --> 0:01:22.880 the way to get people to engage in it, and 0:01:22.920 --> 0:01:28.679 so more specifically in sitification is this theory and description. 0:01:29.160 --> 0:01:31.679 It describes how the platforms that we rely on have 0:01:31.720 --> 0:01:34.840 turned into piles of shit, where first they are good 0:01:34.840 --> 0:01:36.960 to their end users but lock them in, and then 0:01:37.200 --> 0:01:39.400 having locked in the end users and made it difficult 0:01:39.400 --> 0:01:41.000 for them to leave, they make it worse for them 0:01:41.040 --> 0:01:43.480 to make it tempting for businesses. And then when the 0:01:43.480 --> 0:01:47.119 businesses are also locked in, they also extract everything they 0:01:47.120 --> 0:01:50.080 can from those businesses and line their own pockets and 0:01:50.120 --> 0:01:52.760 turn into a pile of shit. But the more interesting 0:01:52.800 --> 0:01:54.800 part of it is why they're doing it, And this 0:01:54.880 --> 0:01:57.600 maybe relates to the conversation we're having today, which is 0:01:57.600 --> 0:02:01.880 about the economics and political economy of monopoly, that when 0:02:01.880 --> 0:02:04.200 you let firms get too big to fail, they become 0:02:04.240 --> 0:02:06.120 too big to jail, and then that makes them too 0:02:06.120 --> 0:02:09.680 big to care, and they destroy our lives for the 0:02:09.680 --> 0:02:12.120 same reason your dog looks its balls because they can 0:02:12.160 --> 0:02:13.240 and no one makes them start. 0:02:13.919 --> 0:02:15.240 I want to get into all of that with you, 0:02:15.320 --> 0:02:17.880 but I've read the acknowledgments of your book, not just 0:02:17.880 --> 0:02:21.040 the acknowledgements. If I'm always interesting acknowledgments, and you had 0:02:21.080 --> 0:02:24.359 the same editor for this book, this new book, the 0:02:24.480 --> 0:02:27.520 Reverse Centaur's Guide, as you did for last year's in 0:02:27.560 --> 0:02:31.920 Certification book, and you kind of credit him with helping you, 0:02:32.400 --> 0:02:34.640 not just with in Certification argument, but with this argument. 0:02:34.639 --> 0:02:38.520 And I'm curious, like, where does the reverse central sit 0:02:39.120 --> 0:02:43.160 in the lineage of inertifications? What is it a development 0:02:43.200 --> 0:02:46.040 of the same argument or is it totally new argument? 0:02:46.040 --> 0:02:49.320 Because in Certification was essentially about the tech platforms of 0:02:49.639 --> 0:02:53.000 the Internet, and this is about the experience of AI. 0:02:53.760 --> 0:02:58.240 The common lineage between in Certification and my critique of 0:02:58.240 --> 0:03:03.680 AI is grounded in the dysfunctions and pathologies of firms 0:03:03.720 --> 0:03:08.959 that conquer their markets of saturate their markets. Monopolies, cartels doopolies. 0:03:10.240 --> 0:03:15.560 And one of the consequences of having monopoly is that 0:03:15.600 --> 0:03:17.920 you become ungovernable in the worst sense, not in the 0:03:17.919 --> 0:03:21.160 cool sense of like a raccoon, but in the bad 0:03:21.200 --> 0:03:26.760 sense of like Donald Trump. And you are so liberated 0:03:26.800 --> 0:03:31.000 from consequence that you can do unlimited bad things to 0:03:31.120 --> 0:03:33.280 lots of people, including people you would think would have 0:03:33.320 --> 0:03:36.840 some power in the economy, and just get away with it. 0:03:36.880 --> 0:03:40.040 And really the worst ideas of the worst people become 0:03:40.720 --> 0:03:44.040 the most profitable way of doing business. So that's like 0:03:44.160 --> 0:03:49.240 in shitification at its theoretical core. But reverse centaur is 0:03:49.240 --> 0:03:54.360 about another epiphenomenon, another thing that a consequence of monopoly, 0:03:55.040 --> 0:03:57.800 which is that if your firm that has saturated its 0:03:57.840 --> 0:04:01.840 market right Google has a ninety percent market share, then 0:04:02.240 --> 0:04:06.560 you can't grow anymore. And you know, there's this saying 0:04:06.840 --> 0:04:11.120 among leftists and crunchy granola types that endless growth is 0:04:11.120 --> 0:04:15.920 the ideology of a tumor, and that grounds it in 0:04:15.960 --> 0:04:18.839 this idea that the reason firms want to continue growing 0:04:18.839 --> 0:04:22.400 as ideological and not practical or concrete. And while there 0:04:22.440 --> 0:04:26.280 is some ideology at work there in the way that 0:04:26.320 --> 0:04:29.919 we think about our economy and in the environmental consequences 0:04:29.920 --> 0:04:33.279 of that, there are very practical, material reasons that companies 0:04:33.360 --> 0:04:36.839 want to be perceived as growing and not as mature 0:04:37.200 --> 0:04:39.960 and at the end of their growth cycle. And that 0:04:40.120 --> 0:04:43.120 is because the share price of a company that is 0:04:43.120 --> 0:04:47.719 growing is much higher based on the total economic turnover 0:04:47.839 --> 0:04:50.760 of that company than the share price of affirm with 0:04:50.800 --> 0:04:55.120 the same turnover. But that is static, and that's not irrational. 0:04:55.680 --> 0:04:58.920 A share and affirm is a claim on its future earnings. 0:05:00.040 --> 0:05:03.600 Firms that are growing have more future earnings to return 0:05:03.720 --> 0:05:08.039 to you than firms that are not growing. And the 0:05:08.240 --> 0:05:11.680 corollary of this is that when your firm stops growing, 0:05:11.680 --> 0:05:16.599 it becomes overvalued because the future earnings have contracted when 0:05:16.600 --> 0:05:19.400 you reach the top of your game and you stop growing, 0:05:19.960 --> 0:05:24.760 and that provokes these mass sell offs by investors. 0:05:25.080 --> 0:05:26.880 You mentioned, I think you mentioned that you know the 0:05:26.880 --> 0:05:30.360 Invidia cell of when the deep Seak moment happened. Obviously, Yeah, 0:05:30.400 --> 0:05:34.359 in video rebounded from that, but it raised the it 0:05:34.480 --> 0:05:37.520 raised a specter that growth may not be unlimited. 0:05:37.760 --> 0:05:41.159 Yeah, the video sell off was the largest decapitalization in 0:05:41.200 --> 0:05:44.320 twenty four hours of any firm. And when your firm 0:05:44.360 --> 0:05:47.560 becomes illiquid, right when it's not, when it's not traded 0:05:47.600 --> 0:05:50.240 at this high multiple and so liquid, you can't grow 0:05:50.279 --> 0:05:52.599 by buying other firms by offering them stock. You have 0:05:52.640 --> 0:05:55.920 to give them money. And money is an exogenous substance 0:05:55.920 --> 0:05:58.080 that's produced outside the firm, and shares are things that 0:05:58.080 --> 0:06:00.080 are produced inside the firm by typing zeros and to 0:06:00.080 --> 0:06:05.280 a spreadsheet. And so it's always preferable to be able 0:06:05.320 --> 0:06:10.479 to acquire firms and talent using shares. This endogenous substance 0:06:10.480 --> 0:06:13.039 you can produce on demand, and you can only do 0:06:13.080 --> 0:06:14.839 that while you're growing. So this is another kind of 0:06:14.839 --> 0:06:16.680 growth paradox, right, which is that it's easy to keep 0:06:16.680 --> 0:06:19.560 growing when you're growing because you can buy other companies, 0:06:19.839 --> 0:06:22.320 but it's hard when you stop growing to start growing again, 0:06:22.880 --> 0:06:25.320 not least because all those key employees that have been 0:06:25.320 --> 0:06:28.320 paid in shares suddenly see their net worthfall off a 0:06:28.360 --> 0:06:31.000 cliff when you when your share price declines, and they 0:06:31.080 --> 0:06:33.240 might go look for work elsewhere, which means they're not 0:06:33.279 --> 0:06:35.240 going to help you start the firm up again. They're 0:06:35.240 --> 0:06:36.359 not going to help you get a back on a 0:06:36.360 --> 0:06:37.200 growth trajectory. 0:06:37.800 --> 0:06:40.880 So that the central argument of the book is that 0:06:41.520 --> 0:06:43.880 a lot of technology companies order the practice order. The 0:06:43.880 --> 0:06:49.120 technology companies have an extraordinary incentive to persuade the white 0:06:49.279 --> 0:06:53.039 world that in a AI is inevitable, and that that 0:06:53.120 --> 0:06:57.720 kind of mythology is of the essence to protecting their 0:06:57.800 --> 0:07:00.560 value and therefore keeping their employees and their ability to 0:07:00.640 --> 0:07:03.359 keep growing by buying other companies with their stock. 0:07:04.120 --> 0:07:07.520 Yes, So what I would say is that the book 0:07:07.600 --> 0:07:10.120 is exploring why tech bubbles exist at all. And this 0:07:10.240 --> 0:07:12.520 is not the first one we've had, you know, metaverse 0:07:12.560 --> 0:07:15.720 and cryptocurrency and web three, all those other things. 0:07:16.080 --> 0:07:20.640 But they were all palpably confections, right, those those were 0:07:20.680 --> 0:07:24.520 not things that people use, whereas, like all you do 0:07:24.640 --> 0:07:27.000 is any conversation you over hear in the subway, in 0:07:27.000 --> 0:07:29.040 a restaurant, whatever, is people telling each other how they 0:07:29.160 --> 0:07:31.720 use AI. So that arguably is a big difference. 0:07:31.800 --> 0:07:34.360 So it's just so those bubbles were smaller. Why is 0:07:34.400 --> 0:07:36.800 the a bubble larger? And some of it is because 0:07:36.840 --> 0:07:40.520 AI is realer than those things. So I'm a fake 0:07:40.600 --> 0:07:43.720 computer scientist. I have an honorary doctor in computer science 0:07:43.720 --> 0:07:46.960 from the Open University, and in my capacity as a 0:07:47.000 --> 0:07:49.640 fake computer scientist, I can tell you that AI is 0:07:49.680 --> 0:07:50.880 pretty interesting. 0:07:50.640 --> 0:07:52.720 And from a user point of view, I would argue 0:07:53.400 --> 0:07:56.280 a feeling of technology magic for the first time in 0:07:56.320 --> 0:07:56.960 a long time. 0:07:57.240 --> 0:08:02.160 Sure, yeah, no, very impressive, although I would also say 0:08:02.200 --> 0:08:05.560 that that feeling regresses to the mean pretty quickly. Right 0:08:05.640 --> 0:08:08.360 that if there was a place where there was like 0:08:08.400 --> 0:08:11.080 some kind of weird localized gravity storm, and you took 0:08:11.080 --> 0:08:12.600 a big pile of leaves and you threw it in 0:08:12.600 --> 0:08:14.960 the air and it fell down, in a sentence, we 0:08:15.000 --> 0:08:17.400 would all go throw leaves in the air and look 0:08:17.440 --> 0:08:20.920 at the sentences they made for a while, and then 0:08:20.960 --> 0:08:23.320 we'd be like, they're just not good sentences. 0:08:25.080 --> 0:08:29.040 Going back to the Internet bubble, obviously Internet bubble crash, 0:08:29.160 --> 0:08:34.240 but then very shortly afterwards, Facebook and Google and Amazon emerged. 0:08:34.280 --> 0:08:36.880 So even if there is like a big froth cycle 0:08:36.920 --> 0:08:42.319 around some of the valuations and companies and new billionaires 0:08:42.320 --> 0:08:45.080 who are emerging, I mean, if it's a bubble like 0:08:45.280 --> 0:08:47.920 the web, it maybe a short term financial bubble that 0:08:47.960 --> 0:08:50.600 will burst, but a fundamental platform shift in terms of 0:08:50.600 --> 0:08:52.920 how we interact with technology in the world, Like do 0:08:52.960 --> 0:08:55.600 you believe that will be true of AI or didn't 0:08:55.640 --> 0:08:58.000 the AI in more similar to web three or to 0:08:58.040 --> 0:08:58.520 the metaverse? 0:08:58.600 --> 0:09:01.559 Yeah? Sure, I think AI has some similarities to the 0:09:01.600 --> 0:09:04.000 Web bubble as compared to other bubbles of the day. 0:09:04.040 --> 0:09:06.200 So you know, a bubble that was roughly concurrent with 0:09:06.240 --> 0:09:10.319 the Web bubble was the Enron bubble, right, the energy 0:09:10.360 --> 0:09:12.920 trading bubble, and that was just accounting fraud, and when 0:09:12.920 --> 0:09:15.480 it was over there was nothing but like indictments and 0:09:15.520 --> 0:09:18.199 a language corpus because they couldn't be arked to pay 0:09:18.200 --> 0:09:20.960 a lawyer to redact their emails before they submitted them 0:09:21.000 --> 0:09:24.400 for discovery, whereas the Web left behind like a couple 0:09:24.520 --> 0:09:28.240 million humanities undergraduates who'd been inveigled to drop out and 0:09:28.440 --> 0:09:31.920 become Pearl and python in HTML jocks, and so you know, 0:09:32.160 --> 0:09:35.679 you say that what succeeded the dot com bubble was 0:09:35.720 --> 0:09:39.560 Amazon and Google and Facebook and so on, but there 0:09:39.600 --> 0:09:42.440 was an intermediate step there, and the intermediate step was 0:09:42.480 --> 0:09:45.319 Web too. It was a ton of little amazing startups 0:09:45.720 --> 0:09:49.600 that were what you got when skill practitioners were liberated 0:09:49.640 --> 0:09:52.240 from the need to indulge the foolish fantasies of their 0:09:52.280 --> 0:09:55.679 bosses who were very good at getting capital and very 0:09:55.720 --> 0:09:58.760 bad at coming up with products. So that was pretty amazing. 0:09:59.520 --> 0:10:02.840 Although we shoul it's say that the Enron bubble, the 0:10:02.880 --> 0:10:06.840 Web bubble, and the AI bubble foundationally are not about 0:10:06.840 --> 0:10:09.600 what they produce. They're about separating suckers from their money, 0:10:10.040 --> 0:10:12.000 right that like the point of a bubble is to 0:10:12.160 --> 0:10:17.040 get insiders to cash out and leave ordinary investors holding 0:10:17.080 --> 0:10:19.440 the bag. So no bubble is good, but some bubbles 0:10:19.440 --> 0:10:23.240 have productive residues. So in that sense, AI, I think 0:10:23.360 --> 0:10:25.600 is like the Web in that we're going to have 0:10:25.720 --> 0:10:27.760 a lot of GPUs at ten cents on the dollar, 0:10:28.040 --> 0:10:30.360 a lot of skilled practitioners looking for work and maybe 0:10:30.559 --> 0:10:33.680 hiring each other. And also these open source models that 0:10:33.760 --> 0:10:36.280 will continue to function whether or not there's a company 0:10:36.320 --> 0:10:40.199 that produce them, because open source software continues to exist 0:10:40.200 --> 0:10:42.640 for as long as it exists. But there's a big difference, 0:10:43.280 --> 0:10:47.240 actually two. So the first is the vibe. So if 0:10:47.240 --> 0:10:49.439 you go back and you read like Harvard Business Review 0:10:49.520 --> 0:10:55.320 articles or NBER reports on the workforce in the era 0:10:55.440 --> 0:10:57.720 of the web, what you see are all these articles 0:10:57.760 --> 0:11:01.520 about like bosses who have to be dragkicking and screaming 0:11:02.240 --> 0:11:05.360 into allowing their workers to use a technology that they 0:11:05.360 --> 0:11:08.520 see as transformative. And when you look at comparable reports today, 0:11:08.520 --> 0:11:11.000 it's full of bosses threatening to fire workers if they 0:11:11.040 --> 0:11:13.760 don't use AI. Right. This is now getting to the 0:11:13.760 --> 0:11:16.600 material way in which the web is different from AI. 0:11:16.679 --> 0:11:19.439 The unit economics of the web were very good. Every 0:11:19.520 --> 0:11:21.760 new user of the web made the web more profitable. 0:11:22.080 --> 0:11:24.640 Every new use of the web made the web more profitable, 0:11:24.760 --> 0:11:27.200 and every generation of the web was more profitable than 0:11:27.200 --> 0:11:29.760 the previous generation. That's the opposite of AI. Right, So 0:11:29.840 --> 0:11:32.600 AI companies are selling hundred dollars bills for a dollar apiece. 0:11:33.200 --> 0:11:36.080 So if you become an AI customer, they start to 0:11:36.120 --> 0:11:38.679 lose money. If you continue to be an AI customer, 0:11:38.760 --> 0:11:41.760 they lose more money. And the next generation of AI 0:11:41.880 --> 0:11:44.240 is going to be selling two one hundred dollars bills 0:11:44.240 --> 0:11:47.559 for a dollar apiece, and so it's going to be worse. 0:11:47.679 --> 0:11:51.439 And so that makes it a qualitatively and quantitatively different 0:11:51.720 --> 0:11:53.560 kind of bubble to the web bubble. 0:11:54.000 --> 0:11:56.319 So what's your role in all of this? And I 0:11:56.320 --> 0:11:59.480 think you mentioned the book or in your speech that 0:11:59.520 --> 0:12:03.080 you gave Washington that you know, as a science fiction 0:12:03.240 --> 0:12:05.360 right to people always want to ask you if your 0:12:05.360 --> 0:12:07.360 opinion on AI because I guess so many of the 0:12:07.360 --> 0:12:11.920 AI overlords talk about how science fiction governs their thoughts 0:12:11.960 --> 0:12:14.840 about what they're building. You know, you've said as well, 0:12:14.960 --> 0:12:18.800 science fiction is anti inevitableist literature, as it was a 0:12:18.840 --> 0:12:21.640 great phrase. I mean, what is what is your role 0:12:21.679 --> 0:12:23.120 in this? What do you what do you want people 0:12:23.360 --> 0:12:24.800 to understand to do differently? 0:12:25.520 --> 0:12:28.880 Yeah, I think that the people who run a bubble, 0:12:30.040 --> 0:12:32.240 whether or not it has a productive residue, you always 0:12:32.280 --> 0:12:35.000 want to style themselves as avatars of the great forces 0:12:35.040 --> 0:12:38.160 of history and the iron laws of economics. Right, they're 0:12:38.240 --> 0:12:43.280 not imposing a technology on you. Rather, they are just 0:12:43.400 --> 0:12:46.040 doing what the circumstances demand. And if it wasn't them, 0:12:46.040 --> 0:12:48.040 it'd be someone else, although it has to be them 0:12:48.080 --> 0:12:50.800 because they're extraordinary. But but you know, this is the 0:12:50.800 --> 0:12:52.680 moment at which we are going to get AI and 0:12:52.679 --> 0:12:54.040 we're all going to use AI, and there is no 0:12:54.120 --> 0:12:56.200 future without AI, and the only II that we're going 0:12:56.200 --> 0:12:57.960 to have in the future is the AI they're making, 0:12:59.000 --> 0:13:03.280 and everybody else just shut up. And I call this 0:13:03.600 --> 0:13:08.080 a kind of species of vulgar Thatcherism, because you know 0:13:08.160 --> 0:13:11.240 Margaret Thatcher, she had this aphorism, there is no alternative, 0:13:11.640 --> 0:13:15.360 and this is the mother of all thoughts stopping cliches, 0:13:15.920 --> 0:13:18.400 because what she meant by there is no alternative is 0:13:18.400 --> 0:13:21.120 don't you dare try and think of an alternative? Right. 0:13:21.200 --> 0:13:24.280 Another way of saying there is no alternative is resistance 0:13:24.360 --> 0:13:29.520 is futile, but there is no alternative implies Also, if 0:13:29.559 --> 0:13:32.840 you're unhappy with this, don't get mad at me. Get 0:13:32.880 --> 0:13:34.960 mad at the iron laws of economics and the great 0:13:35.000 --> 0:13:37.280 forces of history, because it could be no other way, 0:13:38.040 --> 0:13:42.160 and science fiction is entirely concerned with all the other 0:13:42.200 --> 0:13:48.080 ways things could be right and in challenging the idea 0:13:48.600 --> 0:13:51.559 that everything is inevitable and nothing is contingent. 0:13:52.160 --> 0:13:55.480 So what's the alternative vision you're laying out for what 0:13:55.520 --> 0:13:58.080 the future may be? I mean, right now it looks like, 0:13:58.880 --> 0:14:04.360 you know, three trillion dollar IPOs, this year's basic exentropic 0:14:05.040 --> 0:14:09.840 open AI. You know, new trillionaires and billionaires and multimillionaires, 0:14:11.360 --> 0:14:15.240 all of whom will then have the capital. Well, the 0:14:15.640 --> 0:14:18.920 share evaluation is to quiem more companies the cash money 0:14:18.960 --> 0:14:25.880 to support politicians and packs like. What might disrupt the 0:14:25.920 --> 0:14:27.720 futures is currently emerging. 0:14:27.720 --> 0:14:31.400 So I can't predict when the bubble will pop. But 0:14:31.640 --> 0:14:34.840 firms that are making tens of billions of dollars a 0:14:34.960 --> 0:14:37.360 year and spending hundreds of billions of dollars a year 0:14:37.400 --> 0:14:39.600 and that are telling us they need to spend more 0:14:39.640 --> 0:14:42.040 hundreds of billions of dollars a year, but whenever they 0:14:42.040 --> 0:14:44.240 try to raise prices. I mean, Anthropic just tried to 0:14:44.280 --> 0:14:46.360 clean up its balance sheet a little by saying we're 0:14:46.360 --> 0:14:48.600 not selling hundred dollar bills for a dollar apiece anymore. 0:14:48.640 --> 0:14:50.760 They're going to be five dollars apiece, and all their 0:14:50.760 --> 0:14:52.840 best customers are like, I guess we can't use AI 0:14:52.960 --> 0:14:55.640 anymore because at five dollars, these hundred dollar bills are 0:14:55.640 --> 0:15:00.000 not worth it. So when they run out of peace, 0:15:00.000 --> 0:15:04.520 people who will give the money. Because listed are privately held, 0:15:05.080 --> 0:15:07.240 you can't spend more than your taking in unless you 0:15:07.280 --> 0:15:10.280 have a creditor. Right, it doesn't matter how big your 0:15:10.320 --> 0:15:15.200 IPO was, Right, you have operating expenses. I mean, I 0:15:15.240 --> 0:15:18.320 guess you could just like loan the company money from 0:15:18.400 --> 0:15:20.520 your own shares, but you know this kicks off the 0:15:21.440 --> 0:15:25.480 crunch where you flog your shares to get operating capital 0:15:25.520 --> 0:15:28.160 for the company. The supply of shares goes up, the 0:15:28.200 --> 0:15:29.840 value of the shares goes down, and you're in a 0:15:29.880 --> 0:15:32.560 death spiral. So you know, I don't know what's going 0:15:32.600 --> 0:15:37.560 to shatter this fragile system we're in. Maybe it's like 0:15:37.960 --> 0:15:41.800 Golf States say, sorry, Dario, I know you told us 0:15:41.800 --> 0:15:43.960 that we were going to make God here, but we've 0:15:44.000 --> 0:15:47.640 decided that we need to have LNG terminals because Donald 0:15:47.640 --> 0:15:50.920 Trump keeps getting the Iranians to blow ours up, and 0:15:51.040 --> 0:15:53.520 we don't think we have a future without LNG terminals, 0:15:53.560 --> 0:15:56.600 even if we are going to own eight percent of God, 0:15:56.840 --> 0:16:00.560 and then, like without the golf, you lose the the 0:16:00.560 --> 0:16:03.240 rest of the financing collapses. Who knows, right, it could 0:16:03.280 --> 0:16:05.760 be any or all of the above, but anything that 0:16:05.760 --> 0:16:07.440 can't go on forever eventually stops. 0:16:08.000 --> 0:16:10.720 I think, I think, I guess that the pushback would 0:16:10.720 --> 0:16:14.240 be to say, well, maybe right, but you know, if 0:16:14.280 --> 0:16:18.000 there's there are other companies like Google, for example, who 0:16:18.080 --> 0:16:21.200 can continue to fund their plans which are very similar 0:16:21.200 --> 0:16:23.520 to Anthropic and opening eyes into SpaceX is because of 0:16:23.600 --> 0:16:27.480 their you know, they're they're they're profitably if they're operating 0:16:27.600 --> 0:16:31.480 operating business, and that like somebody may they all Elon 0:16:31.560 --> 0:16:33.960 may be able to convince the public and public markets 0:16:34.000 --> 0:16:36.560 believing for long enough that he can bring together space 0:16:36.640 --> 0:16:40.080 and robotics and data centers and all these types of things. 0:16:40.080 --> 0:16:42.080 I mean, I guess like the proof will being the 0:16:42.120 --> 0:16:46.120 pulling is it's just unknowable, but certainly that is what 0:16:46.360 --> 0:16:51.240 is clear is that the it's a historical disconnect between 0:16:51.280 --> 0:16:53.800 the financial performance of these companies and the evaluation and 0:16:54.000 --> 0:17:11.200 and and I mean, that's that's unarguable. After the break, 0:17:11.680 --> 0:17:14.480 what if a computer program decided whether or not you 0:17:14.600 --> 0:17:23.320 deserve a job stay with us. We sort of spent 0:17:23.400 --> 0:17:26.160 quite a lot of time I think setting the stakes, 0:17:26.240 --> 0:17:28.360 the backdrop. There's two more things I want to talk 0:17:28.400 --> 0:17:32.480 about while we're together. One is people and the other 0:17:32.600 --> 0:17:35.280 is politics. But we've had too much fun to define 0:17:35.359 --> 0:17:36.480 what a reverse centaur is. 0:17:38.240 --> 0:17:42.280 Yeah, well, I think that in a long history of automation, 0:17:43.040 --> 0:17:45.639 where automation is driven by workers, it tends to be 0:17:46.119 --> 0:17:50.680 about improving quality, whereas capital driven automation is tilted towards 0:17:50.720 --> 0:17:55.320 improving throughput. Not because bosses are mustache twirling villains, but 0:17:55.480 --> 0:17:57.840 because when you have an asset that's depreciating, you want 0:17:57.880 --> 0:17:59.000 to get as much use out of it as you 0:17:59.080 --> 0:18:02.400 can before it falls off balance sheet. Another aspect that's 0:18:02.440 --> 0:18:06.280 closely related in automation theory is the idea of a centaur, 0:18:07.000 --> 0:18:10.120 which is a person assisted by a machine. So that's 0:18:10.760 --> 0:18:13.480 you on a bicycle, you with a spell checker, you 0:18:13.600 --> 0:18:16.399 with an ide right, Like when I started programming, we 0:18:16.480 --> 0:18:19.200 didn't have debuggers, right, So like that makes you a 0:18:19.280 --> 0:18:24.639 centaur too, whereas a reverse centaur is a person conscripted 0:18:24.840 --> 0:18:27.760 to serve as a peripheral for a machine, and because 0:18:27.800 --> 0:18:32.920 of those dynamics of how capital views its investments, that 0:18:33.359 --> 0:18:36.879 person is usually the bottleneck in the machine's performance. Right. 0:18:37.000 --> 0:18:39.600 People generally can't work as long or as fast as 0:18:39.600 --> 0:18:42.600 a machine, and that person is now worked at the 0:18:42.800 --> 0:18:46.200 edge of their capability and to the edge of their endurance, 0:18:47.000 --> 0:18:49.800 because that's how you maximize the return on the hard 0:18:49.840 --> 0:18:53.320 asset that they are assisting. And again this is not 0:18:53.440 --> 0:18:55.680 a new idea, Like there's a reason that you know, 0:18:55.760 --> 0:18:58.159 whether you're looking at Charlie Chaplin in Modern Times or 0:18:58.520 --> 0:19:00.959 Lucille Ball in that episode where she and Ethel are 0:19:01.000 --> 0:19:03.560 putting chocolates in the chocolate box on the conveyor belt, 0:19:03.880 --> 0:19:08.760 that the motif of automation is the speed up, right, 0:19:08.920 --> 0:19:12.160 and it's it's the worker being used up by the machine. 0:19:13.000 --> 0:19:15.639 And you know, the warehouses that are most automated in 0:19:15.720 --> 0:19:18.800 America are Amazon's warehouses, and they're also the warehouses with 0:19:18.840 --> 0:19:21.879 the highest rate of significant injury. Those aren't coincidences. Those 0:19:21.920 --> 0:19:24.600 are like those are co determined, right, because if you're 0:19:24.600 --> 0:19:26.680 going to put seven figures into warehouse automation, you want 0:19:26.680 --> 0:19:28.520 to work the warehouse automation as quickly as you can. 0:19:29.080 --> 0:19:32.960 So an Amazon driver is there because even if you 0:19:33.000 --> 0:19:35.600 could get the van to drive itself in urban traffic, well, 0:19:36.320 --> 0:19:39.120 it couldn't get the parcel onto your porch or onto 0:19:39.200 --> 0:19:42.960 your mailbox or you know, into your office building, and 0:19:43.280 --> 0:19:45.639 so it needs to have a driver there, and so 0:19:45.800 --> 0:19:48.280 you work that driver at the outer limit of their 0:19:48.400 --> 0:19:52.480 ability and their endurance. And the reason people pee in 0:19:52.520 --> 0:19:55.760 the vans is because you cannot make the quota unless 0:19:55.840 --> 0:19:59.000 you operate at a speed that requires that you not 0:19:59.160 --> 0:19:59.800 have kidneys. 0:20:00.200 --> 0:20:01.640 I mean you said in the book that you get 0:20:01.720 --> 0:20:05.439 penalized for swerving even if you're avoiding Yeah, a danger. 0:20:05.600 --> 0:20:06.680 I mean that does that really true? 0:20:07.080 --> 0:20:10.120 Yeah? So Katie Wells just published a good ethnography. She's 0:20:10.119 --> 0:20:12.000 done a lot of work on Amazon drivers and on 0:20:12.080 --> 0:20:16.280 algorithmic wage discrimination and algorithmic management. She just published some 0:20:16.440 --> 0:20:20.000 ethnographies where she worked with another Amazon a former Amazon driver, 0:20:20.560 --> 0:20:22.880 to interview a bunch of current and past Amazon drivers, 0:20:22.920 --> 0:20:26.720 including one who was fired for swerving when a big 0:20:26.880 --> 0:20:29.800 rig crossed the medium because the driver had fallen asleep 0:20:29.920 --> 0:20:32.520 and was headed for a head on collision. And they 0:20:32.560 --> 0:20:34.960 were fired for making a dangerous maneuver, which is to say, 0:20:35.040 --> 0:20:37.480 swerving out of the way of a truck that would 0:20:37.480 --> 0:20:40.600 have killed them. And there's lots of petty examples too, though, right, 0:20:40.640 --> 0:20:43.720 you can get dinged on your performance if the camera 0:20:43.840 --> 0:20:46.080 thinks that your mouth has opened too much, because singing 0:20:46.200 --> 0:20:50.480 is considered distracted driving. You know, it's something of a 0:20:50.560 --> 0:20:56.120 paradox that these heavily surveilled workers who you would think 0:20:56.280 --> 0:20:59.440 those cameras might be used to vindicate them, find that 0:20:59.480 --> 0:21:02.160 those cameras are only an ever used to penalize them. 0:21:02.680 --> 0:21:04.960 Right that if you've got a million cameras on your truck, 0:21:05.080 --> 0:21:08.200 that's when you show your boss, hey, look I got 0:21:08.359 --> 0:21:11.119 I swerve because the truck was coming straight at me. 0:21:11.840 --> 0:21:15.199 But that's never in the field. In fact, you never 0:21:15.240 --> 0:21:17.120 get to talk to the person who fires you, because 0:21:17.119 --> 0:21:19.359 the way that it works with the DSPs is that 0:21:19.680 --> 0:21:23.040 they don't they're not Amazon, they're they are themselves so 0:21:23.240 --> 0:21:26.840 called independent businesses that hire you. And then the app 0:21:27.040 --> 0:21:30.280 tells your boss to fire you. And then you can 0:21:30.320 --> 0:21:32.919 tell your boss all day long, I swerve to avoid 0:21:32.960 --> 0:21:35.240 a truck, But your boss doesn't get to keep you 0:21:36.040 --> 0:21:39.359 on the payroll if Amazon says that they're not allowed 0:21:39.400 --> 0:21:42.240 to and Amazon makes that determination based on their own 0:21:42.280 --> 0:21:43.040 internal logic. 0:21:44.040 --> 0:21:48.080 Unpack this idea of Amazon telling the DSP to fire 0:21:48.119 --> 0:21:49.600 the driver, and the driver is say, why don't you 0:21:49.600 --> 0:21:52.399 review the well, you know, the video footage and the 0:21:52.480 --> 0:21:55.200 DSP saying, well, even if we could do that, it 0:21:55.200 --> 0:21:59.120 wouldn't change our decision. I mean, how is this being exactive? 0:21:59.160 --> 0:22:03.479 Because the Amazon worker ethnography stuff is haunting, but has 0:22:03.600 --> 0:22:09.160