WEBVTT - The Theory That Explains Why Everyone Went Crazy 0:00:03.120 --> 0:00:18.520 Bloomberg Audio Studios, Podcasts, Radio News. 0:00:20.720 --> 0:00:24.760 Hello and welcome to another episode of the Odd Lots Podcast. 0:00:24.840 --> 0:00:26.560 I'm Joe Wattenthal. 0:00:26.200 --> 0:00:27.320 And I'm Tracy Alloway. 0:00:27.760 --> 0:00:32.040 Tracy, you know, there's a lot of like concerns about 0:00:32.640 --> 0:00:36.640 AI obviously these days, and if anyone who's like reasonably 0:00:36.680 --> 0:00:39.760 intelligent can like list tons like maybe like it's gonna 0:00:39.800 --> 0:00:42.559 go rogue and be smarter than us or maybe whatever, 0:00:42.960 --> 0:00:45.120 But I still think, like what or maybe there's just 0:00:45.159 --> 0:00:49.320 gonna be this like flood of disinformation and deep fakes, 0:00:49.479 --> 0:00:51.600 or maybe it's gonna put all journalists out of business, 0:00:51.640 --> 0:00:54.520 which is certainly plausible, But I think, you know, like 0:00:54.560 --> 0:00:56.920 something I think a lot about is just this idea 0:00:57.040 --> 0:01:00.680 that regardless of what happens, like we're going to be 0:01:00.760 --> 0:01:04.280 increasingly sort of trusting like a black box for answers 0:01:04.319 --> 0:01:08.080 that we really have no idea where those answers, however 0:01:08.080 --> 0:01:09.360 you want to describe that come from. 0:01:09.520 --> 0:01:12.080 Yes, absolutely, And this is something that's come up on 0:01:12.120 --> 0:01:13.720 the podcast a number of times. 0:01:13.760 --> 0:01:13.920 Now. 0:01:13.959 --> 0:01:17.360 I'm thinking way back to an episode we did that 0:01:17.520 --> 0:01:20.640 was basically about the black box of algorithms and how 0:01:20.680 --> 0:01:23.880 difficult it was to understand what goes into them and 0:01:23.920 --> 0:01:27.080 then what comes out. And then of course we recently 0:01:27.080 --> 0:01:31.520 did that episode on pricing and the idea of algorithmic pricing, 0:01:31.880 --> 0:01:36.440 the idea of building proxy consumer profiles, And you're right, 0:01:36.680 --> 0:01:39.880 the issue is we know that there's this new technology, 0:01:39.959 --> 0:01:42.840 We know that there's all this data floating around, but 0:01:43.000 --> 0:01:46.920 we don't entirely know how it is coming to the 0:01:47.000 --> 0:01:50.280 conclusions or creating the output that it actually is. 0:01:50.680 --> 0:01:53.920 You know, the pricing thing is interesting because you know, 0:01:54.640 --> 0:01:57.360 in a market economy, you know, you could argue it's like, oh, 0:01:57.480 --> 0:02:01.040 at any given moment, you know you're being served up 0:02:01.080 --> 0:02:03.520 the like the optimal price, right. And in theory, even 0:02:03.560 --> 0:02:06.160 with the most advanced algorithms and stuff like, maybe there's 0:02:06.160 --> 0:02:08.359 some like this price is happening at the best price 0:02:08.440 --> 0:02:11.000 for both the seller and the buyer, et cetera. But 0:02:11.080 --> 0:02:14.320 I think like people just have a sort of deep 0:02:14.880 --> 0:02:18.480 intuitive distrust about the fact that like, you know, you 0:02:18.520 --> 0:02:20.880 can't go there and like touch it and verify it 0:02:20.919 --> 0:02:23.680 and see like this is why this exists in the 0:02:23.720 --> 0:02:26.040 state that it is. And I think it's going to 0:02:26.120 --> 0:02:29.839 create a lot of I don't know, cultural apprehensions as 0:02:29.919 --> 0:02:32.840 more and more decisions and more and more things that 0:02:32.919 --> 0:02:37.200 are that affect our lives just seem to like emerge 0:02:37.240 --> 0:02:38.760 spontaneously out of the box. 0:02:38.880 --> 0:02:41.840 Absolutely. I'm thinking of all the people working at Chipotle 0:02:42.000 --> 0:02:44.480 who are going to have to answer questions about not 0:02:44.560 --> 0:02:48.000 just portion sizes now, but also questions from customers about 0:02:48.040 --> 0:02:49.360 are they getting the best price. 0:02:49.600 --> 0:02:52.280 Have you seen those awful videos that people are taking 0:02:52.360 --> 0:02:53.840 of the Chipotle workers. 0:02:54.160 --> 0:02:57.080 Yeah, I've seen some of us so vilely imagine, yes. 0:02:56.919 --> 0:03:00.040 It's so vile anyway, that's a that's a that's a 0:03:00.080 --> 0:03:02.960 separate thing. But yes, like all of these things, and 0:03:03.000 --> 0:03:06.440 you know it's it's not just with AI obviously, and 0:03:06.520 --> 0:03:10.000 like this sort of world exists in increasingly black boxes. 0:03:10.080 --> 0:03:12.280 You put in the support ticket, you try to like 0:03:12.360 --> 0:03:15.160 talk to someone in an embassy or a consulate or 0:03:15.200 --> 0:03:17.120 anything you do when you sort of like send out 0:03:17.160 --> 0:03:21.440 some requirement or some request to some bureaucracy or some company, 0:03:21.480 --> 0:03:24.000 and then it moves around. I was on a I 0:03:24.000 --> 0:03:26.840 had a flight recently that was delayed for nine hours, 0:03:26.880 --> 0:03:30.520 and there's like this palpable frustration that everyone feels that 0:03:30.600 --> 0:03:33.360 the person you know, standing at the gate like can't 0:03:33.400 --> 0:03:35.880 answer their questions and they can't get anyone to answer 0:03:35.920 --> 0:03:37.720 their questions, and it just sort of like, you know, 0:03:37.720 --> 0:03:41.200 everyone explodes and everyone knows it's not the gate agent's fault, 0:03:41.600 --> 0:03:44.320 but still, like, you know, there's just this frustration, like 0:03:44.440 --> 0:03:46.680 where is the answer to what's going on? 0:03:47.080 --> 0:03:47.280 Right? 0:03:47.320 --> 0:03:50.120 And you can't ask a single person because that single 0:03:50.160 --> 0:03:52.840 person doesn't, like the gate agent doesn't have the answers. 0:03:53.160 --> 0:03:56.960 I think what's happening is like society has organized itself 0:03:57.040 --> 0:04:02.040 in such a way as to devolve responsibility, and the 0:04:02.200 --> 0:04:07.280 creation of all this new technology is basically going to 0:04:07.600 --> 0:04:10.560 I guess, ramp all of that up, right, Like, so 0:04:10.960 --> 0:04:13.160 it might not even be the gate agent that you're 0:04:13.200 --> 0:04:16.240 asking in the future. It might be you trying to, 0:04:16.320 --> 0:04:19.839 like I don't know, ask the algorithm, like why it 0:04:19.920 --> 0:04:23.680 decided to bump you versus someone else or actually that 0:04:23.720 --> 0:04:26.400 already happens, right, There is an algo that dictates like 0:04:26.640 --> 0:04:29.440 who gets bumped from the plane and who doesn't. So yeah. 0:04:29.520 --> 0:04:31.480 The other big one, of course is health insurance and 0:04:31.560 --> 0:04:34.599 why some claims are suddenly denied and you never get 0:04:34.600 --> 0:04:39.680 this answer anyway. The world is already filled with systems 0:04:40.040 --> 0:04:42.840 in which we have some question and no one actually 0:04:42.880 --> 0:04:45.279 can sort of you know give you the answer. 0:04:45.960 --> 0:04:49.360 Yes, absolutely, and I think we might have the perfect guests. 0:04:49.480 --> 0:04:52.200 We do have the perfect guest. It's someone we've talked 0:04:52.200 --> 0:04:55.920 to multiple times on the podcast, one of the smartest 0:04:55.960 --> 0:04:59.360 guys around, Always interesting, always worth paying attention to. We're 0:04:59.400 --> 0:05:01.480 going to be speaking with Dan Davies. He is the 0:05:01.520 --> 0:05:05.720 author of the new book The Unaccountability Machine, Why Big 0:05:05.800 --> 0:05:09.640 Systems Make Terrible Decisions, and How the World Lost its Mind. 0:05:09.960 --> 0:05:12.520 So this should be really fun. Dan, thank you so 0:05:12.600 --> 0:05:14.240 much for coming back on the podcast. 0:05:14.480 --> 0:05:16.000 Oh thanks very much for inviting me. 0:05:16.400 --> 0:05:19.200 You know, before we get into the meat of your argument, 0:05:19.640 --> 0:05:23.040 The Unaccountability Machine, I'm actually curious about the second half 0:05:23.320 --> 0:05:25.920 of the title, how the World Lost its Mind, because 0:05:25.960 --> 0:05:28.479 I certainly feel like the world lost its mind. But 0:05:28.640 --> 0:05:31.159 I'm like a middle aged boomer, and I feel like, 0:05:31.520 --> 0:05:33.760 you know, anytime you get to my age, you're like, oh, 0:05:33.800 --> 0:05:37.680 the world's gone crazy, the world's gone mad, Like, you know, 0:05:37.800 --> 0:05:40.480 why is everything so nuts these days? Has the world? 0:05:40.480 --> 0:05:42.680 Do we actually know that the world's gone mad? Or 0:05:42.720 --> 0:05:44.640 is it just because like we're all sort of old 0:05:44.680 --> 0:05:47.279 and cranky now everything seems like the world's gone mad. 0:05:47.680 --> 0:05:50.400 Well, from your own perspective, you can never be sure. Yeah, 0:05:50.440 --> 0:05:54.640 but I think there's actually reasonable objective ways that you 0:05:54.680 --> 0:05:58.520 can check up on this, just by noticing that the 0:05:58.560 --> 0:06:02.680 world gets more complicated as it gets bigger, and it 0:06:02.760 --> 0:06:05.840 gets exponentially more complicated. And I mean that're in the 0:06:05.880 --> 0:06:10.240 literal mathematical sense, because the number of connections grows faster 0:06:10.320 --> 0:06:15.000 than the number of things, whereas our capability to understand 0:06:15.000 --> 0:06:19.880 the world, manage it, and make decisions doesn't necessarily grow exponentially. 0:06:20.360 --> 0:06:24.479 So this is the story, I would argue of economics, 0:06:24.480 --> 0:06:27.720 It's the story of any management book that's worth reading, 0:06:28.040 --> 0:06:30.839 because the central problem of management is the world is 0:06:30.839 --> 0:06:36.039 getting faster, more complicated, faster than you can process that complexity. 0:06:36.400 --> 0:06:38.200 What are you going to do about it? How are 0:06:38.200 --> 0:06:40.880 you going to reorganize? And you know, we've just been 0:06:40.920 --> 0:06:44.800 through a global financial crisis, we've just been through a 0:06:44.880 --> 0:06:48.440 political what Aanam two is called poly crisis. I think 0:06:48.440 --> 0:06:50.680 there's decent reasons to believe it's not just because we're 0:06:50.680 --> 0:06:55.120 getting older, and it is actually a crisis of the 0:06:55.200 --> 0:06:59.479 ability to make decisions matched up against the speed and 0:06:59.480 --> 0:07:01.719 complexity see of the decisions that we're having to take. 0:07:02.560 --> 0:07:05.800 So when we talk about the lack of accountability and 0:07:05.839 --> 0:07:10.600 making bad decisions. Give us some concrete examples of things 0:07:10.640 --> 0:07:13.840 that you have spoken about or written about in your book. 0:07:14.000 --> 0:07:15.280 What are you thinking about here? 0:07:15.840 --> 0:07:19.120 Well, I mean there are kind of there's there's little trivial, 0:07:19.120 --> 0:07:24.040 funny examples, and there's big, huge, serious examples. So for example, 0:07:24.320 --> 0:07:28.640 and apologize in advance because this is quite disgusting. 0:07:29.240 --> 0:07:30.520 Oh is it the squirrels? 0:07:30.960 --> 0:07:32.040 Would you rather I didn't talk? 0:07:32.720 --> 0:07:32.920 Yeah? 0:07:33.080 --> 0:07:37.000 You could. This is bad if there are children listening. 0:07:37.280 --> 0:07:39.120 Maybe child. 0:07:41.080 --> 0:07:44.880 At the start of this century there was a craze 0:07:45.160 --> 0:07:50.560 for squirrels as pets in Europe, and squirrels were being 0:07:50.560 --> 0:07:53.560 imported from North America in China to be pets, and 0:07:53.640 --> 0:07:58.040 they had to have the right paperwork. And so one 0:07:58.120 --> 0:08:02.600 day ad of four hundred of the poor little things 0:08:02.880 --> 0:08:07.040 showed up in Chippell Airport in Amsterdam without any paperwork 0:08:07.200 --> 0:08:10.040 and without a return address to send them back to. 0:08:10.880 --> 0:08:14.080 And it's difficult to know what the airline should have done, 0:08:14.560 --> 0:08:17.000 but you can't help thinking that there must have been 0:08:17.040 --> 0:08:20.400 a better solution than what they actually did do, which 0:08:20.560 --> 0:08:24.440 was that they threw all four hundred of them, except 0:08:24.480 --> 0:08:27.840 for one or two that escaped into an industrial shred 0:08:29.240 --> 0:08:34.480 and this caused an outrage. There were questions asked in 0:08:34.520 --> 0:08:38.199 the Dutch Parliament and people immediately started asking how did 0:08:38.240 --> 0:08:43.040 this happen? Who is responsible? And in fact, the press 0:08:43.120 --> 0:08:46.439 release from the airline apologizing for this is studied as 0:08:46.480 --> 0:08:50.280 a masterpiece of crisis pr in business schools. But when 0:08:50.280 --> 0:08:53.960 they went back to inquire, they ended up realizing that 0:08:54.160 --> 0:08:58.000 no one had ever really made the decision that that 0:08:58.160 --> 0:09:02.160 was what they were going to do. The government's biosecurity 0:09:02.200 --> 0:09:07.280 Ministry had set some standards for the importation of small animals, 0:09:08.200 --> 0:09:12.440 the airline had set some standards for compliance with that policy. 0:09:13.040 --> 0:09:15.800 The only people who were expected to make a decision 0:09:16.200 --> 0:09:19.920 about whether this was grotesque and couldn't be done or 0:09:20.000 --> 0:09:23.520 not were some low level employees in a shed at 0:09:23.559 --> 0:09:28.480 Chippell Airport. And frankly, people who work in sheds aren't 0:09:28.600 --> 0:09:31.280 usually going to be thinking that they're meant to be 0:09:31.320 --> 0:09:34.120 second guessing the government. And so what happened is that 0:09:34.240 --> 0:09:37.600 you had this phenomenon that turns up a lot of 0:09:37.640 --> 0:09:41.240 the time at all levels of organization, which is that 0:09:41.320 --> 0:09:47.079 something happened which nobody wanted, but was the predictable output 0:09:47.600 --> 0:09:49.280 of the system that they had created. 0:10:05.080 --> 0:10:07.400 Yeah, this is this is first of all that's grim, 0:10:07.440 --> 0:10:10.000 but also like when you describe it that way, you 0:10:10.040 --> 0:10:12.600 could see it because ultimately, like, all right, here's this 0:10:12.720 --> 0:10:16.600 awful thing that happened to three hundred and ninety eight squirrels. 0:10:17.320 --> 0:10:20.800 I guess in theory, someone had to I don't want 0:10:20.840 --> 0:10:23.680 to get I don't know dumped the bag to the shretter. 0:10:24.400 --> 0:10:25.679 I have not researched them. 0:10:25.760 --> 0:10:30.679 Yeah, but that's not a very satisfying answer. I mean, yes, okay, 0:10:30.720 --> 0:10:33.440 maybe there was someone who did the physical thing, and 0:10:33.520 --> 0:10:37.000 I guess the entire you know, operations of the airline 0:10:37.080 --> 0:10:40.320 and the port and the customs Bureau could like just 0:10:40.360 --> 0:10:43.880 blame that one person. But that's not a very satisfying 0:10:44.120 --> 0:10:47.480 conclusion I guess in terms of like how this actually happened. 0:10:47.520 --> 0:10:49.800 It's funny I brought this documentary up. I was thinking 0:10:49.800 --> 0:10:52.400 about this too, Like I'm with the destruction of the 0:10:52.440 --> 0:10:54.760 old Penn station in New York, which it's like this 0:10:54.920 --> 0:10:58.720 like extraordinary, like I guess, like Roman or Greek building, 0:10:58.760 --> 0:11:01.400 and they just tore it down to like build the 0:11:01.960 --> 0:11:02.800 pretty owes. 0:11:02.600 --> 0:11:04.840 Something hideous, Yeah, something hideous. 0:11:04.480 --> 0:11:07.240 In its place. And now the new Penn station is terrible, 0:11:07.280 --> 0:11:09.160 but it feels like the same thing. It's like, how 0:11:09.200 --> 0:11:11.240 did no one stop and say, like, wait, does this 0:11:11.360 --> 0:11:14.000 make any sense in the long in the big picture. 0:11:14.520 --> 0:11:17.360 Yeah. And the thing is they'd created a system on 0:11:17.440 --> 0:11:20.360 the assumption that squirrels would show up in ones and 0:11:20.400 --> 0:11:24.280 twos and they could be dealt with as individuals, and 0:11:24.320 --> 0:11:26.400 that therefore you would never get into this sort of 0:11:26.440 --> 0:11:30.280 situation because when you build a system, you're always building 0:11:30.320 --> 0:11:33.120 a model of the world, and if something happens which 0:11:33.160 --> 0:11:35.800 doesn't fit into your model in the world, your system 0:11:35.920 --> 0:11:40.000 might do something awful. And there's a sort of symmetry 0:11:40.120 --> 0:11:44.120 and a kind of resemblance here with much bigger and 0:11:44.200 --> 0:11:47.360 more grim things like the Boeing seven three seven max, 0:11:47.600 --> 0:11:51.320 like the light board scandal in financial markets. It's not 0:11:51.480 --> 0:11:55.720 so much that anyone sat down and said, let's form 0:11:55.840 --> 0:11:59.520 a conspiracy to manipulate interest rates, or let's build a 0:11:59.760 --> 0:12:03.800 place that crashes under certain circumstances. It's just that no 0:12:03.840 --> 0:12:06.280 one set things up so that that wouldn't happen. 0:12:07.280 --> 0:12:10.840 Do you think I'm afraid going for the squirrels are 0:12:10.840 --> 0:12:14.680 probably going to be our archetypal example of this, But 0:12:14.920 --> 0:12:18.000 do you think with the squirrels, for instance, the hyper 0:12:18.200 --> 0:12:22.440 specificity of the goals or the job rolls of everyone 0:12:22.760 --> 0:12:26.360 involved contributed to the outcome. So in the sense that 0:12:26.440 --> 0:12:29.960 you have you know, I guess, the Dutch Wildlife Department 0:12:30.080 --> 0:12:34.400 who is trying to protect Dutch wildlife. Then you have 0:12:34.480 --> 0:12:36.560 like the guys at the airport who are charged with 0:12:36.760 --> 0:12:40.440 actually carrying out these orders, and then you have the 0:12:40.520 --> 0:12:44.200 airline which is charged with like looking at the paperwork, 0:12:44.679 --> 0:12:47.480 do those tend to lead to when taken altogether, do 0:12:47.559 --> 0:12:49.599 those tend to lead to worse outcomes? 0:12:50.000 --> 0:12:53.480 Yeah? Absolutely. It's this kind of fragmentation of the decision 0:12:54.040 --> 0:12:57.640 which is a result of the industrialization of the decision 0:12:57.679 --> 0:13:01.240 and its industrialization literally in the atom Smith's sense that 0:13:01.679 --> 0:13:04.480 you don't have anyone in the pin factory building an 0:13:04.640 --> 0:13:09.240 entire pin. You have thirteen guys all performing one simple operation. 0:13:09.760 --> 0:13:12.440 And that's a much more productive way to do things. 0:13:12.720 --> 0:13:15.480 But then when you apply it to decision making, you 0:13:15.559 --> 0:13:19.280 have the problem that everyone assumes that everyone else is 0:13:19.320 --> 0:13:24.400 going to react when something unplanned happens. So this is, 0:13:24.520 --> 0:13:28.760 like I say, it's the central problem of every good 0:13:28.840 --> 0:13:31.800 management sex book. How do you deal with information? How 0:13:31.800 --> 0:13:33.959 do you get a drink from a fire hose. How 0:13:33.960 --> 0:13:37.200 do you stop yourself from being overwhelmed? The answer is 0:13:37.280 --> 0:13:40.079 always in some way or another, you build a system 0:13:40.240 --> 0:13:43.160 to make the decisions for you. But once you've built 0:13:43.200 --> 0:13:46.120 that system to make the decisions for you, you no 0:13:46.200 --> 0:13:49.640 longer feel ownership of that decision. Psychologically, you no longer 0:13:49.640 --> 0:13:52.920 feel like you're accountable for the decision, because if you 0:13:52.960 --> 0:13:55.960 were accountable, you might be able to change it. But 0:13:56.040 --> 0:13:58.680 if you're going to be held accountable for this thing, 0:13:58.960 --> 0:14:01.840 you haven't really moved that thing on, you haven't really 0:14:01.880 --> 0:14:03.160 delegated it to the system. 0:14:04.080 --> 0:14:07.320 You mentioned Boeing, And speaking of Boeing, there was a 0:14:07.360 --> 0:14:11.679 great blog post back in April from Steve Randy Waldman 0:14:11.800 --> 0:14:14.840 inter Fluidity, who I consider another one kind of in 0:14:14.880 --> 0:14:18.640 your category of people who have basically been writing interesting 0:14:18.760 --> 0:14:21.280 things on the internet for a very long time, and 0:14:21.320 --> 0:14:24.040 he has the title was seeing like a CEO and 0:14:24.120 --> 0:14:27.600 this idea that you know, when Boeing merged with McDonald douglas, 0:14:27.720 --> 0:14:30.360 the McDonald douglas CEO became an outside hire and he 0:14:30.480 --> 0:14:35.040 had to essentially gain some legibility into this new organization 0:14:35.120 --> 0:14:37.840 that he inherited, and that was the cause for some 0:14:37.880 --> 0:14:42.120 of this like streamlining and offshore or you know, offshoring, 0:14:42.680 --> 0:14:45.000 things like that. Talk to us like about you know, 0:14:45.040 --> 0:14:49.040 you mentioned Boeing and it's more you know, the crises 0:14:49.080 --> 0:14:51.480 there that's been going on several years. It's a more severe, 0:14:51.560 --> 0:14:56.280 serious issue than the squirrels. But how do you know, 0:14:56.280 --> 0:14:57.800 how do you think about what happened there? 0:14:58.200 --> 0:15:01.360 I think about it in very similar ways to Steve 0:15:01.440 --> 0:15:05.600 Randy Wilburn because the and you see it in Boeing, 0:15:05.720 --> 0:15:08.600 but it's visible to a greater or lesser extent in 0:15:08.960 --> 0:15:12.240 very very many companies, probably the majority of companies today 0:15:12.720 --> 0:15:16.440 that the C suite has an information environment which is 0:15:16.600 --> 0:15:22.560 almost completely composed of financial numbers, because the financial numbers 0:15:23.080 --> 0:15:27.520 are taken by them as objective facts. We can talk 0:15:28.160 --> 0:15:31.680 long and hard with accountants about how objective those financial 0:15:31.760 --> 0:15:34.200 numbers are and how many assumptions go into them, but 0:15:34.320 --> 0:15:38.280 they arrive on a spreadsheet looking very much like objective 0:15:38.280 --> 0:15:43.200 facts about the world. Things like engineering level, in principle 0:15:43.360 --> 0:15:45.560 are objective facts, but you have to do a lot 0:15:45.640 --> 0:15:48.600 more work to find them out and to know what's relevant. 0:15:49.200 --> 0:15:53.680 And then you have issues like culture and kind of 0:15:53.720 --> 0:15:58.360 the social environments which aren't even capable of being quantified, 0:15:58.760 --> 0:16:01.280 So there's always this tender and see, if you're trying 0:16:01.320 --> 0:16:04.160 to do that thing of manage your own capacity to 0:16:04.200 --> 0:16:08.400 manage your information flow, that you're going to concentrate on 0:16:08.480 --> 0:16:12.560 the things that look finite and look manageable, and that's 0:16:12.560 --> 0:16:14.960 always going to be the financial numbers, which can be 0:16:15.120 --> 0:16:20.600 a big problem because financial numbers can mislead. You know, 0:16:20.680 --> 0:16:23.520 you can create illusions in an accounting system the same 0:16:23.520 --> 0:16:24.840 way that you can with anything else. 0:16:25.720 --> 0:16:29.040 How did we end up here? And I'm thinking specifically 0:16:29.360 --> 0:16:33.440 to one particular development in the world of business, which 0:16:33.480 --> 0:16:36.840 is the creation of the limited liability company. And I 0:16:36.840 --> 0:16:39.440 guess the clues kind of in the name there, But 0:16:39.600 --> 0:16:43.600 what were the decisions or the trends that sort of 0:16:43.640 --> 0:16:46.080 came about in creating the current system. 0:16:46.880 --> 0:16:50.600 Well, I mean, the limited liability company is certainly a 0:16:50.760 --> 0:16:54.280 big kind of change if you think about these things 0:16:54.400 --> 0:16:58.960 in feedback terms and information terms, and you know, one 0:16:59.000 --> 0:17:01.040 of the big arguments of the BO is that we 0:17:01.080 --> 0:17:05.400 should be looking to the mathematics of information theory rather 0:17:05.440 --> 0:17:09.359 than the mathematics of optimization to explain and model some 0:17:09.440 --> 0:17:12.840 of these things. But a limited liability company is an 0:17:12.840 --> 0:17:17.040 information filter. It tells you that outcomes below a certain 0:17:17.080 --> 0:17:22.000 amount aren't going to affect you anymore, and that changes 0:17:22.400 --> 0:17:25.600 your information world, It changes what you care about. But 0:17:25.880 --> 0:17:31.199 then what I think really started doing the damage, so 0:17:31.280 --> 0:17:34.720 to speak, was the development of the leverage buyout in 0:17:34.840 --> 0:17:39.919 the nineteen seventies and the shareholder value movement, as it 0:17:40.040 --> 0:17:44.040 was kind of really kicked off by Milton Friedman's essay 0:17:44.040 --> 0:17:46.959 on the social responsibility of a business to increase its 0:17:47.000 --> 0:17:50.720 profits New York Times, nineteen seventy, but then built on 0:17:51.440 --> 0:17:55.439 by just the entire kind of two decades of business 0:17:55.440 --> 0:17:59.680 school research that followed from that. Because again, thinking about 0:17:59.720 --> 0:18:03.960 it in information terms, a leftage buyout is a massive, 0:18:04.240 --> 0:18:08.919 screaming signal. The requirement to make the payments on debts 0:18:09.280 --> 0:18:13.320 becomes a signal that swamps anything else you might be 0:18:13.400 --> 0:18:16.640 thinking of, because if you are a CEO and you've 0:18:16.680 --> 0:18:21.200 got LBO levels of debt, that's your priority. You can't 0:18:21.240 --> 0:18:25.680 think about anything that isn't related to servicing that debt. 0:18:26.160 --> 0:18:28.040 Let's go back and talk about the sort of big 0:18:28.080 --> 0:18:30.919 picture ideas. In your book, you talk a lot about 0:18:30.920 --> 0:18:36.720 this field called cybernetics, and cybernetics is a name sounds 0:18:36.760 --> 0:18:38.480 like something that they would have come up with in 0:18:38.520 --> 0:18:41.640 the early nineties, like the Wired magazine people like would 0:18:41.680 --> 0:18:44.880 get into like cybernetics in ninety one. But actually this 0:18:44.920 --> 0:18:48.720 field has been around at least since the nineteen forties. 0:18:48.960 --> 0:18:51.199 I'm surprised they even had a word like cybernetics in 0:18:51.200 --> 0:18:55.240 the nineteen forties. But what is cybernetics? And talk to 0:18:55.320 --> 0:18:58.080 us about the sort of general framework you use as 0:18:58.080 --> 0:19:00.359 a sort of to start talking about this stories in 0:19:00.400 --> 0:19:00.800 your book. 0:19:00.880 --> 0:19:03.359 Sure, I mean, you're right, it was its Second World 0:19:03.400 --> 0:19:08.119 War kind of talk. It's originally from a word meaning 0:19:08.200 --> 0:19:11.280 the man who steers the boat, So it's a cybernetics 0:19:11.280 --> 0:19:14.640 in that sense. And the first guy to use it 0:19:15.400 --> 0:19:18.960 was a scientist working on creating an automated gun site 0:19:19.640 --> 0:19:24.720 for the United States Air Force. And the idea here 0:19:25.080 --> 0:19:29.600 is that there is some quantity that is preserved in 0:19:29.640 --> 0:19:34.920 an automated gun site between the operator, the radar, the server, 0:19:35.040 --> 0:19:38.160 motors and that and all the components of that system, 0:19:38.280 --> 0:19:43.080 and that component is information. And so at the time 0:19:43.119 --> 0:19:46.800 this was Norbert Viena. I'm talking about the scientist in 0:19:46.840 --> 0:19:50.439 the automated gun sites. Another guy working in the same field. 0:19:50.600 --> 0:19:53.320 You might have heard of was Claude Shannon at Bell Labs, 0:19:53.760 --> 0:19:58.240 who was inventing information theory at Bell Labs and has 0:19:58.240 --> 0:20:02.240 some fundamental theorems there, And in many ways, science of 0:20:02.280 --> 0:20:09.280 cybernetics is information theory applied to control. So you might 0:20:09.400 --> 0:20:11.919 have an information theory kind of pie the piece of 0:20:11.960 --> 0:20:15.479 maths that tells you how much bandwidth you need to 0:20:15.520 --> 0:20:19.959 transmit a given signal, and the cybernetic interpretation of that 0:20:20.040 --> 0:20:24.360 math would be that it's telling you how much capacity 0:20:24.520 --> 0:20:29.680 you need to manage a system that's of similar kind 0:20:29.680 --> 0:20:34.800 of noise. So this was all made huge use of 0:20:35.600 --> 0:20:39.080 in controlling things where you have access to the whole 0:20:39.320 --> 0:20:42.720 information environment. So a lot of that early maths from 0:20:42.760 --> 0:20:46.760 the first cyberneticians has just kind of stayed with us 0:20:47.920 --> 0:20:50.760 through the invention of the electronic computer, and a lot 0:20:50.760 --> 0:20:55.960 of it is actually at work in really modern artificial intelligence. 0:20:56.560 --> 0:21:01.879 Meragai called ben Recht who works on recommendation algorithms, and 0:21:02.040 --> 0:21:05.480 he's quite upfront that a lot of his fundamental mathematical 0:21:05.520 --> 0:21:08.640 techniques are from the four seas and fifties just being 0:21:08.680 --> 0:21:13.280 applied in the context of massively more computing power. What 0:21:13.320 --> 0:21:16.280 I'm interested in is where you take those kinds of 0:21:16.840 --> 0:21:22.320 theorems and apply them in a slightly more unrigorous, slightly 0:21:22.359 --> 0:21:28.280 more metaphorical sense to situations of management and organization where 0:21:28.320 --> 0:21:31.359 you don't have access to the full information environment and 0:21:31.440 --> 0:21:34.879 you just have to say, we're going to think about 0:21:34.920 --> 0:21:40.399 this not in an optimizing economics kind of neoclassical economic sense, 0:21:40.680 --> 0:21:42.680 but we're going to think about this as a system 0:21:43.000 --> 0:21:46.159 that has to be kept under control, and we're going 0:21:46.240 --> 0:21:49.320 to say, well, how much resource do we need in 0:21:49.400 --> 0:21:53.040 order to stabilize this system as a system, Which is 0:21:53.040 --> 0:21:54.879 a kind of abstract way of looking at it, but 0:21:54.960 --> 0:21:58.199 it's the same fundamental problem of management. How do you 0:21:58.200 --> 0:22:01.119 get a drink from a firehouse? How do you match 0:22:01.200 --> 0:22:04.480 your own capacity to manage to the complexity of the 0:22:04.520 --> 0:22:05.560 thing that you're in charge of. 0:22:06.040 --> 0:22:09.720 Wait, can you give us a practical example of the 0:22:09.760 --> 0:22:13.159 application of cybernetics. I guess because it does sound to 0:22:13.200 --> 0:22:15.639 your point earlier, it sounds a little bit abstract in 0:22:15.760 --> 0:22:16.280 my mind. 0:22:17.359 --> 0:22:20.840 Well, I mean practical example I think is the history 0:22:20.920 --> 0:22:24.200 of the development of the corporation. So from the first 0:22:24.280 --> 0:22:28.399 days of the American railroads, which were probably the first 0:22:28.640 --> 0:22:32.919 really big corporate structures, the world ever saw. You have 0:22:33.160 --> 0:22:37.159 this problem that as the network builds out, it gets 0:22:37.280 --> 0:22:41.600 more complicated, and the ability of the head office to 0:22:41.680 --> 0:22:46.520 manage it doesn't grow faster. And you can try and 0:22:46.560 --> 0:22:50.439 solve that by adding more people. You can get a 0:22:50.480 --> 0:22:55.000 great improvement by adding wireless telegraphs. But fundamentally, at some 0:22:55.200 --> 0:22:59.040 point this railroad is going to grow big enough that 0:22:59.119 --> 0:23:03.400 you have to devolve. You have to split it into branches, 0:23:03.800 --> 0:23:07.160 and you have to give autonomy to some of the subsidiaries, 0:23:07.480 --> 0:23:11.720 because that's the only way that you can match the 0:23:11.760 --> 0:23:16.440 bandwidth of the management to the bandwidth of the control problem. 0:23:16.520 --> 0:23:19.720 So I'd say that what we always see in any 0:23:19.720 --> 0:23:24.560 big organization is that it grows, it gets more complicated, 0:23:25.080 --> 0:23:27.720 It tries to deal with that by adding more resources 0:23:27.720 --> 0:23:31.320 at head office, it ends up not being able to 0:23:31.400 --> 0:23:36.800 keep up, and then it reorganizes. And the reorganizations almost 0:23:36.840 --> 0:23:42.320 always either involve pushing responsibility down to the shop floor 0:23:42.480 --> 0:23:45.800 or down to the branches, or they involve spinning off 0:23:45.840 --> 0:23:49.119 parts of the business into a separate organization and giving 0:23:49.160 --> 0:23:51.520 up the task of controlling it at all. 0:24:07.240 --> 0:24:09.399