Dominic Cummings: Vote Leave, Brexit, COVID, and No. 10 with Boris — #28

Dominic Cummings' time in No. 10 Downing Street as Boris Johnson's Chief Advisor was one of the most interesting and impactful periods in modern UK political history.

Show Notes

Dominic Cummings is a major historical figure in UK politics. He helped save the Pound Sterling, led the Vote Leave campaign, Got Brexit Done, and guided the Tories to a landslide general election victory. His time in No. 10 Downing Street as Boris Johnson's Chief Advisor was one of the most interesting and impactful periods in modern UK political history. Dom and Steve discuss all of this and more in this 2-hour episode.

Steve and Dominic discuss:

0:00 Early Life: Oxford, Russia, entering politics

16:49 Keeping the UK out of the Euro

19:41 How Dominic and Steve became acquainted: blogs, 2008 financial crisis, meeting at Google

27:37 Vote Leave, the science of polling

43:46 Cambridge Analytica conspiracy; History is impossible

48:41  Dominic on Benedict Cumberbatch’s portrayal of him and the movie “Brexit: The Uncivil War”

54:05 On joining British Prime Minister Boris Johnson’s office: an ultimatum

1:06:31 The pandemic

1:21:28 The Deep State, talent pipeline for public service

1:47:25 Quants and weirdos invade No.10

1:52:06 Can the Tories win the next election?

1:56:27 Trump in 2024?

References:

Dominic's Substack newsletter: https://dominiccummings.substack.com/


Music used with permission from Blade Runner Blues Livestream improvisation by State Azure.

--

Steve Hsu is Professor of Theoretical Physics and of Computational Mathematics, Science, and Engineering at Michigan State University. Previously, he was Senior Vice President for Research and Innovation at MSU and Director of the Institute of Theoretical Science at the University of Oregon. Hsu is a startup founder (Superfocus.ai, SafeWeb, Genomic Prediction, Othram) and advisor to venture capital and other investment firms. He was educated at Caltech and Berkeley, was a Harvard Junior Fellow, and has held faculty positions at Yale, the University of Oregon, and MSU.

Please send any questions or suggestions to [e-mail] or Steve on Twitter @hsu_steve.

Creators & Guests

Host

Stephen Hsu

Steve Hsu is Professor of Theoretical Physics and of Computational Mathematics, Science, and Engineering at Michigan State University.

What is Manifold?

Steve Hsu is Professor of Theoretical Physics and Computational Mathematics, Science, and Engineering at Michigan State University. Join him for wide-ranging conversations with leading writers, scientists, technologists, academics, entrepreneurs, investors, and more.

Steve Hsu
Welcome to Manifold. My guest today is Dominic Cummings, the former chief adviser to UK Prime Minister Boris Johnson. Dom graduated from Oxford, with a first in ancient and modern history. He ran the campaign to keep the UK out of the euro, thereby preserving pound sterling and ran the Vote Leave campaign that led to Brexit. In Fall 2019 he engineered a landslide general election victory for the Tories and for Boris Johnson. Cummings was portrayed by Benedict Cumberbatch in the 2019 film Brexit, the uncivil war. And in the television political puppet show Spitting Image, Cummings puppet was portrayed as a creepy alien with a pulsating head who drools at the prospect of eating Boris Johnson's Baby.

Dom, welcome to the podcast. Thank you, Steve. Well, we've known each other for a long time, I
think well over 10 years. And I want to start early in your life. What did the world look like
to Dominic Cummings in his early 20s? And what are the main differences between your worldview then, and now?

Dominic Cummings
Oh, my goodness. Well, in my early 20s, I was at Oxford. And then I left to go to Russia. In December 1984, just days before the first war in chess just started, I guess the main thing we felt back to those days is just how amazingly ignorant I was about almost everything, but didn't realize it. And now, I'm still in the scene, Detroit, but I've got a better idea of what I'm worried about, particularly around. Particularly, I think almost everything I thought about politics then was Wu was sort of embarrassingly wrong.

Steve Hsu
And at that point, were you even thinking about politics, because as I understand it, when you went to Russia, you were actually trying to sort of business.

Dominic Cummings
Yeah, I didn't really know what I was doing in Russia, other than I sort of went out of war. Well, I finished Oxford in Psalm 84. And I'd love to do Russian history. And I'd love to write really Russian literature. And I just couldn't really think of any kind of job to do, there was nothing that I'd particularly fancy doing with all my friends, pretty much we were looking for some kind of job in the city in finance, or investment banking, or wealth management consultancy or something like that. And I didn't fancy it. So I ended up going to Russia, really just because I couldn't think of anything better to do. And also I'd had a brilliant tutor, now dead, a guy called Norman Stone at Oxford, who by chance hadn't got to know various KGB people in the very early 90s. And had got access to all these amazing archives in Moscow that turned out to exist that people didn't really know existed. So these huge buildings full of all kinds of things, like some of them famously turned out to be Hitler's actual skull and the sofa for which he was killed and all that sort of thing, along with all these amazing archives needed by the Nazis and then stolen by the NKVD, maybe 45. So he'd been involved with all that. And I'd be talking to him about it at Oxford. And I just thought, Well, other than what else to do. So I will go to Moscow. It sounds crazy. Sounds fun.

Steve Hsu
Was your tutor also a spook?

Dominic Cummings
As far as I know, not. I mean, people like that always end up, you know, having dinner with people in the intelligence services and whatnot. So I'm sure he knew a lot of senior people. And he certainly knew a lot of senior politicians, including Margaret Thatcher. But so I don't think I don't think he was like lots of such people. He knew a lot of senior politicians, including Mrs. Thatcher. He definitely will have known a lot of intelligence services, people, but I've personally doubt that he himself was in any formal sense with them.

Steve Hsu
But he kindled your interest in Russia. And tell us, I mean, my understanding of that period of time in Russia was that it was the Wild West and very rough. And the kinds of crazy things were experienced by people who went there. Yeah, it will.

Dominic Cummings
It was a kind of crazy, wild west time, it was the sort of manic energy in places like Moscow. And politically, it was very weird because obviously, you just had Yeltsin taking over clubs of the Soviet Union, Yeltsin taking over the odd this kind of shift of a huge number of people from the intelligence services into organized crime. You know, it's kind of like imagining a huge faction of the CIA or MI six, six certainly going to work for the mafia. That's essentially what I mean, that's literally what happened in, in, in Moscow. But some of the stories are kind of not. So Moscow is often portrayed as like Sci Fi to be a dangerous place. But that was not true. It wasn't dangerous in the sense of going to parts of South Africa or something or Rio, dangerous for a normal person. But it was incredibly dangerous if you were doing certain kinds of things. So if you were involved in industries like particularly things like aluminum, certain kinds of metals, things like that, then it was extremely dangerous. But it wasn't because it didn't have the feel of an incredibly dangerous city generally.

Steve Hsu
You know, when you read about the occupation in say, Germany right after world war two ended G eyes would write about how they could get a beautiful German girl for half a pack of cigarettes or something like that. And was it like that in Moscow as well?

Dominic Cummings
Yeah, it kinda was sort of crazy like that, because you aren't, you know, all over the country had some kind of economic collapse after that in 91. And therefore, a very rapid emergence of organized crime. These kind of what we both came known as kind of vodka auctions where Credit Suisse working with politicians slash mafia, essentially privatized, in inverted commas, large parts of the Soviet economy, ie effectively handing it out to a bunch of insiders, Gen economic chaos, huge amounts of money being transferred very quickly. And all these families were completely disrupted. So also all these girls coming in from every single part of Russia, into Moscow, desperate to try and make some money in modeling or whatever. So yeah, it was very bizarre. It was a very bizarre time with a bunch of Westerners. They're a bunch of investment bankers and whatnot from the West, the political crooks, the mafia crooks, and all these amazing nightclubs. It was very, it was anybody who was there at that time, remember, it is a very weird, a very weird experience that you never really had again. I wonder if this poor place in China might be similar to that now, but I don't know where it will be? Well,

Steve Hsu
not anymore. Since Xi Jinping took over, he has cleaned it up a lot. But in the period of the early 2000s, you know, roughly until about 2014, or 2010. It was like that. In China, I could tell you all kinds of crazy stories about things that went on. You know, I think the most colorful stuff I've read about Moscow in that period is by Matt Taibbi who worked for a publication called the exile. And he was based in Moscow. I don't know if you guys ever cross paths? No, I didn't. So you were trying to start an airline? And is there an alternate universe where Dominic Cummings became some kind of aviation oligarchy? In Russia?

Dominic Cummings
I don't think so. No, I don't think that was on the cards. But it was a very, very interesting experience. You know how Charlie Munger always says that about the power of incentives and how people don't understand how important they are. That whole thing really taught me that lesson a bit. Essentially, there was a Western guidance from Russian guys who came up with the idea which seemed to be sound. The basic idea was Samira was this city, a former closed city, the Center for Defense in Aerospace Technology, in the Soviet Union, closed as it didn't really appear on maps. And you couldn't go there without special internal passports. And you know, there was a category of place in the old Soviet Union. And the idea was, this place will have to open up, and therefore it should have an airline, which was sound as far as it went. But we went there trying to actually make it into a proper airline, or airlines. But we then realized quite quickly that we were trying to make it into a proper airline so that it worked. Or as the staff there were essentially trying to steal as much money as they possibly could from it, bankrupt it as fast as possible, and move all the money offshore. And the idea of actually building something that will be great in 1020 years time just seemed completely insane to most of the people. And I think that's our you know, our mistake was made many times over by by many Western institutions, and many Western investment banks or various guides,

Steve Hsu
you know, the same situation, the same conditions prevailed in China, when they first opened up their economy. There was very low trust, and everybody was focused on just making a quick buck. Yeah. And take in a few decades for the society to transition to, you know, probably still not quite at Western levels of trust, but higher trust and where people can plan for the long term, and have some confidence that the institutions and the laws are stable. And I'm guessing you were in Russia at a time where none of that was present.

Dominic Cummings
Yeah, I think that's right. And I think there's also huge naivety both amongst Western politicians and also a whole set of kind of Economic Advisors for Western institutions who went over there and made a lot of money themselves personally or advising or privatization. But without really Understanding the culture and without understanding the reality realities on the ground, the privatization wasn't going to be something like the privatization of British Gas or British Airways, it was essentially just a bunch of insiders stealing things and giving it to their friends with a bunch of some investment Western companies get to get cut out of it. For me, it seems to me, you understand this much better than me. But my impression is that blushes not as crazy as it wasn't that maybe for internally, it also isn't, did not evolve in the same way that China did. Obviously, there's huge corruption still in both places. But there's also a lot more kind of long, longer term planning, I think it's possible in China now. Whereas in Russia, everything essentially kind of remained extremely chaotic and extremely arbitrary. Or you just never quite know what on earth can happen from one day to the next. And if someone's just gonna turn up at your offices and say, Yeah, you don't actually own any of this anymore. You've got to leave.

Steve Hsu
Yeah, I think if we talk a little bit later in our conversation about the Ukraine war, and what's going on there, I think I've heard you say that, you know, Russia is still kind of a mafia regime. And I think that it is plausible to me that it's not in the same situation as China, where the average person perceives China as, as you know, the country's back after a, you know, period of decline. And the institutions are solid, and the country will just continue developing. And I'm not sure Russians really believe that about their own country right now.

Dominic Cummings
No, I'm sure No, I mean, when I arrived in late 94, and I would ask Russians, then, you know, taxi drivers or whatever. What do you think of the government, they would use the Russian phrase mafia to vehicles for which basically means mafia government? And from talking to people who still go back there now, my impression is it has changed. It's less chaotic. There's not the same kind of competing power centers. That's obviously because in one sense, like the chief mafia guy was Putin, and he took over and put a lot of other people out of business.

Steve Hsu
Yeah, the key transition is that once somebody takes over like that, can they make a transition to normalcy, some kind of more rule based governance, and I think in China, that's happened. And, you know, we could, people would debate this with me, but you know, when you have 100 million people in the Communist Party, and somewhat rule based processing of regulations and new information, you can't have that sense of normalcy. At least I perceive that when I visit China and talk to people,

Dominic Cummings
yeah, exactly. Versus if you look at Russia, now you can see just even in, you know, in terms of how the Ukraine wars happened, Putin himself has got a lot of terrible shocks, because a lot of investments which people thought in Moscow had been made in building various things turned out to be actually also stolen and things which they thought had been built, turned out to be Potemkin villages, capabilities, which they thought they'd modernized in the armed forces turned out not to really exist. No, all kinds of documents have now leaked out and on the Internet, what about you can see, see some of that? So it's one of the ironies of that think of that kind of Mafia system, that although people appear to have extraordinary power, in some ways, they're also weirdly vulnerable in other ways, as well, to the kind to the suburbs, to the inherent lies and corruption of the whole thing.

Steve Hsu
Yeah, I imagine that, you know, war is one of these things where you can't fake it. You either have the capabilities or you don't. And so I'm sure he received a bunch of nasty shocks in the last year.

Dominic Cummings
Yeah. Sure.

Steve Hsu
So let's go back to the young Dominic Cummings. So the airline didn't work out. You're back in the UK now. Yeah. And what gets you into politics?

Dominic Cummings
So it came back? Like I remember exactly when, but I think towards the end of 96, I then briefly thought about being a lawyer. I had an uncle who was one of the best lawyers in the country, what a lot of interesting constitutional questions. And I was interested in those, but then I realized that I wasn't very good at it. And also, I wasn't particularly interested in it, really. And then I was talking to someone, at the end of 1998, you just set up your organization to try and stop Blair taking us into the euro. And he suggested that I go work for that. And that's what I did. And that's how I got into politics. Now there's

Steve Hsu
a, there's an iconic photo of you. I realize this might be a little embarrassing to you. But there's an iconic photo of a young Dominic Cummings with more hair. She's a very handsome young man with a suit jacket on. And I think smoking a cigarette was that way around that time.

Dominic Cummings
I think it was taken roughly a year later. It was after this campaign had been going for a year or so. And I got to know a guy who was a Time Magazine guy in London. And he was interested in what we did because everyone kind of assumed that Blair was going to prevail and that we'd be joining the euro. And then suddenly, the debate kind of changed. And he said he wanted to do something about it in Time Magazine. And so I ended up meeting up with him. And I think that and that's when those weird photos were taken.

Steve Hsu
Those are great photos. They're almost like a dumb young diamond. Cummings this James Bond kind of photos

Dominic Cummings
very much and were they actually take it very much not like your

Steve Hsu
the wife might disagree. But were those taken at Trafalgar Square or has that been photoshopped to look that way? Some of the ones I've seen looked like they were taken there. I don't

Dominic Cummings
know if there's like people photoshopped them later, the one the ones that would take the time it was kind of right about St. James's St. James's Park and Buckingham Palace, which is where our office was,

Steve Hsu
I see. Yes. Yes. Great. So was that a moment of triumph? So you just said people didn't think you were going to be able to stop the UK from giving up the pound and joining the euro. But you guys prevail. But So was that a moment of triumph where Time magazine was taking your photo?

Dominic Cummings
No, it was before really, it was clear that we'd won. It was clear. At that point, it was clear that things had changed, and that it was actually going to be competitive. And the blend might not get his own way. But it wasn't clear that we'd won. I don't think actually it was clear that we'd won until 911, which is after Blair Well, a second majority in summer 2001. It seemed like the whole battle was about to come roaring back. So we started off in 9899. And it made a lot of progress. And of course, Blair had huge problems on the whole subject. But then he won the second majority, which is massive. And he clearly wanted to have another go at it. And everyone was gearing up for it. And today of 911 he was actually due to give a speech in which he was basically going to kind of go as far as the starting gun for the referendum campaign. These days, it would all have been kind of pre pre leaked on Twitter. But in those days, it wasn't we all kind of knew what was happening. But it didn't really come out to the media. And then of course, the speech was never given. Yeah, then we've got bustled back into a car and driven back to London when the planes hit. So one of the weird kinds of effects of 911 Was it basically killed the Euro campaign in Britain.

Steve Hsu
I see. So was there. I mean, in the wake

of Brexit, there's still tons of argument about whether it was good or bad, or in the short term long term. But when the UK stayed out of the euro, it just I guess it people just accepted it, because it wasn't going to go any other way and how it worked.

Dominic Cummings
I think so. So as I've gotten older, I've realized that people who don't like arguments don't really change very much in politics, really, almost not at all. But events do. And perceptions of success and failure do and I think what happened was he had 9798 99 elite perceptions where the Euro is going to happen, it's going to be a great triumph, it's going to be obvious that Britain is losing from not being part of it. The Poles will all be in this kind of inevitable process similar to what happened with joining the EEC in the first place. But that didn't happen. The Euro hit various problems. And a lot of the promises that were predictions made by the Euro campaign about what would happen simply didn't happen. And in various ways, the Eurozone itself seemed to have various kinds of economic problems. So when we didn't go in, elite opinion changed, but not really because people said, oh, right, okay, you guys were right, we were wrong. It just became part of the background conversation that is well, like the timing doesn't really work. And maybe we'll come back to it in 10 or 15 years time or something.

Steve Hsu
Yeah, I mean, my impression of the post Brexit situation is that people's positions are so dug in, you know, facts are not really going to change people's minds about whether it was a good outcome or Bannock.

Dominic Cummings
Yeah, I think that's right. I think that I think in all sorts of ways. I mean, it didn't have so it didn't have to turn out like this. A big part of the reason for it, obviously, is that because everything became so extremely botched by the Tory party, and by the establishment after the referendum in summer 16. And then the country just kind of drove itself into or rather Westminster drove the country into a cul de sac, it meant that on both sides kind of passions fled. And it became much more a question of political identity. I mean, an interesting thing in polling now is that if you ask people about Brexit and about leaving remain, people leaving remain identities and much stronger in Britain now than political party identities are, there's not really any equivalent of that in in America, where things are kind of hyperpolarized, but on party on party basis. So yeah, I think certainly the kind of insider level in Westminster, there's the two different sets of people and very few of them have really changed their mind about it. And how events play out. Well. Neither side really is persuading the other, but then both sides. See how the worlds worked out subsequently, as evidence that they were right.

Steve Hsu
Here. Let me let me go back because we're a little bit out of chronological order. So back to Brexit and vote leave and stuff like that in just a moment, but just going a little bit more chronologically, so you and I kind of got to know each other, more or less over the internet. Yeah. And it was, I think, some years before we met in person. Am I right about that?

Dominic Cummings
Yes. I mean, I was reading your blog from somewhere around 2004 -2005. I don't think we actually physically met until SICU. In 2014. Yes.

Steve Hsu
I think that one of the things that I blogged about from the very beginning and I started my blog in 22,004, was the housing crisis and predicting that we were going to have a financial crisis based on, you know, the housing bubble in the US. Yeah. And I seem to call you at the time you were working for Michael Gove. And I think you actually tried to warn people in Westminster about an impending financial crisis, is that correct?

Dominic Cummings
I forwarded on your blogs to a lot of people and said, because obviously, you were writing about this way before, it all happened in saying things like, Well, my students are working on this stuff. And they all say that their managers don't understand it, and it can blow up. So I forwarded it on to various friends and family and also on to various media people. But obviously, no one paid any attention. Which was an interesting lesson for me, actually, because it made me realize that there's this huge infrastructure of the modern media that's writing about all of this. But where did I find out about CDs and CDOs? Well, it was on a blog written by a physicist so it's not his main job to be reporting on any of this stuff. That was a very good lesson for me.

Steve Hsu
Yeah, my situation was I had former colleagues and students who worked in finance. And, you know, it also was a coincidence that I was a young professor. And so I was buying my first house. So I was carefully watching the real estate market. But I also knew about these collateralized mortgage obligations and debt obligations that, you know, my friends in finance were themselves building and modeling. And so I kind of had this double whammy where I realized like, house prices are appreciating way beyond what's rationally supportable. And also with these quants are building these crazy instruments that even they think are dangerous. And so I did quite a lot of writing about that on my blog, in the lead up to 2008. So you and I first met in person at sai foo in 2014. And that's an event that's held in the summers by Google, to actually be held at the Googleplex. So it's quite a fun event. Yeah. And they generally they try to invite world class scientists and other leaders my impression that when you and I actually met up there in person we spent I remember, we talked one, at one point, we want to talk really late into the night, my impression was that you had not spent a lot of time around world class scientists before that event. Is that fair? Yeah, that's true.

Could you talk a little bit about that? I

I think you were impressed that the way that these world class scientists talk to each other, and often across disciplinary boundaries, is very different from the way people in politics or in Westminster talk to each other.

Dominic Cummings
Yeah, it was, it was a, it was a fascinating event for me, because as you say, you know, you had all the people like you and George Church, and all these kind of famous people talking about things, but in an extremely different way to the way in which people in politics talk about the world. And it was fascinating in all sorts of ways. It was interesting just to see how their culture worked. It was also interesting to see some of the things that they talked about, for example, one thing I took away from there was clearly how drones were going to change warfare, the fighter blogged about it shortly afterwards, he said, clearly, in a few years time, it'll be possible for some teenager to steer their drones over Westminster, blow things up. People in Westminster thought it was a sign of yet again, how I lost the plot. But it was obvious talking to people about what was kind of late, but at the same time. It was also interesting that I thought that the scientists were generally naive about how politics works. And it's kind of assumed that politics is far more rational, and that the politicians are much more interested in actually trying to solve problems than is actually the case. So I also was rather pessimistic about the chances of that culture affecting politics in the way that they wanted to. Yes, one

Steve Hsu
of the aspects of sigh fu is that, you know, aside from the scientists describing, you know, the recent developments in their own areas to others, which is great. There's always this kind of utopianism, where the scientists, some subset of the scientists decide, like, Oh, we're gonna band together and solve problem x or problem y, where x and y could be climate change, or, you know, social justice or something like that. And that aspect of it, to me, is usually pretty cringe. I mean, it's not like scientists really understand, in practical terms, how the world works. So although idealism is great, oftentimes, I think they're very misguided in what it's actually possible for them to achieve. Yeah, I'll be interested in geophysics

Dominic Cummings
because so one thing I think, is that if I think about two kind of overlapping worlds from that period, so 2014 Take the world of kind of VC and startup world, and then take the world of kind of like basic science To search and kind of elite scientists, I would say that the first world is updated and changed a lot in how it sees about politics is how it sees politics and government. But I don't think that the second world of elite science has. So when I talk to current [unclear] people now, particularly after COVID, I think a lot more of them realize the reality of how politicians work or how big bureaucracies work. Because it was kind of forced, you know, everyone's been kind of forced to face this, right, because not just the COVID disaster in the cloud of spring 2020. But watching all the political parties and the bureaucracies and the old media, kind of very consciously, not updating and not learning, and they just keep repeating the same mistakes. It seems to me that a lot of the startups that you see in the world have watched that and changed their own people in that those worlds have changed their minds. Whereas my impression is that a lot of scientists still don't really want to update and face the reality of how the government works. But I don't know, you're obviously much closer to both worlds than me. Do you? Does that sound right or not?

Steve Hsu
Yeah, what you're saying is completely right now, I can discuss two aspects of this. One is that Silicon Valley was very naive initially, about how important government was. And they gradually learned that because when their companies got bigger, and we're actually, you know, directly impacted by government regulation, they realized they had to have lobbyists and all kinds of other things. And so that was a trend which I think when you were there it was a trend that had just gotten started. The other thing is that startup people are just much more real world in general than scientists, scientists can always live in the ivory tower, and have very abstract notions about what's happening in the world. Whereas startup people have to, they have to really gauge what's happening realistically in their, you know, in their narrow market space, but also in society in general. And so they're much more likely to discover that something in the real world is not actually matching their hypotheses than the average academic or scientist, I think you're spot on about that. And even COVID, didn't write, I think, didn't fully change the, you know, isolated utopian viewpoints of scientists, although although the, you know, plenty of scientists have been radicalized in COVID, because they are watching carefully what epidemiologists were saying and the what actually turned out to be true. And you know, what the government knew or didn't know. And so there's been something happening, but still there are plenty of naive academics. Yeah. So after we got to know each other in person, that's a couple years later, that's really when you got involved in vote leave, is that right?

Dominic Cummings
Yes, actually, probably about a year later. So I think Sophie was 14. And then it was summer 15, just after Cameron won the election, that various people came to me and said, then you've got to start up, you've got to create a startup to deal with the referendum. So yeah, that was almost exactly a year later. And

Steve Hsu
Those are the events that are mainly chronicled in the movie starring Cumberbatch, is that right? Yes. Yes. So I remember communicating with you a little bit about the Senate, I always have to admit, I never really understood, you know, you know why Brexit was or was not a great idea. I mean, certainly at that time, the whole thing to me was, you know, very abstract and I didn't know any of the details. And I just thought, well, if Dobbs is going to do this it would be really awesome. So I'm gonna pay attention. And the day of the vote, I happened to be at this crazy meeting on Lake Como, which is in the Italian Alps, with a bunch of super rich capital allocators, billionaires and such. And the night before we were talking about the vote, and they were pretty certain that Brexit wasn't going to win that vote. And then the following morning, people were running around with their hair on fire for trading, you know, maybe it was in the market. So it's always gonna be very memorable in my mind, because what the moment when it actually happened, I was actually in the, you know, the mansion of a billionaire on Lake Como. And all of them were just totally in shock. But I knew what was coming because I knew you. And when they asked me the markets were way down. And they were trying to decide whether they should liquidate their positions. And they you know, I'm just some physics guy who happened to be there for their entertainment. Right? So they asked me, this billionaire hedge fund guy asked me, Well, what would you do, Steve? And I said, I think this is transitory because we're not really going to know what the full consequences of Brexit are for, you know, years. And so I think the market is oversold right now. I think you're okay. And he actually took my advice, and it saved him a lot of money. Actually, that's my memory

of your victory. Yeah, it

Dominic Cummings
was crazy. I think a lot of well, obviously, the polls were wrong, right. And that and that, inevitably, that inevitably meant that elites all over the world? Well, I think we're about to get a bit of a shock. One of the reasons why the polls were wrong or to do interesting things about it was because our data science team, made up of physicists, predicted the year before in 2016. They said, after the media, various polling companies, they said, you know, what's gonna happen next year, all these polling companies, it's clear that the polling industry are largely legit charlatans, he votes waste. And what's going to happen is that they all live in London, which is the epicenter of Romaine. And they're all going to make a bunch of methodological changes to the polling over the last few weeks. And all the changes will move these neural Lane directions, and they'll fool themselves about what's happening. And therefore, if it's close, like the sweet spot for us is to be like 50 to 48. Behind. Because in the polls, because they're all kids themselves, that's basically exactly exactly what happened. It was close enough for his outfit actually to be passive. But these people didn't really believe it. And the pulse has all made a bunch of adjustments in the last couple of months. And pretty much every single adjustment the pollsters made, pushed the polls in the wrong direction. So yeah, it was an interesting example of how because London was so out of sync with the rest of the country. It meant that hedge funds, the media, everybody was kind of living in a bubble in London and got fooled.

Steve Hsu
Yeah, this, this has kind of become a theme, you know, Trump's victory in 2016, and things like this, but over the years can attest you. And I've had many conversations about the quality of polling, the quality of data science in polling. And you ultimately built your own team that, you know, I think many of them were former physics guys or math guys. They actually in some of the meetings I had with them, they told me they had literally gone through the entire literature on polling theory and voting theory and all this stuff, and classified all the papers in the academic literature as wrong, or probably right. And the ones they had high confidence in were the things that they incorporated into their own methodology. Maybe you can talk a little bit about that. Because you have I know you have your own secret polling data science team,

Dominic Cummings
which has been essentially though, as you say, there were a bunch of physicists and math guys. And then they looked at a bunch of the polling, the public polling in 2015. And we also met various companies and talked to them about how they did things. And the data science guys said, basically, most of these companies just don't really know what they're doing. It's all operating on these 1000 person polls, and simple math, but the way that they correct for these things is not done in any kind of way, properly, scientific way. So then they will often do as you say, they basically scanned around a whole bunch of literature. And they also found some fascinating things, particularly by a guy Professor of Applied Maths in the States called Gelman who's done these experiments on it. There's one paper where he did polling via X box with 300,000 people for the US presidential 2012 campaign. And of course, in any normal sense, you'd say, well, how can you pull the x box because the demographics are incredibly skewed to young to young men. But essentially, what this thing showed was that very large sampling, plus some fancy Bayesian statistics, actually is a just theoretically better way of doing polling than 1000 person polls, and simple math. And so we built devotee VSOs, to build some models using these papers, which they found on the internet. It's kind of remarkable in lots of ways, like, because it turned out that quite a lot of these ideas, actually, I think, go back to the 70s even. But basically, nobody in politics really deployed them and figured it out soon. So one of the most interesting things that we figured out straight away in 2015, was that education was crucial. And if you wait to start your way to polls by education, then you get very different results than what a lot of the public polls are showing. And, of course, this turned out to be very relevant not just to the Brexit results, and why the polls were wrong, but also to what happened in some of the crucial states in 2016. In America with Trump, were similarly a lot of the polling companies, then we're not waiting by education and therefore didn't pick up what was happening with non college whites. And hence why you know, why almost all the models wrong in the states in 2016, for similar reasons to what happened on Brexit. The

Steve Hsu
The guy you're referring to is Andrew Gelman, who also writes a blog and it's a professor at Columbia in statistics in political science. Also, I think he was an undergrad physics major at MIT. Yeah, he definitely is very insightful and these large scale polling efforts with Bayesian correct shins might be the way to go in the future. And its assumption there's more investment in trying to actually do polling in a more scientific way.

Dominic Cummings
Yeah, it also says something, I think very interesting about politics. You know, one thing. So the naive view of politics is that the people in charge actually are sitting in government, then when you actually meet them, and you realize that's obviously not the case that almost all of them have no real interest in government law. But then the slightly updated version of that is okay. Well, they're not interested in the government, but they are at least interested in winning elections. But that also is not correct. One of the most interesting things about politics, I think, is that they're not actually even serious about winning elections. And one of the ways you know, that is because if you're serious about winning elections, then you would take polling seriously. And you can just see as a matter of arguable fact that almost all politicians just don't even take politics seriously. And what they're actually optimizing for almost always is very short term horizons, in terms of how the media operate, and in group signaling that affects their careers, you know, so what the internal shifts, or the internal shifting Coalition on their own side, what they're thinking of very short term calculations about how you send signals to this coalition, and therefore do or do not get opportunities to ascend these hierarchies. And the feedback on these things is very short term, and it pretty much swamps actually calculating rationally, how you would win an election, if you actually want to win an election, that huge amount of Bay View will be radically different than what you actually see

Steve Hsu
Is this situation somewhat different in the UK because you have a parliamentary system because having the support of the other Tories in the long run might be more important than a particular election. Whereas in the US, like the guys who want to become president, they do really care. Now, they may not really understand polling, either, but they do really want to win that election.

Dominic Cummings
So, , no, I don't think I don't think it is different. I'll give you a classic example. Look at the Hillary campaign. If you look at the Hillary campaign, something like a third of the adverts that Hillary's campaign ran, not only were not even just a waste of time, they actively helped Trump and the adverts which got which had most money behind them, and got most attention and retweeting and kind of signal boosting from inside the Democratic coalition were precisely the Yardbirds, which most helped Trump and most harmed Hillary. So if you look at the Hillary campaign, they were absolute. I mean, it was a classic example that we'll be talking about, that they will, effectively the internal culture of the campaign was pulled far more towards, what's Washington thinking? What's the New York Times thinking? What are insiders thinking? What are our activists and our donors thinking? Not? What do non-college whites in Wisconsin think? And that's why they lost the election? Which elect remains that they should never have lost? But is it

Steve Hsu
possibly that they actually have the right motivations and incentives, it's just that they're just overwhelmed by the bubble thinking, you know, the self reinforcement within their own group. And they just don't realize these ads are really going to, you know, that they just can't even imagine the person that wouldn't agree with the ad. So it could be a cognitive limitation as opposed to the wrong motivations and

Dominic Cummings
incentives. But they're not because it's, I mean, it shouldn't be. I mean, conceptually, this is not very complicated, right? Theory, there's a pretty simple way of approaching things. You figure out what, like who the actual critical voters are, they are gonna affect whether you win or lose, what do they actually do, what do they fear? What do they want? What kind of creative communications objectively work with them? Now, of course, this is not perfect. But it doesn't take huge sophistication to go and run a bunch of focus groups and show ads to critical voters in the right states. And determine, Okay, this actually works, or this actually doesn't work, they hate it. But when you actually deal with politicians, you realize that's just that the psychological pressures in elite politics are very strong about what you say to each other, not trying to figure out these problems. And actually doing this was actually literally the very, very first lesson I ever got in politics when it first started the on the Euro campaign at the end of 98, in quarter 119 99, but went out and I did a big market research program about what people actually thought about the euro and various arguments about it. And of course, unsurprisingly, it turned out that what the public actually thought was very surprising and all sorts of ways. And a huge amount of things that kept the pound people believed were true about public opinion, will be false And a huge number of arguments that were making didn't either not work or were actively counterproductive. And the arguments that did work, were a note. So I realized this, and I was really shocked. And I took all the results to these conservative politicians thinking, in my extreme naivety, they'll be very pleased that we've done this work. And they'll realize that everyone on our side will realize that actually, people think completely differently than what we thought. And what we thought works doesn't work. And what does work is completely weird and unexpected. But in fact, in liquid, the exact opposite happened, everyone was incredibly close with me. And like actively angry that I've done this, and basically just dismissed it all. And this was a huge surprise to me in 1989. And to begin with, I thought, well, maybe this is just like a weird aberration. Maybe this is not the kind of standard feature of how this world works, because it's so shocking. But over the years, it's turned out to be completely unchangeable. And exactly the same thing happened on Brexit. Almost all the Tory MPs wanted to use arguments on Brexit, that objectively did not work, either were pointless or the public active he disagreed with. And the things that actually worked best, the pro Brexit MPs, largely completely hated and didn't want to say. And I think it's very similar to the Hillary situation. And they live in a world where they talk to people every day, to each other every day. And that world just ends up dominating the objective world of how regular people outside politics think and feel and talk, which is terrible in lots of ways and depressing. Real and it means a huge, low hanging fruit in various ways as well.

Steve Hsu
Yeah, you know,

I think your superpower is that you are able to actually accept and perceive base reality and react rationally, you know, you know, in running the campaign, whereas I think what you're pointing out is most of the political leaders just can't do that. I also want to say that, you know, your detractors, they have this caricature of you as thinking of yourself as an evil genius, or having a huge ego. But it's actually quite the opposite. I mean, in what you've been telling me, You're, you're always self deprecating saying, Well, I can't believe these people don't understand this. It's so simple. It didn't take much to figure this out. And when I told them, they just got mad. But, you know, you're not saying you're the only one in the world who can do this, because you're a super genius, you're actually kind of shocked that the politicians can't understand these basic points that you're making to them.

Dominic Cummings
You don't think there's anything which I've done in politics, which took really any kind of surgeon take any gray brains, I think I think the thing, which I increasingly think, though, is that the essentially core issue is that politics is such a kind of gang activity, and that in almost all normal ways to get to kind of get ahead in it, you have to kind of be part of the gang. I want to advance in that guy. And that's basically how these party systems work in Western countries. But once you go along with that, then it makes it just extremely hard to face realistically, a whole set of questions. But he faced realistically also questions that similarly makes it extremely hard for you to get along with any kind of gang. And I think probably the main difference between me and a lot of the people that I've ended up politically fighting against and remaining for the election or whatever, is that, for me, for me, it's more important to win. And to figure out, therefore, to figure out what's actually going on, then it is to be friends with people. And I think that's probably, I guess, probably a crucial thing. Before we

Steve Hsu
leave the data science polling sort of thing campaigning sort of thing. I wanted to get one thing on the record. So during vote leave, or after vote, leave one, there are all these charges by your detractors that you had worked with Cambridge Analytica, and that you had actually used Facebook to manipulate the electorate in evil ways. And my understanding talking to your team is that you guys didn't work with Cambridge analytic at all. And that risk, you know, that's become like established wisdom amongst the left, I think in the UK, but it's just totally wrong. Could you just comment on that?

Dominic Cummings
Yeah, I think so. To some extent, this was also a thing, a product of Trump winning, in that the combination of Brexit, then Trump was such a shock, that a lot of the establishment will be looking for some kind of explanation, right? So you had a weird combination of things. The New York Times had done all these stories on Hillary's emails, which objectively was actually quite an important factor in what happened at the end of that campaign. They were looking for someone else to blame. And then suddenly, you had this Carol Cadwallader observer comes along and creates this whole conspiracy theory Sorry, based on the idea that me, Bannon, Trump's guy, a guy called Robert Mercer for machine learning expert, I think who then went to renaissance technologies, the famous hedge fund, and Putin in the KGB. Yet we were actually all essentially working together. And that the huge arguments, supposedly between me and Faraj and the huge fallouts between me and Faraj, and the fact that Faraj was trying to get me fired from vote, leave and all this, she betrayed us all, essentially, kind of information operation scam, KGB style, was actually we were all secretly coordinating behind the scenes via Cambridge Analytica, and this other company called the IQ. And I remember going in 2017, I think it was for this conference, almost all the people there knew who I was. And I remember sitting in these rooms listening to all the academics talk, and they all were talking about this thing, and they all believed it. And they all thought that Trump had won because of Cambridge, Analytica and Facebook, and psychometric, psychometric Facebook targeting and all this. Of course, it's all just quickly, because furiously you're staring obviously, is bullshit, but also the kind of the kind of world which they imagine of how Facebook advertising actually worked, was also just complete nonsense. But the story fit because people wanted to believe that Brexit happened when Trump happened because of a kind of evil conspiracy between capitalist hedge fund people, racists and Putin in the KGB. Why does that story just fit psychologically? So he had this incredible irony of the left saying that, because uneducated masses were fooled by this conspiracy, because they were the ones actually policing this ludicrous. This leader gives Christmas information from The Guardian and The New York Times.

Steve Hsu
Yeah, I mean, that, to me, was a mass delusion that affected millions of people on the left, plenty of thought leaders who are sort of left leaning lots of academics. And now it's just faded just because of time, because time has passed. But it's not because people actually figured out what really happened and then updated properly there. I think they're still confused. Do you

Dominic Cummings
think they were in a weird way? It's kind of lived on? I think that if you look at the whole thing now, with Putin, it seems to me it was one of the things that people in Europe don't understand about America, which I didn't really grasp until I went over to the states in September. Was that for a lot of American elites, what's happening within the war, now they see it as, yes, the war against Putin, but also a war against Trump. And like, why is it that the Bernie Sanders AOC, etc, left, who protested against every wall argued against Iraq, blah, blah, but they're just completely gung ho for this? And I think one of the things that people don't understand in Europe is that part of the reason for this is that they see the war against Putin as part of the war against Trump. And part of the reason why that's the case is these people all still basically believe a large part of the Cadwallader Guardian New York Times Conspiracy. Trump was put there by Putin. Yeah,

Steve Hsu
I think you're absolutely right. Maybe someone who's not an American would not be familiar with this. But I think your take on it is absolutely correct. There's still a remnant of that kind of thinking, of associating Putin with Trump, and therefore, you know, wanting to really smite them in Ukraine. Let's turn to just a lighter topic. Just for a moment, Benedict Cumberbatch played you in a major motion picture. And the last time I discussed this with you, you had not watched it, Have you still not watched it?

Dominic Cummings
I did. So watch it actually in 2021, after I'd left, so the time that would watch it until and unless Brexit actually happened, and after I left government in November 2020, and I watched it sometime in 2021.

Steve Hsu
I want to say to the listeners that you know, it's not it's not completely factual, it's entertainment. But I did find it incredibly entertaining.

Dominic Cummings
Well, it's it's a weird thing, because it's sort of so it's all actually it connects to a previous question, because one of the oddities of it is that the guy who originally wrote it, the guy called James Graham wrote it, having read all of the Cadwallader stuff and the observer and the Guardian stuff. And I thought that this was basically correct. So he wrote the original script of this and it was commissioned by Channel Four, essentially to be a kind of dramatized version of the Cadwallader conspiracy theory. However, I was honest, and he came to us to talk about it. And we explained and I actually sat him down with some of the data science people who explained to him how Facebook advertising actually works. And he realized, holy shit actually So, to cut what was the conspiracy is wrong, but I've sold this story to channel four. So the story ends up getting rewritten in all sorts of ways. So it ends up as a slightly odd thing where the original kind of foundations of it were that this conspiracy theory is true. And there are kind of old traces of it with a weird appearance by Robert Mercer. But the guy writing it basically realized that the story wasn't true, and then rewrote it. So it's a slight, it's an old thing in the end.

Steve Hsu
Yeah, there's

still a hacker character right that you meet with. And that's kind of a composite. I think maybe we have a physicist or two that we know. But yeah, I think they removed most of the Cadwallader conspiracy stuff from the movie, if I recall. Yeah. And

Dominic Cummings
Of course, you know, as you like, talk about flaws in the conspiracy theory version of the story, the interesting technological thing that we did was digital marketing. Whereas in reality, the digital marketing that we did was very bog standard, it wasn't technologically new at all, the interesting technological thing we did was the polling, to try and figure out what people fought and what the real groups were, what the real demographics were, that was actually the interesting thing that occurred that's not at all in the story, which actually ended up being really useful. I think, in other ways, that in 2019, we again, had just far better polling than everybody else in the 2019 election and understood what the true marginal seats were and who the real swing voters were. But no one has really paid any attention to polling since the referendum, because the Guardian fooled everyone into thinking that the interesting thing we didn't do was digital marketing, which wasn't true.

Steve Hsu
You know, we've talked about this a little bit in the past, you know, your degrees in ancient history, this specific incident that we're talking about right now, the Vote Leave campaign, which is a historical event, and heavily documented, I think there have been multiple books written about it, etc. It's a historical event, given how distorted the information that an average person can get about the things you just described, I just can't believe historians can possibly get lots of details, right? I mean, you know, I can look at what supposedly informed journalists and academics are writing about the events of the Vote Leave campaign, and it's just totally wrong. And I don't see how history is going to really fully correct. So that people 20 years from now, looking back on it, we'll have a realistic picture of what actually happened.

Dominic Cummings
Yeah, it's definitely affected my religion. Well, history University, as you say, and having been part of various things, which now people write history books about, it definitely makes me far more skeptical about everything I read, historically, for sure. And far less trusting, you just realize that there were just a hell of a lot of what you read must just be completely wrong. But also, there's just vast amounts of things that actually happened, which are just not recorded anywhere and are completely lost. Also, when the interesting things I have found are, I got very heavily into studying Bismarck, to an unhealthily focused extent, in that I've been two years waking this chronology where possible, like hour by hour of what actually happened. And what are the things that you realize when you do this is that all the main books just have many key dates wrong. And they all kind of contradict each other as the same as you know, so if you sit down and read the top 10 books on World War One, you will quickly see that, like, a huge number of key dates and timings are wrong, which therefore means that a lot of assumptions about causality around because things actually happened after the thing which they're talking about. So yeah. It's totally covered. I think it says all the time that actually you learn far more from travel and you often mystery. I think that's, I think, with the exception of a few books, I think that's probably correct. I still think you should read things like a few CDs, but I think a huge amount of what you read in history is just not correct.

Steve Hsu
I think the Polian referred to history as a fable agreed upon. So let's, let's jump ahead, I'm gonna give you a very strong recollection that I have a summer day in 2019. I'm visiting London for some meetings. And I'm in an air b&b in Camden overlooking the canal. And you come by my Airbnb for beer, and we sit overlooking the canal. And you have a sheet of paper with you, which is a list of demands that you are going to present to Boris Johnson later that evening, which were the conditions for you to join. His team was chief advisor. Yeah, you remember that? I do. So tell us what happened that day or in that meeting with Boris.

Dominic Cummings
It's a slightly complicated affair but simplifying it a bit. Essentially, I'd been abroad. When he was about to join the Tory leadership, it was clear from the polls that he was overwhelmingly ahead, and would therefore be announced as being Prime Minister on Wednesday. And he essentially texted me and said, Can I come around to your house? And I said, Sure. And he came round. And we sat down in my living room. And he essentially said, I need the Prime Minister shortly. Everyone thinks that we're sneakers kind of sucks fine, where there's no escape, there's no way through to get Brexit with our and the government somehow is going to collapse into an election. You guys somehow found a way to win the referendum where everyone thought it was impossible, you should basically assemble your old team into number 10. And somehow, let's find a way through this nightmare and get Brexit done, and move the country on. And essentially, I wrote down some key demands and said, Why don't you sign up to these invitations? They'll Okay, I'll do it. And they were a mix of how number 10 had to work and my role in it. And also things like, you've got to approve a doubling of science research, you've got to agree to create a kind of British version of the old ARPA, you've got to let me get to grips with procurements and some of the Deep State questions and civil service reform and things like that. He essentially just said, Yeah, fine, whatever. I mean, he was just, he really didn't want to be the only Prime Minister for 50 or 100 days and then get spat out. He somehow wanted to win. And he was prepared to sign up to almost anything to try and get us in there to make it happen.

Steve Hsu
And is this what you might later refer to as Boris, in self aware mode?

Dominic Cummings
Yes, he was. Then he said, even though he said, you worked in government, for you understand how Whitehall operates, you got all of these reforms done by the Department of Education, I don't understand why at all, I'm not a detailed person. I don't understand management. I don't even know what the hell the Cabinet Office is, nevermind what it does, and how to maneuver it to get what I want. There's all these technicalities of Brexit that I don't understand. But even the people who've been working on this for 30 years don't understand it's so bloody complicated. And a lot of people in Whitehall don't want us to solve this problem, they actually want the thing to collapse. They want a second referendum and how the hell we have always, we shape. We have a combination of the political problems, we've got the Brexit negotiations that we've got where we've got no parliamentary majority, the whole of Westminster is going crazy. I don't understand how the actual government and the cabinet office work. Now, this is a big complicated set of buttons all to deal with at the same time. But also, another important thing was that his girlfriend wanted us there at the time. So he portrays his workload, but his girlfriend is also software mode. And she called me up and said, No Brussels and got a Scooby Doo on how the government actually works. You've got to go in there. So she was at that time very supportive of the vote leave team going into number 10.

Steve Hsu
Incredible. Is there anything you would do differently now looking back, like different ways to modify the conditions or was all that more or less irrelevant just because the feelings of Boris and his girlfriend turned against you later, a missile sorts of

Dominic Cummings
kind of tactical things that one could say, you know, could have been done differently and Nikolaevna the world so complicated, inevitably when you are in those jobs, you know, people make mistakes every day in various ways. But I think the fundamental big picture on the whole thing was that the relations changed literally from 10 o'clock on the night of the election when the result came, it was not even the result of the exit poll dropped, showing that the ADC majority relations changed. Literally immediately, within seconds, you could feel the room change. And essentially, the girlfriend wanted us there to win and solve the problem. And they showed that they weren't kicked out. But once the majority was one, she thought, I want to control what actually happens downstairs with my friends in key jobs. That's never gonna happen when the vote leave team have their vote leave teams should be got rid of, and me and Boris can do you think I'll wait for the next four years. And he essentially was more frightened of her than he was of us. And so as relations got trickier and trickier during the course of 2020, I mean, it was already very difficult within two weeks of the election. Within two weeks of the election, there were already discussions about how she was trying to force people out, the building wasn't reporting the media then and people didn't realize there were problems really until the summer. But in fact, there were problems within days.

Steve Hsu
Now, I remember having some communication with you in the lead up to the general election. And I would always compare what you told me to what the press was reporting in the UK and the listeners to this podcast. The UK will have all kinds of questions about detailed things like prorogation of Parliament and various tactics. And my recollection is that, you know, you were setting up the other side for this general election and you were going to crush them. And the media just did not get it. They thought you guys were just toast, am I remembering that? Right? Maybe you can just embellish a little bit on what I just said,

Dominic Cummings
Well, I think the time I arrived, and, you know, when we walked in July, late July 2019, there was at the end of three years, or, you know, the whole of Westminster kind of driving itself into a sort of nervous breakdown. When I first went into overdrive, I didn't wait quite, you know, you can, obviously, I had some sense, there's just watching the news. But I didn't really appreciate it until I actually physically walked in. And the first time I walked into parliament, I saw people just kind of weeping and shouting hysterically at each other in the corridors, I had streams of MPs following along behind me dancing, literally dancing with rage, just in my physical presence. So I realized that then a very large fraction of Westminster just basically completely lost their minds in this constitution in this kind of slowly developing constitutional crisis. And we didn't know exactly how we were going to how we were going to get through that in lots of ways the fact that they driven themselves so mad was actually helped turned out to be helpful for us, because they just believed a lot of things that didn't essentially their remit the remain side and labor, very similarly to what happened with remain in 2016. By being in London. And by being surrounded by just themselves and talking to each other. They ended up just really losing connection with what the country thought. And therefore, we were able just to tell a simple story that this problem is going to be resolved. There's only one way of doing it. And it's the only way that the country can move on. And one of the reasons why our slogan wasn't Brexit is brilliant. Our slogan was, get Brexit done. And one of the reasons for that was that a lot of Romaine people who voted remain agreed with it, get Brexit done, led, let's put this behind us. Right. So Brexit, people could support it, because they supported Brexit, and they wanted it to happen. But also a lot of other people could support it, because they realized we can't have a second referendum that way. Last disaster. I voted remain. But yes, I agree with Boris, we have got to get Brexit done so the country can move on. And I think a lot of people in Westminster just didn't understand that feeling just like they didn't understand things in summer 16.

Steve Hsu
You know, one of the things I love about the British people is this. Yeah, fair play. Attitude, which is that even if you were on the other side, you were a Remainer. You a lot of people were just fair minded. They said we lost and we should just carry through because that's how the government works in this country. And so you were roping those guys in as well by saying get Brexit Brexit done.

Dominic Cummings
Yeah, I think that that's ironic, I'd actually suggested in 2050. Now maybe there's an argument for having two referendums, one referendum on the principle of leaving, and then another referendum on the actual deal. And ironically, the entire establishment went crazy with this idea, because they thought it was a trick, and they will and their stuff. They all came out incredibly strongly and completely categorically: the Prime Minister, the chancellor, the whole cabinet, the Labour people, the official Remain campaign, everybody said absolutely not no tricks from Cummings. One referendum, the result must stand, it'll stand for a generation. So they were all kind of on record saying this in summer 16. Because of course, they thought there was no way they could lose. And therefore this was the opportunity to crush the levers for 20 3040 years. But the fact that they did that they've been showing credibly explicit about it, of course then boomeranged when they then tried to say afterwards, oh, well, actually, we didn't mean that. But actually, there should be a second referendum. And that became a big problem for them. And a lot of people get the view. I didn't agree with Brexit, but you can't have a referendum, where the entire political class and Parliament votes explicitly saying, We will respect the result. And then because the result is not what they want, they just change their mind and say, Oh, actually, you know, we're gonna spend three years dicking around and then force you all to vote again. A lot. A lot of the country just regarded that as fundamentally unfair, also dangerous. Because if you really undermining the great irony, the people who were pushing hardest for the second referendum but the people who babble on most about Trump, Trump's coup in inverted commas and anti Democrat, the anti democratic nature of Trump, and of course, they're literally exactly the same people who pushed hardest to say, Yeah, we don't like the referendum so the public should have to vote again.

Steve Hsu
You can't demand logical consistency from these people. but it's way too much. For the record, I recall your data science team made some predictions about how that final general election would turn out. And maybe you could just remind us how accurate, maybe more accurate than they had any right to be, but definitely much more accurate than, you know, the public halls. Yeah, they

Dominic Cummings
probably got a Paul, but he got a bit lucky. But Sven did up, then the model ended up being accurate enough to go to the majority of dirt. Within one, it turns the total number of seats for the conservatives. And so what happened was that in 2018, we thought, Look, there's a reasonable chance that this whole thing is going to collapse in 2019. And maybe there'll be a chance for us to seize control of Downing Street in the chaos. So what we should do is we should ramp up the election models that we've started to build in the referendum, we should ramp these up secretly. So that if we do manage to grab control number 10, then we've got a key bit of technology that will be crucial for the election because you know, you can't build the stuff in a few weeks. So I went off to some hedge funds and billionaires and said, Give me half a million quid, and I'll build this secretly. And then we'll have a substantial advantage if things work out a certain way. And then between our dean, oddly, that's what ended up happening. So while we did screw up controls number 10. We also had these models built secretly in the cloud nine months previous to that, which we then gave us confidence when the election came.

Steve Hsu
So let's jump ahead a little bit. And maybe you can tell us the story of the pandemic, as viewed from number 10.

Dominic Cummings
My goodness, and where to start with that. I mean, obviously, in the first place, it became clear that the planning system thought that Britain was pretty much the best currency in the world for a pandemic and told itself that that obviously turned out to be wrong in almost all important ways. The plan was for the flu virus, not for what actually hid. A lot of the preparations that need to be down hunting hadn't been made, at the heart of it was a kind of group think that there is effectively no alternative to a single wave herd immunity by September approach, that what you were trying was doing in terms of lockdowns, and what we could see in East Asia, with aggressive action plus mass tech plus testing and whatnot that was essentially thought of not just in Britain, but also think elsewhere in the West is just not there's no practical in various ways, partly because completely wrong. So I think one of the interesting lessons is that governments took quote, behavioral science unquote, far more seriously than it should have done. A lot of behavioral science turned out to be complete charlatan, RE and bullshit. For example, one of the things that governments were told not just in Britain, but elsewhere was, while the public won't tolerate various measures being in place for more than a few days, therefore, everything has to be timed perfectly. This turned out to be complete nonsense. So there are all kinds of just completely wild ideas buzzing around. But that was quite common, I think, you know, across the western world, similar mistakes were made, you know, in the States, you had people like Trump and Caprice, saying pose the borders, no Caprice, the Model X supermodel. And of course, a lot of the public health people then said, this is completely racist. And then people said, well look at East Asia, they're wearing masks and no masks don't work. And then suddenly, the whole kind of official tooth completely flipped. And instead of closing borders being racist, the public health world that a lot of the left shifted to, we should close borders, because there's kind of one rule that mistakes happen to the kind of January to spring is one set of questions. Let me know why. I think the more interesting thing is, why were so many of those mistakes repeated. Why was it so hard for the for Western governments, for the political parties, for the media, for the bureaucracies to update? Why did things like Fauci and CDC and everything like why do they just keep being such monumental shows over and over again? Why couldn't they improve?

Steve Hsu
You know, I have a very specific recollection, maybe you can tell me if I have it, right? So when you came into office, there was already a plan for the UK to deal with a kind of pandemic, maybe it was more of a flu like pandemic than what we actually had. But the goal would have been to go for herd immunity. Yep. And that view held the day within sage until you and your team came in and pointed out. This is going to kill an unacceptably large number of people. And we need to shift strategy. Is that a fair recounting of event I

Dominic Cummings
say it's not well, not exactly, I'd say So there was an official plan, which was herd immunity. The plan, though, didn't come from Sage, which is the scientific kind of advisory group to the government. The plan came from a combination of the Department of Health and the Cabinet Office. All those two entities kept thinking that there was no, this was the only way of handling it. But the scale of deaths that this involves, was vast. And it soon became clear that people just hadn't thought through the concept and hadn't thought through the consequences of it, because it effectively meant the Official Plan. And the Official Charts, they presented the prime minister in February and early March, effectively meant that the NHS would just be completely overwhelmed by multiple factors. When we looked at them, there was a day when they presented these graphs of, you know, these waves and various peaks and said, Hey, your Prime Minister, look at this, you can see, here's an interesting capacity. And here's the projected wave. And Boris said, descriptors, nerf the capacity on it to ourselves, was it done by Mr. B, I have to look very carefully at the x axis. Because if you see, it's actually the line of NHS capacity is almost on the x axis, because of the scale, right? Because the peak of the graph was I can't remember now. But like somewhere between 500,800 1000 or something. NHS capacity was in the 1000s. So the line denoting each NHS capacity was almost indistinguishable from the x axis line. That's how just completely blown out of the water the early VHS was going to be. And then we started Miss tearing phone calls and officials coming to tell us things like, yeah, obviously, there's no storage capacity for all the bodies and people were starting to pull around and ask, can the ice weeks or be turned into temporary storage facilities for the bodies because you haven't got the body bags to bury all these people. And you suddenly realize that the whole system was kind of in this groupthink, that, we've just got to accept that there's going to be half a million people killed, actually more than half million, right? Because the official estimates are right about half a million. But then when you actually factored in when we started, yeah, but this also means no whole service for what, three months or something for anything else. So every child that has a serious accident, there's no health service. So as for them either, how many more casualties on top of just the COVID death toll, they're gonna be no one even counted that not or modeled it. And we realized that essentially, the system kind of locked itself into this group thing. But there were other parts of the system, also, including in sage, people realized that this was not sustainable. And we couldn't actually do it. And so over a period in the first half of March, essentially, me and some others tried to, essentially what we try to do is accelerate everything as fast as we could, to bias time. And instead of thinking, the only possible plan is a simple way of herd immunity by September, suppress it for the moment, buy ourselves time, and then start kind of Manhattan Project style enterprises, on vaccines and treatments. And perhaps we'll end up trying to navigate multiple waves over a year or two. And that in that second strategy is where we kind of push things to route about the Wednesday the 18th of March.

Steve Hsu
Right. So thanks. Thanks for going through that. I think I had it slightly wrong what I said to you, but the really key meeting, the dramatic meeting, I recall was you and your team whiteboarding for Boris, these potential casualties and an A against NHS capacity? And that is what carried the day?

Dominic Cummings
Yeah. So we, by round about the kind of Thursday the 12 flooded the 30s were the data science team and number 10 that we built. Plus various outsiders were starting to shout and say, Hang on a second, this is a disaster. People haven't thought through what the consequences of these things are. This is not sustainable. This is not a sustainable plan. We've got to change course. And so the night of Friday the 13th. We rolled out various whiteboards and then on Saturday the 14th we basically got Boris in a room with a very small number of people and talked him through the reality but also explained to him. Things like the Secretary of State of Health, Matt Hancock does not understand this. And he thinks and he is pushing the idea that there is only this one way this single wafer to move The plan, though it's all his fault. He's been told that by various officials, but he is the kind of Minister responsible. And we think that if we go down this path, it's just going to be completely catastrophic. And she's not going to be sustainable. In a few weeks, everyone, the country is going to demand that you change anyway. So given that it's not going to be politically sustainable, and it's the wrong plan, we should basically get ahead of it, ditch the Official Plan and move to plan the, as soon as we possibly can. And we kind of pushed Boris, quite a long way on the foot on the side of the fourteens towards that, but he was constantly trolley backwards and forwards, so you never knew if you'd actually convinced him of anything. And it wasn't really until about five days later that we really succeeded.

Steve Hsu
What was I mean, over the next five days? Was it that he was getting other input that supported your view? Or how did you finally come to a conclusion?

Dominic Cummings
I think, so I think a very important thing was that Patrick Valance, this Chief Scientific Adviser, basically agreed with us and said, the Department of Health Official Plan, the Department of Health slash Cabinet Office official plan is, I agree, wrong. And we should change the plan B. And that happened over the weekend of the COVID 12 13/14 of March. And Pat was Patrick came on board with a change, then Chris Whitty started to support it as well. And it just became, and also people in sage on the 18th. There's really a sage on the 18th, which also moved significantly in that direction. Plus, of course, I think we could see what was happening in Europe, where similar kinds of arguments have been playing out and things were also moving a certain direction, both So fundamentally, Boris shifted because he took the judgment that it was not going to be politically sustainable for him to play the mayor of Jaws, which in various ways he liked the idea of doing, he realized that if he tried to do that he was I put it to him at the time. People will march on Downing Street and hang from the lamppost if you try and be the mayor of jaws, and he realized that's true.

Steve Hsu
Wow, it's an amazing story. I don't know if this has ever been fully fleshed out. I think I follow the media coverage of this. But it's great to hear this from you. When you look back at your time in number 10, which I guess in retrospect was really quite brief in some sense. What are your greatest regrets? And what are the accomplishments that you're most proud of?

Dominic Cummings
Well, we did.

So we solved the constitutional crisis with Brexit and got that out of the way. I think if we hadn't done that, then that completely broken Parliament would have staggered into COVID A few months later, and God only knows what would have happened, then it would have been a terrible situation from which we would still not have nearly recovered. So I think unsurprisingly, I think it's good that we managed to solve that problem. Essentially, there are a bunch of things that get me and my team what you do. There were some issues around science and technology. There's a set of issues around productivity of the economy, things like zoning, planning, reform, skills, training, the whole kind of ecosystem of venture capital, startups, things like that. There was civil service reform. And within the civil service reform, those two kinds of elements of the Deep State are how that works, intelligence services, machine defense, things like that. Those are four big areas. When we started, a lot of things were happening. We created our poor, we doubled the science budget. But also lots of the things we started with essentially got closed down almost straight away, after we left, including, for example, a form of defense procurement, which is still just completely catastrophic. Notwithstanding the fact we've had a year of a year of the war in Ukraine, most things in my area have gone backwards or forwards. But it's, as I said before, it's very tricky to see how things could have worked out differently given that they almost immediately went sour after the election. I think that is a situation where the Prime Minister's girlfriend wants to get rid of you. And she's absolutely determined to do that. And he essentially doesn't want to fight with that and also thinks, on the one hand, I can live being told everyday by Dominic and his team that I'm making mistakes and I can't do what I want, and that I've actually got to make hard decisions and annoy people. And that's the only way to make progress. Or I can get rid of them and enjoy myself and Keir Starmer, my political opponents useless, and I can basically just enjoy myself and not have to worry. I don't I don't think any kind of argument can be made or you've made who made these differently. I mean, essentially, we were relying on the fact that he was too frightened to push this out. But he turned out to be more frightened than before , and to be more frightened of his girlfriend that he was worse, though, of course, we then did push him out afterwards. So he got kind of the worst of all worlds.

Steve Hsu
Would you be interested in having a converse stage with Boris now?

Dominic Cummings
No, not really. I don't think it'll be from a third of them in such a kind of parallel reality? I don't think I don't think it would be. I don't think it will be interesting.

Steve Hsu
Well, let me give you an update on one of the things that you push for. And I did a little bit of work with you on this, it was to improve the high skill, immigration to the UK to make it easier for highly skilled people. Yeah, emigrate. Yeah. And I had occasionally, you know, it's been years now. But I had occasion to look at this, because there was a physicist in Russia, who had fled the country because of the Ukraine war. And, you know, I wanted to emigrate to the UK. So I went online to look at other things that you and I had worked on that had actually been implemented by the, I guess, be the home office. And they have actually quite a positive, very welcoming, well defined set of criteria now by which, for example, this Russian that I was talking to could actually move to the UK and work. So that was also a positive thing that you guys accomplished.

Dominic Cummings
Yeah, I think we did make progress. I think it's still too bureaucratic, as you would expect. And I think if you know, if we'd been there, we could have improved it more. But yes, we definitely made progress there. And that's definitely been that's definitely been, it's definitely been progress. I think even some of the biggest enemies have voted Brexit. Now, except for some of the things we did not have been positive. Yep.

Steve Hsu
So let me ask you some general questions, which I think you're especially well placed to answer. So one thing that I've always been curious about is who really runs the UK shift? Is it big? Is it the deep state? Is it the political parties? And

Dominic Cummings
it's not money? So I think the media presents. So I'm talking very much about Britain. I'm not talking about America, where I think it's different in Britain. Donors really are remarkably little remarkably little influence, I would say. And I was actually very surprised at how little lobbying came my way from donors. I expected it to be more of a problem and then it actually was, I think, a great kind of media and MPs colluded in telling a story about how the system works, which is almost completely false. So the official story, the way the media presents everything, is that serious arguments happen between ministers, and that the cabinet is a very important institution. And those ministers really, you're in charge of departments. And big things are always presented as big. And you COVID was a classic example of this, right? In almost all big things over COVID, the media presented. What was the what was happening as minister, he thinks this administer B says that Mr. C thinks that and there's a great argument between three of them, in fact, almost always mentioned, at the time, these ministers had absolutely nothing to do with anything important. And the decisions were taken almost entirely by officials with almost no ministerial input at all. So overall, the famous British TV show which you have probably seen at least one or two episodes of is called, yes, Minster, yeah. And then the US Prime Minister, is doing to a remarkable extent, that's actually a very realistic portrayal of how the British system works. Now. So even to the extent of people who see an immediate all the time, I suppose the GA things are, you know, Secretary of State X and update. But in fact, if you go into the Prime Minister's private office, in number 10, there's a bunch of what are called Private secretaries, who were recruited from the smartest and best job officials. And these private secretaries' names are never ever in the media. And they are far more influential on how the government actually works. Then really any minister apart from the Prime Minister, and in some ways, the chancellor, so a bunch of 30 year olds who no one's ever heard of, actually, in all sorts of crucial ways, on a day to day basis and much more important than the people who the media presents us actually being important. And I think that's one of the most misunderstood things now. So, if you turn up all decisions of 100%, for more than 99% of things were decided by officials and less than 1% by politicians, but of course, in that fraction of 1%. There's also a lot of big calls and the Prime Minister and the Chancellor do have real power. But again, the exercise is remarkably small. And a large part of how the system works day to day, is a successful attempt by officials to mask from the Prime Minister was his real power is, there was one day, I remember when I was particularly enraged by one of the many insanities no civil service works is that everyone basically has moved on after two years. So you suddenly have the person in charge of disabled education or something is suddenly put in charge of managing some parts of GCHQ, or whatever, it's completely insane. And everyone knows it is insane. And there was one official who one day the document came through to the Prime Minister informing him of the shift. And I said to him, you know, you don't actually have to agree with these things. Over time that sort of, in sort of where the Austro Hungarian empire suddenly works over time, these are presented as if you're not actually in control of them, and you'll just be notified of decisions taken by someone else. But actually, is your decision just the Prime Minister, Tao never actually interferes with them? But you should just not sign off on this and say, No, I refuse to have this official actually moved. And so he did. And of course, it caused absolute chaos, because the Prime Minister actually suddenly asserted his authority to turn down a personnel move, and nothing, absolutely nothing, is of more importance to the permanent bureaucracy than HR and personnel decisions. But again, these are things that the MPs now don't even realize how the system works. And to a remarkable extent, the MPs are not actually interested anymore, in how power actually is exercised in the system.

Steve Hsu
Is it fair to characterize this as a deep state?

Dominic Cummings
I think it's, it's fair in the sense of saying that there are kind of deeply entrenched institutions, which actually practically control huge amounts of what happens with zero to very little democratic insight or even knowledge and understanding that is unarguably the case. I think the problem with the phrase Deep State is wearing morphs more towards the Trumpian. Question of AI, to what extent are there conspiracies going on? Of course, there are often conspiracies going on in the deep state. But I think you have to be careful to disentangle the two things. The deep state is real. The deep state actually controls the vast majority of government. And most of the elected people have no understanding or involvement in what's going on. That's just a fact. But that doesn't mean of course, that all everything, you know, read about this happened because the deep state or this is some kind of conspiracy by officials against Democratic against Democratic politics, that is not you know, you have to treat each of those things on a case by case basis. And I think to some extent, therefore, the phrase can be problematic, because you don't want to encourage the idea that there's a sort of generalized conspiracy of officials to thwart democratic politics, because a lot of what's happening, right, and this is where it gets super, super confusing and difficult. Often what's happening is very good, sensible, intelligent officials with genuine public service, who actually understand what the hell's going on far better than the idiots who are elected, actually trying to stop the idiots who are being elected, doing terrible things. And that happened during COVID. You know, you could, a lot of the things that were best in COVID, was the deep state thwarting Matt Hancock, the election politician for the benefit of the country.

Steve Hsu
You know, I could give some distance to this issue by talking about Japan. So in Japan, the ministries are staffed by the top graduates from University of Tokyo and top universities. They're famously competent, but also very conservative and hidebound. And the politicians, you know, it's openly discussed that the politicians have very little power to override these powerful ministries. So you could take two different views about this one is an anti democratic view that yeah, the power resides in unelected officials. On the other hand, you could take some kind of elitist view saying, Yeah, but these people know what they're doing.

Dominic Cummings
Exactly. You know, there's a phrase, one of the great things about Russia is the kind of cynical phrases that come out of it. And one of them is everybody's right, and everybody's unhappy. And if you think about this problem, okay, as you say, there's, you can say, on the one hand, at least in its own sense, the fact that the quality of the elected people is so desperately bad now, across Western governments, that, you know, brilliant 30 year old women, who no one's heard of, and no one elected, actually wanting things, in many ways is for the good, right. But on the other hand, the long term consequence of this is also that these institutions just become incredibly stale and self reinforcing. And that almost nothing can change in any way, including by the deep state itself. So one of you know, one of the things that would happen with me, which was interesting was that parts of the deep state would come to me to try and break a logjam, but no one else could break, it would effectively say, look, x needs to happen with you know, whatever the intersection of MI six, and the Cabinet Office or whatever. This entity conflicts, if that entity can't fix it, no entity can fix it by itself, the only person, the only force in the British Constitution with the wage to actually change this is the prime minister, because everything else is kind of canceled out in his veto autocracy. So that's one of the other things I think, which is now you know, which is part of this discussion about the deep state is, it itself becomes sort of stable and hard to shift in important ways. That the only people that can break through it, then are potentially some alliance of politicians with parts of the deep state, right. It's just that this is why I think you were talking about before and the whole COVID thing. Why was it? Why do you see things like the CDC, just stay terrible? Or just keep making these catastrophic decisions like on mass testing? Why could it not change a one of the most fascinating things about the whole of the Western world, it's not just a British thing, or an American thing across the whole Western world, you see, effectively the same story, that opposition political parties have a huge political opportunity repeatedly during COVID to stand up and say, What the hell the people doing? Why can't you grasp this bureaucracy and for example, make it do make it procure math tests properly, sensibly fast, and thereby help the economy and help Delve system these things are not in opposition. And effectively, no, no opposition party anywhere really demanded a radical shake up of any of these kinds of big bureaucracies and the CDC, similar regimes in Britain, across Western Europe, they've all basically just stayed in place, and the political parties, the old media, and the bell, bureaucracies effectively have decided, you can see and get revealed preferences, we'd rather just keep going the way we all then actually bite the bullet and try and change and become high performance.

Steve Hsu
Right now. You know, with the Republicans taking control of the House in the US, you're gonna see some hearings, where they're looking into potentially improper behavior by the FBI and interfering with the election and the NIH and funding research that may have had something to do with COVID pandemic or future pandemics. But on the other hand, it's not I wouldn't bet any money that serious changes kind of result from these hearings.

Dominic Cummings
Yeah. I mean, so far, we've just watched what's happened so far. You wouldn't bet on serious change happening, right? I mean, if it even seemed to me that the parts of the American system were trying to slow down the deployment of vaccines because they didn't want the vaccines to come out sooner and help Trump. I mean, I've seen that and it seems plausible to me.

Steve Hsu
I think that definitely happened. I think they delayed approval, or some data release, basically, because they didn't want to help Trump. Yeah. You mentioned mi six. So I have a question for you. Our intelligence briefings are actually useful to the PM. Does the PM actually know more about the state of the world than a hedge fund manager, hedge fund billionaire or Elon Musk?

Dominic Cummings
That's a fascinating question. So let's not read a huge amount of these briefings actually When I was there, because deliberately, one of the reasons why we didn't place in the first place was to try and shake up the system. So they could deal with very big questions like pandemics, nuclear nuclear issues and whatnot. So I spent a lot of time reading these things, I'd say, it's sort of a funny mix. So on one hand, there's no doubt at all that there are very powerful capabilities that Britain and America in particular have. But in general, the Five Eyes system that evolved out of World War Two intelligence agreements, very powerful capabilities, which they'll hedge fund has on which even people like Elon don't have AI, which is illegal for them to have, right. So if you're the British Prime Minister, or the American President, then you're doing you have a kind of vast automated system to do what people like Matt, Nick in Bismarck, were doing ie stealing their enemies bins, you know, they would have spies that would go through the bins and pull up bits of paper and copy them and networks to distribute the intelligence from this. and Western states will have these definitely powerful capabilities. But I'd say they're very badly aimed, you can improve the value of them by 10x, or 100x. In various ways, the trailblaze deprioritization process is appalling where these capabilities are aimed at in all sorts of ways is incredibly bureaucratic and incredibly bad. Also, I think there are lots of ways bureaucracies have become very risk averse, though, in various ways, partly in the States, or happened post Watergate, I'd posted all the investigations into the CIA in the 70s. I know, you could argue in various ways this was a good thing in the States, right, in terms of a whole bunch of murky ships the CIA was up to but it definitely is made a lot in large parts of the system, risk averse. But I think the most important thing is that you have this amazingly valuable material and all sorts of ways. And very powerful capabilities, which you could aim for better than it is aimed. But they don't connect to a wider system for interpreting the world. So a huge amount of the information that comes out of these things is kind of printed out old school, then incinerated, it's not really used, it's not really learned from like the idea of, for example, taking the corpus of a lot of the stuff and then training a large language model on it or something. Like if you were actually cutting edge, and you had a very aggressive and very high performance, deep state working now, then you could imagine all kinds of fascinating things that you would now do with large language models, right on lots of this material that the intelligence services are stealing from around the world. They, in fact, know that the technology for the decision making process throughout doing that is disastrous, huge vetoes everywhere. Politicians don't care enough to Trump to overcome the vetoes. So yeah. So you have all this information and possibilities, but yeah, the kind of core deep state institutions that are trying to sense the world make decisions about it, think it through and apply it and act, I mean, that, I would say the vast majority of possibilities are wasted.

Steve Hsu
Yeah, regarding the LLM thing, we're definitely going to do that. The picture you paint, though, is a little disturbing in that it sounds like there are incentives, maybe or not even that strong, for the people running the Intel services to improve to make their product better, because the politicians, their bosses don't really care that much.

Dominic Cummings
Yeah, I think that I think on the one hand, the politicians just don't really care that much. There's just so much more interest in the bullshit of domestic media things that are happening in the next few days. Also, though, the officials are not incentivized by the politicians to take risks. And that, you know, inevitably then filters down through the system. And then that becomes self reinforcing why people who are places like si es, or GCHQ, or the CIA or whatever, then get discouraged because they think, Oh, do I want to sit around here for 2030 years, just having an idea after idea crashed operation after Operation not approved, then when something does happen, and there was a little bit of blowback in the media, everyone panics and collapses. So then you have the classic kind of vicious circle where a lot of the edgy people with interesting ideas end up leaving, and a lot of the HR lawyer types end up getting promoted. And that's, you know, that's, that's bad. And then it means that you end up in very difficult situations like we are now occasionally coding, then the entities you're dealing with, in some way, right? You see this problem all the time with the military every time a war starts. It's the same cycle, right? You have decades of peace before the war kicks off, and then you start the war, and then you realize the whole bunch of generals promoted are completely useless. And then you need someone like George Marshall, who then fires a bunch of them at the end of World War Two.

Steve Hsu
You know, when I would visit your house and we would have a serious conversation, we would put our phones outside when I went And into number 10, they would take my phone and put it in a little cubby hole to the left of the door when you come in. Yep. When I walked up to your house, sometimes at night, I would see very strange looking minivans parked up around the corner from or just near your house. I'm curious, did you ever see any reports when you weren't number 10? That so famously, they monitored Angela Merkel's phone that Americans might have for a long period of time. Did you ever see any actual point reporting that British intelligence was actually intercepting interesting messages?

Dominic Cummings
Oh, so I would say offensive cyber capabilities are extremely powerful. And anybody listening to this should be extremely cautious about what they use their phone for. But I don't think I'm afraid I can say anything about any particular things that I may or may not have seen.

Steve Hsu
Good. Good. Sorry. It didn't mean to put you on the spot with that. That was the scariest question. Yeah,

Dominic Cummings
I just think I foolishly answered it. Now. That's good.

Steve Hsu
By the way, we're close. We're approaching two hours. And I have a lot of more interesting stuff to discuss. But I want to be conscious of your time and not to tire you out too much. So let me know how many more minutes you think you want to go?

Dominic Cummings
Like you definitely do another half hour? Oh, great.

Steve Hsu
Okay, perfect. Okay, you predicted to me Well, in advance that Liz truss would blow up quotes, like a human hand grenade. Yeah. Why are so many politicians the worst? People tweet

Dominic Cummings
countries? Well, I guess the system is like what they were selected for? Right? I mean, essentially, what these all political parties select for people, he wants to be on TV. And he wants to give speeches that are happy to, to optimize for claiming the greasy pole of the party hierarchy. And that's about it, they're not selected for, can you actually build an organization? Can you actually do something of serious value that's just not fixed, not the kind of people who are selected for by the old political parties across the western world. So that's where we get the trust is a particularly extreme example. And the issue is, potentially, if there's one MP in Parliament, who is even more obsessed with the media, and even less interested in the real world than the media than Boris Johnson, it was less trustworthy. But what does it tell you about how the legal system works, that's who the MPs turned to. You have a complete meltdown of a government and a prime minister, precisely because the guy is so completely absorbed with the media that he just completely loses grip of the actual government, and the grip of reality. And the MPs' response to that is not to go to someone who actually discriminates against the government seriously, or, and is going to prioritize actual reality over media reality, that they pick the one person who arguably is even worse in the same direction than Boris Johnson. And I think that sums up a lot.

Steve Hsu
So you're very detailed, first hand knowledge that this is the case, this very pessimistic scenario prevails in the UK today. Do you think this is a universal phenomenon that the more or less the worst people are leading each of the Western democracies at any given time?

Dominic Cummings
I think I certainly think it's a general problem. I'm not sure it's as quiet as bad in some countries in the West as it is in Aspen in Britain in the last 20 years or so, but it does seem to me that it is a general problem. And I think it's partly connected to there's a few deeper things going on, that are kind of pushed the system in this direction, that in that, if you go back, say 80 years, and you took a kind of regular set of people from places like Oxford or Cambridge, a large selection of them would say they're going to go into public service of one kind or another. Now, though, I think if you go to the smartest people who could do all, you know, a very wide choice, a tiny number of them now have any interest in going into politics or into the civil service. And I think that that long term trend is part of what's working, working itself out. A huge number of these people now go into some combination of mouths, money, technology, hedge funds, VC, etc, etc, right? The worlds that you spend a lot of time with. Very few of them go into the politics of the civil service. And over time, I think that's partly what's responsible for the situation where you have simultaneously, not just terrible quality of people at the top of political parties, but also clearly a much lower caliber, average caliber of people in crucial jobs. You know, like it there Ministry of Defense or whatever it might be,

Steve Hsu
yeah, I feel like the best generalist talent these days is going into tech startups and such. Some are still going into hedge funds and money management, things like that. But of course, there's always academia and science. technology really is sucking in I think a lot of the very best talent these days,

Dominic Cummings
yet, it's going to become sort of selfishly, right because the more politics looks like a clown show, and the more the media works the way it does, and it's kind of destructive way, in general, but also the way it goes up to individual people to try and break careers and destroy their lives. It also means that a lot of very able people who've got a lot of choices are, well, oddly, the most effective. The psychopaths who were just obsessed with being in the media are bothered by that. And they still go into politics, right? But if you're, if you're a very abled person aged well, like Patrick Paulson type person, right in your early 20s, you could go okay, well, I could go into this world where it's very hard to have any actual control of anything and best I've got to kind of sit there playing appalling gay bureaucratic games for 30 years. And then maybe if I don't get unlucky, I get to some senior position where I can have some nonzero effect. Or I can go and start stripping, and build it and build a you know, my own walled garden that can operate in a sensible way. And I can hire great people and etcetera. Because this choice becomes more and more self reinforcing, I think it's a huge problem in America, right? You've got all these extraordinarily able people, a lot of Foom, in the same at 30s. In the 1840s, he volunteered to go to Washington to help build things to solve various problems. But now, the general groves characters are in the valley doing different things, they're not one of them, or very few of them were in Washington,

Steve Hsu
It's a real problem for the way the government, the country is governed. Maybe an economist would say, this is great, we've got the most able people going where they really have leverage to create new things. But I think the government obviously has huge leverage, too. And there's a huge value for having talented people go into it,

Dominic Cummings
I think, especially in a relatively small number of areas, right. So I think where it matters most is things like, around thinking through nuclear war, thinking through nuclear security, you know, even when I often say to people who asked me, you know, what's likely government and what model? And I say, well, always remember, if they don't take nuclear security seriously, then she never assumed that they're taking anything seriously. And people often kind of assume it's a metaphor, but it's not only it's not I don't mean as a metaphor, I mean, it actually, actually is a statement of truth. And then if you look at the states, every few years, there's a kind of report on nuclear security. And you find the same things over and over again, right? There. These places are easily penetrated, you hire some hot Natasha, who goes in there and honey traps people and they always fall for it. There's constant stories of nuclear weapons being left on the airfield overnight, because they got this protocol wrong or whatever. So I think, yeah, in all sorts of ways. It's good that people like Patrick Collison are off doing stripe, and not in some, you know, stupid Department of Education job in Washington. But you do need a set of people like that in critical areas. And to me that's where that's the most worrying and dangerous thing, that we have a combination of very dysfunctional bureaucracies. And this talent problem is truly the most critical area of bio defense, nuclear, etc, etc. At some point, you got to think you gotta run out of luck with that.

Steve Hsu
Yeah, I shouldn't digress into this. But the thing that's gotten me very worried lately is the research on gain of function and creating much more dangerous viruses. It's still ongoing, and I think we're, we're playing Russian roulette with the future of the human species. But that's like going to that. You famously tried to hire quant oriented misfits and weirdos at number 10. Give us an idea how that went.

Dominic Cummings
It went very well. And we started ironically, you know, when I wrote that blog on, I think it was called the exact date but something like the second of January or Sunday 2020 turned out, of course, the COVID was already circulating. You know, COVID was already living in W an at that point, and possibly spreading to Western Europe by then as well. They wrote that blog saying, you know, we need people in here and we need to need to shake up the deep state systems, particularly things around what the EAA community referred to as x risk, you know, these kinds of problems like bioterrorism or whatnot, or we start to recruit people that unfortunately, it wasn't quick enough to deal with the situation in January, February. But we declared the number 10 to date service team to be extremely successful, amazingly cheap, not very big as it shouldn't be, you know, that sort of thing should be small and elite and then outside that we recruited some other people as well. I mean to be generally encouraged people to just go through the mainstream systems rather than to do it kind of ad hoc. So a lot of the people, a lot of the weirdos and misfits are regarded by the Civil Services as normal employees, which is also good. The sad thing is a lot of the very good goes back in a way to what we're just talking about. The sad thing is that a lot of the very smartest people who came in most of them have left now. Because vault by definition, the people with the greatest number of options as to what to do, some of them have left and gone to the valley, inevitably, some of them have gone back to academia. And people like that, you know, they don't want to just sit in rubbish institutions watching things be closed out. And just bad decision after bad decision. And so that's the sad thing. So I think the experiment, unsurprisingly, involved getting very smart, technically able people and putting them in positions to look at data and advice on technical questions. Unsurprisingly, that was successful, and parts of the civil service system welcomed it. I was happy about it and thought it was great. But it's, it hasn't been, it hasn't kind of embedded to the extent that it needs to in the system. And in various ways, Boris kind of diagrams. I mean, ironically, this truss on a very first day moved out the data science teams in number 10, and took a whole set of decisions on her first day that just without her even realizing just destroyed her old power over the Whitehall system. Again, going back to this question that the MPs just don't really understand how power works. One of the things that she did was to remove the data science team from her own office on day one. So yeah, I think, overall, though, a bit more hopeful in that, in that, when I wrote that blog in January 2020, it was seen in Whitehall, as mainly as it's sort of, you know, sort of esoteric, or one of Dominic's ideas. But within weeks, a whole bunch of senior officials had come face to face with the fact that they had no effective data system to deal with COVID. You know, by the end of February, beginning of March, the data system for COVID, the Prime Minister was the head of the NHS reading out numbers that had been faxed in from a bit of paper, and me scribbling them down on a whiteboard, and then hitting tabs to on my phone and then writing the new numbers up. And because of what faculty AI did, in building a new system for the NHS, very quickly, a bunch of C officials saw the difference between the old world and the new world of actual cutting edge data science and actual proper dashboards, etc, etc. So I think one of the one of the positive things about COVID Was that a lot of the old system kind of didn't bring into the idea of upgrading technologically, though, of course, actually doing that in practice is very difficult.

Steve Hsu
Yeah, I was always kind of amazed at how primitive the systems in the number 10 Were like, it really didn't seem to be any kind of sewing room where you could really engage with the entire outside world from that room. When we

Dominic Cummings
started. There wasn't even a cloud system in number 10, there was this loser, there wasn't even a document sharing system like everyone in summer in July 2019, where we walked into the door. The way that number two, the key number 10 staff working for the Prime Minister word was to email attachments of Word documents to each other. That's how, you know, not even rubbish businesses. We're doing that right in summer 2019. Even a rubbish business could use Google Docs if the Prime Minister didn't even have access to something like Google Docs, never mind. Anything else.

Steve Hsu
Incredible. All right, another quick question. If they listened to you, could you win the next general election for the Tories? Oh,

Dominic Cummings
my goodness. Well, I'd say Kiersten was bad enough for politics. And is, is like is so essentially normal for Westminster in the sense that his reality is media reality. That I think if you had top notch campaign team working in numbers, and then you could definitely knock Keir Starmer very far off course, you could break his whole decision making system, you could shake up the game, you definitely have a good shot at shaking the whole thing up enough that you can make it competitive, for sure. And then if you can make it competitive, then you know the best team has got a chance of winning. The fundamental problem, though, is similar to the referendum as soon as 2019. The problem in doing this would not be GSR there. The problem would be that executing the actually rational best strategy would arouse the rage of the Tory MPs and the need. Yeah, so much there, I just I can't see, I can't see the government actually doing it. I think it's more likely that the current government will will just will fail respectably in the sense of they will try that they won't antagonize the inside a world, they'll try and, you know, be seen as the civilized people by the inside a world and therefore, effectively, the government will not control the government, they will actually be able to get very much done, they'll fail, but they'll fail conventionally and lose conventionally. And they'll all basically be happy with that. Because the alternative of actually trying to win would be seen as crazy by the inside world. And it goes back to what we're talking about right at the very start, right? What are you actually optimizing for? Most of these people in politics are optimizing really for kind of getting along with people that they see every day? They're not actually optimizing for winning?

Steve Hsu
Yeah, that's why I said, if they listen to you, could you win the next election? If they don't listen to you? Of course, there's no point. To the extent that this is going to be a winter of discontent within the UK because of soaring energy prices and economic problems. Is that going to be pinned on the current administration? Or is the blame going to be spread across both parties?

Dominic Cummings
No, it'll be blamed, I think overwhelmingly on the conservatives, and to a large extent, reasonably, I mean, you could definitely say that decisions taken 20 times before, particularly on energy, by labor, were also very stupid. But, you know, fundamentally, the public attitude is going to be that you guys have been here for 12 years. And you should have solved these problems out. And instead, you screwed up, you know, almost everything is screwed up by the NHS, you can't even control a bunch of small boats coming across the channel. You put in Liz Truss, who blew everything up and cost us all a fortune in a few weeks. It's your fault. And so I think that's a fundamentally reasonable judgment, right? The fact that labor screwed up a bunch of energy things is interesting for the history books. But I think it's fundamentally reasonable for people to blame the Tories for the mess.

Steve Hsu
Yeah. So I think it's pretty hard for them to overcome that in the next general election.

Dominic Cummings
Very hard. I mean, very, I think, impossible. Barring a very odd approach. That word, as I said, I think would seem pretty crazy. But certainly any kind of conventional, any kind of conventional approach to government or any kind of conventional approach to politics, for the Tories now, is just going to run into the fundamentals and the fundamentals are they going to have been there for 14 years, they're going to be asking for 18 years in power and a fifth term in power. It's never happened in democratic politics. In British history, time for change is going to be extremely powerful. And therefore, to overcome those fundamentals, you'd have to have a very strong counter story. And that's unlikely to develop, I've got a

Steve Hsu
quote to you something you wrote on your sub stack, which by the way, I recommend to everyone who's enjoying this conversation, there's incredible essays that Dom has written over time on his sub stack, the most recent one, I believe it has quite a bit of content about the American political situation. And let me just quote from what you wrote. And here you are describing the way that the consensus view might be changing among the Democrats. So the idea might be that Trump is preferable as the 2024 candidate, because he's more beatable than DeSantis, the crazy Magga candidates got whacked in the midterm. Maybe we'd have a better chance of holding the White House if Trump's the candidate. Yes, much of Trump is Fascism is a fraud. If DeSantis runs, I predict the Democrats in New York Times will forget that Trump is a unique threat to the democracy line, and pivot to DeSantis is even more dangerous than Trump because he's more competent, but believes the same fascist things. And if you're fighting fascism, then you're entitled to do all kinds of things, right? Like deploy the CIA, to spy on campaigns or FBI to spy on campaigns and make sure those pesky Russians aren't stealing the election for their fascist ally. So maybe just react to that and tell us what you think is happening in American politics.

Dominic Cummings
So I mean, obviously I could well be wrong about this. But I got the impression that if you go back a couple of years, I think there was a general kind of mainstream inside of you, which was that Trump's finished the January 6 thing means that like he's just done and now we're all gonna move on from that. And Biden's gonna be a success for Biden, and they won. We thought that January 6, meant that Trump was not a viable candidate anymore. When they started to realize their mistake on that rhetoric then during trade 21 just saw treasury dues all kind of ratcheted up like Trump launched a coup, he's a fascist, it's completely impossible that we can let him have him back, what's the end of it? It's the end of American democracy, etc, etc. Now, when I came over to the states a few months ago, there were a lot of Democrats talking like that. But I have detected an interesting change of tone from them since the midterms. And my impression is that quite a lot of them now, privately, are thinking, actually, it's not so bad for the Democrats. It's actually better for the Democrats if Trump is the GOP candidate, because the research shows that actually, he's much more beatable than DeSantis says. So. Obviously, a lot of people actually have a genuine Democrat side where they say we think Trump is totally evil, and we think he's fascist and it's the end of American democracy. If he's president, again, they believe that. But I think it's also clear that a lot of people say that, but actually don't didn't believe it, don't believe it. Or they're just prioritizing something different and they're prioritizing the Democrats winning. And you can see this to some extent in the argument about there's an interesting internal Democrat argument before the midterms about should we the Democrats help the crazy the craziest maggot candidates. So some elements of the Democrats tried to help the worst, most insane male candidates win the primaries on the basis. Babies used to be in the midterms. Interestingly, one of Obama's closest advisors, David Axelrod, said publicly, you know, I'm pretty uncomfortable about this, and I don't think we should be, I don't think we should be doing it. You know, like, if we're telling everyone that Trump winning is fascism, then why would we be enabling the fascists to win elections? It seems like a very dangerous gamble. But you can see people on the Democratic side now looking at it and saying, actually, this tactic was short term effective. The moment we helped the crazy Miko, and the crazy Miko were indeed, to do much more beatable in the general election than the more mainstream Republican candidates. So I think possibly I've been too cynical. But my hunch is up in a different way. A lot of people were saying to me four or five months ago that the situation is terrible. What else can we do about Trump and stopping Trump becoming president? Again, a lot of those people have gone very quiet. And a lot of them are now saying, reading between the lines, I think quite a lot of them, but will you be happy if Trump's the GOP nominee or do you think I'm crazy? They do consume pressure from anybody?

Steve Hsu
No, I think actually, it's funny the guys that I talked to, I think what you're saying is becoming kind of the conventional wisdom. Well, some of these guys are Republicans saying that we got to get Trump out of there. Because with DeSantis, we can win with Trump, we can't win. Yeah, exactly. So I don't think it's crazy what you're saying, also a thing.

Dominic Cummings
You know, I've been talking to a few people actually doing market research on this, right. And what I've been told by people who are actually starting to research this, in a serious way, is that you can see these dynamics already in focus groups that you can sit down with Republicans and Republicans will say in focus groups. Now, basically, Trump did a great job with the economy. But you are still arguing about 2020? The country wants to move on, you know, my own sister's not gonna vote for Trump again, because he's still going on about 2020. Screw that, you know, if it was Trump versus Biden, obviously, I would vote for Trump again. But no, if DeSantis goes for it, then Sanders has probably got a better chance of winning, the most important thing is to get the crazy Democrats out of the White House. So yeah, if it looks like DeSantis has got the best chance of beating the Democrats, then you know, I'll go for DeSantis even though Trump plans, right? So I think if this is why what I'm being told by people doing this research, then this will filter out as well, right? Because the objective market researchers objective market research, these facts will start to become clearer to people. And therefore, I think they'll spread around the repub they'll spread about on both sides. Right. So the rational response to this from Republicans is, let's try and make this campaign as powerful and sensible as possible. Let's actually try and plan ahead so that if he wins, he can actually, the government controls the government in Washington for the first time in many decades. But conversely, the rational approach drawn from a lot of Democrats might be well, how the hell do we help Trump win the nomination?

Steve Hsu
Yeah, I agree with that. I don't know how actively Democrats are going to pursue that strategy of trying to help Trump win the primary, but I do think DeSantis is the more formidable candidate, I think, to get the nomination.

Dominic Cummings
Yep. I think you know, Trump made as I said, in that piece, I think Trump made a fundamental strategic error, or the term critical is the mid term saying, Look at what I did on the economy. Look at what Biden did on the economy. Here's the way in which he screwed it up. If I was in charge again, here's what I would do to get these turned around right now, that would tap into the fact that a lot of people, including a lot of people who voted against Trump, think that Trump actually did a good job on the economy. And Republicans think that's the single best reason to support him, okay. Instead of like making a mess, basically, his message on what was just empirically, something which got the most support from across the political spectrum. He spent his whole time arguing about the 2020 election, which even his own fans don't want him to do on his own. If you go back, democratic elections are basically always about the future. Okay. Look at the most striking example of this is Churchill in life, we find you can literally defeat Hitler, and then be turfed out if the public thinks that you haven't got the best plan for the future. Right like Joe. So if that's true of Churchill, after literally beating Hitler, it's basically you have to assume it's always true, right. But Trump kept arguing about 2020. No one wants to have arguments about 2020. Everyone wants to move on. So he really shops off on both feet.

Steve Hsu
Yeah, I agree with you. I do think that the primary will be quite challenging for DeSantis if he decides to go for it, because, you know, debating Trump and dealing with all the Trump supporters is not going to be a cakewalk.

Dominic Cummings
totally great. It'll be tough. Sure, tough. I don't know what kind of campaign he's got, or what kind of people he's got. And, you know, this is all the vote where the white campaign is sometimes, like 94, or something with Reagan re election, you could say, actually, the campaign just doesn't really matter very much, he's clearly going to win by a lot. And the campaign's not gonna affect things very much. But other times campaigns are important. And I said big, strategic questions like, to what extent do you attack Trump? To what extent do you not, how do you argue? For example? Here's an interesting question. One of the things that Trump did very well in 2016, was, he didn't do what most Republican candidates do, and promised to privatize Social Security and make huge cuts in Medicare and everything else, you wipe this out? Well, Republicans constantly say they're going to do all these things, but they never actually do it. So it's kind of the worst of all worlds, you end up spoiling the government, but you lose a load of votes from people who think you're going to do something radical. Trump didn't do that. Similar questions will be faced by DeSantis. If he goes for it next year, and figuring out the answers to those kinds of big things, almost everything you do campaigns doesn't really matter very much in the end. But deciding how you handle a few very big questions, ends up being the difference between winning and losing.

Steve Hsu
Well, damn, I kept you a long time. And I realized that you have to go and do some other stuff. So I appreciate your time on this podcast. I know there are a lot of British political junkies who wish that we had covered a lot of other things who we didn't have time for. But in any case, I wish you Merry Christmas, and we'll be in touch later.

Dominic Cummings
Let me thank you so much. It's been such fun, and I hope you do your traditional Christmas Eve blog, which I look at every year. And I always know that that's the official start of Christmas. And maybe we could do this again in 2023, where we can watch what's happening with the emerging primaries in the 2024 campaign.

Steve Hsu
All right. Thank you very much.