Latest / Second Nature / E185 Benedict Macon-Cooney: Winning The AI Century

Transcript

E185 Benedict Macon-Cooney: Winning The AI Century · Second Nature

• 0:06
  SPEAKER_02: You do have some industrial clusters around there already in sort of the clean energy space, typically around Newcastle, the offshore winds.
• 0:13
  SPEAKER_02: All as well in terms of the scenic stuff.
• 0:15
  SPEAKER_02: So I guess if you can begin to build something off the back of that, there is some potential.
• 0:21
  SPEAKER_02: But I don't, you know, I I think the politicians, number one, you know, they get accused of not taking enough of the sort of let's say the other parts of London apart from London seriously enough, right?
• 0:33
  SPEAKER_02: And if you look at some of the economic activity that is happening or is possible in some of these, you know, Manchester obviously being a prime example, Leeds.
• 0:43
  SPEAKER_02: Um, I do think there is some opportunity.
• 0:45
  SPEAKER_02: And I mean, there's a guy, have you ever read Tom Forth before?
• 0:49
  SPEAKER_02: Don't think so.
• 0:49
  SPEAKER_02: He runs a little data company up in Leeds called the Data City.
• 0:53
  SPEAKER_02: And he's very um like he he thinks you're you get more ROI on your investment in things like RD expenditure in the north, etc., because there are so many structural barriers to it within the South, i.e., like you can't build any housing in Cambridge or Oxford or even any lab space, for example, whereas it's much cheaper.
• 1:13
  SPEAKER_02: The problem is still a bit of the talent agglomeration, right?
• 1:17
  SPEAKER_02: Which is just still does cluster, you know, not to I know there's very talented people all across the country, but it does still tend to come around a few of the sort of the core hubs.
• 1:26
  SPEAKER_00: So it's kind of an equivalent argument that uh Demir Sabis m uh made when he fought to keep DeepMind in the UK, arguing that I I've got an asymmetry of talent.
• 1:37
  SPEAKER_02: Yeah, and cheaper talent as well.
• 1:38
  SPEAKER_02: Yep.
• 1:38
  SPEAKER_00: Massively so.
• 1:39
  SPEAKER_00: Yeah.
• 1:40
  SPEAKER_00: Um Bendit, would you like to tell people listening to this who you are and what you do?
• 1:44
  SPEAKER_02: Yes, so I am Benedict Macon Cuney.
• 1:47
  SPEAKER_02: I am the chief AI and innovation officer at the Tony Blair Institute, uh, which uh does a couple of things really, which is we're building AI tools for political leaders and governments around the world, and then help them really with that sort of change and adoption of what is a very sophisticated technology to basically improve how public services and how governments operate.
• 2:13
  SPEAKER_00: Um I mean, it's like the best job to have at the most important time with one of the most influential organizations on the planet.
• 2:19
  SPEAKER_00: So it's a pretty damn cool job to have, right?
• 2:22
  SPEAKER_02: Well, I think so.
• 2:23
  SPEAKER_02: Yeah.
• 2:24
  SPEAKER_02: It's I mean, I've been with Tony now for 13 years and I've been very privileged to go through the whole growth of the organization from you know what was a small institute now into this global place with 45 countries which we work in.
• 2:39
  SPEAKER_02: You know, one of the interesting things I think has always been uh what's been Tony's political talent is you know, is on the modernization, spotting the big macro trends, how the sort of political systems are changing.
• 2:53
  SPEAKER_02: And I think when we saw AI come along, it was this is the thing, you know, and as he wrote in his essay last week, you know, almost everything else is second order.
• 3:04
  SPEAKER_02: And it's not to say that they're not unimportant things, but it's like if you crack this technology challenge, you can then get so many benefits in many other areas, including in climate and energy in particular.
• 3:16
  SPEAKER_00: Can you give people a bit of a whistle stop tour to your background and how you end up in the role you're in now?
• 3:22
  SPEAKER_00: And if you want to weave some of the growth of the institute along with that in parallel, please feel free.
• 3:26
  SPEAKER_02: Yeah, sure.
• 3:27
  SPEAKER_02: Um, so I mean, I started out my career in the Treasury, you know, very fresh out of university, as a lot of people in the Treasury are.
• 3:37
  SPEAKER_02: Uh you know, these young people that are hopefully quite bright, uh thinking about, you know, really, I was doing macroeconomics and microeconomics at different points in that period, as being like on what was called the fast stream economic service.
• 3:52
  SPEAKER_02: Ended up with a bit of speechwriting.
• 3:54
  SPEAKER_02: But I mean, the backdrop to all of that was uh I started during the financial crisis, uh, or just before it started, really, and then sort of cut my teeth through that era, starting from obviously the Labour administration, just as Gordon Brown had taken over as prime minister, then through to uh the coalition government, actually, in sort of the beginning stages of what was the austerity agenda.
• 4:16
  SPEAKER_02: Um and I think at that point, I, you know, been there a few years straight out of university and thought, I want to go and see a bit more of the world, find out what other things are like before I get too institutionalized, despite it being like still one of the most intellectually stimulating and interesting periods of my career, and did a bit of work um overseas, um you know, including then in President Kagami's office in Rwanda on what he was doing with a digital transformation programme, a period of change in government, i.e., things like he did a zipline blood delivery service, which has been quite renowned as a bit of innovative policy making, but also things like uh 4G Korea Telecoms deal, etc.
• 4:59
  SPEAKER_02: So How long did you live outside the UK for?
• 5:01
  SPEAKER_02: So I was four years in total.
• 5:03
  SPEAKER_02: Um and then yeah.
• 5:06
  SPEAKER_00: How um how much did the you that left change as to the you that came back?
• 5:10
  SPEAKER_00: What was that?
• 5:11
  SPEAKER_00: I I'm I asked the question purely self-serving.
• 5:13
  SPEAKER_00: I've always lived in the UK.
• 5:14
  SPEAKER_00: Yeah.
• 5:15
  SPEAKER_00: So I wonder how much an experience like that changes you?
• 5:18
  SPEAKER_02: Um it does, but you know, you do revert to the meme very, very quickly when you come back as well.
• 5:24
  SPEAKER_02: So I I think my experience on a lot of this, because I was living in a very, very, you know, poor country, although growing very, very fast, um was you your primary elements that you're always thinking on there was just very simply how do we make people sort of richer and more prosperous, which is almost sort of it's what we still should still be doing in the UK, but it's not where the primary level of debate is.
• 5:50
  SPEAKER_02: Um there was obviously a big security component in Rwanda as well, because of it was still, you know, it's sort of you know, it's two decades on from the genocide, but that's that's still the recent memory for most of the the population.
• 6:04
  SPEAKER_02: Um, and I think I remember coming back to London at first and sitting with some mates, and I think we were in a bar in Shoreditch on a rooftop, drinking probably overpriced drinks.
• 6:12
  SPEAKER_02: And I remember thinking, like, oh my god, this conversation is so decadent, you know, comparative to what I've been experiencing.
• 6:18
  SPEAKER_02: Um, but you know, it it just I I think the sort of the perspective that I then gave you was just seeing how other parts of the world operate and how they see things.
• 6:28
  SPEAKER_02: I mean, even if you connect this back to something like the AI debate today, you could go to a country, and you know you think they're very poor and you know, you think they're not really thinking about these things, but I could tell you they're thinking deeply about them because they're going, how do we jump up that economic development chart in the same way that during globalization, you know, you a country like Vietnam were saying, right, we're gonna, you know, gun up the manufacturing chart, and France was then having Gilets jaunes protests against it as well.
• 6:54
  SPEAKER_02: And so it just gives you a different perspective on how people are seeing sort of the world and how quickly they're trying to adapt to it, and often in very, very similar ways.
• 7:02
  SPEAKER_02: How do we get richer?
• 7:03
  SPEAKER_02: How do we provide better opportunities for our kids?
• 7:05
  SPEAKER_02: How do we get better education?
• 7:06
  SPEAKER_02: How do we get better health?
• 7:07
  SPEAKER_02: And so these are things that are reasonably universal.
• 7:10
  SPEAKER_02: It's just we have different points on the chart on where we're trying to get to them with.
• 7:14
  SPEAKER_00: And I think almost um facing those challenges at different times give you different opportunities as well.
• 7:20
  SPEAKER_00: So I always remember um, I mean, we'll come on to this because you and I are both massive sports fans, but uh understanding the economic changes in India alongside the growth of the IPL is something that I've loved.
• 7:32
  SPEAKER_00: Um, but they didn't do a huge amount of the investment in subsurface infrastructure for telecoms because by the time they wanted it to explode, mobiles were fine.
• 7:40
  SPEAKER_00: So they could just skip that stage and move on to mass adoption of mobile telemetry instead.
• 7:45
  SPEAKER_02: Yeah, totally.
• 7:46
  SPEAKER_02: Yeah.
• 7:46
  SPEAKER_02: I mean, not to plug another podcast well on your podcast, but there's an acquired, it's called acquired.
• 7:50
  SPEAKER_02: And I don't know if you've listened to the one on the history of the IPL.
• 7:53
  SPEAKER_00: Oh, it's the best thing ever, isn't it?
• 7:55
  SPEAKER_00: It's so good.
• 7:56
  SPEAKER_00: I made the entire penny shaking her head behind the screen.
• 7:58
  SPEAKER_00: I made the entire leadership team listen to it, and most of them hate cricket.
• 8:02
  SPEAKER_00: And and they don't believe I like cricket because it does because I'm northern and tattooed and big.
• 8:06
  SPEAKER_00: They don't believe that fits a cricket persona.
• 8:08
  SPEAKER_00: Okay, and I adore cricket, but it's MMA Are you or something?
• 8:11
  SPEAKER_00: Exactly.
• 8:11
  SPEAKER_00: That's what they think.
• 8:12
  SPEAKER_00: That's what they that's what they assume it would be.
• 8:14
  SPEAKER_00: Stereotype.
• 8:14
  SPEAKER_00: Yeah, terrible, isn't it?
• 8:15
  SPEAKER_00: Yeah.
• 8:15
  SPEAKER_00: Um, but that acquired episode is astonishing.
• 8:18
  SPEAKER_00: Yeah, I I I do agree.
• 8:19
  SPEAKER_02: It is so good.
• 8:19
  SPEAKER_02: Yeah, but I mean it also shows you like it there will be it's an interesting point on the sort of the leapfrog and you know how you think about some of these major shifts, right?
• 8:30
  SPEAKER_02: And obviously, cable TV was a major shift in trans and transmission, and and you know, in the same way that you know, then social media was then a huge, you know, way of building communities and cohorts around the world in information, right?
• 8:43
  SPEAKER_02: And brand, right?
• 8:44
  SPEAKER_02: Which is uh, you know, all lessons for politics as well, because almost every kind of a lot of these opportunities that come have become of leapfrogs in both the information ecosystem and how people have mastered that.
• 8:57
  SPEAKER_02: And it's the same is true often for sort of business as it is for for politics as well.
• 9:01
  SPEAKER_02: And you could put in a Donald Trump, for example, was a master of the modern media ecosystem that managed to just you know, yeah.
• 9:08
  SPEAKER_00: The last election being the podcast-based election when they were first doing long forms.
• 9:12
  SPEAKER_00: As you were speaking, I just thought the next US election will be the AI one.
• 9:17
  SPEAKER_00: It'll be the first that AI has existed as it does in its current form.
• 9:21
  SPEAKER_00: Wow.
• 9:22
  SPEAKER_02: Yeah, and if you have a look at what's happening in the mayoral race in LA at the moment, for example, the sort of the incumbent uh candidate is uh is called Karen Bass.
• 9:31
  SPEAKER_02: And you know, she was there for the wildfires and you know, her approval ratings are are awful.
• 9:37
  SPEAKER_02: She sort of thinks she inherited this horrible legacy in LA with rising homelessness, drugs, etc.
• 9:44
  SPEAKER_02: But her one of her opposition candidates now is a is a former reality TV star called Spencer Pratt.
• 9:50
  SPEAKER_02: And if you look at his his AI-generated videos, they're kind of they're not in the traditional form of what is uh a political video, but they're they're quite gripping.
• 10:00
  SPEAKER_02: And and I think that's a precursor of how things will come.
• 10:03
  SPEAKER_02: And obviously, Donald Trump has also used a lot of that in his uh social media as well.
• 10:08
  SPEAKER_02: And uh, I think we'll see a lot more of this.
• 10:11
  SPEAKER_02: And I think that I think that sort of the traditional side of politics will say we'll be a bit snooty about it, whereas the the people that work out how to harness that and connect with a completely different way will will actually be the sort of the next the next level of victors in the coming future.
• 10:25
  SPEAKER_00: Well, there's our cricket analogy again.
• 10:27
  SPEAKER_00: The test cricket enthusiast would have been awfully snooty about the IPL and um lucky's doing well with the viewing figures now.
• 10:32
  SPEAKER_00: Totally.
• 10:33
  SPEAKER_00: Um so return from Rwanda, did you come back knowing the role in the next chapter you're coming into, or you returned to the UK and thought, now I need to find the next adventure?
• 10:42
  SPEAKER_02: Well, um Tony had actually come out to visit Rwanda, and that's how I sort of uh sort of got well, I was working for one of his charities within the Rwandan President's office and got to know him, um, and then came back into his private office actually.
• 10:56
  SPEAKER_02: And uh I don't think I would have come back to London at that point.
• 11:00
  SPEAKER_02: I think it would have probably been a maybe a few more years away.
• 11:03
  SPEAKER_02: Because I think, you know, once you've done a few little tours, you kind of think, oh, I'll go off and do another one.
• 11:07
  SPEAKER_02: Um, but I the draw of going and working for Tony was you know pretty large.
• 11:12
  SPEAKER_02: So I thought this is a good time to go home.
• 11:14
  SPEAKER_02: And uh, you know, then working in his private office and what was called the Tony Blair office before it was the institute.
• 11:21
  SPEAKER_02: Um, and you know, some of the issues that we're still talking about today were still on the table there.
• 11:27
  SPEAKER_02: Um what year was that?
• 11:28
  SPEAKER_02: 2014.
• 11:30
  SPEAKER_02: So anti-Semitism, extremism, they're you know, they were all sort of a bit of the public discourse, which were very, very similar.
• 11:37
  SPEAKER_02: There's a lot on immigration.
• 11:38
  SPEAKER_02: This was then just a precursor and a backdrop to Brexit at that point as well.
• 11:43
  SPEAKER_00: Um just before we totally come past that topic, um, I was listening to the Moonshot podcast in the gym this morning.
• 11:50
  SPEAKER_00: Do you ever catch that one?
• 11:52
  SPEAKER_00: No, yeah.
• 11:52
  SPEAKER_00: So uh four US tech tech founders um opining on the world over there.
• 11:57
  SPEAKER_00: They they shared an anecdote, and I haven't verified this, but they're pretty good on this stuff usually, that uh a sort of um radicalized anti-AI group has now been recognised as a broad category of terrorism by the US.
• 12:12
  SPEAKER_00: Oh, really?
• 12:12
  SPEAKER_00: People because of like um Sam Altman's house being uh Molotov cocktailed and things like that.
• 12:18
  SPEAKER_00: Yeah.
• 12:18
  SPEAKER_00: So it's so interesting we've had that rapid bounce back in terms of the aggression facing tech and things.
• 12:24
  SPEAKER_02: Yeah, I think I mean I think if I'm correct on this, it was Minnesota as well, which had the very anti-data center backlash and then one of sort of local.
• 12:32
  SPEAKER_02: Yeah, May as well, I think.
• 12:34
  SPEAKER_02: Yeah.
• 12:34
  SPEAKER_02: And I look, I think this is this is both new and not new.
• 12:38
  SPEAKER_02: Um, because if you look at through any kind of period of technological change, you will always have the the people that are very anti it, you know, go back to the Industrial Revolution and the Luddites, um, you know, they've obviously now been cast as the the sort of the villains, as the macro long-term of technology shows the sort of the uplift to humanity writ large.
• 13:01
  SPEAKER_02: But you know, when you're the people losing your jobs, of course, like you're going to be, you know, pretty upset.
• 13:06
  SPEAKER_02: And then even if you take then infrastructure build-outs, uh, if you look at sort of 1930s, 1940s England, when electricity pylons were built across the country, you had a group of romantic uh poets and John Maynard Keynes protesting this.
• 13:22
  SPEAKER_02: They were called the pylon poets, and you know, they were a blight on the English countryside, and you know, that's not that different from these horrible big concrete data centers that they think are sort of spoiling parts of our bucolic landscape today, as well.
• 13:36
  SPEAKER_02: Um, and I think you know, a lot of these things will always have the sort of the rhymes of history.
• 13:41
  SPEAKER_02: But I do think with this one as well, you know, the difference that will end up coming, I think, and this is why it require quite a lot of careful political handling, is that the disruption, I think, probably will be larger because it's be white-collar work as well as you know, the sort of replacement of um displacement of let's say those the more um you know sort of labor-intensive jobs.
• 14:04
  SPEAKER_02: Um, but I do think there'll be a load of opportunity created as every period of history and technology, it just depends on how much, how big, and how fast, I guess, is always the how the question for people and how you just manage them through that change.
• 14:18
  SPEAKER_00: I agree.
• 14:19
  SPEAKER_00: When did the private office start talking about considering the the global institute?
• 14:25
  SPEAKER_00: When did that sort of ambition and process begin?
• 14:28
  SPEAKER_02: Well, I I I I think it was been in the works for a little while, even when I I joined, because there were a few different charities which he'd founded since uh since he left office, which included a Faith Foundation, what was called the Africa Governance Initiative, which is where I was, the Sports Foundation, and then um uh a bit on the Middle East work.
• 14:51
  SPEAKER_02: And I think uh I think Tony would be the the first to say that you know, obviously being a patron of that rather than actually leading it is a very different thing.
• 15:00
  SPEAKER_02: And so bringing them all under one house with his direct leadership as executive chairman just gave him a greater sort of control that he was in charge of things that had, you know, the Tony Blair name on it.
• 15:12
  SPEAKER_02: Um so uh in the first couple of years, I think it was mostly just uh bringing together these different institutions with slightly different cultures, slightly different objectives, and um uh and I think where we are today is, you know, 10 years later, it is kind of extraordinary because uh, you know, I think any companies that around the world that go through mergers and acquisitions, it's it's not an easy thing.
• 15:39
  SPEAKER_02: And this was quite a large one on many levels, bringing together multiple.
• 15:42
  SPEAKER_02: Um, but what we've very quickly got to, I think, actually, is a very coherent mission, which is that of you know equipping leaders with the ability to govern more effectively, um, and an extremely now uh you know quite an impressive footprint around the world.
• 15:59
  SPEAKER_02: Um and so I think you know Tony is obviously a pretty unique political operator, but he's now become a pretty unique uh force for good in both you know building his own business and its you know its influence around the world, which is both important for the UK in my mind, but also for how these governments themselves operate.
• 16:20
  SPEAKER_00: When um I I think the the operational part of what he does is really underappreciated.
• 16:27
  SPEAKER_00: So um last year in uh New York Climate Week, I was fortunate to speak to him for for quite a while about um the delivery unit he put in place when he was in number 10.
• 16:36
  SPEAKER_00: And you could see him beam because it was like it'd been a while since someone had asked him about that.
• 16:40
  SPEAKER_00: And I was like, um, ideas are easy, actually making them happen is is the difficult part.
• 16:47
  SPEAKER_00: And for him to make the sort of infrastructure changes he did in number 10 at the time, which what I do love, you you'll even hear David Cameron reflect now, going, that was his biggest mistake, taking out that delivery unit because he meant he reached for levers to pull to actually make things happen, and it just wasn't there.
• 17:03
  SPEAKER_00: Yeah, and I think that came through really strongly in the in the leading book as well, in terms of how you actually make things happen.
• 17:09
  SPEAKER_02: Sure.
• 17:09
  SPEAKER_02: I mean, I mean, you even seen the problems today, which you know, there is there is an ideas challenge and actually, you know, understanding the world and identifying it, and then actually coming up with solutions to elements, which obviously was part of the essay last week.
• 17:25
  SPEAKER_02: But I I completely agree on the execution and delivery point, which is not as well heralded.
• 17:30
  SPEAKER_02: And if you look at something like the UK government, there are 57,000 policy professionals, I think, uh across it.
• 17:38
  SPEAKER_02: And you know, there's a huge veneration for ideas and for policy papers.
• 17:42
  SPEAKER_02: There is not a commensurate level of attention played to delivery and execution.
• 17:48
  SPEAKER_02: Um, and you know, you you know, you run a company like the you know, the people that are the operators and the get stuff done are the people that you you know you very highly venerate because uh that is really where the action happens.
• 18:01
  SPEAKER_02: And I think Tony, you you know, you're totally right on this in the delivery unit, and it's something we've been constantly updating today.
• 18:08
  SPEAKER_02: And actually, one of the core things that I've been working on the last period has been how you use agent AI to build delivery systems in the same way that companies are going to be using it uh for execution across their business lines.
• 18:20
  SPEAKER_02: Um, and I think it's I think there, if you look at something like the UK as well, actually, just taking this as an example, I think there are only 300 people in sort of the totality around the Prime Minister's office and sort of bits where something like Singapore has 1,600 people.
• 18:37
  SPEAKER_02: Um and there's a lot of talk about the decentralisation kind of aspect, but you know, decentralization can work, but you still need a strong centre.
• 18:47
  SPEAKER_02: Uh, and if the strong centre isn't executing, Liz Truss made this mistake as well.
• 18:51
  SPEAKER_02: Like, I mean, she made other mistakes as well, let's be clear.
• 18:54
  SPEAKER_02: But uh but one of the things immediately said was, I'm gonna get rid of all the staff in number 10.
• 18:57
  SPEAKER_02: And you're like, oh, I mean, as Prime Minister, that is just total madness because you you know, you wouldn't again be a CEO of a company and say, right, get rid of all the like my core executive people that are executing the business for me.
• 19:08
  SPEAKER_00: So There's literally a yes minister episode on that, isn't there?
• 19:11
  SPEAKER_00: Operation hairshirt, where he sacks all his staff around and all of a sudden he asks for something and there's nobody there.
• 19:16
  SPEAKER_02: I haven't seen that one, but yeah, I mean it's Tony always says that uh you know, yes minister is a documentary, not fiction.
• 19:23
  SPEAKER_00: So I I completely agree.
• 19:26
  SPEAKER_00: Um so before we started recording, we reflected on on the essay that Tony shared last week.
• 19:31
  SPEAKER_00: Um going behind the scenes, how does something like that come about?
• 19:36
  SPEAKER_00: What's the process for you and the team in supporting him on that?
• 19:39
  SPEAKER_00: And how far ahead of time was something that seminal being released, would it be thought about and developed?
• 19:45
  SPEAKER_02: I mean, there's a whole policy team in TBI which is constantly trying to build up the ideas for how we think you know you govern in the age of AI as we're talking about.
• 19:55
  SPEAKER_02: And you know, a lot of that is the reform of uh how government itself operates and you know, climate and energy, health, all of these different components.
• 20:05
  SPEAKER_02: It's sort of a constant um, I guess, uh interrogation that Tony has of the team, plus of you know the conversations he's having with prime ministers and politicians around the world to sort of understand where their heads are at, what are the big topics, you know, and I think the unique privilege of somewhere like the institute is it just has this ability to tap into, you know, world leading experts, not just in the the policy domain, but in the private sector as well, because you know, often the private sector people tell you it with a lot of straighter facts where the problems are rather than with all the sort of political guff that comes with it.
• 20:40
  SPEAKER_02: But I mean, in terms of then sort of the process, I mean the uh Tony is very hands-on and he writes these things himself.
• 20:47
  SPEAKER_02: So it was the same when he was in office as I understand, I did obviously work for him then, but you know, big speeches, essays, you know, he's got a very unique and uh uh uh communications style and a voice.
• 20:58
  SPEAKER_02: So um that is you know very true to him.
• 21:00
  SPEAKER_02: And it's you know, it's yeah, he works with the institute, but he is very, very much the the one in charge of those things.
• 21:07
  SPEAKER_00: Love it.
• 21:08
  SPEAKER_00: So one one of the key themes was the call that government-wide needs to be reorganized around AI.
• 21:14
  SPEAKER_00: Unpick that for us a little bit and bring it to life.
• 21:16
  SPEAKER_02: Yeah.
• 21:17
  SPEAKER_02: So I I think uh if you sort of start from the structural diagnosis here and then then then I'll sort of go from there.
• 21:25
  SPEAKER_02: So uh if you look at now UK taxes are at the highest since you know Second World War, uh public service outcomes have been declining.
• 21:38
  SPEAKER_02: Um and at current, sort of a lot of the solutions are well, how do we just increase our tax base?
• 21:44
  SPEAKER_02: Um, and how do we pump more money into the NHS, for example?
• 21:49
  SPEAKER_02: And then if you then sort of look at almost the shape of the state today, um I think it's roughly 44%.
• 22:00
  SPEAKER_02: Of public expenditure will go on some form of redistribution and debt service.
• 22:05
  SPEAKER_02: So, i.e., spending on the past and spending to sort of help those, you know, whether it's you know, on welfare or benefits, etc.
• 22:15
  SPEAKER_02: And I think around 10% will go into anything that would be seen as an investment or growth lever.
• 22:20
  SPEAKER_02: Now, I mean, I don't think it takes uh a magician to work out that there's something wrong with that structurally, right?
• 22:28
  SPEAKER_02: And this problem is not, people will say this problem is also what all our governments are facing, like, and that is true in the OCD, like France, Germany, Japan, they've all hit very similar structural problems.
• 22:40
  SPEAKER_02: But that still means it's a problem, right?
• 22:42
  SPEAKER_02: So when we think about that, the only way to begin to break that chain, because if you look at sort of the history of the world, and particularly in the last few years, if you look at every kind of cost curve, the only ones that ever begin to decline are the ones where you can adopt technology to begin to scale and build quality, right?
• 23:00
  SPEAKER_02: So if you look at something like public expenditure on education and health, it goes like this.
• 23:05
  SPEAKER_02: If you look at something like consumer electronics, etc., it all goes down this way, right?
• 23:09
  SPEAKER_02: Um, and so what we just trying to say that is that you know, AI is not going to be the complete panacea, but it will be the only way that will get you out of this sort of structural problem.
• 23:20
  SPEAKER_02: And what it can do if you get it right is actually it can reduce the cost of uh of public services while increasing their quality, right?
• 23:30
  SPEAKER_02: Um but the same is true then with like the private sector, is if you just begin to whack on AI onto a dysfunctional system, the system's then still dysfunctional.
• 23:40
  SPEAKER_02: So you have to then begin to rethink the system in and of itself.
• 23:43
  SPEAKER_02: And let me give you a few sort of examples where I think that could happen.
• 23:47
  SPEAKER_02: All right, I think there are some very sort of large and boring questions of government, which is a bit like I don't know if you've ever come across the uh the investor Charlie Songhurst, if you've heard of him.
• 24:01
  SPEAKER_02: So he was a former sort of top guy at Strategy and Microsoft, and then is now a very, very successful angel investor who has a very simple matrix of which is on one axis is boring and exciting, and on the other is degrees of complexity.
• 24:19
  SPEAKER_02: And he says, you know, your sweet spot often for investing is boring and complex.
• 24:24
  SPEAKER_02: Um you know, uh exciting and complex, you know.
• 24:28
  SPEAKER_02: I don't know, maybe we do get data centers in space in the near future, probably, but it is really hard, right?
• 24:34
  SPEAKER_00: We'll get on to that.
• 24:34
  SPEAKER_02: Yeah.
• 24:35
  SPEAKER_02: Um, and actually there's a lot of that upside in government on the boring and complex thing.
• 24:39
  SPEAKER_02: So that's you know, that's things like tax fraud, that is like large-scale uh questions which suck up a lot of money.
• 24:46
  SPEAKER_02: That could be welfare reform, etc.
• 24:48
  SPEAKER_02: Even if you take something like um um means testing of energy subsidies during COVID, like we didn't have the ability to target that because we didn't have the right data uh architecture.
• 25:01
  SPEAKER_02: So you ended up spending 38 billion on this, you know, during COVID, when actually, you know, do people like me or you know, other people need a bit of that when it's really the support should be going out.
• 25:12
  SPEAKER_02: So we did a piece of modeling on that, which could save you know 17 billion just through a means testing system.
• 25:17
  SPEAKER_02: So, you know, boring and complex saves you a lot of money.
• 25:21
  SPEAKER_02: Now, there are, I think, where it becomes more interesting is on that sort of very structural questions which you begin to look at, which is let me take another example of a principle of public service in the UK.
• 25:35
  SPEAKER_02: That's the sort of post-World War II settlement of beverage, of universal access, you know, universal access to the NHS, and you know, Bevan obviously setting that up, and then others beginning to look at, you know, where you've got education and other places where, you know, we've decided that's a right for everyone, and everyone should have access.
• 25:53
  SPEAKER_02: Now, that's a good right, I think, on many levels, but they all don't then have the same access to quality, right?
• 25:59
  SPEAKER_02: So you have you have access, but you don't have access to quality.
• 26:03
  SPEAKER_02: Now, I think something like an AI doctor or an AI tutor, if they can begin to come out, you know, writ large across the population, could upend that.
• 26:13
  SPEAKER_02: And now, like, does it mean that schools exist in the same way, or hospitals, or how you deliver care?
• 26:19
  SPEAKER_02: Probably not, right?
• 26:20
  SPEAKER_02: Um, I'm not saying that you then abolish schools and everyone just sits at home on an AI doctor, an AI tutor.
• 26:25
  SPEAKER_02: I think that would be quite depressing.
• 26:27
  SPEAKER_02: But it would get into a structural redesign of how you actually deliver education.
• 26:31
  SPEAKER_02: It would be a you know structural redesign of how you do hospital and capital expenditure on all of these things, if you had these access to things which should be incredibly powerful.
• 26:40
  SPEAKER_02: And I think, you know, I would be of very little doubt that over the next five to ten years, we will be able to have all of those as a universal point of access at extremely high quality and which would actually lead to better outcomes in those domains as well.
• 26:53
  SPEAKER_02: So that would just be a few examples where I could see AI coming in.
• 26:57
  SPEAKER_02: But I think if you take almost every point of government, there'll be some application which, you know, if you're not thinking about AI as at least part of the delivery, you're just not thinking hard enough because you know, AI, a government is a lot of process and paper pushing.
• 27:10
  SPEAKER_02: And that is another area where AI is extremely good at today.
• 27:13
  SPEAKER_00: So, one quick question on the way past before we unpick uh potential government opportunities.
• 27:18
  SPEAKER_00: So you mentioned about doing some modelling around um what that would potentially look like.
• 27:23
  SPEAKER_00: We're all resource constraint.
• 27:25
  SPEAKER_00: We can't we can never dive into every idea we've got.
• 27:28
  SPEAKER_00: Yeah.
• 27:29
  SPEAKER_00: How do you decide which is the one that you say to the team, or give this more of a look, or actually that'll sit in a notebook or a to-do list or whatever?
• 27:36
  SPEAKER_00: How do you make that choice on which ones gets the deep dive?
• 27:39
  SPEAKER_02: That's an interesting question.
• 27:41
  SPEAKER_02: Um I think there are a few different ways that I think about that.
• 27:45
  SPEAKER_02: I mean, number one is that you know, if you are sort of engaged with the world, you're constantly information gathering, right?
• 27:52
  SPEAKER_02: And you're probably you're hearing signals from people, what is important and what is not.
• 27:56
  SPEAKER_02: And you know, when you're in 45 governments, for example, and the if you know, just to put it in a very simple way, if 10, 15 of them said, well, you know, you know, welfare fraud is a huge issue, you're probably like, that might be something we need to look at.
• 28:08
  SPEAKER_02: So it's it it's sort of information processing and sort of signal processing, I think is number one to understand sort of the reality.
• 28:18
  SPEAKER_02: I think number two is then is the second sort of form of that, which is you know, we do have uh lots of information in the public arena, whether that's from news, from the population, to understand what are the issues that they care about.
• 28:35
  SPEAKER_02: Um and I I think one of the problems in politics in recent years is sort of twofold on this, is number one, it often just ignores what people are saying.
• 28:48
  SPEAKER_02: And I think you've seen some of the bubble up of populism sort of not taking this seriously and sort of, and so you know, a lot of people that have then come in have then just kind of almost um, you know, address those angers and grievances not with solutions, but at least they've listened to people.
• 29:03
  SPEAKER_02: So people feel like you know, wow, someone's actually doing that.
• 29:07
  SPEAKER_02: And then uh the sort of the second, almost the flip side of that as well, is that people okay, they hear a signal, but the mistake signal, well, noise for signal is the classic thing, right?
• 29:18
  SPEAKER_02: It's like if you lead by a social media feed, like in terms of just listening, you're never gonna get anywhere.
• 29:25
  SPEAKER_02: Because this comes to the last point, which is you've also just got to work out, you know, what you actually think about the world.
• 29:32
  SPEAKER_02: And you know, I I don't know about you, and same as probably setting up how you came to a business.
• 29:37
  SPEAKER_02: I send I spend a lot of my time just, you know, reading, interrogating, thinking about things and saying, how what do I actually think about that before I push forward the team to get going with it as well.
• 29:48
  SPEAKER_02: So I think sort of, you know, one and two are kind of the information kind of gathering from different forms, but three is just like the, you know, you know, the really hard stuff of like, how am I thinking about this and what do I actually want to do with it?
• 30:00
  SPEAKER_00: Yeah, it's your pre-training and your influence kicking.
• 30:03
  SPEAKER_02: Yeah, I guess so, yeah.
• 30:04
  SPEAKER_00: Um my thinking, um my brain works fast.
• 30:09
  SPEAKER_00: If you give me an idea or a suggestion, it'll give me like five or six options immediately, and I have to pick which one.
• 30:14
  SPEAKER_00: It's why I started writing.
• 30:16
  SPEAKER_00: So I started writing to force me to stay on topic and narrow down to what I actually thought.
• 30:21
  SPEAKER_00: Yeah.
• 30:21
  SPEAKER_00: Not not what a clever sound bite would be, but what I actually thought.
• 30:24
  SPEAKER_00: And um, so yeah, book two literally came out last night.
• 30:28
  SPEAKER_00: The first one's dreadful, second one's not bad.
• 30:30
  SPEAKER_00: Um, but I that's how I do.
• 30:32
  SPEAKER_00: As much I love as much inflow as possible.
• 30:34
  SPEAKER_00: I want all the info I can to craft those ideas, and then the amalgamation of that, I'll pull the nugget of what I actually think and feel about.
• 30:41
  SPEAKER_02: Yeah, I completely agree with that.
• 30:42
  SPEAKER_02: Writing is a is basically the best way.
• 30:45
  SPEAKER_02: It can be so painful and excruciating, but it is always the best, to your point, is the best way to force your thinking down and then pour over each word until you've kind of really got it right.
• 30:55
  SPEAKER_00: So when um my first book was never meant to be a book.
• 30:58
  SPEAKER_00: So it was my reflections and notes on the first two or three years of starting the business.
• 31:03
  SPEAKER_00: And then Triados, the incredible um sustainable bank, reached out to us and said, We really like some of your writing, let's publish something together, which was very kind.
• 31:11
  SPEAKER_00: So we did that.
• 31:12
  SPEAKER_00: But it was dreadful because it was never meant to be a book.
• 31:15
  SPEAKER_00: So when I read bits of that back, which I try and avoid doing, that's why there's no audio book because I didn't want to read it back.
• 31:20
  SPEAKER_00: There's probably like 10 lines in there that I was really proud of where I was like, I like that one.
• 31:25
  SPEAKER_00: Um whereas I was made to read some of book two the other day for um which was meant to be a book and was written over the last nine months to a year of, and it's covered with Wilding Wealth.
• 31:34
  SPEAKER_00: It's about how the capital side of the business works, which just didn't exist five years ago.
• 31:39
  SPEAKER_00: That's a relatively new concept.
• 31:41
  SPEAKER_00: My writing had got so much better, and and I was not embarrassed reading some of those sentences, which was nice.
• 31:46
  SPEAKER_02: I mean, that's also that's totally I mean, I I one of the things that I find human just writing, which is that sort of because everyone can sort of do it in some form, they think they can do it.
• 31:58
  SPEAKER_02: Whereas you you do improve by by you know same with everything.
• 32:02
  SPEAKER_02: Yeah, literally improving by doing it.
• 32:06
  SPEAKER_02: Yeah, exactly.
• 32:06
  SPEAKER_02: And it's just yeah, I mean, I've I'd you know, put stuff into the public domain, I'd probably it's a fair few, I'd look back and go, Oh god, what was I thinking then?
• 32:15
  SPEAKER_02: But you know, you're constantly evolving your thinking and improving as well.
• 32:18
  SPEAKER_00: So I completely agree.
• 32:20
  SPEAKER_00: Um so just returning to the NHS, um, I I don't know if you were citing on this project, but in the early-ish days of um DeepMind, before they were or as they were acquired by Google, I think it crossed over.
• 32:31
  SPEAKER_00: So it might have been in that 14 to 16 range, something like that.
• 32:34
  SPEAKER_00: There was actually a partnership with the NHS and them trying to do some things about that.
• 32:38
  SPEAKER_00: But people's perceptions on uh misconceptions on data security and visibility and things caused real issues.
• 32:46
  SPEAKER_00: What do you think we can do to help people understand the positives that AI can bring and the realities of how we have to operate to benefit from some of these tools and technologies?
• 32:55
  SPEAKER_02: Yeah, so th there were two projects, if I remember right, which is one, the Moorefield's Eye Hospital and the Retina Scanning, um, which was run by a guy called Dr.
• 33:04
  SPEAKER_02: Pierce Keene, who is absolutely fantastic.
• 33:07
  SPEAKER_02: And um, I know is continuing some of that work today because you know, retina scanning is almost a perfect example of what to do for you know large-scale image scanning in AI, right?
• 33:18
  SPEAKER_02: Um uh and so it that unfortunately didn't take off in the way it should within the system, but it thankfully he's continuous.
• 33:26
  SPEAKER_02: And the second one obviously was the Royal Free uh side.
• 33:29
  SPEAKER_02: And I think it quite bluntly, I thought it was a bit of a travesty that number one, we didn't, you know, we didn't build the right public-private partnership there with actually you know having one of the world leading AI companies on our doorstep.
• 33:44
  SPEAKER_02: And secondly, I actually think it then, even worse than that, ended up pushing them out of the uh out of that system for a while because they were too worried about some of the political ramifications on that.
• 33:55
  SPEAKER_02: Um so yeah, I it's something I have thought about.
• 34:00
  SPEAKER_02: Um now there are different principles around some of this, obviously, within the NHS on consent versus non-consent and what the call of the Cold Ecot kind of reviews and sometimes in many ways it's like I I understand where we go through debate on this, and I can understand why people have a lot of the privacy concerns.
• 34:18
  SPEAKER_02: But number one, like you can do this in totally anonymised privacy-preserving ways now.
• 34:25
  SPEAKER_02: Um, and you know, sort of number two, what better area would you not want to be sharing your data in than in healthcare delivery?
• 34:33
  SPEAKER_02: Now, look, there will be things like if you're on a, I don't know, on disability or something where you know, if parts of the system are sharing information, you want to totally ring fence that because you don't want to be penalized for you know different parts uh uh by different parts within this.
• 34:50
  SPEAKER_02: But I basically think that that is one of the biggest areas that we should be gunning for.
• 34:57
  SPEAKER_02: And we've actually slipped because of partly that, but um other reasons as well.
• 35:02
  SPEAKER_02: Now, secure data environments are being currently developed, obviously, by the UK government, the data library as well is one of the aspects that they're thinking about on this.
• 35:11
  SPEAKER_02: We also have the UK Biobank is probably the world's leading uh life sciences resource for genomic data.
• 35:19
  SPEAKER_02: I think all of these things are actually vital parts of statecraft today.
• 35:25
  SPEAKER_02: And, you know, the one thing that I will say to people when I sort of say them is that like in the same way that companies' data is their competitive asset, you know, a nation's data is a strategic asset, right?
• 35:37
  SPEAKER_02: And healthcare is one which provides huge public good.
• 35:41
  SPEAKER_02: Uh, you know, one that you know I personally think of if you know I have a risk of X in my family knowing that they can begin to target that and do clinical trials and do drug development based on all of this stuff, I would be very willing to share my data.
• 35:56
  SPEAKER_02: So I think there are easy solutions to lots of these things.
• 35:59
  SPEAKER_02: Some of it is mostly just the hard part of the political persuasion across the public.
• 36:04
  SPEAKER_00: So I'm gonna dive on a topic that you can absolutely leave and let go by if you wish.
• 36:08
  SPEAKER_00: But I I wish, I wish the current government had used its unbelievable mandate on day one to force some of these things through.
• 36:17
  SPEAKER_00: Yeah.
• 36:18
  SPEAKER_00: To come in and say, actually, we are gonna create a sovereign data set.
• 36:22
  SPEAKER_00: We are telling you we're gonna do this.
• 36:23
  SPEAKER_00: You voted us in, you wanted change, we need to do some of these things that make this possible.
• 36:28
  SPEAKER_00: And if they'd have done that in the first six months, as opposed to cut cutting ridiculously small benefits on fuel.
• 36:35
  SPEAKER_00: I mean, it just what an open goal missed there was just ludicrous.
• 36:39
  SPEAKER_00: Yeah.
• 36:39
  SPEAKER_02: Well, I mean, if you go back to sort of 1997, that government immediately came in with national minimum wage, you know, it had some of the things that were ready to go because you want to shake up the system.
• 36:51
  SPEAKER_02: And I think this is the whole point of like, you know, you need to develop your ideas, come in with a programme of government and completely get going.
• 36:58
  SPEAKER_02: Because, you know, also by the way, I I know how this works.
• 37:02
  SPEAKER_02: If you then spend six, nine months and sort of you know, wait for the system to give the answer, the system will give you the answer it always wants to give you.
• 37:09
  SPEAKER_02: And it will.
• 37:10
  SPEAKER_02: Yeah.
• 37:11
  SPEAKER_02: Uh so yeah, you just need to go and shake up the system, is the honest answer.
• 37:15
  SPEAKER_02: And you see this also in other areas, it's like where you commission a, you know, where I see them let's commission a strategic plan or a or a review, etc.
• 37:23
  SPEAKER_02: It's like, yeah, but like you're just gonna waste a year if you're governing time doing that.
• 37:27
  SPEAKER_02: Like, uh unless there's a really, you know need to do something on that, and most of the time it's it's a it's a pointless exercise.
• 37:35
  SPEAKER_00: It's why I'm vehemently against benchmarking in businesses, because you spend a year trying to get to the point where someone was a year ago and you get the report.
• 37:43
  SPEAKER_00: And it's like, well, they've they've already they've already moved past and gone.
• 37:46
  SPEAKER_02: Totally.
• 37:47
  SPEAKER_02: Totally, yeah.
• 37:48
  SPEAKER_02: Yeah, you're like you're building a business based on a an organizational model of a couple of years ago.
• 37:53
  SPEAKER_02: Yeah, totally.
• 37:54
  SPEAKER_02: Yeah, particularly how world how fast the world is moving right now as well.
• 37:57
  SPEAKER_02: It's like it, you know, people always think this is a bit of a platitude, but like it it's just unbelievable the rate of change right now.
• 38:06
  SPEAKER_02: And you know, particularly if you're looking at something like AI, like you you know, you could build a startup and then it could be dead within three months because of the because of the model developments.
• 38:15
  SPEAKER_02: And so if you aren't moving like at incre you know, increasing pace on this, always people are gonna get more frustrated from the public as well.
• 38:23
  SPEAKER_02: So, like, you know, government just doesn't understand that ability to like you know up the rate of change commensurate with the rate of change in the world around it.
• 38:31
  SPEAKER_00: I I I wrote a piece which is what I'm gonna pass to you that um sort of responded to to Tony's essay.
• 38:37
  SPEAKER_00: And one of the lines in there was Um I I'm just fed up with a country who complains about how fast and chaotic everything is and wishes it would slow down as opposed to just getting better at playing fast.
• 38:47
  SPEAKER_01: Yeah.
• 38:48
  SPEAKER_00: We can't we can't choose the ball's gonna move at this speed, play at that speed, or they don't want to be involved.
• 38:54
  SPEAKER_00: Completely agree.
• 38:54
  SPEAKER_00: Um I again to come back to the essay, I I loved the simplicity, but the depth of context that was in policy before politics.
• 39:04
  SPEAKER_00: And I just don't think we get it anymore.
• 39:05
  SPEAKER_00: Everything that plays out, and um, you know, we're all to blame for this if we read something and click on it and look at it, we're all fueling a system.
• 39:12
  SPEAKER_00: The media is providing what we collectively want and and absorb, but it is politics in the complete absence of policy.
• 39:20
  SPEAKER_00: And in in terms of the government, it certainly felt like they hadn't thought about what they were going to do when they won.
• 39:25
  SPEAKER_00: So even in business, we see this.
• 39:27
  SPEAKER_00: People will some businesses have no plan for anything.
• 39:31
  SPEAKER_00: Half of them have a plan for what happens if stuff goes wrong.
• 39:34
  SPEAKER_00: Very few happen for what happens if it goes right.
• 39:37
  SPEAKER_00: So people be trying to raise money and you say, Oh, fantastic, you've got the raise wave.
• 39:40
  SPEAKER_00: What are you gonna do with it?
• 39:43
  SPEAKER_00: You know, we we sit here with um SpaceX IPO about the go next week.
• 39:49
  SPEAKER_00: Um what what's your views on that and the impact that uh a potential multi-trillion dollar IPO has on global economics?
• 39:57
  SPEAKER_02: Yeah, well there's this is the thing that I going back to the point on the comprehension though of these things, it's like if you are engaged in seeing what's happening in the world, it this stuff just blows your mind, right?
• 40:13
  SPEAKER_02: So, I mean, SpaceX IPO, how many billionaires, how many centimillionaires does it produce?
• 40:22
  SPEAKER_00: Across SpaceX, Anthropic, and I and open air, which is more than currently exist on the planet.
• 40:27
  SPEAKER_02: Yeah, exactly.
• 40:28
  SPEAKER_02: So three companies have now just you know, and yeah, it's slightly mind-boggling statistic, right?
• 40:35
  SPEAKER_02: Um, so not only is that probably the probably the greatest, these are the greatest events of wealth creation, they're also probably one of the greatest events of wealth concentration as well than within a country that is just absolutely, you know, to put it, you know, hand you know, give them, you've got to give it to them.
• 40:55
  SPEAKER_02: They've absolutely nailed this next wave of technology, right?
• 40:57
  SPEAKER_02: Um, there's no resentment to any of these people because it's just like you just built an amazing company.
• 41:02
  SPEAKER_02: Well done.
• 41:03
  SPEAKER_00: And it's only just started.
• 41:04
  SPEAKER_00: Yeah, and the tail of this is gonna be.
• 41:07
  SPEAKER_02: But it this is the bit then you look at a country like the UK and like and you see where the debate begins to go to on this.
• 41:14
  SPEAKER_02: And I think look, you will probably end up with uh you know lots of people with new forms of philanthropy as well, in the same way that the Rockefellers and others came into the world with you know, an ability to solve lots of world problems.
• 41:28
  SPEAKER_02: But there are going to be then this whole raft of politics, you even sort of Bernie Sanders' uh uh essay yesterday in the New York Times on let's create a sovereign wealth fund based on the ups uh you know upsides of this and redistribute all that money that other people have made to other people again as well.
• 41:44
  SPEAKER_02: And now that system is gonna play out large.
• 41:47
  SPEAKER_02: So I think the only thing that I really care about in many ways right now is how other countries become more competitive.
• 41:54
  SPEAKER_02: And the the the way they don't become more competitive is then by sort of you know doubling down on regulation or etc.
• 42:01
  SPEAKER_02: Like, like you, you know, the only way that you contend with these now huge hypersystemic sort of forces are becoming more competitive and powerful yourself.
• 42:10
  SPEAKER_02: Like, and there's gonna be a whole politics of resentment and envy that go on on this, it's gonna be a whole push for it, uh um uh redistribution, etc.
• 42:20
  SPEAKER_02: Now, that I think there will be some policy responses to this that will become inevitable.
• 42:24
  SPEAKER_02: Competition policy will definitely change at this at this point, and probably there will be some areas where you need that, right?
• 42:30
  SPEAKER_02: Because you don't want all these now completely mega companies just able to gobble up everyone else because that's not also how economic dynamism works.
• 42:39
  SPEAKER_02: But at the same point, my my sort of lesson to everyone else is that's what wealth creation engine looks like if you get it right, and that's what you kind of want within your country to begin to get going on this.
• 42:50
  SPEAKER_00: People are often surprised to hear me.
• 42:52
  SPEAKER_00: I like I run one of the country's leading conservation businesses, yet my view is there's far too much environmental regulation.
• 42:59
  SPEAKER_00: Yeah.
• 42:59
  SPEAKER_00: And and people can't square that circle.
• 43:02
  SPEAKER_00: The problem is we're at the point now whereby, for every positive thing a a government, irrespective of its colour or policy, wants to do, there's already 20 documents that tell them they can't.
• 43:14
  SPEAKER_00: So you look in Scotland right now, the government year on year are failing to hit its tree planting targets because a local community will come along and say, ah, there's 20 really small rules locally that mean you can't plant trees anywhere you want to.
• 43:29
  SPEAKER_00: And there will inevitably be a retired barrister who goes, I'm gonna take you to court.
• 43:33
  SPEAKER_00: So Scottish Forester have been sued multiple times in the last few months, which make doing the thing they want to do for the planet, for the environment, for the local communities, for business, nigh on impossible.
• 43:43
  SPEAKER_00: And all it does is ramp up the costs associated with doing more and more strategy documents and impact, and nothing happens.
• 43:50
  SPEAKER_00: And it makes lawyers very rich and nothing else.
• 43:52
  SPEAKER_00: And they're not struggling.
• 43:54
  SPEAKER_00: The lawyers are doing um e even if you believe the argument that AI is bad for lawyers.
• 43:58
  SPEAKER_00: I I was at an event with um Salesforce and Anthropic, actually, a few months ago.
• 44:03
  SPEAKER_00: And there was a law professor who turned around and said, I hate to break this to you.
• 44:07
  SPEAKER_00: AI is creating that much bad law, because anyone can be a lawyer now using a chatbot of some kind.
• 44:14
  SPEAKER_00: All of a sudden, actually, we're going to need more lawyers.
• 44:17
  SPEAKER_00: So, false alarm, there's going to be more lawyers, not less.
• 44:20
  SPEAKER_00: Yep.
• 44:21
  SPEAKER_00: Everyone's paradox playing out.
• 44:23
  SPEAKER_00: One thing that we've looked at as well is my hypothesis goes, and you mentioned about Minnesota and some of the other states.
• 44:31
  SPEAKER_00: If we are going to get the scale of physical data centers, because we're talking about cloud computing, but these clouds are big metal buildings that have concrete and water and footprints and energy and a disturbing environment.
• 44:43
  SPEAKER_00: For those to be built at the scale we need, we are going to need to secure the social and environmental license to operate.
• 44:50
  SPEAKER_00: I do not see a scenario that Anthropic, OpenAI, Google themselves, NVIDIA are not the biggest investors in real nature assets that ever existed.
• 45:01
  SPEAKER_00: Yeah.
• 45:01
  SPEAKER_00: Because in the absence of anthropic national parks, they're not going to let them build anywhere because they're going to have to demonstrate we are having a net positive impact on the environment, therefore we're fine to do what we're doing.
• 45:12
  SPEAKER_00: So that's ultimately, I think, where our business ends up aligning with being the physical delivery body for the digital world.
• 45:19
  SPEAKER_00: The more digital we become, the more reliant it is on nature.
• 45:22
  SPEAKER_02: Interesting.
• 45:22
  SPEAKER_02: Yeah.
• 45:23
  SPEAKER_02: I mean, and I mean, we looked at a lot about what I would call, what I have called the twin transition between particularly on the clean energy side.
• 45:34
  SPEAKER_02: I mean, land obviously feels pretty obvious to me now that you you say it.
• 45:39
  SPEAKER_02: Um, but when you're looking at some of the I mean, what, seven trillion was committed in capital expenditure for AI infrastructure in recent years.
• 45:50
  SPEAKER_02: So a lot of that has then led to a commensurate need for energy and seen huge investments in um energy alongside it then, including, you know, for example, there's a company in the US called Form Energy, which had its first of a kind, geothermal, um, alongside one of the hyperscalers.
• 46:11
  SPEAKER_02: You know, you're looking at you know, uh nuclear becoming back on the table in very, very deep way.
• 46:17
  SPEAKER_02: Um, and I think fusion, of course, like almost every one of these uh sort of uh tech millionaires now has a um an investment in a in a fusion company of some sort.
• 46:29
  SPEAKER_00: Yeah, fusion and space.
• 46:30
  SPEAKER_00: Yeah.
• 46:30
  SPEAKER_00: Tick those two in.
• 46:31
  SPEAKER_02: Exactly.
• 46:32
  SPEAKER_02: And you know, the the the funny thing about the spacing is obviously if you say that to someone, it sort of sounds slightly you know mad and sci-fi-y, but I don't see why in five, ten years' time you wouldn't be beginning to look at this as well.
• 46:46
  SPEAKER_02: And it what feels very unusual to us right now will probably feel very, very normal to a future generation on this.
• 46:54
  SPEAKER_02: Um, and you know, there is lots of space up there, so um yeah, I you know, we've seen for a few years debates on obviously whether you could get solar up into space, and obviously within deep difficulties around transmission and the loss of elements around that.
• 47:13
  SPEAKER_02: But I I mean I I wouldn't I would genuinely not be surprised, and I I would put it in my future forecast that we will have something with this within you know the not too distant lifetime as well.
• 47:24
  SPEAKER_00: And even um, are you familiar with the work of um John Elkington?
• 47:29
  SPEAKER_00: Um he wrote he's he's written 20 odd books, but he literally invented the word sustainability and triple bottom line thinking.
• 47:36
  SPEAKER_00: Yeah, amazing guy in his uh I'll say mid-70s, so he doesn't slap me around because he's coming on the podcast again soon.
• 47:42
  SPEAKER_00: Um and bear in mind what he's seen in his working life in terms.
• 47:46
  SPEAKER_00: So we recorded a podcast with him in Canary Wharf, and he told a story about standing at Canary Wharf when there was nothing there in like the 70s or 80s, and he was part of the team, did the master planning around it, and he came back last year and saw it in in all its glory now.
• 48:00
  SPEAKER_00: And I said, a little bit like you did about space, but oh, in the next five or ten years, he went, No, Rich, in AI, that's two.
• 48:08
  SPEAKER_00: Two, yeah.
• 48:08
  SPEAKER_00: Yeah, and it the and the the challenges, the difficulties around everything from um protein folding and medicine and scenario planning all the way through to space.
• 48:20
  SPEAKER_00: Maybe these challenges we need to solve in terms of energy transition, AI will solve them for us in terms of that equation.
• 48:26
  SPEAKER_00: And that's how without a shadow of a doubt, we will have data centers in space, we will have material manufacture like our extraction from the moon, we're gonna get all those things.
• 48:35
  SPEAKER_00: They're gonna come.
• 48:36
  SPEAKER_02: But there's what I mean, some manufacturing happens in space already because how slowly it crystallizes.
• 48:42
  SPEAKER_02: Um metals, for example, uh, because you can in low gravity, you can crystallize them slower, right?
• 48:49
  SPEAKER_02: So unfortunately, I think the main uh uh consumer product around this at the current stage is golf clubs, because you know, there's uh rich people can pay for those things, but uh uh definitely been looked at in drug development and other areas where if you slow down some of the process of crystallization, you can get some new forms of how you actually develop this.
• 49:10
  SPEAKER_02: So, as I understand it is happening in some places, just in very niche areas of where you get in the demand right now.
• 49:16
  SPEAKER_00: It's one of the things I think people are are sleeping on a little bit is the um isomorphic labs that DeepMind have recently started.
• 49:25
  SPEAKER_00: Um I so I I sense within the room we have a shared love and appreciation for DeepMind.
• 49:30
  SPEAKER_00: I I uh I adore those things.
• 49:32
  SPEAKER_02: It's a great, great national success story, even when everyone always puts the caveat on it being bought by Google.
• 49:39
  SPEAKER_02: It's that was great for the company in many ways, uh, because it gave them access to huge amounts of compute capital that we wouldn't have been able to provide in the UK.
• 49:49
  SPEAKER_02: I understand the counterfactual argument and the the story, but it was always because now being based in King's Cross, it was always going to be this massive talent multiplier, and now you'll begin to see that in places like you know, obviously you're isomorphics with his house, but you know, David Silver's ineffable labs, you know, the great the biggest pre-seed round in Europe, others that are spinning out left, right, and centre.
• 50:13
  SPEAKER_02: I you know, uh basically, yeah, I'm very much share your love, and I think Demis has been a you know basically a national hero for game all of us.
• 50:21
  SPEAKER_00: Yeah, come come completely agree.
• 50:24
  SPEAKER_00: Um I want I heard it described recently, and what's true of AI in this sense is is very true of natural capital as well.
• 50:31
  SPEAKER_00: So um AI are now providing an exceptional product to people.
• 50:35
  SPEAKER_00: Yeah, everyone can see what is there is brilliant.
• 50:38
  SPEAKER_00: So um we have we have twice in the last year broke the world record for the highest price anyone's paid for a nature-based carbon credit.
• 50:45
  SPEAKER_00: So objectively, we have put a great product out there.
• 50:48
  SPEAKER_00: The argument I heard, I think it was on a Harry Stebbings 20 VC podcast, was that the product's great, but we haven't really solidified the business model on AEI yet.
• 50:58
  SPEAKER_00: People are trying lots and lots of different things.
• 50:59
  SPEAKER_00: I think the one they're sleeping on is um the the pharmaceutical, the drug part of DeepMind may well end up being as profitable as anything else.
• 51:09
  SPEAKER_00: And no one else is um, and that was Demis' core, right?
• 51:12
  SPEAKER_00: This was a science enterprise for him.
• 51:13
  SPEAKER_00: This was a learning and discovery part.
• 51:15
  SPEAKER_00: It's why OpenAI ended up putting out the first transformer technology, despite the fact DeepMind came up with it because he wanted it to be science-based.
• 51:23
  SPEAKER_00: We're having that same battle at the moment in terms of what is the right model for natural capital.
• 51:27
  SPEAKER_00: You know, it's been pushed down this carbon credit route, BNG units, which are great.
• 51:32
  SPEAKER_00: But I don't think anyone ultimately thinks they're the right answer in the end.
• 51:36
  SPEAKER_00: There will be paid ecosystem services, there will be things like the relationship that we just spoke about in terms of data centers and a sort of holistic environmental route.
• 51:44
  SPEAKER_00: But I think with AI or natural capital, we've they're so early in terms of being industries and businesses.
• 51:49
  SPEAKER_00: We've got 10 or 20 years where we will figure out these financial instruments and business models and things.
• 51:53
  SPEAKER_02: Yeah, I think that's right.
• 51:54
  SPEAKER_02: I mean, you could even look at the labs themselves are adapting this as they go along.
• 51:59
  SPEAKER_02: You know, you had Sam Altman last week also saying that he thinks that AI might become a utility over the long time and you pay for your sort of token usage when you pay for the water, right?
• 52:08
  SPEAKER_00: Back in Feb, they were going to do advertising and everything and decided it wasn't for them.
• 52:12
  SPEAKER_00: And you you mentioned acquired earlier.
• 52:14
  SPEAKER_00: I'm actually listening to the three-part Google series again.
• 52:17
  SPEAKER_00: And um, I'd I'd forgotten this.
• 52:19
  SPEAKER_00: I was broadly aware of some of the bits, but Google had like 10 or 12 different runs at getting their ad model right.
• 52:25
  SPEAKER_00: And some of the early ones they ended up costing themselves money or they were so gamifiable that if you owned a business and you just kept clicking on things because it wasn't pay per click, you could ramp up your rating by just having like interns clicking on things.
• 52:37
  SPEAKER_02: So yeah, well, that was there's there's a long way to go on lost.
• 52:40
  SPEAKER_02: I mean, I don't know if you even saw that story last week on the the company that didn't put credit limits on its anthropic use and the guy ran up for billions.
• 52:50
  SPEAKER_02: Um yeah, you know, I think there's this is the thing about this the change right now, more broadly, as well, right?
• 52:57
  SPEAKER_02: Is you know, yeah, and I mentioned this earlier, you know, if you're building a startup right now, you're sort of almost your existential fear often is like it's just it's one of the big guys just gonna immediately do it, right?
• 53:08
  SPEAKER_02: Um but you know there is still a long way to go on some of the economics and sort of business models around that.
• 53:17
  SPEAKER_02: Now, uh there's been debates of like the bubble and like scaling laws breaking down, all these things as well.
• 53:23
  SPEAKER_02: You know, the sort of natural limits on the you know training runs as well.
• 53:26
  SPEAKER_02: If you begin to get to like a trillion dollars on a training run, you know, where's your sort of you begin to get your you know your revenue returns?
• 53:33
  SPEAKER_02: Like it's one of those areas more broadly, which is like I mean, because I've been living in it since 2022, it's still it feels quite old in some ways, but then you realize it's it's just so new, you know.
• 53:48
  SPEAKER_02: And like I know the meme is now we're still so early, but it you know, you you both I I constantly feel both early and late within this one, um, because of the speed that's happening, and you sort of you you veer between these two states sometimes.
• 54:04
  SPEAKER_02: Uh uh, but you know, it's but then you realize these questions, and a lot of people are just trying to figure it out as they go along in in the same way.
• 54:12
  SPEAKER_00: I um so you'll get immersed in something, you'll be building, working with the team.
• 54:16
  SPEAKER_00: So um, one of my favorite facts um we currently in the team have more PhDs and master's degrees than DeepMind had when they sold to Google.
• 54:26
  SPEAKER_01: Oh wow.
• 54:26
  SPEAKER_00: Now that's that's probably the only logical alignment and comparator at the time, but I love that, and it's just a tip of the hat to what they've achieved as has been astonishing.
• 54:35
  SPEAKER_00: But you you spend a lot of time within the business talking to people about tech and AI and data and our approach to we've got this amazing um in the time we've been doing this, we've collected 500 billion data points about the portfolio we own, and we can now visualize carbon stocks to individual trees by the centimetre level.
• 54:53
  SPEAKER_00: So change changing things dramatically there.
• 54:55
  SPEAKER_00: So you're in that day-to-day, and then you step into the sort of real world, and all of a sudden, oh AI is not going to be a thing.
• 55:01
  SPEAKER_00: So it's like absolute 80% of the time exposed to this, and then you catch yourself, oh wow, no, there's wider.
• 55:07
  SPEAKER_00: Do you have similar moments?
• 55:08
  SPEAKER_02: Totally, totally, because it's it you know, it you you begin to almost it's the same with any thing where you begin to get so very immersed in it, and you know, you begin to have these conversations, and you you build up the lexicon and you're speaking these terms which are just totally alien to other people.
• 55:26
  SPEAKER_00: Neural networks, reinforcement learning, yeah, yeah.
• 55:29
  SPEAKER_02: Yeah, exactly.
• 55:30
  SPEAKER_02: The one thing I do the the thing I do think is different with AI though, which I do find interesting, like it if you'd asked a if you've gone uh yeah, the public are never stupid, right?
• 55:45
  SPEAKER_02: Uh but if you've gone back sort of 10 years ago and asked them, like, you know, what's cloud computing, is it important, and they're kind of like, maybe they didn't sort of they engage with it because they store their you know their photo their photos on their phone on it or etc.
• 55:58
  SPEAKER_02: But it wouldn't have been something that sort of deeply sort of resonated in how they sort of feel about the world.
• 56:05
  SPEAKER_02: Even if people haven't aren't engaging with AI sometimes on that sort of, let's say the technical level, or the I think they're very they're very common sense understanding of what this means like actually for sort of the world and the and the potential size of it.
• 56:21
  SPEAKER_02: And consumer statistics show how much people have sort of adopted this and how much they sort of use it on a day-to-day.
• 56:29
  SPEAKER_02: But this is why the debate on things like jobs, data centers, all these things, they they are they will resonate because people have absorbed that this is a really important issue.
• 56:42
  SPEAKER_02: Uh, and so even though you know they might not know the neural network, or they might they they they know that this is gonna this is gonna mean something different for them and their kids.
• 56:51
  SPEAKER_00: And that's almost the the fear, the concern is a signal that people are aware and taking it seriously, whereas you'd be met by apathy otherwise.
• 56:59
  SPEAKER_01: Yeah.
• 56:59
  SPEAKER_00: And um that's really interesting.
• 57:02
  SPEAKER_00: I hadn't thought about it that way.
• 57:04
  SPEAKER_00: Um move moving on to um some more of the work you've been looking at.
• 57:08
  SPEAKER_00: You've been thinking about you know the need for more disruptive innovation in the UK and how we might stimulate some of those things.
• 57:14
  SPEAKER_00: So can you talk to us a little bit about that?
• 57:17
  SPEAKER_02: Yeah, well uh I I think sort of on a macro level my view of society history economics is is one of uh quite a deep love of science and technologies just because of what I think it has done to change the course of humanity, right?
• 57:36
  SPEAKER_00: Do you remember where that first came from?
• 57:37
  SPEAKER_00: Do you remember when you were first sort of excited about technology?
• 57:41
  SPEAKER_02: Um I think I always have been to some degree.
• 57:44
  SPEAKER_02: Like um like my dad was very into like gadgets and like you know, he was an architect.
• 57:51
  SPEAKER_02: Um I remember when you know we got CAD on the the on the on the computer who's like one of the I was definitely one of the first kids at school to have a computer, and then we used to just hack around on the back end on MS DOS and you know that's probably almost I'm gonna betray my age here, you know.
• 58:08
  SPEAKER_02: Of course, every kid of my age playing around on things like Encarta, uh, you know, it's probably the first form of Chat GPT to do your to do your homework, you know.
• 58:16
  SPEAKER_02: That you know, we're talking about plagiarism again, but you know, the amount of kids that would copy and paste stuff off Encarta just to you know dial it in.
• 58:24
  SPEAKER_02: So uh again, there are rhymes of history through all of these things.
• 58:27
  SPEAKER_02: Um so I think that it's sort of been almost a precondition of how I thought about politics and government, all those things anyway.
• 58:39
  SPEAKER_02: And I think it's because people sort of compute they can get quite narrow on how they think about technology, right?
• 58:46
  SPEAKER_02: As well, which is yeah, uh because it's not you're talking about it's more the sort of the philosophical sense of it being about human progress and how we you know how we develop things and we develop new tools, like in the same way in the Bronze Age, you develop like the tools for that, like you know, we are continuous continually as a you know species developing tools that improve what we can do and and expand the bounds of humanity.
• 59:11
  SPEAKER_02: And now, like if you take the last 200 years from the industrial revolution on, like some of those breakthroughs are you know, some of them are industrial, obviously, things like the steam engine and um uh and components like that, with the you know, but then it's also penicillin, it's like you know, it's air travel, it's like all of these things that compound over time, but they just you know they create the path, right?
• 59:36
  SPEAKER_02: And I I think it's weird that we to me that almost feels obvious.
• 59:42
  SPEAKER_02: Um uh that you know, that next wave of economic growth, opportunity will be by creating the new tools and systems and ideas that like will be there for it.
• 59:54
  SPEAKER_02: Now, it doesn't always have to be science and technology, by the way.
• 59:57
  SPEAKER_02: I think it's just almost the broad term that you you begin to use them for.
• 1:00:00
  SPEAKER_00: But zero to one Peter Thiel articulates it that way that when he says technology, it's any form of innovation of any kind, whereas people have now fixated that it's electronic computers as opposed to exactly, and I and I think maybe some of the physical world stuff is beginning to come back a bit.
• 1:00:16
  SPEAKER_00: I I think um There's been a lot of money raised there recently, hasn't there?
• 1:00:20
  SPEAKER_00: That that rush to because we're seeing that the biggest players move towards the data centers, the room is it's the picks and shovels argument.
• 1:00:27
  SPEAKER_02: Yeah, exactly.
• 1:00:28
  SPEAKER_02: And uh like people always understand picks and shovels rather than sort of data and AI and etc.
• 1:00:35
  SPEAKER_02: But you know, some of the I guess the macro trends on the physical world are things like obviously defense and energy and all these things which are becoming more important again to sort of national debates and obviously the reindustrialization debate that we sort of briefly touched on a bit earlier, um, which has come back in particular in advanced manufacturing.
• 1:00:54
  SPEAKER_02: But in many ways, also all of these things are only becoming possible because of them, again, the compounding effects of the breakthroughs of other technologies, because it it even if you take space travel, like you know, the ability to do that at low cost is because of SpaceX, but there are a load of other you know elements that have come into that which have driven that.
• 1:01:13
  SPEAKER_02: I mean, there's a satellite company in the US called Planet Labs, which does you know real-time observation of the Earth all the time.
• 1:01:21
  SPEAKER_02: And uh Will Marshall, who founded that, was a former NASA scientist.
• 1:01:25
  SPEAKER_02: And you know, it's a slightly like clipped version of the story, but I love it, which was like when they were sending a NASA rocket up into space, one of you know, him and another NASA colleague were wondering, does consumer tech work in space?
• 1:01:38
  SPEAKER_02: So they like you know, strapped an iPhone to a rocket and then had it take pictures of the Earth and went, oh lo and behold, like why are we spending like 50 million on every satellite launch?
• 1:01:50
  SPEAKER_02: We can probably send this up for 200 bucks.
• 1:01:52
  SPEAKER_02: So, like that was the fundamental breakthrough there as well.
• 1:01:55
  SPEAKER_00: So you've seen that visualization that shows the number of satellites in space by year, haven't they?
• 1:02:00
  SPEAKER_00: So it's one of those things that plays plays through on LinkedIn and it's like one a year, one a year, one a year, and it gets towards when um our Elon's uh satellite um the Starlink.
• 1:02:11
  SPEAKER_00: Starlink, thank you, and it's suddenly like a thousand a year, bang, bang, bang, bang, bang, and all of a sudden you've got this constellation up there.
• 1:02:17
  SPEAKER_02: Yeah, amazing.
• 1:02:18
  SPEAKER_02: I mean completely the point as well.
• 1:02:20
  SPEAKER_02: Yeah, and it's you know, these things all just and they begin to collide and they offer you new opportunities, then you think about them, and then you you know you can play around with all these things in different ways.
• 1:02:30
  SPEAKER_00: And I think that's sort of just what I find quite exciting, and I think it's you're you're articulating the um one of the driving philosophies about what what we did as a business.
• 1:02:40
  SPEAKER_00: So if I take you back to 2021, so it was our fifth birthday yesterday, ahead of 2021, we've we've had some amazing, very generous funders that have allowed us to do what we're doing.
• 1:02:52
  SPEAKER_00: But their thesis was there's no ambition, audacity, scale in the environment sector uh at the time.
• 1:02:59
  SPEAKER_00: And actually, embarrassingly, still today, you will see a press release for a 10,000 pound environmental project.
• 1:03:04
  SPEAKER_01: Yeah.
• 1:03:05
  SPEAKER_00: And I and I said, right, we are gonna build a 100 million pound portfolio of land for conservation and natural capital, and we were literally laughed at, and we should have been, because it was a moonshot at the time.
• 1:03:14
  SPEAKER_00: To say that from where it was, um I sat on a stage yesterday with multi-billion pounds of investment in this sector for businesses that didn't exist five years ago.
• 1:03:23
  SPEAKER_00: And what we wanted to do, and we we didn't know if we'd be alone or there'd be others around it, but there now are some great and British companies doing this.
• 1:03:31
  SPEAKER_00: There's been enough scale that's now in the sector, you have got those second and third and fourth order businesses that are coming as a result of that.
• 1:03:38
  SPEAKER_00: Because there is now billions of pounds worth of carbon credits and things, you have now got it, there's a really great bubbling UK insurance industry for natural capital products.
• 1:03:47
  SPEAKER_00: Now, if you've got an insurer, you also need incredible data for them to understand what they're doing.
• 1:03:51
  SPEAKER_00: So there is a whole plethora of data and tech.
• 1:03:54
  SPEAKER_00: And you've got people like Mosaic Earth, who are a essentially distributed ecology service that uses consumer products to take photographs of things and feeds those into AI and into experts to identify things that are out there.
• 1:04:08
  SPEAKER_00: You've got amazing machine learning algorithms that are scanning and evaluating the environment.
• 1:04:12
  SPEAKER_00: None of those would have existed without the catalyst of the investment into the sector to create all those other businesses.
• 1:04:17
  SPEAKER_00: So that's why I'm so excited about what AI can bring.
• 1:04:20
  SPEAKER_02: Yeah, for me, a lot of the time as well, it just increases the surface space of what you can do.
• 1:04:25
  SPEAKER_02: Like in a very simple way, it's just like you know, a lot of people be talking about this AI psychosis, and I did worry that I had it for a little while because it was like you know, things that would sometimes could take weeks, months, etc., I could just go around and play with and do in a couple of days.
• 1:04:44
  SPEAKER_02: And I remember this little period where for about two weeks I was like, I don't want any meetings, I don't want anything, I don't want to, I just need to get my head around this, and I just want to build and play around with some things and came out of it like, you know, probably a bit like sort of you know Tom Hanks on you know castaway or something.
• 1:05:01
  SPEAKER_02: But it was like it it, you know, you just sometimes need that time to go and play and understand and work out what you can do and to do and I just sort of came away from that being, you know, I was totally convinced, but then I was like, okay, like there is basically nothing else that you you want to be focused on right now, because this will just be what completely sort of 10x is, 100x's, 1000x is what you want to do in the next few years, and I yeah, I will be on top of it.
• 1:05:26
  SPEAKER_00: I have a similar cycle, so because the nature of what I do, I'll I'll get two or three days at home, then I'll be traveling for a couple of weeks and doing the actual CEO part of the job.
• 1:05:35
  SPEAKER_00: And I look so forward to those two or three days at home of being behind a big screen with bulletproof Wi Fi.
• 1:05:40
  SPEAKER_00: I get to build and play.
• 1:05:41
  SPEAKER_00: The slightly annoying bit is some of these incredibly gifted people in the team, they're at home getting to do that all the time.
• 1:05:47
  SPEAKER_00: So I catch up a little bit when I go home, and then I step away again.
• 1:05:50
  SPEAKER_00: And when I come back, I'm like, oh Christ, I'm so far behind again.
• 1:05:53
  SPEAKER_02: Yeah.
• 1:05:54
  SPEAKER_02: But you know, that's why you hire brilliant people who are smarter than people and ask them what to do.
• 1:05:58
  SPEAKER_02: I totally agree.
• 1:06:00
  SPEAKER_02: Let go of that sound.
• 1:06:00
  SPEAKER_02: Palmalocke talked about this actually with Andrew.
• 1:06:05
  SPEAKER_02: Whereas because he was obviously an extremely good VR guy.
• 1:06:08
  SPEAKER_02: And I mean, you know, pretty much up there.
• 1:06:10
  SPEAKER_02: And then he's like, you know, he then hired someone to run that division who was like obviously better than him.
• 1:06:15
  SPEAKER_02: And at what point you then just sort of you decide, you know, leading now is just about allowing them just to go off and maximize all this, and you just helping them to do that.
• 1:06:26
  SPEAKER_02: You know, and it's like you don't hire, you know, even if you become CEO and you, you know, you came through the CFO route, you don't then try and still be like the best CEO, a CFO within the company.
• 1:06:34
  SPEAKER_02: You just like you get the the people that are better than you and then you you work on that.
• 1:06:38
  SPEAKER_02: And I think it's always a good lesson.
• 1:06:40
  SPEAKER_02: I have some amazing people who just you know make me feel quite stupid a lot of the time, but uh, you know, they just absolutely just allow them to own their domains and crush it.
• 1:06:50
  SPEAKER_00: So the phrase I like is I've sacked myself 50 times and replaced it.
• 1:06:54
  SPEAKER_00: Um so I guess the best example we've got of that is so the first concept of our proprietary model that values every square meter of land, uh I put together.
• 1:07:02
  SPEAKER_00: Um, it is now evolved so much I'm not even allowed to open it.
• 1:07:06
  SPEAKER_00: So it's so far away from me.
• 1:07:07
  SPEAKER_00: I do not have access to the folders it sits in because the team are running it at such a level.
• 1:07:12
  SPEAKER_00: So uh shout out to the intelligence team, they're doing some amazing stuff.
• 1:07:15
  SPEAKER_00: Um I will pick your brains on recruitment a little bit, actually.
• 1:07:18
  SPEAKER_00: So um it must be really hard for you guys to choose because um are there any other equivalent to the institute globally?
• 1:07:27
  SPEAKER_00: Are are there yeah, what are the nearest comparators in terms of organizations?
• 1:07:33
  SPEAKER_02: Um we we have thought about this.
• 1:07:35
  SPEAKER_02: And there are obviously people that do similar things in diff in similar areas to some degree.
• 1:07:42
  SPEAKER_02: So there's obviously in our advisory work, you know, you look at the you know, there's the advisory from firms around the world, your McKinsey's, etc.
• 1:07:50
  SPEAKER_02: Then probably in some of the more sort of philanthropic space, there have been people like the Gates Foundation that have done things.
• 1:07:56
  SPEAKER_02: But the honest answer is no one that really does anything that aligns with our sort of core USP on working mostly with prime ministers, presidents, political leaders, bringing both you know, policy and thought leadership as well as now technical and delivery skills as well.
• 1:08:16
  SPEAKER_02: You know, I think it's quite a unique business on that, and it melds a few different components all together, which is almost a complete way to help, I guess, import state capacity if you're uh uh a government around the world as well.
• 1:08:33
  SPEAKER_02: And you know, we are constantly thinking about how we improve that and and do it.
• 1:08:41
  SPEAKER_02: Um, and sort of you know, as with every business, what is your USP within the market?
• 1:08:46
  SPEAKER_02: Um, but in some ways, actually, one of the reflections that we had recently is that the mission hasn't actually changed very much almost since day one.
• 1:08:53
  SPEAKER_02: It's just how you go about doing that, has and that idea of just helping governments with their reform agenda, helping leaders lead, it's kind of always what Tony has done post office based on his own experience of leadership.
• 1:09:05
  SPEAKER_00: The onus on certain tools will change a lot.
• 1:09:07
  SPEAKER_02: Exactly.
• 1:09:07
  SPEAKER_02: It's just how you do it, just changes.
• 1:09:09
  SPEAKER_02: And so you, you know, you raise delivery units earlier, like you know, a a delivery unit in you know, these sort of the olden days, let's say, you know, would have been a few really bright people and like you know, a couple of Excel spreadsheets sending out emails, like filling in the details, uh, you know, and then just you know, probably doing, you know, well, or definitely doing like a stock take meeting every week.
• 1:09:32
  SPEAKER_02: Like, is that on track?
• 1:09:34
  SPEAKER_02: Like, can we put a little traffic light system in to show where it is?
• 1:09:38
  SPEAKER_02: Where are we at risk on that?
• 1:09:39
  SPEAKER_02: And it's like, yeah, it's these are very brilliant people, but it's like it's a it's a labor-intensive way of doing it.
• 1:09:46
  SPEAKER_02: Now, delivery unit today is okay, you still need some people that can understand that, but they're you know, they're not doing an Excel spreadsheet and whipping around via research.
• 1:09:59
  SPEAKER_02: And so you're you're you in many ways, you've kind of codified the software version of that human side, but then you've also then had to reimagine the system about how it actually does that as well.
• 1:10:13
  SPEAKER_02: So um, and you know, to that point, it definitely removes some of the cost, but it also dramatically improves the outcomes as well.
• 1:10:22
  SPEAKER_02: So, yeah, so the mission, many of the evidence is much the same.
• 1:10:26
  SPEAKER_02: It's just there are just so many better ways of doing what we were able to do even a few years ago.
• 1:10:30
  SPEAKER_00: So when you and you you touched on it then briefly, but when you think about recruitment now, so I I don't even want to guess the number of applicants you have in terms of roles.
• 1:10:40
  SPEAKER_00: Has what you're looking for changed in the era of AI in terms of talent, or are you still looking for the same fundamental principles you were previously?
• 1:10:48
  SPEAKER_02: I think still a lot of the fundamental principles are the type of people that you want.
• 1:10:52
  SPEAKER_02: And because I think probably everyone sort of will probably say the same thing, like you know, problem solvers of high agency and could get stuff done, you know.
• 1:11:01
  SPEAKER_00: Yeah, universally good everywhere.
• 1:11:02
  SPEAKER_02: Yeah, yeah, exactly.
• 1:11:04
  SPEAKER_02: Um I I think the the difference that we have right now is sort of you know, we do look for more technical skills, obviously, particularly in the engineering team.
• 1:11:14
  SPEAKER_02: Um but also the one thing that I'm constantly thinking about right now is that the AI will commoditize a lot of like the sort of the simple intelligence, let's say.
• 1:11:29
  SPEAKER_02: Um so you know, we've been building internal tools like our kind of internal brains for you know specific domains, like that's a sort of geopolitical advisory or uh uh you know, you know, policy thinking as well.
• 1:11:45
  SPEAKER_02: Um and what you then really highly value are the people that just have such deep domain knowledge and expertise that know how to then operate the system and sort of add on the value add that AI can't do because you can just say to it, can you write me a quick slide deck on supply chain risks around the world?
• 1:12:06
  SPEAKER_02: And it can, you know, it's not a an amazing job, but it's like you can turn that around pretty quickly in a in a pretty good way.
• 1:12:12
  SPEAKER_02: Uh and then it in many ways, then just extremely bright, high-agency young people that come in with new ideas and they're pretty like you know, on top of the tech as well.
• 1:12:23
  SPEAKER_02: And you know, that's it's a little bit of a gap there sometimes, if you can see where the where the missing gap is on the sort of the recruitment side.
• 1:12:31
  SPEAKER_02: Um, but that's that's sort of a lot of where I think the composition is at the moment.
• 1:12:36
  SPEAKER_02: And now you know, so the the products change and the tools change, so you the shift might change over time as well.
• 1:12:42
  SPEAKER_02: But that's definitely the high value that we find.
• 1:12:44
  SPEAKER_02: So it's so interesting.
• 1:12:46
  SPEAKER_00: Um during our San Francisco Climate Week, I was really fortunate to meet with Sastra, the company.
• 1:12:53
  SPEAKER_00: Um, I don't know if you know those guys at all.
• 1:12:55
  SPEAKER_00: So uh Jason, the chap off the 20 VC podcast, that's his company.
• 1:12:59
  SPEAKER_02: Oh, yeah, okay.
• 1:12:59
  SPEAKER_00: So um Oh, do you know who you mean then?
• 1:13:02
  SPEAKER_00: They they told this incredible story about how there were about 30 people doing a couple of billion a year, and they lost half of their team to Salesforce and LinkedIn for 10x salary gross because they were so talented.
• 1:13:16
  SPEAKER_00: And they sat and went, we can't recruit in Silicon Valley, like it is not viable to get talent because of how it's pulled away.
• 1:13:22
  SPEAKER_00: So they were the first, it's about two or three years ago, to go into they replaced every one of those roles they lost with an agent of some kind, and them talking about making that a reality as opposed to just the concept of it.
• 1:13:33
  SPEAKER_00: And the key learnings I took from that conversation was they have agents at the time and it's always moving, but that did very, very specific things.
• 1:13:41
  SPEAKER_00: So they would say one agent, for example, literally just went through every inbox and found any lead that had gone cold.
• 1:13:47
  SPEAKER_00: Yeah, and they said the beauty of that is we know how to do that.
• 1:13:51
  SPEAKER_00: So if you're gonna build an agent, you need to know exactly how to codify it and build it.
• 1:13:55
  SPEAKER_00: Otherwise, it's gonna be a very bad agent.
• 1:13:57
  SPEAKER_00: Yeah, exactly.
• 1:13:58
  SPEAKER_00: So having those people that can replace themselves specifically up to a function, um, which was amazing.
• 1:14:03
  SPEAKER_00: And then they said, so it would the other beautiful point on that was a human would be embarrassed to point out someone else's error or mistake, or doesn't want to throw their friend under the bus and go, you didn't follow up on that email.
• 1:14:14
  SPEAKER_00: The agent doesn't care, is able to do it.
• 1:14:15
  SPEAKER_00: And the other thing that I picked up from them that massively changed my workflow.
• 1:14:19
  SPEAKER_00: So I so I write a couple of pieces a week that goes out to a few thousand people.
• 1:14:22
  SPEAKER_00: Um, I've now changed when that's released to a Wednesday evening and a Sunday morning.
• 1:14:28
  SPEAKER_00: Okay, yeah.
• 1:14:28
  SPEAKER_00: So everyone teaches you, well, Monday mornings when you want to send stuff out there.
• 1:14:32
  SPEAKER_00: They learn because they tested hundreds of thousands of pieces of agent-driven correspondence, that actually the decision makers in Silicon Valley go home on a Wednesday night because it's the middle of the week and check their inbox.
• 1:14:42
  SPEAKER_00: And on a Sunday morning, they've done the family stuff on a Saturday, they clear their inbox on a Sunday morning.
• 1:14:47
  SPEAKER_00: I've almost doubled the open rate of our weekly newsletter by going to Wednesday and Sunday.
• 1:14:51
  SPEAKER_02: It's definitely that would align with my cadence of work as well.
• 1:14:55
  SPEAKER_02: So yeah.
• 1:14:56
  SPEAKER_00: It's it's so interesting, isn't it?
• 1:14:57
  SPEAKER_02: Yeah, yeah.
• 1:14:58
  SPEAKER_02: I mean, on the that agent agentic point, I think, is bang on as well.
• 1:15:02
  SPEAKER_02: And I think but I think one of the things that also gets sometimes getting a little bit confused in the vape right now, and particularly on the automation stuff, is that that is an extreme example of task automation, not job automation.
• 1:15:15
  SPEAKER_01: Yes, yes.
• 1:15:16
  SPEAKER_02: And and and this is where I mean you'll see so many, so much ability of a as a company to automate specific tasks.
• 1:15:23
  SPEAKER_02: And often they're the really again, the boring ones that no one wants to do because they're you know, uh, but that won't negate sometimes the need for new jobs, which then get creative for some of the you probably still want then that human to then go out and speak to the customers once they've done the follow-up, etc.
• 1:15:38
  SPEAKER_02: And it might shift some of the composition to that, but you you're automating tasks often, not jobs as well.
• 1:15:44
  SPEAKER_00: The um the guest we recorded with before you is a specialist in evaluating senior talent for VC firms.
• 1:15:51
  SPEAKER_00: Yeah.
• 1:15:51
  SPEAKER_00: And and I asked him about how that's changed what the most in-demand talent is, and he said the more charismatic personal people are more valuable than ever before because they're your closers.
• 1:16:01
  SPEAKER_00: They're your they're they're they're your closers in cricket, they're your goal scorers in football, you need them and they'll attract a premium as a result.
• 1:16:06
  SPEAKER_02: Yeah, interesting.
• 1:16:08
  SPEAKER_02: Yeah, there's a there's a guy on um an open AI um researcher, one of the early guys, he's uh he's called Rune on Twitter, he's brilliant.
• 1:16:19
  SPEAKER_02: Um he had this theory a few years ago, which were the um God, I'm gonna try and remember, it's the shapeshift, not shapeshifters, it's something about shapeshifters versus word cells, um, which is obviously the people that could talk eloquent prose were like gonna be completely annihilated by the the people that could manipulate um uh code and all these things, and uh it sort of feels like the balance then swings back as well because you actually can automate a lot of the code and some of the software side, but what you can't replicate is that extremely soft skills and sort of the the stuff, the persuasion that people sometimes need as well.
• 1:16:59
  SPEAKER_00: Well, and if if sports taught us anything, we will shape rotator, yeah.
• 1:17:04
  SPEAKER_00: So it's not wrong, he's he's brilliant, he's very pivot to one aspect.
• 1:17:09
  SPEAKER_00: So cricket will teach us that all of uh the West Indies had five fastbowlers, everyone should have five fastbowlers, we're all gonna do that.
• 1:17:15
  SPEAKER_00: All of a sudden, a team reinvents and suddenly brings a load of leg spinners in, and the game changes dramatically.
• 1:17:20
  SPEAKER_00: You go back through the history of cricket, you have those chapters.
• 1:17:23
  SPEAKER_00: As soon as everyone goes one way, the counterfactual becomes more valuable, rarer, more unique, and and more impactful.
• 1:17:29
  SPEAKER_02: It also forces you to get better, by the way.
• 1:17:30
  SPEAKER_02: If you talk about the history of DeepMind as well, that okay, like Go is now completely changed as a game.
• 1:17:36
  SPEAKER_02: Yeah, but if you look at what happened, the Go players became better because the AI was so good.
• 1:17:41
  SPEAKER_00: And it's more popular than it's ever been, isn't it?
• 1:17:43
  SPEAKER_00: People were worried that, and the same with chess that it was going to destroy it, and actually it's forced us to another level of understanding and you know, we're we're quite responsive to competitive forces as a species, so yeah.
• 1:17:54
  SPEAKER_00: I love it.
• 1:17:54
  SPEAKER_00: Um what are you excited about for the the rest of this year, both personally and as as part of TBI?
• 1:18:01
  SPEAKER_02: Um number one, I guess.
• 1:18:05
  SPEAKER_02: Like I I will just continually be excited about some of the breakthroughs in AI because you know, that point on the rate of change that I've made, it's like it it is quite exhilarating.
• 1:18:17
  SPEAKER_02: There's a bit of a there is I think the ups and downs on this is say are sort of almost like the dopamine hits as well, and you kind of get the like uh but you know, it is this is one of the things I also just I can't quite again comprehend sometimes when some of the people that are let's say I I I'll just the polite way I'll call them incurious about AI because it's like even even if you don't fully buy into it and you think it's all a bit overhyped, etc.
• 1:18:49
  SPEAKER_02: Like there is this incredibly consequential thing happening in the world right now, of which like this has happened a few times throughout history, and to almost sort of be willfully ignore it seems very odd to me.
• 1:19:08
  SPEAKER_02: And my wife seems very diplomatic away from that.
• 1:19:11
  SPEAKER_00: Yeah, so my my wife first in maths, yeah, brains of the outfit without a shadow of a doubt, zero interest in AI.
• 1:19:17
  SPEAKER_00: Interesting.
• 1:19:18
  SPEAKER_00: To the point whereby um, and then I kind of reconcile maybe I haven't given her the access use case.
• 1:19:24
  SPEAKER_02: No, and it won't be for everyone, by the way.
• 1:19:26
  SPEAKER_02: That's also fine.
• 1:19:28
  SPEAKER_02: I think conscientious objectors feels, yeah.
• 1:19:31
  SPEAKER_02: Fine.
• 1:19:32
  SPEAKER_02: I will have a few of them.
• 1:19:34
  SPEAKER_02: Um yeah, but it it's I I think for me, the the fundamental point around some of this is like even if you don't like something, you still have you, you know, you're gonna have to sometimes engage with gravity, yeah.
• 1:19:52
  SPEAKER_00: You don't get to choose, yeah.
• 1:19:54
  SPEAKER_00: It's gonna happen anyway.
• 1:19:54
  SPEAKER_00: Like people say, I don't like change, yeah, time's gonna pass by.
• 1:19:58
  SPEAKER_02: Yeah, and I think I'm I'm I'm obviously more aiming this at some sort of like politics and government, because like you either way you're gonna have to respond to it.
• 1:20:08
  SPEAKER_02: So you might as well respond to it from a point of understanding rather than a point of ignorance, um, because you know, I know which which side is gonna get you the better outcomes as well.
• 1:20:20
  SPEAKER_02: Um now I and I think this is sort of almost a structural issue sometimes with politics.
• 1:20:28
  SPEAKER_02: Now, there are reasons why when some of our most prominent politicians are asked do you use Chat GPT, they say no, like because you know they'll be hammered.
• 1:20:39
  SPEAKER_02: There's also the FOI laws, which mean that they can, you know, the things they're using it for can be be searched.
• 1:20:45
  SPEAKER_02: Now, but I think that's also kind of a only part of the answer because I do think one of the things that has often happened in politics in this way is that the people that are regulating or making policy on things that they have just no experience with, and I think that is just uh yeah, I just I I cannot ever believe that form of policy making as well, because it's we got close last week in the US, right?
• 1:21:13
  SPEAKER_00: There was almost the presidential intervention, and Trump himself went, I am not signing this, and they were going to mandate that all new models had to be reviewed by government for 90 days.
• 1:21:23
  SPEAKER_00: Who do they think's reviewing these things?
• 1:21:25
  SPEAKER_00: The people that built them are going to all be billionaires at some point in the future, or certainly multimillionaires, are getting paid insane numbers, and you're asking a civil servant to audit these models.
• 1:21:35
  SPEAKER_00: Yeah.
• 1:21:36
  SPEAKER_00: It's just not a plausible scenario.
• 1:21:39
  SPEAKER_02: Yeah, well the the the there is one version of this which I which I'm actually think we should be quite proud of as the country, actually, the security institute that was set up within the UK government.
• 1:21:52
  SPEAKER_02: Um so this was it was previously called the Safety Institute, but it got rebranded as the Security Institute, and that does a lot of model evaluation.
• 1:22:01
  SPEAKER_02: But it what it does is and it is probably I would say, in terms of like state capacity around the world, it's it's probably probably the highest form.
• 1:22:13
  SPEAKER_02: I've never heard of this.
• 1:22:14
  SPEAKER_02: Yeah, so um it it's and it was run by um well Ian Hogarth, who is the um who is the founder of Songkick and then is runs Plural Platform, or was one of the partners of Plural Platform, the VC is the chair of it.
• 1:22:29
  SPEAKER_02: So an extremely technically competent, you know, technologically competent person as well.
• 1:22:34
  SPEAKER_02: But it what it managed to do was draw in world-leading researchers from DeepMind, Anthropic, and other places.
• 1:22:42
  SPEAKER_02: Um, some of them who you know maybe only come for two years, etc., but care quite deeply about um the state having um understanding and and control over it.
• 1:22:52
  SPEAKER_02: And I think that has been the best response to that kind of element today, which is like actually building state capacity that can understand the technology, evaluate it, and then come up with responses.
• 1:23:05
  SPEAKER_02: And yeah, so there is one version of that.
• 1:23:07
  SPEAKER_02: Uh Singapore has looked at a similar model as well, and Japan and South Korea have.
• 1:23:13
  SPEAKER_02: But I I think the UK actually, I say it because you know someone's quite hard on the government, but it is one area we actually got very, very right on that on that on that question.
• 1:23:24
  SPEAKER_00: Love it.
• 1:23:24
  SPEAKER_00: And and what about TBI in general?
• 1:23:26
  SPEAKER_00: What does the rest of this year look like that's particularly exciting?
• 1:23:29
  SPEAKER_00: And you can deviate away from AI if you wish.
• 1:23:31
  SPEAKER_02: Yeah, no, it's um I mean, I think there are you've got this huge technological revolution happening right now, you've got this huge geopolitical disruption happening right now.
• 1:23:43
  SPEAKER_02: Uh you you know, you've got um obviously the war in the Middle East or multiple sort of conflicts in the Middle East.
• 1:23:51
  SPEAKER_02: Um, you know, you've got war on European doorstep.
• 1:23:55
  SPEAKER_02: This is not in you know, this is not an easy period for the world in many levels, but you know, for obviously an organization like us, which is sort of helping political leaders navigate that, it's it, you know, it's an intellectually challenging thing for us to constantly think through.
• 1:24:16
  SPEAKER_02: And um, you know, when often when as with many organizations, sometimes you know, you're at your best when you're dealing with crises, right?
• 1:24:25
  SPEAKER_02: Because it's forcing you to think harder, more innovatively, and do things differently.
• 1:24:30
  SPEAKER_02: And I think actually, one of the sort of kind of these crises often then are quite exciting in some ways.
• 1:24:38
  SPEAKER_02: And it's not to just, you know, there's like it's called causing deep difficulty and and like you know, human suffering on many, many levels as well, but it it also just forces you to think differently about how to do things as well.
• 1:24:49
  SPEAKER_02: And so I think that's one of the interesting periods that we're in right now.
• 1:24:53
  SPEAKER_02: And I don't think it's uh you know, Tony wrote about this as an essay, these two big structural things colliding at the same time.
• 1:25:01
  SPEAKER_02: Um it it you know, it's someone that kind of relishes intellectual stimulation, it's sort of there is a point of me which is like just thinking about you know, how do we answer this?
• 1:25:16
  SPEAKER_02: Because I I think the last point on that is that for a few years that we've been saying that there needs to be this big structural reform to how the state actually operates.
• 1:25:27
  SPEAKER_02: Um, you're kind of seeing like the sort of almost the end point of the manifestation of that to me.
• 1:25:34
  SPEAKER_02: And what I mean by that is like if you take the sort of something like the Bretton Woods solution, it was kind of uh, you know, you know, you had the sort of the post-World War II settlement, etc.
• 1:25:47
  SPEAKER_02: These sort of big big structural changes then lead to new kind of like solutions.
• 1:25:54
  SPEAKER_02: And I think the fundamental thing at this will be actually how the state operates, because like it cannot keep going in the same way that it does today.
• 1:26:03
  SPEAKER_02: Like, number two, actually, the sort of your multilateral order will change um as well.
• 1:26:09
  SPEAKER_02: And again, some of that is already beginning to see, you know, some of that is turning into protectionism and nationalism, but some of it is also leading to other kind of components, you know, particularly something in like in Europe, where you're looking now at more, you know, there's obviously a bit more of an anti-American tech view, but how so how would you stimulate your own innovation ecosystem?
• 1:26:31
  SPEAKER_02: They're not going very fast on Draghi, but you know, it all of these things, you know, lead to different sort of counterforces.
• 1:26:39
  SPEAKER_02: And I think that's it's actually quite an interesting period in many ways for that.
• 1:26:42
  SPEAKER_00: I've sat with a number of significant European institutional investors, so large pension funds, who have said the the structure of governments now are not appropriate for the development of infrastructure anymore.
• 1:26:56
  SPEAKER_00: So, what would have underpinned railroads and things like that?
• 1:27:00
  SPEAKER_00: Someone's gonna have to fill that void.
• 1:27:01
  SPEAKER_00: Now, whether it is institutional investors, whether it is the tech companies of the future, whoever it is, we are gonna have to have other people build the infrastructure because the future's not gonna operate with the current infrastructure that we have.
• 1:27:13
  SPEAKER_02: Yeah, I mean, you see that in the UK.
• 1:27:14
  SPEAKER_02: I mean, we're totally incompetent in building infrastructure today.
• 1:27:17
  SPEAKER_02: I mean, HS2 was a whether you agree with the decision or not to build it at the outset, it was a farce.
• 1:27:22
  SPEAKER_02: Like everything about it then compounded the, you know, like on top of each other.
• 1:27:27
  SPEAKER_02: And it's like, and then like, you know, and again, if you take something like that, if you're a if you're a UK taxpayer, why would you support a government like building huge infrastructure if it ends up costing multiples of what they said it was and doesn't get delivered on time into no one's kind of clear benefits?
• 1:27:44
  SPEAKER_02: So that is you know, uh it you know, this country now is massively underinvesting in it in in infrastructure.
• 1:27:51
  SPEAKER_02: But until you can actually begin to solve some of those structural issues, you can also understand why that will continue, because if the price is too high, people won't want it.
• 1:28:00
  SPEAKER_00: Well, it's nine impossible to plant a tree.
• 1:28:01
  SPEAKER_00: Yeah.
• 1:28:02
  SPEAKER_00: Never mind build a railway or a data center.
• 1:28:04
  SPEAKER_00: Yeah.
• 1:28:04
  SPEAKER_00: I I forget the gentleman's name, but the one that just recently did the government's AI strategy.
• 1:28:09
  SPEAKER_00: Um Clifford?
• 1:28:10
  SPEAKER_00: Yes, thank you.
• 1:28:11
  SPEAKER_00: He spoke wonderfully on a podcast about how, look, let's just compare what it's actually like to do things in the country.
• 1:28:17
  SPEAKER_00: He's like, it's impossible.
• 1:28:18
  SPEAKER_00: Yeah.
• 1:28:18
  SPEAKER_00: Compared to what you see around the world.
• 1:28:21
  SPEAKER_00: You know, I think one of the stats he or facts he shared was that China had built the equivalent of HS2 every year for 20 years or something, and we just hadn't got there.
• 1:28:29
  SPEAKER_00: I always like the sphere example.
• 1:28:31
  SPEAKER_00: Are you familiar with the sphere in Las Vegas?
• 1:28:33
  SPEAKER_00: Yeah, I've actually been there.
• 1:28:34
  SPEAKER_00: Have you?
• 1:28:34
  SPEAKER_00: Yeah.
• 1:28:35
  SPEAKER_00: I never have.
• 1:28:36
  SPEAKER_00: So the story goes that the planning permission for Vegas and London went in at the same time.
• 1:28:43
  SPEAKER_00: Vegas has been open four years, something like that.
• 1:28:46
  SPEAKER_00: We haven't heard the planning determination yet for the London one.
• 1:28:50
  unknown: Oh my god.
• 1:28:50
  SPEAKER_00: Yeah.
• 1:28:51
  SPEAKER_00: So fact check me on that.
• 1:28:53
  SPEAKER_00: But that's the story that I that I heard that backs it up.
• 1:28:56
  SPEAKER_00: Um, Benett, you've been absolutely generous with your time.
• 1:28:59
  SPEAKER_00: I've loved every minute.
• 1:29:00
  SPEAKER_00: This has been occasionally the podcasts are really decadent gifts for me, and the AI deep dive is definitely one of those.
• 1:29:06
  SPEAKER_00: I finish every show with the same question.
• 1:29:08
  SPEAKER_00: What's the one question I haven't asked you you wish I had?
• 1:29:13
  SPEAKER_02: Well, you haven't asked me anything about sport yet, so I would love to do that.
• 1:29:18
  SPEAKER_00: What's on your mind particularly about sport at the moment?
• 1:29:20
  SPEAKER_00: What's catching you up?
• 1:29:21
  SPEAKER_00: Well, we've got the World Cup coming up, which is uh and the biggest one ever, like a hundred and something fixtures, I think.
• 1:29:27
  SPEAKER_02: Yeah, which is incredible.
• 1:29:29
  SPEAKER_02: And I do I do think the England might have a good chance this year.
• 1:29:34
  SPEAKER_02: We say this all the time.
• 1:29:35
  SPEAKER_02: Of course, we're we're a nation of it's actually we're a nation of optimists and believers.
• 1:29:39
  SPEAKER_02: Like we have an oversized capacity to believe that England is uh always going to uh triumph here, but then obviously there's the hope that kills you always on on this side as well.
• 1:29:51
  SPEAKER_02: Um I think one of the things that um I mean obviously this is we we've sort of touched upon this as well a little bit on the revolution in AI and in in sport as well.
• 1:30:10
  SPEAKER_02: And it there are elements of this which I can understand, like you know, VAR in particular becomes a very contentious topic.
• 1:30:20
  SPEAKER_02: And I can I can understand some of the the balance between I guess what can be sort of an attempt towards absolutism and optimization against, I guess, a bit of sometimes the messy randomness of the world.
• 1:30:37
  SPEAKER_02: Um and I I think it's uh for people like you and I who think a lot about data and optimization and all those kind of things, I think there is always an interesting sort of human element that always has to be thought out in all of these things.
• 1:30:51
  SPEAKER_02: And I think you know, you even see this in some of the these are strange sort of analogies on this to some degree, but you know, if you take you take me at 30 years old on wearables, I was like, they are totally the future and we can optimize everything.
• 1:31:05
  SPEAKER_02: And then there's a little bit of me which is like, well, like it's just telling me my heart rate's gone up for a couple of glasses of wine, and it's just a bit annoying.
• 1:31:14
  SPEAKER_02: So you just sort of create the the balance between like I guess what is your sort of you know, living a sort of a more optimised, better life, and but also then uh appreciating some of like I guess what would be some of the the randomness that humanity can bring as well.
• 1:31:31
  SPEAKER_00: Do you think the review system's made cricket better or worse?
• 1:31:34
  SPEAKER_02: Um I think better with cricket, actually.
• 1:31:37
  SPEAKER_02: So do I.
• 1:31:38
  SPEAKER_00: Because I think it's because it's both mostly quick and obviously on quite easy decision, you know, Snicomata, for example, is the fact it stays with the um the unpacked as privacy and they've actually made in some ways T20, I think it's enhanced the game, yeah, but because it fits in with the cadence and the timing and the things like that.
• 1:31:58
  SPEAKER_00: Actually, there's echoes of that with AI as well, right?
• 1:32:00
  SPEAKER_00: We just talked about agents that you know you shouldn't try and automate everything, you should use agents for a specific role at a specific time.
• 1:32:06
  SPEAKER_00: Um and the other bit I think about is how um I do I do love data and the determination of it, but also I love the fact that in rugby they explicitly are not expecting the game to be refereed the same way.
• 1:32:19
  SPEAKER_00: Whatever that person decides is the right answer, and you respect that.
• 1:32:23
  SPEAKER_00: Yeah, and I think because it's authentic and owned, and people know what the rules are and how that's fine.
• 1:32:28
  SPEAKER_00: I think one of the problems I'm not sure.
• 1:32:30
  SPEAKER_02: Alex explained as well.
• 1:32:31
  SPEAKER_02: Yes, yes, the referees, the dialogue is really, really important there as well.
• 1:32:35
  SPEAKER_00: May I've enjoyed every minute.
• 1:32:36
  SPEAKER_00: Thank you.
• 1:32:37
  SPEAKER_00: Really appreciate it.
• 1:32:38
  SPEAKER_00: Yeah, thank you.