# Greg Jensen - Co-CIO of Bridgewater | Podcast | In Good Company | Norges Bank Investment Management

https://www.youtube.com/watch?v=zRkNO9yNYoI

[00:01] Hi everybody and welcome to In Good Company.
[00:03] And today I'm here with Greg Yensen who is the co-CIO of Bridgewwater.
[00:07] Now Bridgewater is uh just an incredible hedge fund, the world largest and best I think we can say and um it's an honor to have you here Greg.
[00:18] Thanks for having me.
[00:19] What are the most important things you are spending time on just now?
[00:22] What are you looking at?
[00:24] Yeah, if I said sort of at the the biggest level, the three big themes that I'm concerned about or focused on are a the change in how the global and US economy uh is being managed.
[00:39] Essentially, what we call a shift to modern mercantalism that's been a reaction to China's rise.
[00:47] It's created political changes across the west mo probably most clearly in the US leading to a whole different philosophy on how to run the economy.
[00:56] So that's one big theme modern mercalism the way the US is operating both with respect to economics but also with
[01:02] respect to geopolitical conditions huge change really important to understand.
[01:08] The second is the technological change that we're in the midst of.
[01:12] I've been as I'm sure we'll get into thinking about and and working with AI for a very long time now machine learning for 15 years and more generally my whole career and I remember people used to say um Greg why are you talking about AI all the time nobody says that anymore um and um and that it's nice to be right at the end is well there's a lot of being wrong along the way plenty to be wrong about but the um but that that it it's so important right it's a third to half of everything geop politics, markets, everything you you have to understand it to understand the macro world.
[01:45] And um and third, that all is happening in a world where capital is more concentrated in the US than ever, more concentrated in equity and illquid assets.
[01:56] Um so a kind of risky setup.
[01:59] So understanding where money has flowed to, why it's flowed
[02:03] there, and how likely that is to change.
[02:06] Those are the big three headlines.
[02:07] Well, let's unpack um let's unpack them um one after the other.
[02:13] Um what are the implications for investors that we are moving towards mercantalism?
[02:17] Yeah.
[02:20] Well, I think a it's it's a big shift if you're if you're thinking the economy works the way it used to work.
[02:27] It's it's a really important change, right?
[02:29] If you take there's a major change in the 1980s, a shift towards small government, a lot of capitalism, a lot of like kind of freedom of the private sector to a very different view, right?
[02:41] That because China has risen, because you've hollowed out the middle class in in a lot of the West, you've got a political reaction takes different forms, but you've got a political reaction in the rest of the world.
[02:50] in Trumpian kind of mindset.
[02:52] The basic issues are okay, you need the government to step in and stop what's been going on in terms of the neoliberal consensus on trade.
[03:01] You need this to
[03:05] basically protect US industry in order to be geopolitically strong.
[03:12] You can't be dependent on other countries.
[03:14] So you need to be independent looking at trade as a zero- sum game where um where if you have a trade deficit that's actually a problem for your wealth and um and that that along with other elements that come with more government control on these things corruption and other things have have changed the way the game is being played.
[03:34] So for investors, you've got to adapt, right?
[03:37] But but if we if if I 12 months ago had told you what the world would look like now in terms of mercantalism, uh you know, tariffs and so on, you would what would you have expected markets to have done?
[03:49] Yeah. Well, if it's only one thing changing, right, you would think that this would have led to a a bit of a shift away from the US.
[03:57] You would have thought that people would have taken down their massive risk concentration in the US because it's changing so quickly.
[04:03] You need a more of a risk premium on those assets and you're getting a more
[04:06] like the US from an exceptional place to invest to a more normal place to invest.
[04:10] But it hasn't quite happened, right?
[04:11] Exactly. Because it's not the only thing going on.
[04:13] Okay. Though if you look underneath the system right in one way it did in the sense that if you look at US equities for the first time since American exceptionalism really emerged post the global financial crisis the last 15 years you've had the worst performance of US equities relative to the rest of the world equities in common currency terms that you've had over that 15-ear period.
[04:32] So under the surface you're seeing it but the US is of course buffeted by the other big thing we talked about.
[04:38] Yeah.
[04:39] AI is sucking up incredible amount of capital and that's about to enter a new phase.
[04:42] We'll get to that. But um but you've got this like yeah modern mercantalism people are questioning the wisdom and the the the rule of law in the United States.
[04:51] You see that across institutional investors etc. But you have an offset in that where else are you going to invest in the technological revolution that's going on.
[04:59] You've seen that massive offset. And so what looks like calm on the surface, right?
[05:03] If you said S&P 500's having perfectly good
[05:07] year, very similar to every year.
[05:10] And if you look at bonds too, everything looks fine on the surface.
[05:13] If you look underneath that, split up the equity market.
[05:15] Well, where's that equity coming from?
[05:16] How is the US equity market X the AI names doing relative to investments in the rest of the world?
[05:23] If you look at gold as another example where you're seeing the geopolitics play out.
[05:27] Um, those places you see it.
[05:30] And so I think you can make a mistake of thinking it's not happening by looking just at the calm of the of the averages versus what's happening beneath those averages.
[05:39] So you've been talking about AI for a long period of time and now it's happening and and so how is it panning out compared to how you thought it would could pan out?
[05:47] Well on the big trajectory pretty much in line.
[05:49] So for me if I go back I um I came to Bridgewwater 30 years ago and one of the things that attracted me is Bridgewwater's now just celebrated 50th birthday.
[05:58] So Bridgewwater had been around for 20 years at that point.
[06:01] Well you you you came straight from Dartmouth in straight from Dartmouth in '96.
[06:04] In '96 straight from undergrad and I was attracted to Bridgewwater.
[06:06] It
[06:08] was a tiny place, 40 people, tiny place at the time.
[06:10] But I was attracted to the basic concepts, which was a that you would have the discipline to take all of the things you believe, write them down, and stress test whether they're actually true.
[06:21] What a important thing that fit really with my personality in many ways.
[06:25] But um but I said, "This really makes sense to me."
[06:29] And on what questions?
[06:31] On the questions of how does the macroeconomy work?
[06:32] How do markets set prices?
[06:35] take all of your human intuition, translate that into um algorithms.
[06:38] And at that phase, Bridgewater and I believe it's probably the most profitable expert system that was ever created.
[06:44] We built an expert system for how you do that.
[06:45] Everything was human intuition.
[06:47] I worked on how do you take hundreds of people, really smart people, how do you get them to take their intuitions, put them into algorithms, do that?
[06:54] that my expertise was generating those intuitions and building a community of people who deposited their intuitions into our compound understanding massively.
[07:03] Then about in 2012 I said okay how are we
[07:09] going to when will machines write the rules not just support the rules which we've been doing have great technology support human intuition but when will their intuition actually be better than human intuition?
[07:22] right and that I started on that journey in 2012 and they were nothing like that.
[07:26] there's big data great pattern matching but there was nothing like intuition but I was looking for the pieces that I knew we need.
[07:33] we needed a reasoning to generate intuition and we needed certain pieces.
[07:38] You needed um language models to come along.
[07:41] You needed diagnosibility to um come along.
[07:42] You needed to figure out how to deal with small data problems where you don't have a lot of data.
[07:51] And I went through the world and that that journey brought me to Open AI.
[07:55] I was in the first round of investors in OpenAI because they were doing unique things.
[07:59] When was this?
[08:00] That was I think it was about 2016.
[08:03] Um and so they just Elon Musk had just stepped aside and they needed funding and they were deciding to um
[08:10] you invested in it on behalf of Bridgewwater or
[08:13] No, personally we had different views at Bridgewwater at the time on how to how to handle AI but so I invested in that personally and I um because I was interested I wanted to see how can we build these pieces together.
[08:24] I was also very interested in AI safety issues and so anyway I was there for that got to know a lot of the scientists and the thinking which brought me to anthropic later on when some of the best scientists I knew started anthropic and was literally the first check there but all that was a journey
[08:39] to find how can we find these pieces we need to create reasoning that could compete with humans and most importantly along that journey I met Jazz Econ who's our chief scientist at Bridgewater and now we've we've built an artificial investor and continue to make it better
[08:53] when did hire him
[08:54] 2018. So he was a professor at Berkeley working on some of these small data problems.
[09:01] He was working with private companies as well but but he was it wasn't in our field but he was what he was doing was exactly what I was looking for and um he's just incredible genius and so brought him into
[09:12] Bridgewwater and we started working together and still the technology wasn't ready for what we wanted to do.
[09:17] 2022 I said the technology is almost ready that the pieces we needed we could start building an artificial investor.
[09:23] started that March in 2022.
[09:25] That was chat basically the same time.
[09:27] Right.
[09:27] Right.
[09:27] Right.
[09:27] Right.
[09:28] On then all the tools were coming together at that time that I thought we could get all our pieces and we could actually build an artificial investor that could compete with me and you on how do you how do you think about the world?
[09:38] How do you actually generate intuitions about the world?
[09:40] How do you write rules to say okay this is how I would apply that intuition.
[09:45] And we then by 2024 thought we had a smart enough investor to to generate alpha in the world.
[09:52] and um and have been doing that.
[09:53] We're with some of our incredible clients.
[09:56] Um and so right now my role is running um Pure Alpha as managing CIO of Pure Alpha where all the human intuition sits but also running this separate entity that's designed around a machine learning agent.
[10:08] But but if but we'll come back to this a bit but just um so
[10:13] from there um looking at what's been happening to the market the AI stocks uh you know the chip producers uh and so on how is how is the market reaction to this panning out compared to what you thought?
[10:28] Yeah.
[10:28] And I'd say um if that's where it's panning out.
[10:31] I think people were way underestimating.
[10:34] I the phase that we were in before I used to say the bubble's ahead of us, not behind us.
[10:39] I get the question starting with Hatchp moment more or less.
[10:42] Is this a bubble?
[10:44] And my thought was I think there will be a bubble but we're nowhere near the bubble phase.
[10:48] we were in the phase where people have no idea what's hitting them like meaning how important this is and how much is going to get invested because this is not a typical cycle when you when you have people like Elon Musk and Sam Alman and and Google and so on whose whose businesses are threatened and believe that the power to control Earth and the universe is only a couple years away.
[11:13] They're not motivated by
[11:16] this normal profit incentives of the typical cycle.
[11:17] It's not a capex cycle that's the same as other capex cycles.
[11:21] This money is going to get spent, right?
[11:23] I'm thinking that in 2022, 2023, they're going to build this out.
[11:27] Maybe someday they'll be proven this is a deadend path, but we're so far from any possibility of that happening that the the data centers were going to get built out, the semiconductors were going to get used.
[11:37] That has played out.
[11:40] Um, and so where so where are we now?
[11:42] So now we've just entered what I I've said is a more dangerous phase.
[11:44] I still don't think we're in a bubble, but we are in a more dangerous phase for the following reasons.
[11:49] That a we're in the resource grab phase now to do AI.
[11:52] There isn't enough resources to go around.
[11:55] So, everybody's trying to grab their resources.
[11:58] Microsoft's got all the land where you can get power on the grid.
[12:01] They've they've done a great job of getting that land.
[12:05] Other people need to figure out places off the grid to get power.
[12:11] the Nvidia's supply has been like bought up for years to come in the future and
[12:17] So on and so forth.
[12:18] So you're in this resource grab phase.
[12:22] Compound growth is easy to do when it's in cyber like in the tech world you can do that.
[12:26] But now it's in the physical world you to continue growing 50 60% a year.
[12:31] The um tell me about the land grab a bit more before we we we also going to cover capital here as a third point.
[12:36] But tell me about the land grab.
[12:38] So what what part what are the important elements of land grab?
[12:41] Basically to control this you need power, you need chips and you need scientists.
[12:46] Yeah.
[12:48] Look at what's going on with all of those.
[12:50] Everywhere you could get p power.
[12:51] I don't know last time you were in Abu Dhabi, but it's just incredible.
[12:54] If you go to Abu Dhabi and you sit in a cab, like the cab driver and his brother are building a data center.
[12:59] Like the anywhere you can build power, they are building data centers.
[13:03] Maybe Europe's the exception, but everywhere else you get power.
[13:05] You put chips in a in there and you're building a data center.
[13:10] And the rate of that is huge.
[13:12] So you have to get where can you get the cheap power?
[13:14] How do you how who gets it?
[13:17] Where do you get the chips?
[13:18] And so the power so you don't think so the power you don't think it's uh kind of overhyped?
[13:26] Do you think that's that whole?
[13:27] I don't think it's overhyped.
[13:29] I mean I understand like the one of the real ch the whole thing has this risk that the depreciation schedule is probably going to be quite fast and you hope it has to be in a sense that if you take anthropics mission or open AI and Google like the idea of building AI that builds faster AI one of the things they have to do is figure out how to make the chips more efficient make the energy more efficient and they're trying to use AI to do those things and given what Google achieved on protein folding etc I think there some of the scientific advantages advances that will depreciate the current assets will come from those assets themselves from the AI will generate better ways to do this.
[14:05] Yeah.
[14:06] So that's so that's the power and then the chips.
[14:08] But but just to finish that but in the in the meantime everybody needs power so desperately and there's such a shortage in the west that it's a huge problem.
[14:15] China is a different picture but but that is I don't think it's um
[14:20] overrated right now for the reason that I just described.
[14:24] People really believe many people believe including myself that incredible power is at the other side of this meaning like power to control outcomes on earth the fountain of youth etc these types of thoughts then man people are going to use every resource available to get there and so I don't think it's overhyped is first point.
[14:41] the second on chips you were going to ask.
[14:45] same thing.
[14:45] tell me about the land grab on and also the the kind of the circularity and the vendor financing and so on that we're seeing here.
[14:50] yeah good so the first part I'd say the the land grab on chips is look, everybody needs if you want to get to the next level in the next the next two levels of models, right?
[15:00] This is what exponential growth is really hard to keep up with, right?
[15:04] That how do you grow it was one thing to grow 60% a year when you're small.
[15:08] Now to grow compute 60% a year.
[15:11] If you extrapolate that out for three models, you've got data centers everywhere on Earth.
[15:14] That's obviously not going to happen.
[15:15] Things have to change to prevent that from happening.
[15:17] But but um but that exponential growth is incredibly hard.
[15:20] and the chips are so scarce that you need that Nvidia is now in this position and this goes to vendor financing.
[15:29] People look at this vendor financing and think it's because of normal bubble dynamics.
[15:33] This is how they're going to get their revenue.
[15:34] Not at all.
[15:36] Nvidia can get as much revenue as it wants.
[15:37] They have no problem selling the chips, but they don't want to set up a system where they lose their competitive edge.
[15:43] Their problem is Google is a true competitor.
[15:45] Google's trying to go the whole stack all the way down to the chips.
[15:48] They don't want Google to win or they don't want Google to be the the final win.
[15:52] So, they're trying to control the ecosystem.
[15:53] They're like Standard Oil in the Gilded Age trying to create monopolistic control on things.
[15:59] So, create their own e ecosystem and sell their chips to people who need their chips who will not create an alternative to their chips.
[16:07] So what you're seeing is a design of the ecosystem where there's a Google ecosystem there's somewhat of an Amazon ecosystem and then there's an Nvidia controlled ecosystem that has let's say open AI
[16:21] Certain models on the top and goes all the way through all the steps that Google has and that's what you're seeing when you're seeing all these deals is everybody's got to lock up who do I partner with where am I going to get my chips and power and if I don't do it I'm going to die.
[16:36] Yeah.
[16:36] What about the land grab of scientists?
[16:39] Yeah. Well, you're seeing that's like the toughest one, right?
[16:42] Because there are not that many cutting edge scientists.
[16:45] How many are there?
[16:45] How many how many really good ones?
[16:48] I mean, it's probably not a great question for me, but but I think less than a thousand.
[16:51] Um, and so that's really constrained.
[16:55] And if you're if you're meta and you don't feel like you have the scientists, but you have the chips and the power, you get the scientists, right?
[17:03] You're seeing this actually really
[17:05] and you can't buy them at a very high price.
[17:07] You buy them, right?
[17:07] But then you're buying the ones who are viable.
[17:10] That's always culturally as you know like that's a cultural issue and um and of course then they're also viable again.
[17:15] Meaning like this is one of the problems if you take how fast the if you're like
[17:21] things that are going badly in the ecosystem is too many of the scientists.
[17:26] too many people naturally are drawn to like where do I jump to get the next paycheck and that when you're trying soccer players and the kind of the transfer season.
[17:35] exactly so that's really bad now you're seeing and this is where I would think you see differences in the labs you know there are people that went to anthropic because the mission there many of them are there I'd say like their ability to maintain their talent is unique because the people that went there are there for a mission that's a little bit different.
[17:51] Now, everybody has and I don't blame anybody. Everybody has greed in in them to some degree, but um but you're seeing differences in the different places of who can hold on to scientists.
[18:00] and so so in in a different entropy. Who else can hold on?
[18:03] You think I think Google is doing a good job of holding on to scientists and have been fighting this culture for a long time and have a lot of raw material. That's incredible.
[18:17] Um otherwise it's just very very hard. You're seeing some startups like I'm I'm very impressed with thinking machines. I think they have great people there but they just got
[18:23] there and who knows.
[18:25] Um and so I think when you look around this is sadly slowing down scientific progress in a big way.
[18:32] This this sort of everybody jumping around and not it's going to take some of the big breakthroughs are going to take a team working together for a extended period of time to get through.
[18:41] I think they're very possible the breakthroughs that are missing in AI today but but with everybody jumping around that's certainly slowing it down.
[18:49] Okay.
[18:49] So just to remind uh everybody of the structure here we talked about the economy mercalism we talked about tech where we touched on power chips and scientists and then the last of the three points is capital and all the capital that's gone to the US.
[19:01] So what are your thinking here?
[19:03] Yeah, it's super interesting because these two other pressures cut against each other to some degree, but I think the economy has changed.
[19:11] Just back on mercalism for a second.
[19:13] The US is no longer the US that it's been post World War II.
[19:18] The idea of global institutions and even what the US currently sees in its interest has
[19:24] changed radically.
[19:27] the US was always pursuing its interest but but what it saw is in its interest of international cooperation and things like that um have have changed in a significant way.
[19:36] So I think you are seeing that start to play out.
[19:38] We're still in the very early phase of what the next steps are.
[19:43] What is the retaliation from the rest of the world against what the US has happened?
[19:48] One of the things that surprised me this year is that the US came out as a bully in a sense, like came out and said, "Okay, we're going to raise your tariffs."
[19:55] And you're not going to do anything about it.
[19:57] And everybody except for China went along with that.
[20:01] I was surprised.
[20:03] Honestly, I remember conversations with members of the um administration saying like, "How are you going to do this? like they're going to hit back and we are um desperately in need of the rest of the world not because of trade but because of capital."
[20:17] So why are they not hitting back?
[20:20] Well, I think people first thing is Trump has been successful in picking people off, right?
[20:24] If you try to punch
[20:26] back, he raises the tariffs even more
[20:28] and there is like a a problem of all the
[20:30] countries being smaller and so they
[20:32] can't punch in the same size and and
[20:33] they take the risk of being picked off.
[20:35] So that's one reason. I think beneath
[20:37] the surface though on things that are
[20:39] less obvious like how do you invest etc.
[20:42] You are seeing countries, you're seeing
[20:44] certainly the lawmakers in Canada
[20:45] saying, "Okay, we got to invest in
[20:46] Canada." You're seeing home bias
[20:48] everywhere as a reaction to this. I
[20:50] think you will see more of that. But the
[20:52] punching back is difficult with Trump in
[20:55] the presidency because he's going to
[20:56] punch again and you have to be up for
[20:58] that fight. But what you see in the
[21:00] politics which will lead to this is the
[21:03] US goodwill among the rest of the
[21:06] world's voters has collapsed. We have
[21:09] saw from 20 um 24 to 2025 the biggest
[21:13] collapse in support for the United
[21:15] States in the rest of the world that
[21:16] we've ever seen.
[21:17] >> And what are the what are the
[21:18] implications of that?
[21:20] >> What you're seeing is populist
[21:23] mercalist um candidates gaining ground
[21:27] everywhere.
[21:27] >> So I think you're going to get the first
[21:29] wave is a shift to the populist right
[21:32] everywhere. And you're seeing that if
[21:34] it's the AFD in Germany, if you I mean
[21:36] shocking, but if the UK had election
[21:38] today, I think I might be out of date by
[21:40] a couple weeks here, but reform party
[21:42] would win. I mean, take that compared to
[21:44] 5 years ago, that's like an insane
[21:45] thought. And same thing, France, etc.
[21:48] So, a shift towards populist to people
[21:50] that are going to take care of their own
[21:51] countries in in their in that way of
[21:55] populist right kind of thing. And of
[21:57] course, what the populist right then
[21:58] sets off is more strength than the
[22:00] populist left. New York City mayoral
[22:03] election or whatever, but you see the
[22:05] shift towards populist right, populist
[22:07] left, and the continued just like you
[22:09] had the hollowing out of the middle
[22:10] class. You've had the total hollowing
[22:12] out of the middle of the economic of the
[22:14] political spectrum.
[22:15] >> And that's the world we're going into is
[22:17] this probably shift to populist right.
[22:19] If those policies don't work, shift to
[22:21] populist left. And that's the dynamic.
[22:24] and the populist right, the hitback,
[22:26] even though they're aligned to Trump, is
[22:29] just more domestic
[22:32] um more focus on their domestic issues
[22:36] and less international cooperation.
[22:38] >> So now so now we have so now we have
[22:40] painted the world, right?
[22:42] >> Yeah.
[22:42] >> What are so when you then look at what
[22:44] it's going to lead to in terms of
[22:47] economic indicators and developments,
[22:49] let's kick off for instance with in what
[22:51] are the implications for inflation for
[22:52] instance? So if you take the the things
[22:55] that just happened from a merculus
[22:57] perspective, they're clearly
[22:58] inflationary. Two big things, right? You
[23:00] get tariffs, which creates rather than
[23:02] being able to secure your pipeline in
[23:04] the cheapest way, you you have to secure
[23:06] a pipeline of goods that's more
[23:10] um
[23:12] resilient and more domestic than cheap,
[23:14] right? So a that's inflationary. the
[23:16] product one of the huge benefits over
[23:18] the last 30 years was the productivity
[23:20] change by taking advantage of the the
[23:22] cheapest and most efficient places to
[23:24] build things that's totally wrecked in
[23:26] the world right even you thought okay
[23:28] we'll get out of China and go to India
[23:29] well India's tariffs are now higher than
[23:31] China like there's if you're an
[23:32] international company this is incredibly
[23:34] hard to build a build that so that's
[23:37] inflationary secondly the big
[23:39] inflationary push from this is
[23:41] everywhere you're seeing a fiscal
[23:42] reaction right that Germany most extreme
[23:45] but But a great example a you got to
[23:48] build your own military now. You have to
[23:50] get off the dependence on the US. You
[23:51] got to build your own military and that
[23:54] there's nothing more inflationary really
[23:56] than military spending because it
[23:59] creates demand for labor and demand for
[24:00] goods with no supply into the real
[24:02] economy. So you've got military spending
[24:05] surging at incred in nonwart time
[24:08] incredibly fast rates. And you've got um
[24:12] so military spending plus you've got the
[24:15] need to rebuild the infrastructure so
[24:16] that in Canada's case but just in
[24:18] Europe's case too can't ship to the US
[24:20] as much anymore. How do I get my goods
[24:22] to other places and those pieces are
[24:27] inflationary.
[24:27] >> So the and the magnitude of this. So how
[24:30] much worse do you think inflation will
[24:32] be than the general consensus? Let me
[24:35] say the flip side of that is on the
[24:38] other hand growth is dominated by this
[24:40] AI
[24:42] >> investment AI investment while starting
[24:44] to take up power capacity and whatever
[24:46] is very low labor intensity relative to
[24:49] the unit of GDP
[24:51] >> so you have a disinflationary weak labor
[24:54] market because so much of the growth if
[24:56] you take US growth this year right
[24:58] you're going to have a normal growth
[24:58] year this is the same thing in the
[25:00] averages 2% 2 and a half% growth but
[25:03] without AI it would be 1% % right so and
[25:06] that 1% of growth that's coming from AI
[25:09] investment it's um it's very non- labor
[25:13] intensive
[25:14] >> so do you think
[25:15] >> so that's a disinflationary effect
[25:16] >> and net net
[25:17] >> net net I think we are inflation
[25:21] particularly for the next couple years
[25:23] is going to be a significant constraint
[25:25] on policy makers that you're running
[25:26] around like break even inflation is
[25:29] around 2 and a2 we think the fed is very
[25:32] comfortable above two and a half and
[25:34] probably we're in the three range as a
[25:36] base right now
[25:38] >> with the risk that these things push
[25:40] higher
[25:40] >> m
[25:41] >> um but but the AI story right and this
[25:45] depends all these investments have a Jc
[25:46] curve in the first phase you don't get
[25:48] you spend a lot of resources you don't
[25:50] get a lot of output other than the
[25:52] investment itself but then you get the
[25:55] um the part of AI that will eventually
[25:58] be highly disinflationary although I
[26:00] think that's further out and in meantime
[26:02] central banks
[26:04] It depends a lot on what central banks
[26:05] choose to do. You know, the US part of
[26:07] populism certainly in the US is getting
[26:09] control of the central bank, taking it
[26:11] into executive power and and getting
[26:13] them to lower real rates, which lowering
[26:15] real rates into a boom like this is
[26:18] particularly inflationary if it ends up
[26:19] going down that path.
[26:20] >> Economic growth.
[26:21] >> Yeah. So growth um is going to be this
[26:25] two-track economy, right? I think you're
[26:27] going to have good growth in the United
[26:28] States and generally decent growth
[26:31] because you've got this fiscal thing in
[26:32] Europe and you've got this um in the US
[26:35] you've got AI where growth next year is
[26:38] going to be further pushed by this AI
[26:41] push and um and so growth in the US we
[26:44] think will be a little bit above um you
[26:46] know above potential two two and a like
[26:48] two and a half kind of percent growth
[26:50] and um and but a lot of that one and a
[26:52] half percent or so is going to be AI so
[26:54] you're going to have a weak
[26:56] probably weak labor market, weak economy
[26:59] in many places while you have this huge
[27:02] boom in a very concentrated sector. So
[27:04] >> what about budget deficits
[27:06] and the government debt situation? Yeah,
[27:08] that's another aspect of what's changing
[27:10] in the world is you've got you've put
[27:12] yourself where
[27:15] let me just say one thing at the
[27:16] beginning of the you know at the
[27:19] beginning of this year a little before
[27:20] the beginning of this year I wrote a
[27:21] piece called we're all merciless now and
[27:24] I'm thinking about the piece for next
[27:25] year which is a little overstated I'm
[27:27] not quite there but something like we're
[27:28] all Brazil now and what I what I mean by
[27:30] that is Brazil for a long time has been
[27:32] constrained despite having their debt
[27:34] mostly in domestic currency now it's not
[27:35] like Brazil in the '90s where the
[27:37] problem was having dollar denominated
[27:39] debt that they couldn't fund, but
[27:40] despite having most of their debt in
[27:41] domestic currency, they're very limited
[27:43] in what they can do. You're starting to
[27:45] see that the UK is in that position. Um,
[27:48] and that changes the dynamic. When you
[27:50] have fiscal policy and you have a lot of
[27:52] room, like Germany does as an example,
[27:54] and you announce a big fiscal policy,
[27:56] what happens? Your currency goes up,
[27:58] your stock market goes up. Um, when you
[28:02] have
[28:04] and interest rates go up a little, when
[28:06] you're the UK, if the UK said, "I'm
[28:08] going to do what Germany's going to do.
[28:09] I'm going to write a big fiscal check."
[28:10] Currency would go down. The equity
[28:12] market would probably go down, although
[28:14] that's closer call, and interest rates
[28:15] would rise a lot. And so, what you see
[28:18] now is more developed world countries
[28:20] constrained by hitting the limits of
[28:22] fiscal policy. Um, and the limits of
[28:25] fiscal policy are complicated. It's not
[28:27] like a simple number like you could take
[28:29] Japan they could have 300% of GDP in
[28:31] terms of their budget their accumulated
[28:33] debt for the government and in Brazil
[28:36] you might hit that limit at 60 80%. The
[28:38] difference is how much domestic savings
[28:40] you are, how much productivity you have,
[28:41] how much willingness there is to save in
[28:43] your currency. But all countries have
[28:45] limits and we are testing those limits
[28:48] in certain countries. Which means you're
[28:50] moving from a world where policy makers
[28:52] were unconstrained. when something went
[28:54] bad they could lower interest rates and
[28:57] print and spend money
[28:59] >> to a world where more and more countries
[29:01] including by our measures getting close
[29:03] in the US are constrained where actually
[29:06] if you when you get constrained spending
[29:07] money becomes counterproductive rather
[29:09] than productive that's the case in the
[29:11] UK and Brazil today um and the US is
[29:14] drifting towards that line as well I
[29:16] think we're in this death march on
[29:19] institutions you've already seen that
[29:20] the World Trade Organization what does
[29:22] it even do now I mean every
[29:24] international institutions are gone and
[29:26] in the US the domestic institutions are
[29:28] fading fairly quickly. Although I will
[29:30] say it's been interesting and I do think
[29:33] we're going to face this next year that
[29:35] they're going to put in a chairman who's
[29:37] going to for the first time in a very
[29:39] very long time. I don't know if this
[29:40] ever happened in the US where the
[29:41] chairman's going to get outvoted
[29:43] >> a lot. That'll be an interesting period
[29:45] while the while the government while the
[29:49] Trump administration tries to get more
[29:50] governors in who will vote in line with
[29:54] the new chairman. But um but for a while
[29:56] you're going to have this very divided
[29:57] Fed and where the chairman may be in the
[29:59] minority. So now we have set the scene
[30:03] for what the world looks like and what
[30:06] it's going to look like. So here here I
[30:08] give you all my money. Greg,
[30:12] >> could you please take look after our
[30:14] money?
[30:15] >> Yeah.
[30:15] >> What do you do with it?
[30:17] >> Well, just so you so let's say it's a
[30:20] hundred. Just where do you put the money
[30:21] and how do you invest it?
[30:23] >> Yeah. And the main thing that I'd say is
[30:24] different in our philosophy than most is
[30:27] I I would say the way I would take your
[30:30] $100 is is to say how do we survive in
[30:33] the wide range of possible worlds
[30:35] >> rather than try to pick the best thing.
[30:37] Yeah. And um and that is very different,
[30:40] right? If you take basically the last 15
[30:41] years.
[30:41] >> I'm sorry. Is that the way you always
[30:43] think? Let's survive rather than let's
[30:44] get rich.
[30:45] >> Um
[30:46] >> or do you just think that's the way to
[30:47] get rich in the long term?
[30:48] >> I think it's the way to get rich in the
[30:49] long term. Um and that because basically
[30:54] there's a lot of ways to make money in
[30:55] the world. The main thing you want to do
[30:56] is avoid really bad outcomes and that's
[30:59] how compounding wealth works. If you
[31:01] keep earning more and you're in the
[31:03] game, you'll you'll be able to compound
[31:05] wealth in an incredible way. So for me
[31:08] right now, if I look at the world, I
[31:09] think it's very dangerous. And this
[31:10] depends like because you have this home
[31:11] bias move and a lot of the geopolitics
[31:13] like literally where you're located.
[31:14] Being located in Norway is an advantage,
[31:16] a disadvantage in certain ways. Being
[31:18] located in the US advantage and
[31:19] disadvantage in certain ways in terms of
[31:20] what you should do. But to me, what
[31:23] where I think most people the last 15
[31:25] years are a trap. Most people have moved
[31:28] away from diversification because it
[31:31] hasn't worked for 15 years. All you
[31:32] needed was the US US equities and more
[31:35] liquid the better. Like that has worked
[31:37] incredibly well. I think it's mostly a
[31:40] trap. So the ways I would do this is
[31:42] look get a much more globally
[31:44] diversified portfolio and
[31:47] than most people have and because you
[31:49] don't know what the winners and losers
[31:50] are going to be and the US the change of
[31:52] the US are are quite radical.
[31:54] >> And what about Europe? Yeah. Look, I
[31:56] think Europe is investable like even
[32:00] though um as we agree that
[32:02] >> hope so we got 25% of the money there.
[32:04] >> Yeah. Yeah. And that the fiscal changes
[32:06] look there's two things that have
[32:07] changed a lot in Europe. The fiscal move
[32:10] the move like look there you can't build
[32:13] this economy on exporting to the US and
[32:15] you can't build this economy on
[32:16] exporting to China even. Um that you've
[32:19] got to create a more domestic economy a
[32:21] shift towards fiscal policy. fiscal
[32:23] policy works in the profits of domestic
[32:26] companies. Um, and you need to move to a
[32:29] more independent economy in Europe and I
[32:32] think they'll do that and companies in
[32:34] Europe have seen what's happened to the
[32:36] US and in some ways unfortunately not in
[32:39] the technological domain which is a
[32:41] massive problem but on treating
[32:43] shareholders better on on returning
[32:45] money to shareholders buybacks etc. that
[32:49] Europe has actually moved to a more
[32:52] shareholder friendly place in many ways
[32:54] than it was for a while. And the
[32:56] companies that are bigger, if you look
[32:58] at companies that directly compete
[33:00] between Europe and the US, they're
[33:01] generally priced cheap. In a lot of
[33:02] cases, the European companies are
[33:04] actually have better earnings and and
[33:06] better um management than the comparable
[33:08] companies in the US.
[33:10] >> When does China come in?
[33:11] >> Again, this depends a little bit where
[33:12] you're located because the biggest risk
[33:13] of being invested in China is your
[33:15] government saying you can't be invested
[33:16] in China. than the risk of being in
[33:18] China. Look, I think China is really
[33:20] important to global diversification.
[33:21] They have taken a big step. You you were
[33:23] in this big deleveraging in China. The
[33:26] deleveraging was playing a certain
[33:27] course. You also have obviously the
[33:29] challenge that the the party and she in
[33:32] particular
[33:34] >> needs the party to be more powerful than
[33:35] the the companies. And so you went
[33:37] through that phase really badly. But
[33:39] what what they have seen is if they want
[33:41] to compete with the US on AI and this is
[33:43] worth taking a minute on. Um which they
[33:45] now know they need to. It's existential.
[33:47] They determined this about a year ago.
[33:49] It's existential to compete on AI. They
[33:52] need the private sector to do it. They
[33:54] watch what happens in the US. They see
[33:55] the US has so much more funding to AI
[33:57] because the markets work well. The
[33:59] equity market works well. It gets a
[34:00] tremendous amount of funding beyond what
[34:02] any government could do. And um and
[34:05] they've shifted. You've seen the shift
[34:06] bringing Alibaba back in. bringing
[34:09] saying okay we've got to go do this and
[34:12] um and they need the equity market to
[34:14] function to do it they know it and
[34:16] you've seen the shift there and they
[34:18] have incredible
[34:20] unlike Europe there's incredible
[34:23] technological ingenuity there that while
[34:25] they're behind in cutting edge AI
[34:27] they're probably on the cutting edge of
[34:30] applying AI in businesses in China
[34:33] >> and bonds bonds and equities how you
[34:35] split it
[34:36] >> so bonds I I think are again depends a
[34:40] little bit on your starting point. I
[34:41] think most people got out of bonds when
[34:44] many of investors got out of bonds when
[34:45] interest rates were zero and interest
[34:47] rates have risen and real interest rates
[34:48] have risen such that there is a place
[34:50] for bonds in portfolios. But but I would
[34:53] say you really bonds are not what they
[34:56] were for the last 30 years that these
[34:58] fiscal limits are really important. All
[34:59] of a sudden the correlation between
[35:01] bonds and equities and currency will
[35:04] shift when you hit limits. In the UK,
[35:06] you have a different a very different
[35:07] diversifying instrument than you had
[35:09] before. And B, when fiscal policy is
[35:11] such a powerful lever, the big thing
[35:14] bonds diversify against our major
[35:16] deflations and disinflations. But if you
[35:18] believe there if there's a disinflation,
[35:20] they're going to just push on fiscal.
[35:22] They don't they don't hedge as well. So
[35:24] I think bonds are risky into this and
[35:28] particularly the massive supply that's
[35:30] coming that the fiscal supply massive
[35:33] the need now for the next wave of AI AI
[35:37] ecosystem used to be a net capital
[35:39] provider of the world the Microsofts the
[35:40] Googles they were so profitable they're
[35:42] they were buying back assets while
[35:45] investing a lot now you're about out of
[35:48] that they're basically spending their
[35:49] cash flow when you look across that
[35:51] ecosystem they need a lot of money
[35:52] you're seeing that in private credit and
[35:54] that's going to be sucking in money. So,
[35:56] you're going to have the shortage of
[35:57] capital both between fiscal and AI
[36:01] that's going to create, I think,
[36:02] problems, potential problems for real
[36:04] rates. Although, what's fighting against
[36:06] that is central banks in the US that
[36:07] want to drive down real rates. Um but uh
[36:11] but the the picture I would say is that
[36:14] um that you're going to have this big
[36:15] competition for capital between fiscal
[36:17] and AI. Moving from a world where you
[36:20] had an excess supply of capital because
[36:21] you had huge savings in Europe and in in
[36:24] Asia to a world where Europe, Asia, and
[36:26] the big tech companies were big savers
[36:28] to where Europe, Asia, and the big tech
[36:31] companies are becoming spenders.
[36:32] >> Real estate.
[36:33] >> What do you do with real estate?
[36:35] >> I don't know enough. Um
[36:37] >> crypto. What do you do with crypto?
[36:38] Look, I think it's most there my view on
[36:41] this. I I started studying crypto in 20
[36:43] around the same time I started studying
[36:44] AI. So 2012, 2013, 2014. My my thought
[36:47] then which hasn't changed much is this
[36:50] is not very helpful technology. This is
[36:52] a complicating thing and it's
[36:54] >> do you understand it?
[36:55] >> I think so. And my basic view of the use
[36:58] cases back then were mostly speculation
[37:01] and and and this the thing and the use
[37:04] cases have largely worked out that way.
[37:06] Um, and I thought at the time like AI is
[37:10] a much better bet on how the future will
[37:11] go than than crypto is with like I think
[37:15] there's obviously a place for Bitcoin in
[37:17] the sense that you um you if in this
[37:21] world if you don't trust governments and
[37:23] such having a way to move money around
[37:25] the world, Bitcoin is obviously the best
[37:27] choice. It is a replacement in some ways
[37:29] to gold although in many ways it's not.
[37:31] We could get into um and so I think that
[37:34] has some merit to it. You have the two
[37:35] problems of you have a very intense
[37:38] bubble corruption
[37:41] thing going on in that area and you have
[37:44] some
[37:45] >> So what do you mean in that area? What
[37:46] do you mean by that?
[37:47] >> Cryptocurrency was like a it was a
[37:50] magnet for the most corrupt people in
[37:53] finance. Um because there's low reg I
[37:56] mean just for the reasons low
[37:57] regulation.
[37:57] >> Is it is it still?
[37:58] >> Yeah. Um I think so. I mean bubbles are
[38:02] always but I I would say um the crypto
[38:05] area is just focused for that edge of
[38:10] corruption. If you look at the companies
[38:12] that just buy the Bitcoin and put in
[38:14] their treasury or whatever like the
[38:16] things going on there I don't think
[38:18] they're supported by the use cases and I
[38:21] think then the idea like it is possible
[38:23] because there is some good to the idea
[38:25] of taking a database and then
[38:28] distributing it so that no one person
[38:29] has that power. No one entity has that
[38:31] power. You get the need for that in this
[38:33] world, particularly in the world, but
[38:35] that's what it does. That's a more
[38:37] inefficient database. It may be more, it
[38:40] may be necessary because you can't trust
[38:41] anyone. But it's basically built on a
[38:44] technology that deals with a trust issue
[38:46] that makes it more inefficient. Living
[38:48] in a world with no trust is a very
[38:50] inefficient world. And crypto represents
[38:52] that.
[38:52] >> Bridgewater. Why has Bridgewater been so
[38:55] successful? I think the most important
[38:59] thing that bridge order did well was to
[39:02] say
[39:04] a we're focused two things. How do you
[39:06] deeply understand how the global
[39:08] financial system works and how do you
[39:10] build great portfolios? Those are the
[39:11] two things that we do. And um and then
[39:16] to drive that
[39:18] the idea that you have to do that by
[39:21] compounding understanding. You have to
[39:22] have the discipline to write down what
[39:25] you believe and why you believe it.
[39:26] Share that with others so that they
[39:28] could assess what's wrong about that,
[39:29] what's right about that. Build that out
[39:32] and keep compounding understanding. So
[39:34] to me, basically the focus on those
[39:36] things, getting a culture of people who
[39:38] people who care deeply about how those
[39:40] two things work and then taking
[39:43] everything we ever learned and having
[39:44] the discipline to write it down,
[39:46] translate it into algorithms and keep
[39:49] moving forward that way. Those are the
[39:51] the the magical pieces.
[39:53] >> But is it possible to compound knowledge
[39:55] within an organization like that?
[39:57] >> I think it's very very difficult but
[39:59] yeah I definitely think it is because
[40:01] >> I mean so now let's say now I joined
[40:02] Bridgewater. So how do so then I have
[40:05] access to everything you have ever
[40:06] thought?
[40:07] >> No because of security but but you but
[40:09] you could like meaning like if we if we
[40:12] didn't have security constraints
[40:13] >> we could make that all you could
[40:15] literally see everything. And one of the
[40:16] amazing things that we built, which I
[40:18] think is super important, is not only
[40:19] everything, how it changed, every time
[40:21] we learned something new, what we
[40:23] replaced with what we built, that's all
[40:24] available, right? I can go in there and
[40:26] see everything. And for what it's worth,
[40:27] AI could read over everything we've ever
[40:29] believed about markets, all the rules we
[40:31] had, how we changed them, why we changed
[40:33] them.
[40:34] >> Um, and why is that good to have?
[40:37] because
[40:39] because you can't
[40:41] um you can't carry in your brain all of
[40:45] the things that you've thought about and
[40:46] forgotten not you not even just that
[40:48] you've done now we have hundreds of
[40:50] people that have studied these issues in
[40:52] China that if you start and you start
[40:54] from somebody's understanding that's
[40:56] been what we say is to compound it what
[40:59] I mean by that is it's got to be human
[41:01] readable and computer readable meaning
[41:02] like a human's got to be able to pick
[41:04] this thing up and make sense of it and a
[41:06] computer's got to be able to run it. Um,
[41:08] and that that is extremely extremely
[41:12] powerful to have that everything that
[41:14] we've ever believed.
[41:15] >> Where else in society have do they use a
[41:18] similar type of system?
[41:20] >> Well, I think if you look at I mean this
[41:22] is how society progressed, right? Like
[41:24] writing was such a if you take major um
[41:28] major technological advances, right? the
[41:31] printing press and once you started
[41:32] being able to compound human knowledge
[41:35] through the printing press etc that is
[41:39] like at a society level we're doing that
[41:41] we do that inefficiently we do that in
[41:42] certain ways but um institutions rarely
[41:45] do people come in like how much
[41:47] >> why don't they do it
[41:48] >> because it's so damn hard you know this
[41:50] was one of the great things
[41:51] >> what's the hardest thing doing it
[41:52] >> the discipline
[41:54] >> like you don't need it that day why am I
[41:56] writing down what I'm doing um and it
[41:58] just takes so much goddamn discipline
[42:00] and people are lazy. Um, and so
[42:03] >> how do you force people to do it?
[42:04] >> Yeah, it's super hard. So, the thing
[42:06] that I um did in 2012 versus maybe a
[42:09] little before that and I said it doesn't
[42:11] exist unless it's compounded. You get
[42:13] no, you don't get any bonus. Nothing
[42:15] that you did. I don't care how great
[42:17] your results were. I don't care what
[42:19] happened. If that knowledge is not in
[42:22] what we call our secure garden, but if
[42:23] it hasn't been translated into
[42:24] compounded knowledge, it does not count.
[42:26] Zero. You get nothing. So do people have
[42:28] to write it down or do they tape record
[42:30] it or they transcribe it or what
[42:31] happens?
[42:32] >> It's all written. I mean we have tools
[42:33] to make it better but meaning written
[42:35] and translated into algorithms. So
[42:37] meaning like there is a plain English
[42:38] version of everything we did and then
[42:40] there algorithms that represent that so
[42:42] that when you're done with it you have
[42:43] this idea. I want to go buy this thing.
[42:45] I want to go do that thing that that the
[42:48] reasoning is written and the
[42:50] manifestation of that reasoning in code
[42:53] exists and otherwise you get no credit
[42:55] for it
[42:56] >> and secure. So you call it secure
[42:57] garden.
[42:58] >> Yeah.
[42:59] >> What what So secure garden just tell me
[43:01] what is it? How is it like uh uh is it
[43:04] in the cloud or you have a computer down
[43:06] in the cellar or basement? It's
[43:07] >> in the cloud. Um and so what it is
[43:09] >> how big how big how how many
[43:12] flips or flops or whatever?
[43:14] >> Huge. I mean some stuff we've started to
[43:17] try to cut back. We used to keep all of
[43:19] the videos and everything and so we cut
[43:21] back some of it but mass massive amount.
[43:22] But but what it is but it's curated,
[43:24] right? So you can go in there and say,
[43:26] "Okay, I'm interested in modern
[43:28] mercurism. I'm interested in oil. I'm
[43:30] interested let me get every
[43:31] manifestation of that
[43:33] >> organized for me, right?" And um and so
[43:36] that it's both computer and human
[43:37] readable. So I can go in and see the
[43:38] algorithm that's making our decisions on
[43:40] this process or I could see the charts
[43:42] that that generates to help me think
[43:44] about those problems. And um and that
[43:46] that thinking was like how do we take it
[43:48] because if you were at Bridgewwater
[43:49] before that, right? One person would do
[43:51] something in Excel, another person would
[43:54] do it in C++, another person would do it
[43:55] over here. It was impossible, right?
[43:57] Like to go you have to go like nine
[43:59] different languages and all and all of
[44:01] these different ways to manifest it. By
[44:03] by agreeing that no, we will build the
[44:06] best way to do that across the whole
[44:07] company and everybody will work on that
[44:09] platform, you change that and you
[44:11] compound it across the whole thing in a
[44:13] better way. We're always trying to do
[44:14] that. But manifesting your culture both
[44:17] in technology and in process is super
[44:21] important and that's what we did there.
[44:24] >> And which also sets us up for what it's
[44:26] worth because of that is sets you up for
[44:29] this world that we're now in
[44:30] >> where look it's very hard for humans to
[44:33] go dissect all of that. AI is built to
[44:35] dissect that stuff. Um
[44:37] >> but you set up the AI as a separate kind
[44:40] of venture. Why why did you do that
[44:43] instead of applying it to your secret
[44:45] garden?
[44:46] >> Yeah. So the the reason was a couple of
[44:48] reasons but most importantly is the
[44:50] nature of the technology at this point
[44:52] although this is changing but let's say
[44:54] in 2022 is you needed people who wanted
[44:57] to do it that were 100% in this and it
[45:00] had to be the whole thing. So we
[45:01] designed a factory
[45:03] a idea factory in ya where AI is at the
[45:07] center and the people are trying to help
[45:09] make the AI work. Bridgewater's not like
[45:11] that, right? Bridgewwater like what
[45:13] people want from technology is they want
[45:14] it to help them um work. And so to me, I
[45:19] wanted to you to do do this where you
[45:22] put AI at the center and you you hire
[45:24] the different kind of people and you
[45:27] build a team that's going to say, "Okay,
[45:28] how do I make the AI work and how do I
[45:30] fill in the things AI is bad?" The way
[45:33] humans tend to look at it is like how do
[45:35] I get this technology to help me? So two
[45:38] different factors. Now more and more
[45:40] we're using AI in the other factory
[45:41] because it does actually allow us to to
[45:45] um systemize ideas we couldn't do
[45:47] before. It allows us to accelerate that
[45:50] process. So it's coming more in there.
[45:52] The the second reason is the security
[45:54] challenge. Like look what you couldn't
[45:56] do now then we're much closer now is put
[45:59] this over all of your proprietary stuff.
[46:01] So that I had just had access to the
[46:03] data had a lot of benefits of being in
[46:05] midst of Bridgewater but it didn't have
[46:06] access to all you know 50 years of
[46:09] intellectual property the the now AI
[46:12] being able to but but I was setting it
[46:14] up so that we would design AI that
[46:16] eventually could suck in that
[46:17] intellectual property and be that much
[46:19] better. And the third reason was I don't
[46:22] want a copy of us. We're so flawed.
[46:24] We're so bad at so many things. I want
[46:26] something better than us. And if you
[46:28] start particularly before, but if you
[46:30] start AI and you give it too much stuff
[46:34] that it can cling on to, it gets stuck
[46:36] there. Um, you really want it to be able
[46:39] to generate its own independent ideas.
[46:41] So I wanted ya to be uncorrelated to
[46:44] pure alpha. Um, I wanted it to generate
[46:46] >> when you look at it now, do you think it
[46:47] was the right thing to do?
[46:48] >> Yeah, I I amum I really do. I think it's
[46:51] changing now. I think we are bringing it
[46:53] back together again now. But at that
[46:55] time, and I'd still say for everybody,
[46:57] this is like I love this lesson from
[47:00] Amazon, but when they started trying to
[47:01] build robots for their warehouses,
[47:04] it sucked in the beginning because they
[47:05] tried to have robots do what people were
[47:07] doing. As soon as they were like, "No,
[47:08] no, no. Let's let robots do what they
[47:11] do. I don't need a robot that's human
[47:12] shape. I need the robots just like move
[47:14] the stuff in the warehouse." Once they
[47:16] designed it around the robots,
[47:18] incredibly efficiency gaining. You see
[47:19] this in China all the time. If you try
[47:21] to get robots to do what humans do,
[47:22] you're in a mess. that I think is the
[47:25] same thing in business. One of the
[47:27] reasons that AI hasn't swept through
[47:30] corporate America is people keep trying
[47:31] to replace humans with it rather than
[47:34] design around the technology and reset
[47:37] their processes.
[47:39] >> So how many people are you
[47:40] >> Bridgewater in total? So Bridgewater
[47:42] 1300 we have 50 in IM um
[47:44] >> 1300 and what do these people do
[47:46] >> most like at the center of it is our
[47:48] alpha engine where there's a couple
[47:50] hundred people thinking about looking at
[47:52] everything that we've ever figured out
[47:53] in the world the 50 years of compounded
[47:55] understanding and because that's all
[47:57] systemized they could just focus on what
[47:59] we're missing what is happening in the
[48:00] world today that you're worried that
[48:02] process doesn't have reflected right and
[48:04] that's there's a lot modern mercalism
[48:06] all that like there's a little
[48:07] >> okay so that's a couple hundred people
[48:08] >> that's a couple hundred people we have
[48:10] people than building the technology that
[48:12] support that. Another couple hundred
[48:13] people that build the infrastructure
[48:15] that support that. Then we back to our
[48:18] mission to build portfolios and
[48:19] understand the world. We have a big team
[48:21] of people supporting our partners, our
[48:23] clients all over the world who who we
[48:26] have great relationships with where we
[48:28] share our understanding. They share what
[48:29] they're seeing in the world. So that's
[48:31] another, you know, big chunk. And then
[48:33] you're running a large company. So you
[48:34] have internal
[48:37] um processes that are super important to
[48:38] us of how do you hire the best people?
[48:40] How do you um how do you motivate them?
[48:42] Well, how do you how do you support that
[48:44] whole ecosystem?
[48:45] >> You got like one big fund in at the core
[48:48] here or do you tailor make all the
[48:50] portfolios to your clients?
[48:52] >> Well, I'd say one pure alpha is the
[48:55] center of it. It's like if you were just
[48:56] like how do you say take everything that
[48:58] we've ever learned and apply that that's
[49:00] pure alpha, right? We also have bestow
[49:03] bespoke solutions that take advantage of
[49:06] some of that alpha but also our strength
[49:09] in how to build great portfolios.
[49:11] >> We've thought about that question for a
[49:13] long time. If you say why did we survive
[49:16] >> Bridgewater as a hedge fund in 1991, if
[49:18] you looked at our competitors in 1991,
[49:20] none of them exist anymore, right? That
[49:22] um that the reason we survived was
[49:25] partially this process that I'm saying
[49:27] is good and you make good decisions, but
[49:28] much more important is risk control.
[49:30] much more important is how you take the
[49:32] fact that we're all flawed and survive.
[49:35] Um and
[49:36] >> and are you and are you in charge of
[49:38] Pure Alpha?
[49:39] >> I am the um managing CIO of Pure Alpha.
[49:42] So yes, I'm in charge of
[49:42] >> So if it's a good year, you it's kind of
[49:45] you.
[49:46] >> No. Um I'm running
[49:47] >> because you've had a good year, right?
[49:48] >> We're having a great year. But but it's
[49:50] because
[49:51] >> is that why you're quite happy
[49:56] today? people do think um you know I
[49:58] think like look performance has luck in
[50:00] it but mainly it's the it's the work of
[50:02] so many people I sit in this incredibly
[50:05] lucky seat to have 50 years of compound
[50:08] understanding and a couple hundred
[50:10] people who are trying to make that
[50:11] better
[50:12] >> that's what's doing it I am um trying to
[50:15] make that meritocracy work as well as
[50:17] possible evolve that process as well as
[50:20] possible but those are the things that
[50:21] make this performance this year happen
[50:23] >> do you overrule the machine from time to
[50:26] Yeah. So pure alpha. Yeah. So yes in
[50:28] both cases. Look I am the reason to have
[50:30] all this compound understanding is so
[50:33] that you can reflect your thinking in a
[50:35] way much better than you can but it
[50:38] can't be a straight jacket
[50:40] >> on average. Has it made sense to
[50:42] overrule? So when you overrule on
[50:43] average is it a is it sensible thing to
[50:45] do?
[50:46] >> Barely but yes. Um, and that's why we do
[50:48] it so only because we don't do it very
[50:50] often and we're very disciplined in when
[50:53] to do it and and there's different types
[50:55] of things, right? The one thing that
[50:57] we've done a lot better is what do you
[50:59] do when you have research in the lab,
[51:01] like actually building it out so that
[51:02] it's sustainable or whatever takes some
[51:04] time. You've research in the lab, it's
[51:06] not quite ready yet, but you think it
[51:08] would change your positions, right?
[51:09] We've we've created a process to allow
[51:11] that to happen much more quickly. So
[51:12] when new things are happening, if you're
[51:13] like the worst moment at Bridgewwater,
[51:15] >> worst moment in my career and where I
[51:18] failed the most was
[51:20] >> COVID. We had the best research on CO.
[51:22] We were incredibly on top of it from the
[51:24] very beginning. We have a huge
[51:26] >> um presence in China. We had a very good
[51:28] understanding what was going on and we
[51:31] totally screwed it up because we were
[51:33] trying to systemize. We got caught in
[51:35] our straight jacket that if you couldn't
[51:37] systemize it, okay, we had done work on
[51:39] the 1917 Spanish flu, but the economic
[51:42] impact of the Spanish flu because it was
[51:44] so disperate was so different than what
[51:46] ended up happening and we knew it and we
[51:48] didn't
[51:49] >> we got stuck with our process and we
[51:52] didn't take in what we knew and apply
[51:54] that quick enough. Since then I was like
[51:56] we a bunch of things that we did
[51:58] organizationally to say okay that can't
[52:00] happen again and um and that led to
[52:04] things where our systematic ideas get in
[52:06] much quicker our researchers have a lot
[52:08] more independence to move faster and um
[52:11] and that that led to where quote unquote
[52:14] overruling the system although all I
[52:16] would I would really more describe it as
[52:18] accelerating new research such that you
[52:21] can use it in an intermediate phase
[52:23] before it's totally done when nec
[52:25] necessary is a big part of what has made
[52:27] pure alpha better since co than it was
[52:30] before.
[52:31] >> You play poker.
[52:32] >> I do. I um I love poker and uh it is an
[52:36] interesting
[52:37] game that's somewhat connected. Right.
[52:39] One of the reasons I
[52:40] >> How how good are you?
[52:41] >> Not good enough. I did win a a bracelet
[52:44] in the World Series of Poker. I've I've
[52:46] was I had an edge for a while,
[52:49] particularly when when the machines had
[52:52] learned more about poker than humans had
[52:53] and many humans hadn't caught up. Um,
[52:55] but I knew a lot about what machines had
[52:58] learn about poker and I was good.
[53:00] >> But like in the world, how would you
[53:02] rank yourself? Top
[53:04] >> I don't play enough to be the top of the
[53:06] top anymore because
[53:07] >> But what's the best when you were at
[53:09] your best? You were top what? I think
[53:11] there was a time, I mean this is a bold
[53:13] statement, but there was definitely a
[53:14] time when I think I was in the top
[53:16] couple hundred people at poker. Um, and
[53:19] um,
[53:19] >> how does how does poker help how does it
[53:22] help you
[53:23] >> in your day-to-day investing business?
[53:25] >> Well, I'd say a couple things. So, I the
[53:30] things I like about poker, first up,
[53:33] it's it's very meditational to me
[53:34] because it's when I'm playing poker, I
[53:36] can zero in on poker and everything else
[53:38] can go away. That's a very rare mo
[53:40] usually there's I'm talking to you. I've
[53:41] got 12 different ideas going on in my
[53:43] head and I'm distracted and poker allows
[53:45] like it I need to and when I do it I
[53:47] could just totally focus on it.
[53:50] >> So you think about 12 things in parallel
[53:53] when you talk to me at the same time.
[53:54] >> Yeah. I mean you're interested.
[53:55] >> I kind of thought I was I thought I was
[53:57] kind of engaging here. Gee,
[53:59] >> but um but I but poker for whatever
[54:02] reason I could do that when I'm
[54:05] playing poker. the things like
[54:06] >> I thought men I thought men could only
[54:07] do one thing at a time.
[54:09] >> Probably one thing well at a time. I can
[54:11] do this badly and um and think about 12
[54:13] other things. But um but the but poker
[54:16] look there's
[54:18] >> I like poker before I like markets and
[54:21] what I think the markets and the thing
[54:22] that we do is much better much bigger
[54:24] much more impactful. So I but it was it
[54:27] was a good stepping stone for me. The
[54:29] dealing with incomplete information, the
[54:31] dealing with probability, the dealing
[54:33] with um making good moves and being
[54:36] wrong and learning how to handle that
[54:38] and then also being able to find your
[54:40] bad moves. Um all of those things are
[54:43] are really helpful and
[54:45] >> is investing a game?
[54:47] >> Well, I think predicting the future,
[54:49] right? In the end, markets are really
[54:51] tough. How do you win in markets, right?
[54:53] you have to know the future better than
[54:56] other people do in otherwise you should
[55:00] just be at the benchmark which is
[55:01] totally reasonable um and um and that's
[55:04] very very hard to do. So first off in
[55:06] that way it's a probabilistic exercise
[55:08] where you have to be used to being
[55:10] wrong. You have to get more right than
[55:13] you're wrong about. You have to make
[55:14] decisions um in high quality ways with a
[55:18] lot of uncertainty. And so in those ways
[55:20] it's sort of like a game. Um and um and
[55:24] that that's what is it takes to be good
[55:26] over a long period of time.
[55:29] >> And there's some of that in poker.
[55:31] There's a lot of that in poker. Um but I
[55:35] would say like my interest and what drew
[55:36] me back to poker was I thought it was at
[55:38] a very good intersection which is also
[55:40] happening insting of where
[55:41] >> you had this breakthrough around eight
[55:44] years ago where machines got better than
[55:46] poker than people and most people didn't
[55:48] know it. And that was a huge opportunity
[55:50] because if you could conceptually see
[55:52] what computers had discovered, people
[55:53] have played poker for hundreds of years
[55:55] and they were really bad at like if you
[55:58] play poker 20 years ago, you play poker
[56:00] today, totally different despite the
[56:02] fact hundreds of years of trying to get
[56:03] good at poker, people had no idea what
[56:05] they were doing. And um and that is like
[56:08] a pretty amazing thing. I mean, I think
[56:11] go was
[56:11] >> What's the biggest What's the biggest
[56:12] change uh in poker from 20 years ago?
[56:15] Well, I think people the biggest thing
[56:17] is
[56:18] let me see a and this this maybe
[56:21] happened 15 years ago or so, but really
[56:23] starting to understand how to think
[56:25] about the the range rather than think
[56:27] about your cards. Think about the range
[56:29] of cards you could have and that your um
[56:32] opponent can have and how to handle
[56:34] that. That was kind of building block
[56:35] number one. That was a big deal of how
[56:37] to play ranges rather than playing the
[56:38] individual cards you have. How to get to
[56:40] the math of bluffing, right? A lot of
[56:43] people like they were just building on
[56:45] their intuition of when and how often to
[56:47] bluff. There is a real math to how to
[56:50] bluff
[56:51] >> that um
[56:53] >> that people just didn't understand. It
[56:55] was a hard thing to take out. Everybody
[56:57] knew you had to bluff a certain amount.
[56:58] But but
[56:59] >> are you are you a good bluffer?
[57:02] >> Um in poker because literally there's
[57:04] like good theory on how to handle it.
[57:06] Now you need to modify it for who you're
[57:07] playing against, but there's good theory
[57:09] on how to bluff. Absolutely. I mean you
[57:11] cannot play poker if you don't know how
[57:15] and when to bluff. Um
[57:16] >> in poker you hide your cards. Uh in
[57:19] Bridgewater you are known for radical
[57:21] transparency, right? Is radical
[57:23] transparency real or is it like a
[57:25] storytelling
[57:27] thing?
[57:27] >> No, I think radical transparency is
[57:29] critical. If you come back to what's
[57:31] necessary to compound understanding in a
[57:33] community, right? Very different than
[57:34] it's an poker is an individual game.
[57:35] This is the kind of the problem. But if
[57:37] you want a compound understanding in a
[57:38] community,
[57:40] you need transparency. Now, we say
[57:42] radical transparency, and I want to
[57:43] differentiate that from complete
[57:45] transparency. That doesn't mean
[57:46] everything's complete transparency. It's
[57:47] just radically transparent, meaning like
[57:49] compared to most other organizations.
[57:51] And if you're trying to get to the best
[57:53] ideas, you have to fight against certain
[57:56] human things. Like humans, most humans
[57:59] like order. They like the person at the
[58:00] top to make the decision and whatever.
[58:02] You need to fight against the political
[58:04] things that are natural and allow the
[58:07] best ideas to win. And one part of that
[58:09] that's so critical is take all the
[58:11] decisions you make at the top and make
[58:13] them as transparent as you possibly can
[58:15] so people can push back on all of these
[58:18] things. And so for us, the radical
[58:20] transparency of saying, "Hey, we've got
[58:23] to share why we're doing what we're
[58:25] doing and take feedback on those
[58:28] things." I think that if you go back to
[58:31] things that made Bir successful, that is
[58:33] really important. We've screwed that up
[58:34] in many ways through history. So, it's
[58:36] not like we get that totally right. But
[58:38] >> when did you screw it up?
[58:39] >> Well, for a while I'd say the demand of
[58:44] transparency was and and the demand of
[58:47] feedback was all kind of top down
[58:48] pushing down. You must be transparent.
[58:50] You must get this feedback. You must and
[58:52] actually transparency at the top wasn't
[58:54] working. So what we did,
[58:56] >> what we realized and I given Ner Barde
[58:59] our current CEO a lot of credit for this
[59:00] but we had to focus the arrow of
[59:02] transparency and feedback up rather than
[59:04] down saying the people that need the
[59:06] feedback the most the people that for
[59:08] which the standard of transparency has
[59:10] to be the highest are the people at the
[59:11] top of the organization that it doesn't
[59:13] really matter. You can give truly honest
[59:16] feedback to the to a junior member of
[59:18] your team. It's easy to do and it might
[59:20] help them a little bit but it's not the
[59:21] important thing. The important thing is
[59:23] that the leaders get the feedback. It's
[59:25] okay if a young person's arrogant. It is
[59:27] a huge problem if the leaders.
[59:28] >> Do you think do you think people are
[59:29] honest with you?
[59:30] >> I think it's a huge um problem, right?
[59:33] People that's back to you have to fight
[59:35] human nature, right? Look, I'm powerful.
[59:37] They they they they worry about
[59:41] insulting me or whatever and worry about
[59:42] me being mad at them. So, how do you
[59:44] fight that? Right? A you measure and you
[59:46] treasure it, right? So a we look at the
[59:49] best managers in the company are the
[59:50] ones that get the most negative
[59:51] feedback. We try to encourage this. The
[59:53] only people think terrible things about
[59:55] you. They must they're independent
[59:57] thinkers. They must think you're
[59:58] screwing things up. They're smart
[59:59] people. The people that are able to draw
[01:00:01] that out, which we measure. We have ways
[01:00:03] of measuring are you drawing out the
[01:00:05] criticism about you. The worst managers
[01:00:07] in the company are the ones that their
[01:00:08] people say all good things about. So I
[01:00:11] could whip out our tool here, but you
[01:00:12] could see the incredible amount of
[01:00:14] negative feedback I get. And I'm still
[01:00:16] sure that people don't tell the total
[01:00:20] truth or anything like that, but you are
[01:00:22] hurt. Are you ever hurt by feedback?
[01:00:24] >> Sure. You you know, you find it like
[01:00:26] >> when were you last hurt by feedback?
[01:00:28] >> This morning. Um meaning like I was just
[01:00:30] sort of annoyed with like cuz I made a
[01:00:32] decision to say, okay, somebody could
[01:00:33] talk to the press about this thing and I
[01:00:35] didn't feel follow our process. I
[01:00:36] thought it was a small thing like
[01:00:37] meaning like and and my like I and and
[01:00:41] just yesterday one of the members of the
[01:00:43] team was talking about how the things
[01:00:45] some of the things I've been wrong about
[01:00:47] that I mercurialism being one like I
[01:00:49] thought it would have bigger impact than
[01:00:50] it has as an example and that I kind of
[01:00:53] can explain it away um a little bit like
[01:00:55] I did today well there was AI on this
[01:00:56] side and um and so I get this feedback
[01:00:59] and I
[01:01:01] >> do you ever think you know what I've
[01:01:03] been doing this for 30 years I'm
[01:01:04] stinking rich I uh I know what I talk
[01:01:07] about and here you are 26 year old and
[01:01:10] you're telling me what to do.
[01:01:11] >> Exactly. And so and that's the thing
[01:01:13] back like and that's the worst of me.
[01:01:17] Right. And if you know it and you expose
[01:01:19] it and you help other people show it to
[01:01:21] you when you're doing it, that's great,
[01:01:23] right? The same thing as the
[01:01:24] 26-year-old. Like you got to flip that
[01:01:25] on the other side. The 26-year-old's
[01:01:27] unlikely to say it to you. Even though
[01:01:31] man, how much better off are you to have
[01:01:32] these observations about yourself?
[01:01:35] There's no actual downside. There's only
[01:01:36] downside because you're a weak, fallible
[01:01:39] human. Like, what's what's the actual
[01:01:40] downside? If I listen and think hard
[01:01:42] about, wait, what am I missing? I'm
[01:01:44] still rich. I'm still like um and if I
[01:01:46] don't, I miss out on all of these
[01:01:48] jewels. So, I am definitely flawed in
[01:01:51] that. I feel it. I feel that anger. I
[01:01:53] feel that like, who the are you?
[01:01:54] Uh, sorry about that. But um but I feel
[01:01:57] that and I and I have so much experience
[01:02:01] with this and say when I feel that
[01:02:04] treasure it
[01:02:06] >> know that that's because they're saying
[01:02:07] something that you're trying to block
[01:02:09] out.
[01:02:09] >> Yeah.
[01:02:10] >> Um and
[01:02:11] >> that's a great that's a great point.
[01:02:13] >> That's a great point.
[01:02:15] >> How do you um who do you hire? What kind
[01:02:18] of people fit into your Well, I you know
[01:02:21] I read the book about you which is kind
[01:02:22] of about the firm which is kind of a bit
[01:02:25] one-sided um one could argue but um
[01:02:29] either people sur people who don't
[01:02:31] survive they leave very quickly right
[01:02:33] >> yeah I think um also evolved right and
[01:02:36] so I'd say there's a lot written in the
[01:02:39] press that has nothing to do with what
[01:02:40] happens at Bridgewater as I'm sure you
[01:02:41] can imagine press written about you but
[01:02:43] the um but the basic picture is look
[01:02:46] it's a very unique place that's not for
[01:02:47] everyone. This thing that we're talking
[01:02:49] about transparencies like some people
[01:02:51] just the pain of that is too much. I'd
[01:02:54] say this balance between the thing that
[01:02:58] the two failure modes for people are
[01:03:00] either are getting are are either being
[01:03:04] in unwilling to say what they believe or
[01:03:07] being unwilling to listen to other
[01:03:08] people. So we try to balance open-minded
[01:03:11] and assertiveness. Like this is the if
[01:03:13] you're say the two things you want in
[01:03:15] yourself as much as possible is to be
[01:03:17] really open-minded and be really
[01:03:19] assertive. And those two things are
[01:03:20] somewhat in conflict but it's the way to
[01:03:23] get to improvement because if you're
[01:03:24] open-minded to people listening but
[01:03:25] you're willing to say what you believe.
[01:03:27] >> Yeah.
[01:03:28] >> That's the mix, right? And I think two
[01:03:30] failure modes are people who are too
[01:03:32] open-minded and won't say what they
[01:03:33] believe and people who are too assertive
[01:03:35] and won't actually take in what's coming
[01:03:37] at them.
[01:03:38] >> How does it work to have a to have had a
[01:03:41] a really kind of dominant founder and
[01:03:44] figure in the firm?
[01:03:46] >> Yeah, it's a really hard dynamic in many
[01:03:50] ways and a great dynamic. Of course, I I
[01:03:53] benefit from the impact Ray had for 50
[01:03:56] years because we compounded it all and
[01:03:58] Ray had some great cornerstone ideas.
[01:04:01] >> Yeah. So, for the people who don't know,
[01:04:02] Ray Alio started this and he is, you
[01:04:04] know, like bit of a legend a legend in
[01:04:06] the industry like like you are
[01:04:08] increasingly becoming.
[01:04:09] >> Yeah. And so, and and for great reason,
[01:04:11] right, he came up with so many of the
[01:04:13] cutting edge ideas on portfolio
[01:04:14] management. We had a a hedge fund for
[01:04:17] institutions to create uncorrelated
[01:04:20] alpha way before that existed virtually
[01:04:22] anywhere and currency overlay and if the
[01:04:25] inflation index bonds and all of these
[01:04:27] things that that Ry was a huge part of
[01:04:30] that and probably most importantly Ray
[01:04:32] was a huge um believer in culture and in
[01:04:36] building something that would outlast
[01:04:37] yourself. So I such valuable things and
[01:04:41] he was
[01:04:43] as as most great people are. He had a
[01:04:46] huge impact and shadow on everybody
[01:04:48] else. And so we Ry decided, you know,
[01:04:51] this was like 15 years ago that to
[01:04:54] replace me it will take a decade, right?
[01:04:56] And um that sounded kind of crazy and
[01:04:58] arrogant at the time, but but it it did.
[01:05:01] And um and to his great credit, I
[01:05:04] believe he did let go, you know, like
[01:05:06] meaning it was hard to a lot of time and
[01:05:09] a lot of pain.
[01:05:10] >> But here you had people like Bob Prince
[01:05:13] who's been at Bridgewater 40 years, me
[01:05:14] Bridgewater 30 years, Ray had been at
[01:05:16] Bridgewater for 50 years, all totally
[01:05:18] committed to making Bridgewater
[01:05:20] long-term successful, different views on
[01:05:22] how to do it. But if you have people are
[01:05:24] committed to the same goal and they're
[01:05:26] good people, you will work your way
[01:05:28] through it. And I don't know, I mean,
[01:05:31] I'm not sure of a place like
[01:05:32] Bridgewwater that has successfully
[01:05:35] transitioned. The fact that we finished,
[01:05:37] we started this process 15 years ago.
[01:05:38] Ray stepped out of the CIO man CIO thing
[01:05:42] right after COVID and let us change how
[01:05:44] we did the investing in a big way and
[01:05:46] then, you know, stepped out of
[01:05:48] management and then stepped off the
[01:05:50] board and totally out of Bridgewater in
[01:05:53] June.
[01:05:54] >> That is an incredible success story. So
[01:05:56] the press loves to pick up on the pain,
[01:05:58] but if you look at that, that's a huge
[01:06:00] success story.
[01:06:00] >> What's the best piece of advice he gave
[01:06:02] you?
[01:06:03] >> I'd say I learned more from like
[01:06:06] watching him than like literal advice.
[01:06:08] But but I would say
[01:06:10] u pain plus reflection equals progress
[01:06:12] is great advice. Um meaning like when
[01:06:14] just as I was describing before, every
[01:06:16] time you're in pain,
[01:06:18] reflect on the pain. Don't run from it.
[01:06:20] And that's how you get progress. I think
[01:06:22] that's incredible advice. I think his
[01:06:24] view on compounding understanding on the
[01:06:27] fact that we should be disciplined and
[01:06:29] the and the um incredible energy it
[01:06:33] takes to be excellent. He was a role
[01:06:35] model in those things and I was very
[01:06:37] lucky to know him as well and for as
[01:06:39] long as I did.
[01:06:41] >> What's the hardest decision you made as
[01:06:43] a co-CIO?
[01:06:45] the hardest decision. Um, I'd say
[01:06:49] probably need to give this a a little
[01:06:50] bit of thought, but we've done a lot of
[01:06:54] reorganizing
[01:06:55] the investment engine that's been very
[01:06:58] hard like that. I mentioned the mistake
[01:07:00] pre-COVID, but how to get comfortable
[01:07:02] distributing more of the decision-m
[01:07:06] empowering people when you're managing a
[01:07:08] lot of security and intellectual
[01:07:10] property. How do you actually build that
[01:07:11] out? So we take big steps to that that
[01:07:15] required taking a lot of people who had
[01:07:17] been at Bridgewwater for life and
[01:07:19] removing them from a process because
[01:07:21] they were slowing down the progress and
[01:07:25] that not purposely like they were great
[01:07:27] but they were they were built in a mode
[01:07:30] where we centralized the decision-m
[01:07:32] largely at the CIO level and we had a
[01:07:34] lot of like people working for the CIOS
[01:07:37] versus freeing up great investors making
[01:07:39] them independent and doing that that
[01:07:42] transition which re you know kind of
[01:07:45] happened postcoid in a couple of waves
[01:07:48] has been extremely painful and extremely
[01:07:50] powerful at the same time. Those were
[01:07:52] probably the hardest decisions to take
[01:07:53] people who are great in the old mode and
[01:07:57] move them away as we move to a new mode
[01:07:59] of how we how we generate the ideas.
[01:08:02] >> Last question. What's your advice to
[01:08:03] young people?
[01:08:04] >> Yeah, I'm really bad at this. Um because
[01:08:07] for me I all I could say is what it was
[01:08:09] like for me to be happy and to get
[01:08:12] contentment from life. I was lucky in
[01:08:14] that I am graduate college come to a
[01:08:18] place that I fall in love with that I
[01:08:19] want to build you know and and the the
[01:08:21] the things for me were a a um passion
[01:08:27] for what you're doing which I had
[01:08:30] combined with a community of people that
[01:08:33] draw out the best in you. back to like
[01:08:35] Ry like Ry no getting feedback over and
[01:08:38] over again shapes you if you're willing
[01:08:41] to take it and I I was surrounded by
[01:08:43] people that were willing to give me
[01:08:44] feedback all the time that made me
[01:08:46] better and um and this problem of trying
[01:08:50] to understand the world those things
[01:08:52] came together for me in a way that I
[01:08:55] don't know that everybody wants those
[01:08:57] same things or whatever so I see it more
[01:08:59] like I could share my journey and what
[01:09:01] makes me happy that and want to raise
[01:09:04] great terms is that I totally believe
[01:09:07] meaningful relationships and meaningful
[01:09:08] work if you my wife my f my three kids
[01:09:11] etc like having meaningful relationships
[01:09:13] that support you in the difficulty of
[01:09:15] the things you're trying to do and
[01:09:17] having meaningful work which to me that
[01:09:19] means interesting work with people that
[01:09:22] push you to be the best
[01:09:24] things that worked for me and I was
[01:09:26] lucky to get them in spades and I don't
[01:09:29] know that I had any like other than
[01:09:32] knowing it when I felt it I don't know
[01:09:33] that I had any um great advice on how to
[01:09:36] get there. I I was I found these things
[01:09:39] that I loved and was able to pursue them
[01:09:41] with great passion.
[01:09:42] >> Very good. It's been away for a show.
[01:09:45] >> All right. Thank you so much.