Trade Wars, AI, and Bonds: What Investors Need to Know ft. Alex Gurevich

29:58And that's not good liquidity. No. Somehow it feels to me that liquidity needs to be injected directly into the bond market. Somehow it feels like there's a liquidity problem in the bond market that's stopping the usual buyers' being able to buy? Well, I think the change in banking regulation would be in effect injecting liquidity directly into the bond market. But the Fed cutting rates is also injecting liquidity into the bond market. They've been dragging their heels. They're so slow. I mean, you can see the forward-looking inflation stuff is falling and growth is slow, yet J-PAL doesn't want to cut rates yet.

30:39Well, paradoxically now, they almost cannot, because there was so much political pressure on them to they almost can't it's like this stubborn persevering child when when the parents tell them you have to go 40 and they were like no i don't need to go 40

30:59yeah true and i don't even know which side of this is being the stubborn child by the way no exactly no exactly maybe both no political statements here i'm just saying that there is a little bit of that dynamics going on now that how it seems like very logical would be pre-logical for them to cut rates in May but it doesn't seem like that's happening I don't know let's see how the economic data comes out I feel like it would be an outlier it would have to be probably an outlier development of situation for them to cut rates this time but it could happen but it just seems to be slightly unlikely right now because i feel like the the credibility will be weakened if they do in their mind what's your favored trade on the macro side currently and obviously that can change over a week or or several months or whatever it is is it tips or is it rates outright or you know what what's your favorite bet well long i love long duration anywhere on the curve i do have right now i do like tips because they're just really underperformed and they especially on the tips and they really kind of cover all possible scenarios high inflation low inflation is a kind of trade that works in a broad range yeah so in interest trade world long duration but honestly being long like red your adult your red see i am so old i'm saying still your dollar so for futures right i can't i still can't get my head around so for futures i just think of your dollars like being long the reds it seems to be not a bad trade because I think there will be more cuts than projected in the market.

32:37The reds is which year? Well, it's like 2026 kind of scenario. Just think that the next 12 months, there will be more than 100 basis points of cuts in this way. And I do think that's the case. So there is a lot of space there for you could get a pretty big deal if they cut to zero. That's like still 300 basis points to go. And do you trade that via options, call spreads, or just straight outright? Well, it depends on wall. You could do it outright. You could do option structures if you think wall is low. Recently, wall has come up, so it's harder to trade options. But you could do interest spreads if you want to reduce the wall.

33:21But, of course, then you have a lot of good offers. It's more like I don't have a strong right now advice on this is how you should structure your trades to people. But I'm just saying something with this thematics is pretty good in the interest rate world.

33:36But there are trades in other worlds. In the currency world, I think yen still has a way to go in terms of appreciation relative to other currencies. The yen appreciate. Yeah, I agree. I'm looking at the yuan-yen chart, which I think is one of the most important charts in the world. and it's a beautiful head and shoulders top that suggests that the yen needs to strengthen and the yuan can weaken somewhat. I definitely like both legs of that trade. Yeah. And there are some other currencies. So I did like in the currency world, I would not like so much better on the dollar just itself as I would look at various currencies which are very expensive and very cheap.

34:18Like for example, Swiss franc is very expensive. Yen is still very cheap. Chinese currency, it's kind of hard to tell whether it's expensive or cheap, but given the interest rate differential and policy divergence at the record, it seems like it could get cheaper. Yeah, and it doesn't seem like they will allow it to cheapen against the dollar, but they really want it cheaper against the yen, because the yen has been so weak in their export competitors. Right. And so it's probably not going to be like, it's hard to see a totally catastrophic devaluation, but this kind of grind of carry plus maybe slow depreciation probably is going to be happening in China.

35:00So there are like what I'm saying in currencies, there are pockets of things you can do, pockets of buying things which are cheap and selling things which are expensive, which seem to make a lot of sense in the long run. Again, I'm not going to be like super specific about, oh, you have to do all of this and this and options, but more like just, I think good teams are short, long yen. If you want to be a long dollar, I would probably be, you could balance it with some hard assets, and hard assets have been doing really well. I mean, I really don't know how to puzzle the gold performance fully.

35:39I mean, of course, like it's easy to even like various things that we discussed to attach a narrative to gold, but I have to say gold a little. So my target on gold was 3 ,000. It a little bit surprised me by going to 3 ,400. I had like a 3 ,000 target on gold since like 2011, right? I'll tell you what I found. I was just working with Julian Patel, and we were shooting some ideas around because writing GMI. We found that, and I've never realized this, that gold just follows with a month's lag financial conditions. And financial conditions leads liquidity, leads crypto, leads all of the other things.

36:16So gold is kind of living in current financial conditions. And we classify financial conditions as the dollar, rates, and oil.

36:29So you're saying that financial conditions, also financial conditions lead to higher gold? Yes. And it's pretty decently highly correlated. Financial conditions are not terribly loose right now because of the higher rates and the fall in the stock market. It's the dollar. No, forget the stock market. That's a lag. The one we use is just three variables. And it has the highest correlation, has a 97.5 % correlation to the NASDAQ, all of these things. And it is the dollar, rates, and oil. Now, rates have come off a tad this year after going up a lot last year. The dollar has obviously had one of the fastest moves in the shortest period of time we've had in the last decade, decade and a half.

37:13and that's really moved financial conditions and oil obviously came down quite sharply as well um so financial conditions are loosing quickly and and gold seems to be just following that and has done for the last decade well i would be careful of um there's not being what i call a tautological correlation because gold is clearly correlated to dollar because it's gold denominated in dollars like what i would like to look at gold and euros as correlated to dollar yeah so we looked at the correlation of them individually as factors and we found combined was better than individually so anyway that that's what that's my current thinking is i don't know i'll write about it uh this is it's to do with financial conditions or you can just say that there's more buyers and sellers.

38:07That's true too. But there's clearly more buyers and sellers of gold. What is interesting, gold mining stocks were doing well, but lagging gold performance. And silver is lagging. And platinum is really lagging. Yeah. And actually, my experience with platinum and palladium is that you just have to be patient. And one way or another, platinum will go to$3 ,000 as well.

38:37but you have to be really, really patient. It's like a decade-long trade you're talking about. But those trades, however, do give adequate return on capital. If you bought Palladium in 2003 at a couple of hundred dollars, you would have watched it go to like 3 ,000 over almost 3 ,000 over the course of next, whatever, 15 years. And I don't have like the exact years right now. And it would be a really good return on capital. Yeah, the annualized return over time would have been fine. Yeah, so the strategies of trading the Platinum Palladium complex have been yielding good results. They just have to be extremely patient.

39:15And silver, I think silver has space to go, obviously based on historical patterns. And cryptocurrencies obviously have space to go. I don't have like a very as you know I'm not very passionate about like hyper-Bitcoinization or anything like this it's more like just look at the historical charts Do you trade it for the fund or do you just hold it yourself? I can trade it for the fund I don't like discussing my specific positions but I definitely have the volubilis to trade it from the farm. You have the ability to, or you hold it yourself. Yeah, so I have the ability, and I can express the pros and cons that I see on crypto as much as on gold or silver, right?

40:05Yeah. Like, I can tell you the big picture of what I'm thinking about there, but you could, what I'm saying is, like, being long dollar against something worse, like China could be offset by, if you have the appropriate view, by being long hard assets. and what about the commodity complex like copper stuff like that which is a trade that's on everybody's radar screen and also what about equities so for me copper is an interesting trade because it's a it's a trade like when you it's one of those trades that i'm fairly confident that if you go see through all the cycles 10 years forward i think copper has much higher to go because I don't think they will cancel electricity.

40:51And I don't know if they will have transmit all the electricity through microwaves, beams yet, 10 years from now, but the need for electricity is increasing and I think they probably will need copper. In the next up swing of the cycle, copper could go probably higher in price. And, however, the counterflow is that If you do think of China's slowdown and global cycles, cyclically, there could be some headwind to copper because it's an industrial metal, it has a strong industrial component. So you'd rather buy low, sell high? I'd rather buy low and sell high, yes. And equities, any strong views here or not really?

41:35I'm leaning negative on equities. I've been negative equities for a while now. probably I don't like being overly short stock market no it's a very good thing for me learn that lesson but you know what however I remember there was a convention some kind of investor convention around 2018 I think maybe like even before before that big correction in 2018 right and I remember I was giving some kind of presentation I was in front of 100 people or whatever 200 people and asked them hey raise your hands if you think that within the next five years you have better than 90 chance of buying stocks significantly cheaper than they are now and almost every hand went out and i was like then why are you long

42:24so what i feel like what on one hand yes in the very long run stock market is always cheap but what i don't know any exceptions to the rule that once stock market breaks out and makes a good run to the new at the money highs like significant new at the money highs then there will be a point in the future that you can buy it significantly cheaper it's never like so it goes through various like like if you just look from a chart perspective it goes through various like it breaks right barriers it goes rallies 30 more many people think that it cannot rally again but it rallies 30 more but at some point it still goes back to test the previous breakout points.

43:11And it just happens every single time. And one time it's COVID, another time it's tariffs, another time it's global financial crisis, another time it's September 11th. It's always something. So I feel like the odds are not so much that I feel like, oh, stock market has to go down tomorrow, but I feel like the odds are skewed towards, I have felt that stock market being significantly lower at some point in the next two years. And

43:43every year something happens. First of all, the rule is something happens. And this is what happened now. Now the whole tariff kerfuffle happens, right? Like a year ago, I wouldn't know what it is that is going to happen. But this is what happened now and maybe this is it. or maybe, but I think my intuition is that there's a good chance of things like this winter, it was a strong view right now because there was depth like, you're still waiting for something to happen. Now that something happened, you're like, okay, now do I start buying the dips at 5 ,000 S &P? But maybe that's not really a dip yet.

44:18Now, when it goes to 3 ,200, that will be a dip. That's how kind of look. However, I don't want to be too short because if you think about investor or money manager myself, the greatest opportunity lies not in timing to short stock market. The greatest opportunity for me historically lies in buying stock market in March 2009 and March 2020. And honestly, the reason I am successful is very much based on those things. Even Stan Druckenmiller said the same. It's like I've never really made money out of being short the market. But if I look back, I've basically made all of my money being long stocks.

45:02Long, but long, long stocks, long risk assets, but investing at the right times. I mean, in some of the crises, I was, sometimes it's a matter of luck because every crisis develops in their own unpredictable ways. And I think even in an earlier masterclass, I don't feel like in 2001, I was positioned well for it. in 2007 I had a lot of trouble I thought I was well positioned for it turned out I was not because things didn't go exactly the way and eventually if all positions held they would have been correct but they were not correct for the beginning of GFC right in 2020 I was well positioned for that like when in 2022 when there was like the bond market collapse and the war in Europe I was not necessarily well positioned for that so with every new crisis so this is a little bit of a Russian roulette every crisis yeah but what does make money what is a very big portion of making money for investors is having the courage to gradually become fully invested in beta at the bottoms at the bottoms being not the person who is selling but the person who is buying and a lot of successful performance comes from that so it's kind of like if you have a crisis and it didn't go really well but it didn't lose all your money and just kind of left standing that's also okay because that's kind of what happened to me in 2007 2008 yes it was kind of painful but it was also a buyer of stocks throughout 2008 and 2009 him yeah i got that one wrong because i was doing the wrong thing i was still bearish in march of 2009 i'd had money through the whole crisis and then gave a whole lot back by just being negative too long yeah well it depends on your style there is nothing wrong with being like um investors on the short side could be asex like as we decided before there's no such thing as bad markets they're only wrong bets right that's right that's right so like if you can be bad on shot and being making money that's all the more power with you i wrote a book about how it's easier to swim with the tide but not everybody has to swim with the tide no it's a matter of personal preference so i'm not going to judge because i've been plenty wrong both throughout my career and recently to like the judge of like oh you should have known better on that year because that would come to bite me pretty badly.

47:48Yeah, exactly. Hubris never wins. So listen, let's move on. And I want to hear your thoughts about this singularity because you have been not only a writer of science fiction, but a reader of science fiction. You've been thinking about a lot of these things and now suddenly it all seems to be happening. What is your mental framework for this right now? Because I don't think people get it yet how big this is. no they don't and paradoxically even the most strong proponents of singularity like people like ray kurzweil they almost like moderate their speech because they cannot really spell out what are the real consequences of that because you cannot easily spell it out because you'll be crazy now when i was first introduced to the concept of singularity was almost 20 years ago and since then i've been thinking like i read the a couple of books, including by Ray Kurzweil and some other stuff around singularity, right?

48:44Maybe like three books or so on that subject, right? Back in the arts, right? And I thought of it as a really cool idea. I definitely bought in into the way of thinking, but bought in as like a speculative macro trader. Yeah, this might be a good trade, but it could go either way. Like, and it's always, this is one thing I also know. If you look at the asset performance in the past, you can derive a lot from looking at the graphs. But you never really feel the asset until you invested in it. So from that moment when I read this stuff, I started tracking this thing a little bit and thinking about this regularly.

49:27And so many people denied singularity. So many professionals in artificial intelligence were saying, oh, this is all bullshit. This is not going to happen. Trust me. Trust me. Trust me. I know all of this is all nonsense, right? Not in our time, in 200 years or whatever, right? I heard this so much. But what I've noticed, and this is what I've been steadily noticing, and those are not the words I'm saying now. This is the words I was saying 15 years ago. Of course, I could be lying now, but let's assume that I'm telling the truth, right? That the proponents of singularity as a theory were showing charts, graphs, math, calculations, while their opponents were just dismissing it uphand without any data.

50:07Narrative only. Yeah, and what I noticed is that data tracks the Ray Korsvalium path of singularity that everything has been tracking. And it's been tracking and continuing to track and you could see like, okay, this is going a little slow, this is going a little faster. And then exactly as predicted at some point, things started to go vertical and this is now.

50:37and and and and i know and now i'm kind of already to me the debate is over people who say like singularity is not happening i'm just laughing them off people who start saying about that well yeah ai can do this but cannot do that it'll never be able to do this i think i'm just this is i'm no longer having this argument and i'm not i don't believe even the stupid about what is agi and what's not agi and they keep shifting the goalposts exactly and i'm not everyone i'm no longer seriously engaging with people who think like AI cannot do something. No. AI cannot be caring or AI cannot be creative or AI cannot write good poetry or this or that.

51:16Of course, people are entitled to their opinions, but this is no longer a matter of debate. I was willing to debate it a few years ago. I'm not willing to debate it anymore. I think it's just hilarious at this point. I just want to pat people on the head and say, oh, you're so cute. You think AI cannot do this? Well, maybe for the next three months, it won't be able to do that. Okay. I mean, I've never seen a technology increase in its power and abilities at this pace ever. I mean, don't forget, it went from using, let's say, using IQ as a measure, which is not a great measure, but it's a measure.

51:51Last year, IQ doubled, and now the latest model, the O3, has an IQ of 150, let's say. okay so now it's very very very smart now if it doubles again next year and it has been doubling every year it'll be at 300 it'll be smarter than any human that's ever existed in history and then the next year it'll be 600 and the next year it'll be 1200 and then it's like it's beyond human comprehension yeah well this is actually this is where i want to talk about this because i keep this this iq and it's actually this is has a very interesting implication what i want to say about to NFG industry too. Yep. So the growth, you know, like I know that you have a lot of thesis of like exponential age and I want the title is the end of exponential age.

52:41We've had been in exponential age for a couple of hundred years. Now it's ended. Right. Because people, a lot of think of exponential, exponential is just a lot. There's some people, mathematicians, for mathematicians, exponential growth is a certain type of growth. Yeah. And it both has to accelerate and both come to an end. And I want to explain why. So can I go math? I don't know how you're doing with that. Can I go total math geek on you for the next five minutes? You do it. Let me unleash my math geekiness. You do it. You got the PhD in maths. Go for it. Yeah, so let's unleash it on the unsuspecting audience.

53:21A lot of stuff. So a lot of computer computation theory stuff and calculations based on iteration that you do the procedure and repeat the procedure and repeat the procedure. So in math, the first level of iterating or procedure is addition. Plus one, plus one, plus one, plus one. Think of it as in terms of population growth. That's probably the easiest way to describe it. You have a family and every five years they have a kid. That's like the one level of first operation. That's the first level of operation math addition. The next level of operation is multiplication, which is you have every family in the world have a kid in five years and the world population gets multiplied by whatever, 0 point, whatever, 1.5, 1.25 or whatever, every certain period of time.

54:10Now, when you reiterate it over generations, you get what you call exponential growth. Now, we've had this experience, exponential growth in population, even though it's slowed down now. Notice that it's slowing down, that peak can slow down. We had exponential growth in GDP. Yep. And many other growth. We've had exponential growth, for example, in speed of computation. Yep. Now that, what I'm trying to explain is when it's driven by these factors, it's pretty stable, could be like 2 % exponential growth for a long time. But now the factors changes. Now imagine, because computers, AI works very differently.

54:51Because up to this point, we've had this paradigm. Every generation, certain family has certain amount of kids, and the population grows by a little bit. But now imagine that the world is interconnected, and not people just having kids not with each other, but with every single person in the world. So this is now not like 2 billion couples are having kids, but for quintillion couples having kids. But furthermore, it doesn't have to be in couples because interconnectedness rises in complexity. It's actually two to the power of four billion possible combinations. Every subset of human, imagine that every subset of human population can create a family and have children.

55:34Now, imagine also that the amount of children they can have is growing also with the size of population. that is yes the population is four billion one billion people then you don't have one kid but if a population is quintillion people they didn't have quintillion children and that is how ai i am not in like this is not science fiction this is how ai actually works yeah i think of it i think of it is if if most exponentials are metcalfe's law is how you look at them this is reed's law which is Metcalfe's law. Yeah, well, but this is way below, beyond the Metcalfe's law. But next is Reed's law, is Metcalfe's law squared?

56:13Yeah, but it's not squared, what I'm trying to say. It's a completely different thing because what happens next is the amount of time it has to take children declines proportionally to the amount of people in the world. And that's true too because each new generation of computation takes less time because it becomes... And then, so what happens is First, you iterate multiplication to create exponentiation. Then you iterate exponentiation to create what the process called penetration, the powers of exponents. Then you iterated this to create plantation, hexation. And then you have like, but then what happens next?

56:51This is what people are missing. Then we move to a completely different paradigm, the paradigm which we cannot even conceive right now. what all of this that i described to you is still within a paradigm that we can see right now but what what happens in actual math when you start doing this process which is called fast growing hierarchies eventually they move to like completely different totally wild paradigms which you just you need to not only have p if you have pg in mathematics in about 20 hours you can understand what's happening right so how to get to numbers that big and something's going to happen that we cannot even conceive of because we are only using our very limited intelligence to conceive it compared to transcendent intelligence.

57:35And this would be what, this would be the move towards ASI, the artificial superintelligence. Correct. And when, but as this happening, what we have to, those numbers, like fast growing hierarchies, they cannot exist in real universe. Because within two seconds, you can move to surpassing the computational capacity of all multiverse, right? And there are quantum physics restrictions of what you can do. So what is the practical thinking about this? The practical thinking about this is that energy might be a serious constraint to our growth. Right now, we're seeing the growth in energy consumption from data centers and computation, and it's very rapid.

58:24But I think that when I'm talking about all this process to you right can you imagine how much energy would take to supply all this process so what i do can we be a kardash kardash of level one society or do we need to loot to level two level three i is the sun enough elon's point is the sun is enough to power all of this i i do i do not think that well first of all what i what i think is our energy demand will way outpace our technological capacity. It's much easier to create artificial superintelligence than to dice on the sun to become level two. I was just talking about this with my friends who is a physicist.

59:08How do you dice on the sun? And their principal problems are so difficult. You don't even know how to do that with superintelligence or without superintelligence. It's just not an easy project. Let's put it like this. However, I can allow the possibility that in 20 or 30 years, our need, our demand for computational capacity will surpass the energy output of the sun. Which is staggering to think about, right? But that's why people don't want to throw such statements because that's why it creates crazy science fiction. But I think this is the direction we're going into. Which means that energy unavoidably, I don't see how energy will not become a bottleneck.

59:53but at what point does it become the bottleneck i mean at universal scale or solar system scale sure but when do we get there when does the multiplication of the numbers get us to a point where we can't use the energy that we can capture on earth first well that's a good question but i wait do we get past a level one and can we keep up with that rate of growth in energy supply as the demand is required? Well, it's almost like I would almost look at this problem from the opposite direction. I would argue that energy has always been and will remain a bottleneck. Because if you think about this, a lot of energy costs, whichever they are, one of the major constraints of current global growth.

1:00:43Yes. If energy was free, maybe each of us would have... I don't think everyone will become private planes, but you can imagine a very different paradigm, right? Energy is a constraint, but it's just constraining different parts of life. What I think, the question is not when energy will become a bottleneck, but will energy will bottleneck specifically for computation.

1:01:10And my guess is that, for example, my guess is that thermonuclear energy will come sooner than people think because of AI, because AI is hungry. And when people, and it needs energy, it wants to eat. It will figure out to feed itself thermonuclear energy. So, and all the problems that we're having with how to harness it will be solved, I think. But that's not an expert view. This is my science fiction hat on, right? Yeah, but I can't believe it's directionally wrong that the AGI itself will solve its energy problems to the best of its abilities. Correct. That's what I'm feeling. But however, just as humans solve their food problems and resource problems, they improve their situations, but they never have sufficiency of the resources.

1:02:02They always fight over resources, always ration them somehow. Same thing will happen with AGI. It will have to ration energy, fight for resources, and challenge them. Is the fight going to be between, and there's two parts to this. One is right now we've got all these different models. They're all being trained. They're all operating. They all use energy. So there's a competition amongst these new AI nation states that are rising. Okay, fine. But at some point, all of these models are cross-polluting themselves by the internet, which I can see happening. So they all become a super cluster of these things.

1:02:42is there going to be a conflict between the AIs? I don't know if there'll be a conflict, but I think there'll be a competition. I don't know if it will be a conflict or a competition. But if they're smart enough, surely what they would do is try and kill their competitor. Not all humans try to kill their competitors. We just hype it up. For example, if you have a store and there is another store across the street, majority of humans try not to let go kill the person across the street by just put them out of business is good enough, right? True. So we also have wars for energy, which is... We do.

1:03:17So I think just as in a human world, in AI world, there will be good actors and bad actors. Yeah, makes sense. So there will be people who compete for energy economically and the people who resort to... And there will be not people, but agents, we'll say, who resort to dirty tricks like viruses and sabotages and whatever, right? And is it possible, is it even remotely possible that fossil fuels can drive the next 10 years of this? Or are we really going to have to double down on every source of energy we can possibly find? I mean, how do we solve it? Let's assume that compute can keep going further than the energy allows.

1:03:55So that's the dragging factor that slows it down. What energy sources are we going to use? Where is the hyperscale here? Nuclear, obviously, is one. Well, nuclear, I think we're going to have, I think, well, solar is growing pretty well, but I think eventually it's not going to grow fast enough. I have mixed feelings on fossil fuels because I guess my science fiction writer's head says, like, well, we're going to still need fossil fuels for quite a while yet, but eventually they're not even going to move the needle. So they're not even going to matter. No, that's right. Because if you're, you know, to complete that whole Kardashev scale civilization, it's a lot more than just the fossil fuels.

1:04:35It's kind of, we have to carry the entire energy of the sun. Yeah, well, like you're talking about level two, which again, no matter how powerfully I feel about singularity and like medical and computational field. Again, I am not as optimistic about hard technology. Like, for example, robotics, I think will even robotics my leg a little bit. It's much easier to teach artificial intelligence to be a lawyer than to be a gardener. Yeah. but building like a sphere around the sun or something like this it's just like and making it like not cost more energy than it consumes and what is this sphere just if you really think that what it should be make of that it doesn't collapse out so it doesn't collapse under the gravity of the sun like the technological problems again maybe i'm thinking with my limited human brain that as i will say will solve easily but my limited human brain does not see even an inkling of a practical way how to get there.

1:05:35So, okay, we've got this limiting factor that may slow down the exponentiality or the, not even exponentiality, but even further rate of growth. What does it do to the economy? Because I think of it like an economic singularity where we get to a point, I don't think it's far away, where economies don't function in ways we understand anymore. Yeah, well, you know, it is interesting. You've been talking about economic singularity recently, but I don't know if you know, I wrote an article, Orbiting Economic Singularity, somewhere like in 2016 or 2017. No. You can look it up. There's an article called Orbiting Economic Singularity.

1:06:15No way. I didn't know that at all. And one of the things that got some pushback on people because I argued as one of the reasons to stay long stock market. And people were just kind of telling me that I'm just using bullshit to justify my overvalued positions. But excuse me, in terms of long stock market. So what does happen to the economy and when does it start breaking apart in ways that we don't? Yeah, like a lot of things. Economic singularity is a very interesting thing. There's so many aspects to that. One of the factors like that, the job loss starting from higher paid jobs, not from lower paid jobs, as I know it is.

1:06:56Every time in the past, people are comparing to previous technological shifts, but previous technological shifts took out jobs from this is not new idea like that's been percolating now for a while by the way what i'm telling you is not my insight like previous technological shifts will take out like the least desirable the least paid jobs right and everybody else could take other jobs which are and would actually create this way more wealth more opportunities for people and more new jobs will create be created right but this time it's like the doctors and the lawyers are being taken out. So the high paid people are being taken out and the gardeners still have their jobs in the foreseeable future.

1:07:38That might be taken out by robots, but it doesn't seem to be like an immediate horizon. And then so people so what happens when people just don't have marketable skills anymore? Human beings don't have marketable skills skills anymore. There's like one aspect, like almost an innovatability of some sort of UBI and very different fiscal policy, very different societal structures. And on the other hand, again, this kind of energy rush, gold rush to energy, which will have to end up, what I see is some sort of like gold rush survival battle for the energy and then some kind of stabilization and actually somewhat of like explosive growth potentially and then flatlining growth potentially when energy capacities will run out.

1:08:27And it will happen because I feel like this software will run ahead of hardware, so to say. So for example, when I said we're going to get to thermonuclear energy, maybe sooner than people think, but probably not soon enough. Because to put it on big scale, it's just like, it's a long process, no matter how you look at it, with all the regulations and politics. Even if tomorrow they have a thermonuclear reactor, that can viable it's still a decade before it's gonna make a difference right so so i feel like there'll be those slowdown periods when they'll be like oh it wolf uh how do we so as a quote opposed to being liquidity as the constraint on the system which is the framework that you and i have used for for many years in macro it becomes energy is constraint on the system yes that's what I'm thinking that that might become the case because liquidity in the end, central banks have proven that they will provide it.

1:09:28And we can't provide. We don't have the ability to keep growing the economy by increasing energy output fast enough. Yeah. That's what I see. Sorry, I'm just mentally getting the image. So we get the ramp up as all of this happens and the investment piles in and we get the ramp up in the usage of energy. But eventually it stops growth because growth is a rate of change. And if you can't, you've got no energy to create the rate of change. Everything slows down until we catch up. Yes, that's kind of what I paradoxically see. As paradoxically, I see that at the end of exponential age through the super exponential burst and computation, and then everything kind of having to slow down.

1:10:14Fascinating. Because, you know, I thought that after 2030 or so, things get complicated. And I was kind of imagining that we will probably be in a bubble all the way through until that with some cyclicality, I'm sure. And in my mind, and I didn't know why, I had this idea that we were flatlined for a while. whether it was because the markets were being run by AGI and therefore they'd become just cyclical. But your thesis really plays into that because it's about by 2030 where we'll start hitting these constraints because we just can't build it fast enough versus the speed that this stuff is scaling.

1:10:49Yeah, it could happen. Well, again, it could happen by 2030. It could take 10 years. It could take 20 years, or it could take two years. It's like really hard to predict because we don't really know everything that's going in AGI, right? Because what we see is just the tip of the icebergs. What we see is what Google is doing. We see the Gemini's, the ChatGBT's, the Groks of the world. Those are open source things, not open source things, which are kind of being showed to the world and free or cheap. Now, I would be very surprised if major hedge funds, and I don't have any inside information telling you about this right now, but I'll be very surprised if major hedge funds are not working on serious AIs to boost both of the investment process, right?

1:11:42Yeah, because we think Renaissance Capital had it for a while, right? In some way, shape or form. Well, yeah, but they did not have the capacity to, there were various AI systems, but LLM didn't really, But now that LLM exists, who knows what people are doing with it, right? And they are definitely not publishing their results. Now, who knows what the governments are doing with respect to military, right? Again, they're probably not publishing the results. So the stuff that might be going under the surface there could be staggering. And when the stuff starts using results...

1:12:21you know there is this joke about Douglas Adams novel in which there was this girl who could she kept like reciting the stock ticker in her head but she was reciting she was sitting wherever she was she had this psychic power to recite the stock ticker but she was reciting the current stock ticker and nobody really wanted to spend any time funding the research on this girl but somebody said like if she starts for reciting tomorrow's stock ticker. We'll have no trouble funding this research.

1:12:55Exactly right. So this is one of those things. Once these things, as these things become more and more efficient, they'll start sucking more and more. So, look, knowledge is being disrupted, particularly the more expensive knowledge. So the high-end margin of knowledge is lawyers, accountants. And it's also hedge fund managers. How are you thinking through this, right? If you prove to be good with your knowledge, you capture a large premium versus the energy that you use as a human, and AI is going to disrupt that. Okay, so this is a very good question because I was really seriously thinking about this, and this was kind of like a senior therapist-level identity crisis.

1:13:43Yes. For hedge fund managers. I mean, Alex, just to explain, yesterday, well, on Friday, the new O3 model came out with a deep reasoning. It is ludicrous. It does probabilistic future analysis of macro. And it will break down all of its logic and why it got there with references. And I'm like, okay, we really are six months away from nobody ever needing me. You just need to ask the right question. And I'm like, you know, we are six months away from it being able to run a macro portfolio. brilliantly yes there'll always be some person who's better but i'm just it's it's starting to shock me and i knew this is coming i've been talking about it for a long time but now i'm seeing it i'm like holy shit i mean this is better than 99 of all economists and market strategists that i know of well so here is my my here's what i came to thinking and this is a big identity crisis for people like you and me because i grew up on very part of my identities this is what i'm good at this is who i am more than like i don't think of myself like my identity by maybe because i'm under a privileged class or whatever but i don't think of my identity as like oh i'm a person of certain gender i'm like a parent or i'm this or i'm that first and foremost i'm a person who has certain strategic skills right that's my answer to the question of who i am right and when if that gets negated then i still have many things left in my life but that's a very big important part of it And it's also implication for investment.

1:15:16But one of the things that is important is imagine what happens when AGI takes over running hedge funds. And everybody gets convinced that they're better than managers. An infinite amount of capital runs to those hedge funds. Every market price will start rapidly going to balance. Yeah, that's my theory. Where is the imbalances if the AGI is in charge of money? Well, yeah, so there will be like a huge market shock, right? And they will adjust for the fact that there will be gyration as you approach the shock. It's like when you're crashing into a black hole, you actually can come out on the other side, right?

1:15:52So you can crash through the prices that are reasonable prices and go through them, right? So, but what I'm saying is that if you right now are macro manager and you can identify significant long horizon imbalances, the advantage of it is that you can get payoffs much sooner than you would have otherwise. when AI takes over. Because your positions will still be valid. If your positions are valid and AI will come and validate them, the reallocation of capital can very quickly move them in your favor. So it is actually a good time to say that. We talked a lot about narratives and false narratives and narratives that create reflexivity.

1:16:34Surely there are no narratives in an AGI investing world. And so a lot of how we think about market sentiment analysis, none of it matters. It might not. It might not. It might be, but they can be like the agents can create their own kind of society around this and they will be trying to guess what the other ones are doing and trying to front round them and game it and they will be using all sorts of game theoretical methodologies because even if they see that the balance price is better than we do, the biggest problem would constrain macro trading. I mean, it's not our ability to identify value.

1:17:13It's like our ability to manage risk and manage liquidity. Because it's not really, you don't have to be a genius as a macro trader to identify. There are a few trades that in the long run make sense economically. Even though people will disagree on many things, but there's a lot of trades that people say like, oh yeah, if you hold this thing over 20 years, that will probably, it makes money in every case. okay alex well then let's go back to another thing you said is the stock market is always underpriced over time right well surely an agi will realize that calculate the expected future returns and just ramp it to there now exactly but then another agi will calculate that this agi is over leveraged and squeeze them out they're just playing the same games that we play with each other yes they're just going to play them at a higher level i think because they'll have the same problem they'll still have to deal with leverage and liquidity yeah they just don't need to deal with emotion i don't think maybe they do they won't have to deal with emotion but they will have their own like tidal effects of like how they interact with each other how they affect with each other how again it's hard to predict the paradigms could be different but there would be various feedback loops could still happen they're like all the games theoretical things like i know that i know that she knows that she's gonna do this and then i'm gonna do that it's just going to be much more complex levels but I still they still will mess with each other just probably in a less emotional way but that whole thing of like calculating what are the chance that this other AI is over leveraged and if we push the price down they're going to force them to sell and buy it cheaper right and the other AI will throw out all sorts of false flags and pretend to be short when they're long and whatever all i do know is you know we've skirted around it really but this is something this what is about to happen is of civilizational importance the scale of what happens when you bring in infinite human knowledge or infinite knowledge with infinite physical labor the robots is something we will never and can never understand.

1:19:30And what it's going to do to our economies, what it does to the world, what it does to the universe, what is the universe? There's a lot of people, and I've spent a lot of time on this, thinking that the universe, the fundamental principle of the universe is not space-time, but consciousness. And that seems to be coming to the forefront as people understand more about AI. We're just going to go through a really strange period where all of the stuff you read through in science fiction becomes hyperstition, i.e. it kind of happens. And we're seeing that everywhere. And in many aspects, we're already leaving a lot of science fiction behind.

1:20:07Like a lot of stuff which some science fiction presupposed could not happen is already happening. Yeah. And you can already see people talking about ASI and saying, yeah, but that's going to be another 50 years away. And I'm like... three yeah as a gi i think is here asi i mean yeah i mean if it's three or six or whatever then you might hit the energy constraint that it will need to solve because it will be the super the super creature but um people are wildly underestimating the speed of this as you said at the beginning yeah it's um it's it's it's it's very wild and it's very difficult as a manager or money manager to keep like, you have to be very thoughtful about this and think there are still cycles, there are still considerations that we're used to, there's still all the procedures that we're used to.

1:21:01And then you think he has a fadden, like a fadden, you don't know if it's business as usual or will all be some kind of ball of energy spreading around across the universe. One thing I do know is that very soon this is a deflationary nuclear bomb. This is the single most deflationary thing ever to happen. and i would agree with this it does appear to be that way yeah and that will play out when that starts when people really realize how deflationary it is when every lawyer's you know lawyers charge 600 bucks an hour and it goes to 29 a month or 20 a month i mean that's insane deflation at the most premium part of the economy so we've got all this fun to be had listen i've got to jump But as ever, a fabulous conversation.

1:21:48We need to do this over a glass of wine because it's several hours. Yeah, I would like to. I hope that we have a chance to meet more socially and just have a chat about life. Yeah, absolutely. I look forward to it. Anyway, thank you for coming back to Real Vision. I really appreciate it. I think it was a pleasure. So a great conversation with Alex, as ever. I think we have some differences. He seems more thinking towards a recession is coming now. I don't think that will be the case. In fact, I don't think we have recessions again for a while just because of how the Federal Reserve use liquidity to remove risks from the economy.

1:22:25But either way, growth slowdowns, that kind of stuff. We have some differences. We have similarities. We're both disinflationists. We both think that the trends going forward is lower inflation. We're both looking at the bond market thinking, you know, there's probably opportunities there. I might look at how that drives other assets. He might trade bonds directly. But overall, fantastic conversation. Then going into the future, there's many schools of thought about how this plays out with AI. And Alex and I really dug into the idea that energy is the constraint. Now, we're also talked about the fact that AI, AGI and ASI will obviously solve its own energy problems as well.

1:23:04But he's not sure that we can get there because of the law of big maths. I don't know about that. And obviously Alex doesn't either. We're just trying to think about it, think it through. What could slow this down? What could change the trajectory? All we know is up until a while, this will continue. And that trend is one of the biggest things we have ever seen. And in the end, it is the biggest thing humanity will have ever lived through. Anyway, exciting times. What a time to be alive. I'll see you next time. If you liked this episode, I'd love for you to head over to realvision.com forward slash join for a free membership.

1:23:43Start your journey today to unfuck your future. Just one click away.

1:24:05Feel ready? You can move to real money with as little as$100 once your account is approved. And the great thing is that in addition to crypto, Plus500 gives you access to a wide range of instruments. S &P 500, NASDAQ, gas, and much more. Explore equity indices, energy, metals, Forex, and beyond. With a simple and intuitive platform, you could trade anytime, anywhere. Experience the fast, accessible futures trading you've been waiting for with Plus500. With over 20 years of experience, Plus500 is your gateway to the markets. Visit us.plus500.com to learn more. Trading in futures involves the risk of loss and is not suitable for everyone.

1:24:41Not all applicants will qualify. Plus 500. It's trading with a plus.