Transcription
Is the at at base is the problem that uh faces American AI? Uh one of the major problems is that the the valuations we currently see. Typically hyped up almost like the valuations of a bubble. Yeah.
But they are predicated on uh when this is up and running and we're in the market, we're going to be able to charge a an arm and a leg for this wonderful technology. But it didn't take into account that they are likely or could well be undercut when it comes to price by the Chinese. And therefore, if that's the case, the the pricing models holding up these valuations, well, they collapse. Do they not?
>> They're going to collapse. I completely agree. And uh if like a factor of 10 fall on the cost on the pricing is quite likely. And I saw one commentator saying that a lot of those data centers are going to turn into pickleball uh halls at some point, you know.
>> They make more money out of pickleball. It's very popular these days.
>> Be more fun at least anyway. But yeah, I I think this is a huge over investment. It's what typically happens in any transformative technology coming. And there will be over investment first of all, and a boom and bust caused by that cycle. But the double what's going to double this up is as you mentioned private credit.
Now, a lot of this people think private organizations give money to somebody else money that already exists. That's less dangerous than if those private private credit had then taken the money they pooled and then gone to a bank and leave it it up. Now, most likely they have done that with not getting any transparency on that at all. So, as well as uh the amount of money is vaster than is actually being saved to be reinvested by the Microsofts and the and the uh Facebooks and so on. Uh and it's also been levered. And that's exactly what gave us the for the global financial crisis, excessive private sector leverage. So, we're getting that again plus a a scale bust uh and at the same time as the unavailability of many of the physical inputs that are necessary, the strike of almost affecting the supply of of sulfur dioxide and therefore of cable of copper for cabling inside these data centers. So, it's in many ways it's it's the it's the biggest transformative technology challenge we've faced coming at a time when the resources, the availability of resources to produce the ordinary goods and services we expect are themselves in in challenge. So, I wouldn't like to be an investor in AI right now, let alone somebody who started one of these companies.
>> I can understand that. And I want to come on to the financial risk and financial consequences a minute cuz that is my next big topic. But before I do, let me just finish off on the China-American dynamic. I- If everything you say is largely true, uh but also I guess one should never underestimate the resilience of American enterprise, its ability to adapt, uh and to go in different directions. It's very agile. But even taking that into account, how where do you see things going in the next couple of years with AI in into the next decade, let's say, given the China-American dynamic that you've just outlined for us?
>> Yeah. Well, I expect most of the AI companies in America wants to go bust uh because the difference between their revenue and their cost is huge. I've been 5 to 10 times, cost being 5 to 10 times revenue. Remember remember the the internet bubble, you and I obviously lived through that one. And do you remember pets.com? Uh and I think it was valued at more than the the the entire pet industry in America. There were there were there were Amazon was valued at more than the transportation, you know, the the airline systems. Valuations get crazy and then they crash after the you know putting in the technology transforms society but doesn't lead the profit that the initial people who rush in expect. So we're getting that in on steroids this time round and we will see I think maybe you know one in 10 of the companies surviving but the wild card as we were just discussing now is that there's another country that's in this game in a way that America hasn't experienced before. So in the telecommunications bubble or the internet bubble occurred was exclusively American technology doing it. This time round they've done it, they've over invested, they've over borrowed, they're levered, they're going to crash and on top of that they might lose all the competition to China anyway.
>> Let me move on to the whole financial implications of what's going on. Put aside it let's supposing there wasn't a China problem to worry about. Um but that all all so let's freeze that for the minute and isolate the American AI even within its own ecosystem. Uh this massive this uh multi-trillion dollar investment in AI and associated technology and technology companies of course like all these things it's being done on debt. And particularly the new kid on the block compared with previous decades is this thing called private credit which you know I think four five six years ago most of us didn't talk about at all. It's a it's become a a new thing to take into account even if it was there before it's about scale. I emphasize in my monologue the that the private credit itself was a potential systemic risk in the financing of American AI. Do you agree with that?
>> Absolutely. Like this is why we we ignore the private sector. We we we get paranoid about level of government debt and never look at the level of private debt. Now I'm the exact opposite. I have come from the perspective of what's called Minsky's Hyman Minsky's financial instability hypothesis and that argues that it's the dynamics of private debt that cause booms and busts. And the public sector basically picks up the pieces afterwards. And that's why I saw the global financial crisis coming, cuz I looked at the mere level of American private debt. It was already about 150% of GDP in 2005. It was growing at up to 16% per annum. I said that growth rate of growth can't be sustained. When it turns negative, there'll be a financial crash, and that's what's happened. So, private credit coming from banks at that stage went from being plus 15% of GDP in 2006 to minus 5% in 2009. And that 20% of GDP turnaround in bank created credit is what caused the global financial crisis.
Now, what we're all talking about here is that a lot of large holders of large amounts of money, so the Microsofts and the and the Facebooks with enormous cash reserves, they've then pulled that into their own distribution of private credit, and they're lending out money that they've had idle in the sense on their bank balances. That itself is not as dangerous as bank money, because you're not creating new money. But I'm sure, knowing how the financial sector behaves, they saw a potential profit. They'd probably leave it up the amount of money that has been saved by these institutions. And so, you might I mean, the amount of money might be, you know, three or five times what we're seeing in terms of the private credit pools because of the leverage from bank loans as well, which are likely to be off off-book vehicles. So, you can't actually see them in the in the recorded data. So, I can we we're going to get a triple whammy. Overvaluations that mean you're going to have massive bankruptcies, and the bankruptcies cascading through the financial system. The loss of the money that's be these huge companies have poured in. So, the Microsofts and the and the Facebooks will find they haven't got the cash reserves they used to have. They'll try to find some way to recuperate those. And the the the the the loss of money that's going to come out of this, the bankruptcies that'll flow through the system, as well. So, we are facing a gigantic financial crisis. Uh and again, it's being made even worse by the straight of hormones and everything is happening in the Middle East.
>> Yes. Indeed, which drags everything down. Uh although private credit is is opaque, as you say, it's off balance sheet, it's in these special purpose vehicles, which are quite hard to to be able to penetrate, we are right in thinking that although it's not a majority of the debt going into AI, it is a a big enough chunk, if it goes sour, to cause problems. That private credit is a big player in the expansion of AI.
>> Absolutely. And and I they again, it's the enthusiasm This is part of what the capitalism's vitality comes from the things that also causes crises, because people see a potential enormous profit, and everybody who sees it dives in, and they get funding, then others get funding in the same area. So, you get massive over investment is a natural of the way a capitalist economy functions, and then you get a slump in the aftermath. So, that's that's the standard picture. Schumpeter covered that brilliantly back in 1907 in his theory of economic development. That's the best explanation for this boom and bust cycle in technological innovations in capitalism. But we're added to it again, because we have an enormous level of leverage at the private sector with private debt that was built up in the subprime bubble and the telecommunications bubble and the bubble before that. So, the high level of private debt, we already had fragility that way, because to service your private debt, you've got to be making a sale out of cash flow. Now, if your cash flow is negative, which is the case for these companies, you've got to go back and borrow again to remain in business. So, all these things mean it's we we've got a a Schumpeterian downturn on steroids hitting us with the AI sector. And the ultimate winners, as you say, potentially being China rather than America. There's going to be lots of bankrupt billionaires in America coming out of this.
>> Now, we talk about the opaque nature of private credit, and it certainly is, but it's not entirely opaque, Professor, is it? We do know, I I think, tell me if I'm wrong, that it actually comes from institutions that we know very well. It actually comes from banks. Banks have been using it as a way around the regulations. Uh private equity has been involved in it as well. You mentioned it, see, Microsoft and other companies are are are going in through the putting their money in through the private credit route. There is a kind of there's a kind of doom loop blowback danger here, is there not?
>> There is, and this this this happens all the time. This is one reason I emphasize we have to study far more the private financial system than we would do with the government financial system, because it's the private financial system that leads to these booms and busts. And by ignoring that, mainstream economics completely ignores private debt. By ignoring it, we're not saying what actually causes the real booms and busts. And this is a bit like a wake-up call coming along saying, "Hello, you've been ignoring me for two for 100 years. See if you can ignore this one."
>> Well, of course, we always think we can, cuz as we always know, this time will be different.
>> Hope [laughter] springs eternal in the human breast. You know that wonderful line? And that's And capitalism amplifies that, which is one of its great creativities, but it also has the great downside. And that we need to manage the financial sector and not allow these booms and busts to be as extreme. And when we're obsessed about government debt and don't look at the private sector, we take our eyes off what really matters. So, I hope this is a wake-up call to people to say, "Take a look at the level of private debt, both what the banks do, and what private non-financial institutions like the very well cashed up ones once they were well cashed up. Now they've blown all their money. So, the cash might get distributed, but we're going to see so much of a wreckage as a result of it."
>> [snorts]
>> So, you and I can sit here, and can see the structural imbalances. We can see the risks. Uh we can see the perhaps dark clouds gathering on the horizon. Um and you saw that in the run-up to the great crash in 2008. Others saw it with the dot-com uh boom. People saw it even with the great railway crash in 1873. But we saw it in the run-up to 1929. But
>> That's right.
>> But what triggers it? What what causes There I can see the dark clouds. What causes the thunderstorm? Talk us through how this manifests itself.
>> That's good question. Mainly it is that you take out you take on debt and therefore you've got to service the debt to be able to remain in business. So, when you're building the data centers and laying out the systems and so on, you don't actually have the cash flow yet. And so you you've got a timing issue. Can you get the cash flow to the level that's makes make your investment profitable before you accumulate too much losses in the meantime. Now, what happened in every of those bubbles you've mentioned, the time element broke. So, like in 1929, uh people don't realize but the down the downturn in private credit based demand in America began in 1927, 2 years beforehand. So, suddenly the cash wasn't turning up to finance all the share market purchases. People were running up losses servicing their um margin debt at the time. They could get away with it until suddenly people were forced to sell and then bang, when some people are forced, the whole system cascades and collapses. So, we really need to have a a a overview of the financial sector and not let these bubbles get as extreme. And what happens after every one of these crises, mainstream economics comes out and basically sort of tells us, "Don't worry about private debt. Get obsessed about government debt as well." We take our eye off the private sector and they do the damn same damn thing again because, you know, hope springs eternal in the human breast. And the the bankers all dive into exactly the same obsession at the same time. One of my best friends in this whole area is a economist and financier called Richard Vague. I'd love you to have a talk to Richard as well. He was a banker in the in the in the 1980s in America in Texas and saw the massive investment in oil rigs at the time and then when they when all the banks did it, they all price collapsed from $40 to $10. His company his bank collapsed. He bought out the consumer wing and became a very successful banker and he said this obsession of driving into the what's currently hot is a natural thing for bankers and unless we restrain it, we're going to continue getting booms and busts.
>> I'm going to ask you it's probably an impossible question, but I'll ask it anyway. When does the bubble burst?
>> I'd say the next 2 years. I I can't be more precise than that. You you can say it's going to inevitably going to happen, but as to when timing is always the hard bit, but certainly I think the next 2 years it's got to come crashing down.
>> And when you take together the China competition, the China threat if I can put it that way, the private credit risk we've been talking about, the energy crunch that will probably stop a lot of the even if they've got the money, will stop these data centers from coming online. Add that that all up, it's a kind of triple whammy of bleakness, is it not?
>> It is and that's the fourth way we we we're not spending the ridiculous war on Iran and and Ukraine for that matter. We're damaging our physical facilities at the same time as we're causing a financial bubble. So that that's what we actually worries me in many ways more than just the AI in terms of the most immediate crisis we face because we're destroying so many of the channels of producing goods and services that are necessary to service the debt that private enterprises have. So there's going to be a whole range of companies who suddenly can't sell the products they need to sell to service their debt and we'll get bankruptcies popping up all across the system and they're going to precede what happens with AI.
>> Mhm.