Is AI About to Crash the Global Economy?

0:00David McWilliams: Today's podcast is all about the AI boom. Is it a boom or is it a bubble? We go back in history to 1873, to a massive railway investment-linked collapse in the American stock markets. We say, is that going to happen again? It's AI, it's America, it's a bit of history, and it's lots of futurology. It's all coming up in a couple of minutes.

0:30your 67th security blind spot. The good news? The Vanta agent works like a GRC engineer in the background, finding every app your team uses, scoring the risk, and drafting fixes for you. Vanta is the platform used by over 16 ,000 fast-moving companies like Ramp, Cursor, and Harvey, who are shaping the future with AI and staying ahead of AI risk. Get started at Vanta.com. Shipping, billing, admin, payroll, marketing. You're managing all the things. So why waste time sending important documents the old-fashioned way? Mail and ship when you want, how you want with Stamps.com. Print postage on demand 24-7 and schedule pickups from your office or home.

1:10Save up to 90 % with automated rate shopping. That's why over 1 million small businesses trust Stamps.com. Go to Stamps.com and use code podcast to try Stamps.com risk-free for 60 days.

1:25David McWilliams: To understand the economy, you have to understand human nature. This podcast is powered by Acast. How you doing there? How is your midsummer ahead? Are you still baking away in Balmy, Dublin, Cork, Galway, Belfast, Limerick, Waterford, wherever you happen to be today? Today's podcast is all going to be about the risk in the United States of a significant turnabout of events due to the overinvestment of the Americans in AI technology, which promises to change the future, but has yet to deliver any real returns for the people who put their money in. The question is, can AI deliver before the companies that borrowed huge amounts in order to build the AI infrastructure need to pay that money back?

2:23David McWilliams: We're going to look at history. We're going to look at the future. We're going to look at the present. And we're going to make a little connection with Ireland because you'll have seen that in the last couple of days, Intel has announced a big increase in its investment. They're looking for tech bros, etc., and tech professionals in Leak Slip. However, this is all part of the same idea. There is a massive, massive, massive surge in demand for all things technological, for all things chip-based. The question is, is it all sustainable? We're going to take you on a little journey to history. we're going to look at the present and we're going to have a stab at the future.

3:10David McWilliams: Johnny boy, how are you? I'm good. I'm good. I'm enjoying my nice little break from the World Cup. It means that I get to sleep at night. I know, I know. Dear listener, poor old John, from the very beginning, from the very beginning. I am a super fan of the World Cup. I didn't think I would be actually, Mac. I was kind of secretly hoping, like an awful lot of people, that the World Cup wouldn't be as successful. But it has been hugely entertaining, that's for sure. And also hugely controversial. But that's part of the joy and excitement of it. Yeah, no, I get the controversial stuff. I get the controversial stuff.

3:47David McWilliams: But dear listeners, you're right, I just confided in you, John has an addictive personality and is now emerging in his addiction to late nights watching television from the 12th of June until today. Yeah. He hasn't got a wink of sleep. And now he is delighted that the World Cup is winding down so he can get a little bit of kip. But like all Irish people now, he's complained he can't sleep for the heat. That's right. Exactly. So talk about heat. You're off in Europe. Where are you this time, Mike? I am in Slovenia. I am in Ljubljana, the capital of the beautiful, almost Alpine Republic of Slovenia.

4:24David McWilliams: was one of the breakaway states in the former Yugoslavia. The Slovenes always regarded themselves as quite different to their Croat, certainly their Serb and Bosnian neighbours. And they have turned this little place in the last 30 years into, in effect, a small version of Switzerland. It looks the same geographically, topologically. It's got mountains, lakes, etc. It's got the Julian Alps, John, that the spine of the Alps, the coming down. And it also has, and this is the amazing thing about this place, I think I might have said it in the podcast before, I think we did a podcast from Slovenia about three years ago, it has a sense of central European industry that is not shared by its Balkan neighbours down the road.

5:13David McWilliams: What do you mean? Every small Slovenian town seems to have a little fabricat, fabricat, a little place, a little manufacturing place. Almost all Slovenian houses seem to have little, almost tool sheds in the back where everyone's making things. The farms are impeccably maintained. The barns, John, I've got into barns, right? I didn't realise I might have a barn fetish. I swear to Jesus, I didn't realise I had a barn fetish, but I am now walking around rural Slovenia. I'll do that now in the next couple of days where I was yesterday, marvelling at the impeccable state and upkeep of the barns in this part of the world.

5:55David McWilliams: But what I mean by that is that they are an industrious race. They are a Germanic Austrian sort of race. They are very good in the same way as the German economy was maintained by what the Germans called Mittelstand, which are small family-run businesses that actually have become, in the German case, big family-run businesses. you get the same sense here in Slovenia that it's very, very well grounded. And I'm waltzing around Ljubljana, John. And Ljubljana is a absolutely beautiful, beautiful city. If Dublin wanted to... So I believe. If Dublin wanted to take a leaf out of any book in terms of getting rid of cars, updating the infrastructure, making sure that the old buildings were maintained, making sure there's no vacancy, There's no dereliction.

6:47David McWilliams: Ljubljana would be the place to come. Small, it is manageable, it is walkable, it is an impeccable little city. And here is where we're going to kick off the podcast, John. And I'm sure it's full of Irish students as well, because Ljubljana is definitely on the Eurorail destination list. And I know that Ruby, my daughter, ended up there and Lake Bled and stuff like that last year and had a blast. No, it's a great place and there could be a hell of a lot of our listeners here because we have loads of listeners amongst the Irish student population. So who knows who I'll bump into later on this evening, John.

7:25David McWilliams: Yeah. While I was here, I was struck by the architecture of the place. Right. I was here about 30 years ago. And then because it was post-Yugoslavia, post-communist, post-socialist, and it was during the war that was raging down the road here between Croatia and Serbia and Serbia and Bosnia, The place obviously needed a lick of paint. And now that it has had that lick of paint, what you see is this amazing sort of neoclassical architecture built by the Austrian-Hungarian Empire. And you kind of forget that this was a big city in the Austrian-Hungarian Empire. Right. And in fact, the Slovenes were an integral part of that particular empire and very much part of the, I wouldn't say the upper crust, but, you know, part of the middle classes of that empire.

8:15David McWilliams: And, of course, their capital city, Ljubljana, reflects that integration that they had with the Austrian empire, the Hungarian empire, over the years. So Slovenia's kind of put a lot of investment into upgrading and giving a lick of paint and all that, and making it look kind of nice. Huge. The place was amazing. It was already a beautiful city. It was already a very, very well-planned city. It was already a very intricate city. It has got a huge amount of very, very adventurous architecture. Obviously, the bridges over the river here are what it's known for, but they put massive efforts into making it walkable and making it livable in the last 20 years.

8:54David McWilliams: And that has profoundly changed the experience of this city, the feel of it. You can walk around everywhere. Dublin could take a leaf out of the Slovenes' book then. Completely. In fact, you know, this new government task force, which is charged with redeveloping Dublin One, could well, but, you know, the problem with taking a leaf out of John, in Ireland that usually means a whole bunch of public servants and county councillors going on a jolly to a place like Ljubljana, right? And they go and they get radarsed here, they come home, they then commission somebody at great expense to write a report, the sort of report that we're giving them on this podcast for free, and then that report constitutes the research that they have done and that then sits in a file and nobody ever hears of it again.

9:43And basically, these fact-finding missions are just a jolly paid for by us. Yeah, they leave their notes in the pub.

9:50David McWilliams: They leave their notes everywhere, yeah, exactly. So rather than do that, just send one person over and say, yeah, we're going to do that and get the people from here who did that and bring them over to Ireland and say, what should we do? And do it. But anyway, that's neither here nor there. Okay, talk to me about Ljubljana from the Austrian Empire perspective. Well, I want to link here. The Austrians in the 1850s, 60s and 70s built enormous amounts of architecture, buildings, railways, linking the entire Austrian Empire together. and I want to end up talking about AI, John, but I want to frame it all in the context of Central European and American railways, the railway expansion of the 1850s, 60s and 70s.

10:44David McWilliams: What happened there? And are there any lessons that we can learn from the extraordinary American expansion in AI? The huge sums of money that have been put into AI because the same sort of effervescence existed 150 odd years ago around here, which was that railroads was the technology of the future. The railways would open up huge parts of Europe and huge parts of America to trade and commerce. They would link previously unlinked cities together. They would create the network. The network would create what they call cluster effects. and everything from culture, art, industry, manufacturing would flourish and would be gelled together by railways.

11:32David McWilliams: Now, the upshot of that was that the perception of railways being risky dissipated dramatically because people said, how can you possibly say these are risky because aren't these going to transform the way we live? And in fact, railways did transform the way we live, so much so that things like distances, time, all these sort of things changed, our perception of these things changed dramatically because of the railways. But on the 9th of May, 1873, John, I know it's an auspicious day, very few people noticed that the stock market in Vienna, which had been the source of finance for all these railway and architectural and development initiatives, which the Austrians had become addicted to, the stock market began to wobble.

12:29David McWilliams: Now, most people thought, well, this is a kind of an adjustment in a country, an empire, that has been going gangbusters for a couple of years. And you know what? It's not a big deal. So this is just happening in Vienna? Or was it a more global thing? No, it's just Vienna, Budapest, Ljubljana, and Prague. These four major cities of the Austrian Empire. So the stock market begins to wobble on the 9th of May, 1873. Most people outside, they said, don't worry about it. It's grand. It's not going to affect us. In fact, they've been on a bit of a bender. And this is really a, as the American banker said at the time, this is a normal adjustment.

13:12David McWilliams: the sort of thing we hear all the time when there is... Soft landing, normal adjustment. Soft landing, normal adjustment, all that sort of thing. And, you know, basically, not to worry about it. Until four months later, John, in early September, when these financial troubles seemed to migrate from Austria to New York, and stock prices started to fall in New York. And in particular, a company, which was maybe one of the biggest companies in America, finance companies called Jay Cook & Co, preeminent with offices in New York, Philadelphia, all those things, that went bust. It was like the Lehman Brothers moment of the time.

13:56David McWilliams: And what happened was people couldn't understand because Jay Cook had made his fortune in lending money to railroads. And people thought, hold on a second, aren't these railroads almost without risk because they are actually the future? But what happened was once Jay Cook, this huge bank, like JP Morgan, like, you know, one of those massive, massive banks, when that went bust, it had a Lehman effect. And suddenly people started to think, wait a second, the psychology of punters changed dramatically. And investors, big and small, said, well, if Jay Cook can go bust and its bankruptcy is related to a crisis in Vienna, which is miles away.

14:40David McWilliams: Is anybody's money safe? By the 20th of September, John, of 1873, European investors, German and Dutch and British, started to take their money out of America, out of American railroads, because the act of contagion, like the virus, like COVID, the same sort of way, maybe COVID came to China and then to Italy, and people said, no, it's in Italy, but will it come here? well, maybe not, it won't. And then before you know it, it's everywhere, right? That same sort of physical virus manifests itself emotionally and psychologically in financial markets. And what you have is a full-on crisis. So much so that the city of London, which had been the central transmission mechanism of all finance around the world, by 1876, which is three years later, This shows the depth of the crisis.

15:35Lending from the city of London around the world had fallen by 80%.

15:41David McWilliams: So you have all emerging markets in crisis. You have all railroads in crisis. You have a stock market crisis, all of which came seemingly out of the blue. And I'll just leave you one last little footnote here, John. the two countries that experienced the biggest ramifications, political and economic from this, were the Ottoman Empire and Egypt, which was a protectorate of the Ottoman Empire. Because both these countries had borrowed enormously in the 1860s and 1870s to modernize first their military, then their railroads, then their infrastructure. They were the ones that paid for, remember we talked about the Suez Canal?

16:25David McWilliams: all that sort of stuff, right? They went bust. And it's probably not a massive stretch to say that the source of the end of the Ottoman Empire, which ended in 1918, was the 1873 crash. Right. That just put the wheels in motion. It put the wheels in motion. It changed the whole world. We can talk about it. In actual fact, I'm reading at the moment a book about the crash by a guy called Liyakwat Ahmed. Great name, wonderful man, won a Pulitzer Prize winner. for his book about the central bankers that caused the Great Depression. He's got a new book out called 1873. That's what I'm reading as I meander around Slovenia.

17:05David McWilliams: But what it really signals to me is the similarities between then and now. And the 1873 crash, how does that compare? I know totally different times and eras and all this, but how does that compare with like the 1929 crash or the 2008 crash? Well, fascinatingly, and we will leave it here, but we'll go on, because it's a really good question, and we'll probably require an entire podcast on its own. The reaction to this 1873 crash, which actually most people now have forgotten about, but it was the financial experience that dominated people's minds until the First World War. The reaction to it, and this is interesting, the governments at the time in France, in Germany, in Britain, etc., they decided that what the financial system needed was stability.

18:04Yeah.

18:04David McWilliams: And they would require countries to hold gold and tie their exchange rates to gold in order to bolster that stability or guarantee that stability. Were they not already tied to gold? No, you see, lots of them were tied to silver. So this is when the gold standard emerges, right? The problem with the gold standard, as I've always argued, is the gold standard is inherently deflationary. Why? Because it stands to reason, if your money is tied to silver and there's more silver around, then if you tie it to gold, there'll be less money around because there's less gold around. So by definition, it's deflationary.

18:42David McWilliams: That's the first thing. And the second thing is a deflationary environment always and ever benefits creditors, not debtors. Because debtors want inflation to inflate away their debts. Now, that will have switched the body politic of Europe over a period and will have enraged the middle classes of Europe over a period. Why? Because the middle classes were the people who took out the debts that caused the boom and the bust in 1873 in the first place. So there's a huge amount of fascinating political, economic, cultural stuff that stems from that. So to answer your question, I get the impression now with the benefit of hindsight that the 1870 crash was even more debilitating than 2008.

19:31David McWilliams: And as debilitating as the Great Depression that came after the 1929 crash. The reason I want to talk about it, John, is I'm looking at the numbers of how much money has been spent on AI now in America. And what I see is an entirely irrational boom grounded in a rational idea that AI will change the world, but driven by extraordinary exuberance and irrationality, which is driving people to do mad things. And those mad things are being put down as investment when in fact lots of those things are speculated. What do you mean by that? Give us a few numbers to put it into perspective. So total investment in the United States in AI last year was$256 billion.

20:27David McWilliams: That's up from$95 billion in 2024. It could be by the end of this year as much as another$650 or$800 billion spent on AI infrastructure and AI betting on the AI future, right? And this, according to Forbes, could go as high as$1 trillion by 2027 and$3 trillion by 2035. Now, AI spend is on track to be more than 2 % of total US GDP this year. To put that in context, higher education, all of higher education in the United States is 2.3 % of GDP and defense is 2.9%. So AI spending by private companies is now on track this year to be close to and clearly in the years ahead outspend the US defense budget, which is phenomenal.

21:28And this would be on a par of the amount of spend on the railroads back in the 19th century.

21:36David McWilliams: Exactly. So what we're looking at is the rerun of the railroad boom. And the fascinating thing was the railroad boom and the bonds that were issued by railroad companies to build the infrastructure. Because they said at the time, as they say now about AI, it's so expensive to roll out the infrastructure, we need to borrow phenomenal amounts. And people got overexcited and overexuberant. So it's exactly the same dynamic at play, except for the fact that the concentration in AI is even greater than the concentration that was the case in the railroads. Because there were many railroad companies in terms of the AI boom, there's only a few companies, right?

22:21Yeah.

22:22David McWilliams: So you're talking about your Microsofts, your Oracles, your NVIDIAs, your Open AIs, all these companies, these are the ones that are spending the absolute fortunes. But that's not the only problem, John. There is a dilemma of circular financing. So NVIDIA invests in OpenAI. OpenAI uses the money to buy NVIDIA's chips via Oracle or Microsoft as the intermediary. So think about this. They're all related together. Oracle borrows to build these big data centers that it's hoping that OpenAI is going to rent from them. So Oracle's ability to repay depends on AI or OpenAI paying its cloud bill via the data centers, which in turn depends on OpenAI's own funding round closing.

23:12David McWilliams: So think about it. All these companies are investing in each other, getting dependent on each other, all of which is dependent on there being money, infinite amount of money at the end to buy up all this stock and buy up all these bonds, particularly bonds now. So financing has switched from equity and shares and private equity and private money to the bond market. It seems to me exactly the same as the railway boom. If you look at what happened in the 1850s in America, it was privateers, came in and they said, we'll finance it. Then by the 1860s, it became, and certainly after the American Civil War, there was punters and it was bonds they were buying because they didn't like equities because their equities were too volatile, so the civil bonds are long-term places, it's fine.

24:04David McWilliams: And you get to a situation by 1873 where everybody's leveraged on the mirage of the railroads. Yeah. And you get a small prick to that bubble thinking, which started in Vienna. And before you know it, one of the great American investment houses is bust. And before you know it, 80 % of global finance that used to go out of the city of London has stopped. And that's the risk, John. So here we are again with another bubble, an AI bubble. But can I just ask you quickly about circular financing? Because it seems really complicated to me. But are we talking about real money here? Like real money changing hands?

24:44Or is it just kind of little arrangements and value is being just passed on via services or whatever.

24:52David McWilliams: No, it's real because each company is investing in each other, parts of each other's infrastructure in order to create the overall ecosystem. So it means that the ecosystem from the data centers to the chips, to the cloud, to the original financing, all are interlinked, all are interwoven and therefore all are interdependent. And when you get that level of interdependency, what you do is, as Nassim Taler would say, you increase the level of potential fragility because everybody's dependent on everything else. And they're all borrowing from the same people as well, from the same banks. Borrowing from the same people and now, of course, borrowing from the bond market and the bonds are going to be issued to American pension funds.

25:39David McWilliams: They're going to be issued to all sorts of significant pillars of the American financial system. So at the beginning of this, it was just basically private equity, high net worth individuals, rich people. But now as these IPOs ramp up, as the bond salesmen ramp up to sell all the bonds, where they're going to go is to ordinary investors. And the speech is going to be the same. You are buying into the future. You are buying into the next generation of America. And of course, so... But what they're actually doing is offloading the risk. They're offloading the risk. Exactly, John. And then, of course, there's one final piece of the jigsaw, which is the following, which is that the price of using the AI models charged by open AI, for example, needs to rise in order for open AI to generate the revenue to pay back the debts.

26:39David McWilliams: Yes. The assumption is that all of us will use AI in some shape or form. All of us will pay for it. All of us will pay a hefty price for it. And that hefty price will make everybody's borrowings make sense because a huge amount of revenue is going to come into the AI companies and then they're going to be able to pay everyone. Now, of course, because a lot of the tech companies are involved and the tech companies don't know what to do with their money, there is a presumption that this will always and ever be the case, that there'll always be a surplus of cash. But this is not the way economies work.

27:16David McWilliams: Economies work usually in a very unusual way, which is that something small can tip the balance, particularly in a highly leveraged situation. But I mean, I do think that, you know, you get a company like Meta, whose credit rating is now better than the American governments. They pay less for their debt, the American governments. So there's a counter argument, which is a lot of these massive tech companies are so well sourced and funded that it would take a monumental crisis and a monumental disappearance of liquidity to knock this particular house of cards down. But be that as it may, the direction of travel appears to me to be so obvious.

27:56David McWilliams: And then one final difference is the following. You can say it in your Trump accent, China. China. Exactly. So despite all the investment of US tech companies in AI, China owns six of the 10 most popular AI models in the world. Okay, yeah. So the US was initially the leader in AI. We know all that. But then the Chinese announced DeepSeek. Major AI company released the flagship model a year ago called V4. But the most interesting thing, John, and this is where I think the risk to America lies most profoundly, is the Chinese are open sourcing their AI. What's the impact of that? So basically what the Americans are saying is, open AI will scour the world for technology graduates and brainy technologists.

28:49David McWilliams: It will then pay them an absolute fortune. They will come into a closed system and they will make the next iteration of AI models for open AI in a private closed source. Like a normal private company. What the Chinese are doing, subsidized by the Chinese government, is the Chinese are open sourcing their AI. So what they're basically saying is that anyone can build it. So they are turning the world, but particularly Chinese tech, into an army of engineers who are doing this for the good of the world. So this is, as they call it, democratizing AI, which is ironic coming from China. Now, it's not ironic if you think that China's idea is to undermine the states at every iteration.

29:40Yeah, yeah.

29:41David McWilliams: So what the Chinese are doing is they're saying, if we open source AI, it means the price to the consumer is going to fall dramatically for Chinese open sourced AI models and therefore AI product. The world has no loyalty to American tech. In fact, the world doesn't like American tech. And if a Chinese AI model was infinitely cheaper, and even 80 % as good, people would use the Chinese model. In the same way as people are buying Chinese cars now. Not because they're better than Mercedes, but because they're cheaper than Mercedes. So what the Chinese are doing, they're playing that great geostrategic game.

30:24David McWilliams: So as the Americans get into more debt to finance closed source AI, to make people like Sam Altman really rich, and to allow him hang around at Davos, the Chinese are doing something quite, quite different. They're saying, let's use AI as yet another front in our war against the United States. And what we can do by open sourcing the AI is we can guarantee that it's actually going to be made cheaper. And we can therefore make sure that Chinese AI companies and models are cheaper and made universally available to the third world, not to the first world, to the global south, as they call it, right?

31:08David McWilliams: An expression I don't like. I must admit we can come back to that. Okay? Okay. That, John, is the interesting geostrategic vulnerability of the United States right now, which is the very technological creature, the invention, that is supposed to make America strong and to lock in America preeminence for the next generation. May, with this combination of the bubble a la 1873 and the fact the Chinese have a number on the Americans in disregard by open sourcing the AI production, this could well be the very thing, John, that makes America not strong, but weak. Yeah, you know, the funny thing about AI and this whole AI bubble for me is the fact that I never asked for it.

32:05And it's kind of been rammed down our throats and it's kind of this is the future and all the rest. But it's also hugely unpopular because, you know, it's having a massive impact on employment, particularly first time jobbers, huge layoffs across all the tech industry. So what are the implications of the future of AI for the global economy? But before you answer that... I love that. Small question. Just a little one. Oh, just ask ChatGPT. All right. Talk to you in a sec.

32:54businesses have skipped the line with stamps.com. Join them to save up to 90 % off carrier rates from your computer or phone right now. Print postage for certified mail, registered mail, and packages in seconds. Then schedule a pickup right from your home or office. For a limited time, go to stamps.com and use code podcast for a free welcome gift. Taxes and fees apply. Your team just added its 67th AI tool and also your 67th security blind spot. The good news? The Vanta agent works like a GRC engineer in the background, finding every app your team uses, scoring the risk, and drafting fixes for you.

33:29Vanta is the platform used by over 16 ,000 fast-moving companies like Ramp, Cursor, and Harvey, who are shaping the future with AI and staying ahead of AI risk. Get started at vanta.com. We all belong outside. We're drawn to nature, whether it's the recorded sounds of the ocean we doze off to or the succulents that adorn our homes. Nature makes all of our lives, well, better. Despite all this, we often go about our busy lives removed from it. But the outdoors is closer than we realize. With AllTrails, you can discover trails nearby and explore confidently with offline maps and on-trail navigation.

34:08Download the free app today and make the most of your summer with AllTrails.

34:17So, Mike, the big promise of AI was a beautiful, big, bright future for everyone. But in actual fact, it's been rammed down our throats. It's cutting jobs, particularly for the first time jobbers. And it's having an impact on pretty much every aspect of our society, including distorting politics. and at the risk of being accused of being a Luddite of some sort, it's not that welcome in certain areas and is becoming quite unpopular. So what is the implication for the future of the global economy?

34:53David McWilliams: Well, I think the first thing to consider is this unpleasant fact that people never asked for technology. Nobody asked for the printing press, right? Nobody said, well, let's have a plebiscite in Mainz in 1453, and let's see whether or not a fellow called Gutenberg can come up with a printing press, you know, the steam engine, all those things. So the interesting thing is nobody ever asks for technology, and nobody mandates technologists or entrepreneurs to actually make the stuff. So that's the first thing. That's fair enough. And I think it is the elixir of invention, that once something is invented, it unbalances people because the whole point of the invention is to make something more efficient or less efficient or more innovative or less innovative or worse, to use that awful management expression, to take out cost.

35:51David McWilliams: Taking out cost, John, is basically employment. Yeah. Right? Because people are 62 % of the cost of any business, right? like predicting normal businesses. So the first thing to appreciate is nobody ever asks for technological change. It is part and parcel of the Schumpterian process, John, of the gales of creative destruction. It is what the capitalist economy is about, is technological innovation. So that's the first thing. But the second thing, the point you make is right, is that in democracies there is a reaction. Now, I don't know if you've seen in the last few months the way in which American students and graduates have booed lots of people who have gone up and said, get used to AI, right?

36:36David McWilliams: Because they understand that particularly entry jobs have been hit dramatically. There's apparently junior hiring has fallen by 8 % in the last six quarters, right? And Oracle, for example, one of the largest tech companies said that AI cost the company 21 ,000 jobs in the first six months. That's very, very brilliantly put. That use of language is appalling. It didn't cost the company. Larry Ellison is the person who's actually orchestrating at the top, who's worth 202 billion. And he is the person who is actually using AI to cut those jobs. Lots of the money will flow up into his arse pocket, right?

37:18David McWilliams: Yeah. But I think that Forbes, again, in June, showed that AI was responsible for an estimated 38 ,579 job cuts in May alone in the United States and 87 ,714 in the year to date. Meta itself laid off 10 % of its workforce in April. You'll have known that TikTok as well laid off lots of people in Ireland. So the optics are not good for the people who are shouting and screaming about AI. In fact, a Molotov cocktail was thrown, I think, through the window of Sam Altman's apartment. Yes. Not that long ago. So, all of these suggest simmering anger towards AI and companies, right? Overwhelmingly, Americans in the latest Pew survey believe that AI will have negative effects for themselves and for society.

38:14David McWilliams: Now, these are huge numbers, John. That's the first thing. The second thing is the impact of AI on data centers and the impact of data centers on electricity prices. So what is happening is that part and parcel of all this huge expansion of AI is the expansion of data centers. Data centers are unbelievable users of electricity. At the moment, all across the world, data centers are simply plugging into the national grid of the country. but they're using a huge amount more electricity. This is pushing up prices for Joe Punter. And Joe Punter is saying, well, hold on a second. Not only am I not benefiting from this increase, but I might actually be paying for my own redundancy.

39:00David McWilliams: Yes. So, you know. I've lost my job and now I have to pay for more electricity. Exactly, exactly. So what I would say, John, is that there are many historical incidences which seem to me to be quite similar. our 1873 moment is, I think, very instructive. Dear listener, if you could bear with it, we might even do a series on the 1873 crash once I have finished with this wonderful book. And I think it's a fascinating and incisive look at what happens when you overlook the risk in large technological investments and people get too optimistic. and that optimism not only feeds into investors and entrepreneurs and business people, but actually feeds into the state itself.

39:52David McWilliams: And certainly the state itself begins to deny the fact that there's maybe a risk out there. Yeah. Last thing I'll leave you with, John, of course, and then we have, as you said, the human backlash, the democratic backlash against these companies. The fact that AI and its leaders seem to be very wealthy people lecturing reasonable people about how they should live their lives and how they should adjust to this great technology, which answers your question, is who wants it in the first place? But I think what is much more likely to occur, John, is not a democratic or street-level protest against AI stopping it.

40:32David McWilliams: I think what's much more likely is that we will see a market reversal in AI, not unlike the 1873. Now, will it happen in the immediate future? I'm not too sure, but I'll leave you with this idea, John. Summer is an unusual time in financial markets, right? Because everyone goes on holidays, markets are thin, there's very little liquidity, and things that typically might have happened at a different time of the year happen during the summer and end up manifesting themselves in September or October. The Russian crisis, 1997-98, which was part of the Asian crisis. Exactly a summer event. The Gulf War, a summer event of 1990.

41:21David McWilliams: The American recession of 2002, also a summer event. 2008, a summer event. Even looking back, John, 1928, the S &P started to fall in September, right? Lehman's collapsed in September. My point is, John, the period between August and October in financial markets is always the most dangerous. Why that is, I have no real idea other than to say that these things have happened. The question I want to leave you with and the listener with is will that period, which usually is the harvest period, August, September, October, it's a buoyant, positive, we have sowed, we now reap, all that sort of stuff.

42:13David McWilliams: Will this 2026, the harvest season between August and October, be the one that is remembered for the end of, or the beginning of the end of, the world's biggest, biggest technological boom, which is the AI boom? Questions on the back of a postcard. Or answered by ChatGPT. Exactly. Or DeepSeek, I'm not sure which. Whichever one is left standing. Whichever one's left standing. It's a question we will ask and we will answer at Kilkenomics on the 6th of November, just after the Halloween break. Anyway, we will talk to you on Tuesday.

43:13Shipping, billing, admin, payroll, marketing. You're managing all the things. So why waste time sending important documents the old-fashioned way? Mail and ship when you want, how you want with Stamps.com. Print postage on demand 24-7 and schedule pickups from your office or home. Save up to 90 % with automated rate shopping. That's why over 1 million small businesses trust Stamps.com. Go to Stamps.com and use code PODCAST to try Stamps.com risk-free for 60 days. Your team just added its 67th AI tool and also your 67th security blind spot. The good news? The Vanta agent works like a GRC engineer in the background, finding every app your team uses, scoring the risk, and drafting fixes for you.

43:57Vanta is the platform used by over 16 ,000 fast-moving companies like Ramp, Cursor, and Harvey, who are shaping the future with AI and staying ahead of AI risk. Get started at Vanta.com.