podcasts / Best of the Spectator / episode
Best of the Spectator cover art
podcast
📅 2026-11-16 ⏱ 32 min episode 📝 Transcript · 5,513 words

Reality Check: is the AI boom a bubble – or the next big revolution?

Best of the Spectator · · Transcript & summary by WhipScribe

Read the transcript Episode page ↗
🔔 Get every new Best of the Spectator transcript by emailOne email per new episode. Unsubscribe in every message.
One email per new episode, unsubscribe in every message. Not shared with the publisher.

Summary, chapters and key takeaways for this episode are on their way; the transcript below is complete.

▸Transcript

Our own transcript of the public episode, with a timestamp about every 90 seconds. Tap a timestamp to play the episode from there in the show's own player; the link button on a paragraph copies a link that opens this page there.

0:00Hey, I'm Farnoosh Tarabi, host of the So Money podcast. One thing I've learned from talking to entrepreneurs, your time is one of your most valuable resources. With Google Workspace with Gemini, Google's AI is built right into the premium versions of tools like Gmail, Google Docs, and Sheets. No new tools or extra logins, just AI to help you work smarter. Because less time spent on busy work means more time building the business you believe in. Get AI tools for work. because less time spent on busy work means more time building the business you believe in. Get AI tools for work. Upgrade to Google Workspace with Gemini at goo.gle slash workspace with Gemini. Is the AI bubble about to burst? Are we even in an AI bubble?

0:46That seems to be the question everyone in markets is asking this week. Well, to answer that question, I'm delighted to be joined by Giulio Bertini, who has written a book about 10 moments that defined markets. Giulio is a former rates trader for the ECB and now writes about markets in his Andromeda reports. Delighted to say that Julio joins me on Reality Check now. Julio, thank you for joining Reality Check. Now, we quite often have friends of the show on this podcast, but we don't often have friends of the host on the podcast. And I should declare that you and I went to university together. We used to live together in Edinburgh. And from what I know about you from when we were flatmates, I'm slightly

1:30surprised that you have produced a book, which I'm holding up for the cameras to see here. So why did you write this book? We'll get on to the book itself. Why did you write this book? And what have you been doing since we graduated university? Well, first of all, thanks a lot for having me on the show. It's a real pleasure. So the idea of the book itself came from a passion of history. As you probably remember, I've always been interested in history and the history of economics. And I was discussing at a dinner table earlier this year about comparisons of the dot-com era with the current AI era today. And by discussing this as well as the energy crisis today and its similarities with the energy crisis of the 1970s from a conversation like this,

2:23I thought what would be interesting is to tell these events that we just discussed at the dinner table over a short book. After university, I began working for the European Central Bank and I spent a couple of years working on the quantitative easing program. So designing and implementing... Printing money. Yeah, printing money, printing money basically buying uh scooping up government bonds and corporate bonds all across europe and that was really interesting especially because it was at the time of covid and that was the time when central bank's balance sheet were increasing and um and so it was it gave me the opportunity to learn about bond markets something which i then carried on onto the next experiences,

3:05which were in bond market trading, first for a large investment bank and then ultimately for a hedge fund based in New York, but I was working from the London office. And I was trading interest rates and government bonds. And that passion for markets combined with a passion for history is what really led to the book so the book didn't come out just of a passion for history but it's it's from a work experience in the interest rate and government bond sector combined with a passion for history that eventually led to the book um and i thought that when i was in hedge funds or in a bank, things were getting explained to me in a difficult way. But so my view for the book is try to simplify everything as much as I can.

3:57So take episodes like the dot-com bubble that we will discuss later or take episodes like the 2008 financial crisis. Everybody knows about these events, but not many people probably know the details of these events very much. I thought that explaining these moments in a simple way could be engaging and bring more people to the economic space or the history space behind these events. So you mentioned how the book is about moments that defined markets. And this is what your book is. You've picked 10 moments. Yes. And I wanted today, as you've alluded to, to talk about a specific chapter, which you've called Warning, Internet Companies Are Running Out of Cash. Yes. And in that, you tell the tale of the dot-com crash.

4:48And we'll get into it a bit later on, about whether there's similarities with what's going on with AI and chip makers and stuff today. But for the uninitiated, what was dot-com mania and what was the dot-com crash? Yeah, so first of all, the title is not, I didn't make up the title the title itself was the same title that appeared in the cover of baron's magazine and that eventually led to the first crash um so i thought that was interesting to use as a title for the for the chapter but essentially the story goes along the following lines um in the mid-1990s, you had the creation of the internet. It was something, it was a technology that was very new to people. And people were left wondering,

5:34what can we do with the internet? At the time, it was not much. It was buying books online, it was reading the news from all over the world, But it was very simple and rudimentary things. And people were dazzled by the wonders of the internet to the point that they were researching different companies and buying them. And there was a bit of a hype around these companies because people thought, oh, what the internet could do tomorrow could be potentially very profitable and um and so you started having um these companies that were um advertising themselves as internet companies uh one of being netscape next netscape is a is a really important example because it was one of the first biggest iposOs of the dot-com era in 1985.

6:25And Netscape made sort of the original web browser. Yes, exactly. Netscape was the, think of it as a Google, but it's not a full-on search engine like Google. But it acted a bit of a Google back in the day. At the same time, this has to be understood in the context of more people from home had access to online training platforms uh whereas this was something that is that was not the case in the past well this is this is a really interesting point because there's two things i picked out from how you've written about dot com one is sort of just like the mania of this is the new fangled thing yeah and i mean there's a remarkable remarkable statistic you point to where any company that literally had .com in the name

7:09would see like a price benefit in terms of their share price, which shows you how much it was almost like, wow, this is the new fangled thing. But you also point out that this is the same time where you could buy stocks online. People will be familiar with various platforms now, but there was a couple emerging. How much is it sort of like chicken and egg? Do you think we get dot-com mania if you didn't have online retail investing? Or is it just a coincidence that both happen at the same time? Yeah, sure. I mean, it's one thing that feeds off each other. That's how I would put it. So if you didn't have as much as many retail investors perhaps you would still had a mania but it would have been to a lesser extent

7:51and the both in the in the upwards part as well as in the crash to give you some some numbers the the nasdaq went up by 450 percent in five years between 1995 and 2000, just before the bubble burst. And how does that compare to like historical manias? It's a, I mean, compared to historical manias, it's very much in line with history in a sense that before a bubble bursts, you really do see a huge run up in performance in these indices. But what I wanted to mention as well is that the dot-com era was quite unique because you had a combination of shared interests. So you had retail investors who wanted their stocks to go up, all the dot-com stocks to go up. Then you had the institutional investors who also wanted their own stock to go up. And then you had a third

8:45element, which is not present today, which is investment banks had an incentive also for these stocks to go up because they were generating revenues from all the IPOs. So the more IPOs, the more fees the banks would receive. And in addition to to that research departments in banks had all the incentives to put a buy recommendation to these stocks because the research departments were running on the same fees investment banks would generate the investment banking divisions would generate and is that that to me from reading your history there is where it becomes clear it's a mania you you rent reference one um bank researcher who went to his boss and said um there's there's absolutely no maths that

9:33justifies these valuations is that when it became clear that something was going to burst it didn't become become clear to people but i mean to the in hindsight that's when it becomes clear but it's it's very hard if you capitulate if you put yourself in that moment uh people had no reference points in the past so they didn't know that this was a completely different time so they didn't know um when the bubble board burst but it was it was starting to get clear even i in the in the chapter i quote an email exchange and that's when inter insider people became aware of the um speculation around the avanda.com companies yeah the uh analyst that i mentioned in in the book he himself is fictitional it's a fictitious character

10:21but uh it is true that um henry blotjet who's a real person put a recommendation on amazon stock uh which seemed outrageous at the time three weeks later this stock reached the target and then this analyst got hired by bank of america merlin for a huge compensation so there's a bit there's a big reward if you're the analyst that said buy this stock even though it's a crazy valuation yeah there's a there is definitely you're the analyst that said, buy this stock, even though it's a crazy valuation. Yeah, there is definitely a reward. And that really shows you how all the intense incentives were aligned. Not only retail, so retail investors, banks and researchers within banks, they all had all the incentives to keep pushing their share prices higher.

11:03And so we got to the mania take us to the crash when did it become clear that the bottom was about to fall out of this market and what happened so there were a few catalysts one of them being the cover of the magazine that i whose title the chapter is as well that was on the cover of barons yeah they sort of called bullshit on this market yes exactly they they basically run an in-depth study of uh on these companies and they said hey hold up these companies are not actually generating many revenues um and they um why it was so shocking to people is because they they revealed how most of these companies that went public or about to go public actually were about to go bankrupt. And so that's when the first cracks started appearing.

11:52And that's when the first slump also appeared in the Nasdaq. We had a big correction of 35% in a week. And that started to scare some people. But then what happened was that because the market held up strongly for so many years, for half a decade, it just went up. Investors, they thought, OK, let's buy the dip. Buy the dip. And so that's what they started doing. But then eventually what happened is the Fed increased rates yet again. And so you had rates at 6.5% at the time, which were really starting to bite on the economy. And at the same time, Microsoft was found guilty of running an illegal monopoly. And that really did it for the market. So a lot of the people who bought the dip got stopped out

12:38of the dip and the market kept going down. And eventually that's when the bubble started to burst but what's interesting is it's not a kind of burst that we think of today it's not like a v-shaped recovery like we had in covid or we had it last year with the tariffs this was a two-year sell-off and it was it was brutal because every time it would it would kill people metaphorically but every time it would kill people metaphorically, but every time the market would start going up, investors would buy again thinking, oh, the market has bottomed. And then it would crash again and these people would get stopped out again. And this affected real people, right? And by that, I mean, not just traders

13:19who live by the sword and die by the sword. This will have affected pension funds. Companies went bust. There were real economic consequences. Yeah, especially because a lot of people were still holding onto these losses. Eventually, they sold out at a huge loss. And so the average person stared at ruined retirements or ruined college funds for their kids. So it's something, it's a bubble that really hit not only institutional investors, but the average household as well, as well as pension funds. And so that's why it had huge implications on people as well, not only on the economy and on financial markets. In an increasingly standardized wealth management industry, Raymond James's high-touch, service-first way of working sets it apart.

14:08It puts people at the heart of everything it does. Advice that's as unique as you. That's the power of personal. Please remember, investing involves risk, and you may get back less than you invested. Well, fast forward to where we are today in 2026. Whilst listening to you explaining the dot-com history, alarm bells are going off in my head because I'm sensing certain similarities, right? of the last couple of weeks has been going up and up and up whatever the rest of the economy has been doing and every retail investor for at least a couple of years now has been piling into the now very newsworthy nvidia yeah the chip maker um but not just that we have ipos with crazy valuations i mean if you think to the spacex ipo we've talked about this on the podcast before, it implies revenue kind of multiples that haven't been seen in the history of capitalism.

15:11And what are central banks doing at the same time? They're raising rates. So are we in a mania, but it's not dot com? Are we an AI mania? Yeah, I mean, that's a great question, Michael. And I've heard lots of journalists, lots of writers writing about this. And it's very a topical thing at the moment. One can definitely draw many similarities and main differences. I will stick to three core similarities with the dot-com era. One of them being you have something which is very new it's a ai is a technology that is still not very well understood we still don't really know well what the use cases are in a very similar way we didn't know what the extent of the internet's use cases and also there is preliminary data on adoption of ai at companies, but we still, again, don't really have hard facts about how AI is helping companies generate revenues or AI in healthcare, for example, how, like what it will deliver in terms of drug discovery.

16:17So it's a very new technology, very much like the internet was, that is still not very well understood. So that's definitely a similarity. A second similarity is is a concentration so you have top 10 companies in the S&P now accounting for 40% of the performance of the S&P 500 this is very much similar to what was happening in the in the dot-com era where the top 10 companies at the time they were accounting for a huge chunk of the performance in the S&P, but before it was announced like the major index. Today it's 40% before it was 25%. But what's important to note is that in both cases, that's when you reach the highest concentration of performance of these top 10 companies. And a third similarity is that, as you also mentioned,

17:11central banks have been hiking into the dot-com bubble, but also hiking into this new AI rally. So back in the day, in 1990 and 2000, you had six interest rate hikes, three in 1999 and three in 2000 for a total of 175 basis points. Today, we've had one hike in September, one most likely in December next month, and one, possibly two more next year. So in both cases, you had a scenario where we were in an interest rate hiking path. So I'd say these are three key points. So, but I noticed in your answer, you're very careful not to use my word, which is bubble. You said the AI rally. Does that mean you don't think we're in a bubble in the way that .com was a bubble? Yeah, so I think that is not to say there might not be uh in the next couple of years an ai bubble um but we're not there yet but yeah from from my view uh we're not yet in an ai bubble and i'll

18:18give you just a few reasons i give you three reasons um One of them is valuation. So it is true that companies, AI companies today trade at very high multiples. However, except for... And by that, you mean the multiple of the stock price to the revenues? Yes. I mean, that's one way of measuring multiples. We could have many other measures, but let's say for simplicity, let's say the revenues over the market capitalization of the company. Back in the day, if you think of even the company I mentioned in the book, the globe.com, when it got IPO'd, it was valued at 600 million, but it barely generated any revenues. Today, so that's a very high multiple. it was valued at 600 million but it barely generated any revenues yeah um today um so

19:07that's a very high multiple today except for at the most extreme corners of speculation in ai in most cases uh you have valuations which are more compressed they're still stretched by historical standards but uh in no way shape or form are as as stretched as they were in the dot-com bubble on average the second point is is speculate is speculations the third difference for between now and the dot-com era is that um i mentioned it before but usually typically when you are in a bubble you have a huge performance in the index in the run up to the burst of the bubble. So in the case of the dot-com era, you had the Nasdaq going up 450% in the five years leading up to the crash. Today, although performance has been strong since the global financial crisis, there's not been one or two years where we've had the same extent of returns.

20:07The Nasdaq year to date is up 8%. The Nasdaq year to date at the height of the dot-com year went up 85%. So you can see how the extent the stock market has gone up is significantly different. And that to me could be a measure that we are actually not in in in a bubble territory at the moment there's loads of that right i mean whenever i go on tiktok there's all these like finance influencers being like i've bought nvidia i've got spacex i'm gonna buy anthropic there is but back in the day more than, I guess it's like fundamentals. Like back in the day, these companies that were going public, they not only they were highly speculative and they barely generated any revenues, they were also financing themselves through a huge amount of equity and debt issuance.

21:01Today, the fundamentals of these companies are much stronger so take a for example anthropic that you mentioned sure it's trading at very high multiples the ipo might be um the highest ever ipo in the world um at the same time it is generating revenues like it the projection is 60 billion revenues this year uh with having grown from 5 billion revenues last year. And a lot of these hyperscalers, like the big companies that are spending a lot in data centers and infrastructure, they are generating a lot of revenues and a lot of cash. And most of their financing, most of the way they finance these expenditures is by using their own operating cash flows. And this is very different from in the past where it was all IPOs and debt issuance.

21:53Sure, debt issuance is increasing and it will continue to increase. But up until now, most of the financing was through the operating cash flows of the other companies well i want to um pin you down a bit on on that point because even if the the stock market is not going to crash yeah it seems to me that there is a big amount of private credit involved here and what i mean by that is for your big hardware rollouts for your AI company, for data centers, for buying your chips, you do need to raise finance for that. And a lot of that is on private credit, on corporate bonds. A lot of companies borrow from NVIDIA actually. But it strikes me that the Western AI bet is based on the idea that the proprietary model wins out that you know you have to pay for a subscription for chat

22:45gpt um or a clod or whatever your ai of choice is but if the chinese model ends up proving more effective this model where you they're all open source and the wisdom of crowds makes them smarter even if they don't have access to the the western chips if that out, surely the whole bet on the Western companies fail and we've got this mountain of debt that could collapse. Sure, you could have a repricing in share prices. I think the debt that these companies have, although large, it's not yet at alarming levels. It might become alarming. But where we do see alarming levels is in the corners of the AI market, in the most speculative corners of the AI market. For the moment, companies like Amazon or Alphabet,

23:44they don't yet show unsustainable levels of debt. So sure, there might be a repricing if these Chinese models come on the market. And sure, these companies' bonds trade at at the moment, they don't yet trade at a high enough premium that would suggest a debt crisis within these companies is yet coming. Another kind of slightly reassuring thought is that even if it's not that China wins, another argument why some people who say it's a bubble put forward is that we assume that AI is getting better and better towards the singularity where either none of us ever have to work again or the AI kills us all. But some people say, what if we stop improving? What if the models don't get better and better?

24:43And then we've we've met we've kind of reached peak performance would the bottom of the stock fall out then but i guess the counter to that is even if the current model of chat gpt and claude or whatever is the best it's going to get to be there's presumably still big upsides from them being rolled out wider in the rest of the economy because most people are still not using AI, right? Sure. I think that's it. So you would have to distinguish between two things, the impact on companies and the impact on the stock market. So if you have a stagnating market, a stagnating capacity of ChatGPT or Claude or the likes. Sure, these companies might improve their profitability. These companies might operate more efficiently.

25:33But the stock market might not like it. The stock market is currently pricing for a continuous evolution of these models. And therefore, an improvement in efficiency in these companies to a greater extent that we have at the moment. At the moment, it's also difficult to quantify how, to what extent these companies are improving their efficiency using AI. So there would definitely be a huge readjustment in share prices on the back of, okay, although these companies might operate more efficiently, revenues might not grow as much as we wanted because the new models are not coming out. Okay, I want to widen the discussion out into Britain-American economies just in a minute. But just before we finish this segment

26:16on this AI bubble or not, I'm going to make you answer in kind of one word. If we come back here in a year's time, what is your bet? Will AI still be booming or will we be in an AI market bust? Personally, I think AI will still be booming. Okay. That is not to say there is not going to be or there may not be a crash in the future. But I still think AI, like the internet back then, it's still in its infancy and it has room to go up further because i i don't think we're yet seeing a bubble yeah um in the future we might see um let's say we start having human and robots at home and people start ordering a lot of them even if they're not yet produced that's when we could have a situation like the dot-com era where people are getting ahead of themselves and therefore you see

27:10these stock prices move way ahead of what actually is is going on that was a very long one word answer but i'll i'll forgive you and i want to talk briefly about america and then because we're in britain we'll finish talking in Britain. So despite all of this, we are seeing probably, from what you were saying, a temporary dip in US stocks in reaction to the midterms. And that's because for a huge, we've done an episode about this before, about data centers and objections to data centers in Britain. But in America, that objection is like hyperpowered. There's lots of hatred, not just for big tech, but for data centers. So there's this idea that because the Democrats have got political power now again, they're going to restrict them. And how much of a problem is that going to be for

28:02the American economy? And then if you can move your answer to Britain, is that something that we could take advantage of here? Yes, that's a very good question. So I think while there may be a slower rollout of data centers and related infrastructure in the U.S. as a result of the democratic sweep, I do think eventually this infrastructure will be built be it in the US or be it in other countries but I think at the moment the market has reacted to quite materially in the last few days over a potential delay in this rollout and as opposed to a halt in building new infrastructure I just think there'd be maybe a delay or a slight scale back in the extent of the infrastructure being built in the US.

28:53And Britain could benefit from this insofar as data centers could be built in the UK or in Ireland. And the UK specifically would benefit from an increase in employment as as a result of these data centers um but it's it's not clear to me how and to what extent the um like britain would would benefit from um from an impairment in the us uh from this point of view so i always like to end by asking people if they're positive or negative about Britain. And I'll ask you about that as an outsider. And I'll tell a funny story about why I'm asking you specifically, because when we used to live together, I remember you used to be quite down on Britain for people can probably guess when we were at university,

29:42some of the things going on, you would be like, this is a shit country. You're making so many mistakes. But then we were at university, some of the things going on, you'd be like, this is a shit country, you're making so many mistakes. But then we were having a conversation last year, we were just chit-chatting and I was surprised about how positive you were about Britain and our economy. Is that still your view now? Yeah, I mean, I wouldn't say rosy picture, but I would say one needs to detach themselves from the gossip or the political gossip versus the economic fundamentals while britain has suffered from i think from from brexit um i think fundamentally it it is an improving it is improving economically and i think out of every downfall, out of every difficult moment, there is a chance for things to improve. Britain has had political chaos in the last few years. I think if the current government can steady the ship and remain in power long enough, the optimism will return even within the brits i think markets as a whole are optimistic if you if

30:47you see the uk stock market at the moment is is is close to ultimate highs i still view the uk in a positive light i think one needs to detach themselves from all the all the political upheaval which which there is but now i think it's it mostly been, it's mostly in the past and out of every dip, I think this could be an interesting opportunity to invest in the UK actually. And I'm quite optimistic about the UK. I think the UK has some issues that it needs to solve, like every country. I think one of the issues the UK is dealing with is demographics and how to make everything work. I think that's one of the key challenges Britain has at the moment is how to integrate all this new influx of immigration

31:38that it has had in the last few years. But your positive message is by Britain. You are long Britain. Yeah, I mean, I would say, I mean, I'm a contrarian in this point of view. I think kind of there's been too much negative talk. And when now I think things are starting to improve slowly, but steadily, sure. So on the whole, I have a positive outlook. Julio, thank you very much for joining Reality Check. Thank you very much, Michael. It's been a real pleasure. Thank you for listening to that episode of Reality Check. If you want to buy Julio's book, which I'm holding up here, you can get it on Amazon. And if you're watching on YouTube, there's a link in the description if you're watching on YouTube,

32:26there's a link in the description. If you are on YouTube, don't forget to subscribe and tap the bell so that you never miss an episode. If you're following along on a podcast platform, give us a follow there too. And for our archive of episodes, as ever, go to spectator.com forward slash reality check. search, email, spreadsheets, and our shared Google Doc, the headline hammer. With Google Workspace with Gemini, Google's AI is built right into the premium versions of tools we already use, like Gmail, Docs, and Sheets. No new tools or extra logins, just personalized AI for the way we work. That means less time on the work, around the work, and more time finding the stories worth talking about.

33:19Get AI tools for work. Upgrade to Google Workspace with Gemini at goo.gle. That's goo.gle slash workspace with Gemini. That's goo.gle slash workspace with Gemini. £12. Not only that, but we'll also send you a £20 Amazon gift card, absolutely free. Go to www.spectator.com forward slash voucher to claim this offer now. you

▸More episode summaries

Best of the Spectator cover art
Best of the SpectatorHoly Smoke: Pope Leo's homecoming – the strange Catholicism of Peru2026-11-15 · 24 min
Best of the Spectator cover art
Best of the SpectatorQuite right!: Zia Yusuf interview2026-11-18 · 52 min
Transcribe your links. Paste a YouTube, Google Drive or podcast link and WhipScribe gives you the transcript. Transcribe a link →
New Best of the Spectator transcripts by email
One email per new episode, unsubscribe in every message. Not shared with the publisher.