Collective hallucination; has the AI bubble now burst?

What’s the first thing that comes to mind when you think about the “AI technology revolution”?

It’s an apt time to ask, considering that we are right at the start of the US stock market earnings season.

And for those who reported early, the numbers are staggering. Many AI stocks are reporting record revenue and growth for their businesses.

So, what did you think of?

Large hyperscalers? Applied AI leading to new medical breakthroughs? AI agents doing all your work for you, such as managing your diary, handling client relationships or booking your holidays?

Let me tell you what I think.

I think about…. steam railroads.

Wait, what? How can I possibly be thinking that? What does such an archaic technology as a steam powered engine have to do with AI?

Let me explain.

Whilst the numbers reported are simply incredible in some instances, regardless almost every AI stock price is – falling!

Here’s an example of what I mean.

Beyond the headlines, there is something fundamental about what’s going on that you need to understand.

We have entered a place where the narrative may be completely right, yet the stock prices of these companies may be completely wrong.

And the explanation for this paradox is to be found in a long-forgotten boom almost two centuries ago.

Let me show you once again just how incredibly powerful knowledge of history be.

No one probably appreciated it at the time, but when Queen Victoria took her first ever train trip on June 13^th, 1842, from Slough railway station just outside London, arriving at her destination 25 minutes later at Paddington station, it sparked the imagination of an entire country.

Believe it or not, before the Queen’s train ride, the introduction of the railway, this world-changing technology, was greeted with outright hostility.

Steam-powered engines were regarded as a public nuisance, with outcry about the amount of pollution they spewed out across the idyllic countryside as they passed through.

The noise they make was said to upset livestock and, in the most ironic twist, rural landowners said the noise and pollution would ‘harm land values.’ (How wrong that point of view would turn out to be).

That 25-minute royal journey not only changed the prevailing attitude towards railways, but it also attracted private investors who were quick to seize upon the new popularity of the railways (or railroads, as they came to be known in the United States).

Britain had just recovered from recession, and the news about the Queen’s rail trip was seen as one way to really turbocharge the recovery.

The 1844 Railway Act set the platform for existing and new railway projects, capping the returns these near-monopoly businesses could charge paying passengers and the dividends they could return to investors.

Then the subsequent 1844 Bank Charter Act saw official interest rates fall from 4% to 2.5%. In an environment of improving economic conditions, investor enthusiasm for all things rail and now cheap money flooding the system, all the ingredients necessary for what is probably history’s the biggest boom were there.

It came to be known as the British Railway Mania.

What really set things alight, though, was the fact most of these new rail companies were listed on the London stock exchange.

Everywhere you looked, official and unofficial publications were only talking about one thing, railway stocks and the incredible amount of money investors were making from them.

The demand to get in was so great that shares were denominated lower to allow the majority to buy. The average stock appreciated over 100% in three short years. But many others rose ten times this.

Finally, think now of all the land these railroads required.

Once authorised, the price of the land on which these railways were built increased immediately. Speculators forged ahead of parliamentary authorisation to secure the best possible sites for themselves before on-selling them to railway companies.

Simply put, where the rails were laid, the gains for the landowner were prodigious.

In fact, so enormous were the gains that this one bubble drove not just one but two complete 18.6-year Real Estate Cycles in Britain (and three full cycles in the US!).

Now, you might be thinking “Ah-ok, now I get Darren’s analogy. I see the relationship between that bubble in the 19^th century and the AI bubble now”.

But that would be a mistake. Because whilst there are very strong parallels between them, it doesn’t address our core issue for today’s newsletter.

Why, after posting record profits and revenue, are AI stock prices falling?

Prepare yourself for the single best lesson you’ll learn this decade.

Who pays the bill, and who cashes the cheque from it? That is what the AI trade has now become. One side’s spending is the other one’s order book. It’s all a neat and nice little revolving circle of revenue for everyone. At least, that is what we have been told.

With the US earning season upon us, the truth now drops. And the truth hurts! Look at some of the biggest names in tech. Taiwanese Semiconductor Manufacturing Company ($TSMC) posted its fifth straight record quarter. The effect? The stock price fell.

Micron’s ($MU) revenue for the quarter was 388% higher than this time last year. Net income rose by an astonishing 1,400%. But the stock still sold off.

There have been plenty more stocks just like that in the last week or so. So, what on earth is happening?

The answer to me is obvious. It’s all a hallucination.

A collective hallucination is what brought down the British railway mania 180 years ago. From around 2000 miles of track in 1830, by the time of the ensuing bust in 1845 Britain had laid 10,000 miles of track.

The longest routes were called direct lines, which were railway routes directly to and from terminals. Huge amounts of money and effort went into planning and promoting such lines (and comparable amounts into fighting them by non-direct lines that perceived them as threats).

Investors paid a premium to get involved in these as the engineering costs were astronomical.

And the formula used to anticipate potential returns was far too optimistic.

But what really killed the bubble was a simple human trait: we like to take short trips.

Pretty much no one took the direct line from London to Manchester in its entirety; instead, people travelled much shorter sections between the numerous stops on the route. The money in railways was to be made in these short trips, not the long direct routes that were so costly to build.

All that debt, the colossal overcapacity built, the overinflated stock prices, greed, corruption, fraud – all of it predicated on pure fantasy – a hallucination.

Railway tracks laid down almost side to side by competing companies led to fierce competition. This was great for travellers, who paid pennies on the pound for this travel, but it was a nightmare for investors.

The cost cutting got so bad that many railways offered free travel, hoping to make their money on the connecting rides from terminals. It was the dreaded race to the bottom.

And so, when economic conditions inevitably started to worsen, when the revenue to pay dividends vaporised, and banks and lenders asked for their money back, the bubble finally burst for good.

This is where we are in the AI tech bubble – a race to the bottom.

The competition is undercutting each other, and that competition is now more likely to come from China. China’s Moonshot dropped Kimi K3 a few days ago, a cheap open model media desks are calling a “DeepSeek 2.0” moment.

And the fear it incites is simple: if good models get cheap, the trillions of dollars in buildout capex gets harder to justify. That lands hardest on the picks-and-shovels names in memory and AI hardware, where the spending must pay off for the stock to work.

Folks – it’s not happening. The railroad mania IS the future of the AI tech bubble. And your very first signs that big money is asking themselves the same question is in the performance of those stock prices today.

When the narrative may be completely right, but the stock prices are completely wrong.

And so, dear reader, if I can explain the true reason behind the stock price performance of the world’s biggest tech companies by using a little history lesson, what else can history teach you?

It’s something our current membership of the Boom Bust Bulletin can tell you, so why not become our newest member? Learn the history of the 18.6-year Real Estate Cycle and uncover the truth behind why the world is the way it is today.

With written monthly editions and regular video updates bringing you the very latest news on the cycle, you’ll discover for yourself, like I have here, just how powerful a little knowledge can be for you.

Semiconductors now represent 20 percent of the entire S&P 500. One industry, one fifth of the American stock market. This is the highest weighting ever recorded, above even the dot-com peak.

Estimates put the American AI data centre buildout at $11 trillion between 2024 and 2029, with roughly $7 trillion of it financed through debt. And yet, open-source models (such as Kimi K3) can reach 90 percent of peak performance within months, at a fraction of the price.

Wall Street just raised its long-term earnings growth forecast for the S&P 500 to 25.5 percent a year, the highest ever recorded. Analysts now assume chipmakers will keep 50.3 cents of every revenue dollar as profit, forever, more than triple the market’s margin.

Can you spot a collective hallucination when you see one?

All bubbles eventually end, but super cycles, like the real estate cycle, endure.

The time to make your financial knowledge enduring is now.

Sign up now.