Is AI in a bubble (Part 3)?

The $725 Billion Bet: Is AI the Greatest Investment Opportunity of Our Lifetime or the Biggest Bubble Ever Created?

By now, one thing should be abundantly clear.

The AI story is far more complicated than the headlines suggest.

On one side, we have the largest and most profitable companies in history collectively committing close to $725 billion annually towards AI infrastructure. On the other, we have growing evidence that enterprise adoption is not yet producing returns commensurate with the investment being made.

History tells us that these situations rarely end neatly.

The question is no longer whether AI is transformative.

The question is whether investors have become too optimistic, too early.

This is perhaps the most difficult question facing investors today.

Several seasoned investors—including Ray Dalio, Michael Burry and Jeff Bezos—have, in different ways, cautioned against blindly extrapolating today's AI enthusiasm indefinitely.

Yet it is equally dangerous to conclude that AI is "just another dot-com bubble."

The reality lies somewhere between those two extremes.

To understand why, it helps to compare today's environment with previous speculative cycles.

The easiest comparison people make is with the dot-com bubble.

Technology stocks are rallying.

Capital is flooding into infrastructure.

Valuations are expanding.

Naturally, investors assume history is repeating itself.

But there are crucial differences.

During the internet bubble, many of the companies commanding astronomical valuations generated little or no revenue.

Several telecom operators financed expansion almost entirely through debt.

Many internet startups had no viable business models.

Today's leaders look very different.

Nvidia generated extraordinary profits over the past year.

Microsoft, Alphabet, Amazon and Meta remain among the most profitable businesses ever created.

They generate enormous free cash flow before investing heavily into AI.

Even valuation metrics are less extreme than those witnessed during the peak of the dot-com era.

At the height of the internet mania, the NASDAQ traded at forward earnings multiples approaching 60x. Today, that figure is closer to 26x.

That doesn't necessarily make markets cheap.

But it does suggest today's environment is considerably more grounded than 1999.

This distinction matters.

The companies funding today's infrastructure are financially resilient.

If AI demand disappoints, they will almost certainly experience slower growth.

They are far less likely to disappear altogether.

Having said that, bubbles are not defined solely by valuations.

They are defined by expectations.

Every bubble begins with a genuine technological breakthrough.

The mistake investors make is assuming that because the technology succeeds, every investment surrounding it must also succeed.

History repeatedly disproves this assumption.

Railroads transformed transportation.

Thousands of railway investors still lost fortunes.

The internet transformed communication.

Countless internet companies disappeared.

Electricity transformed manufacturing.

Many early electricity companies failed.

The technology survived.

Capital did not.

This is precisely the risk investors must keep in mind today.

The current AI infrastructure build-out assumes demand will eventually justify hundreds of billions of dollars in annual investment.

If demand arrives more slowly than expected—or if pricing power erodes due to competition—the returns on that infrastructure could disappoint.

The infrastructure itself may remain incredibly valuable.

The investments made at today's prices may not.

Perhaps the single most important lesson from history is this:

Being correct about the future does not automatically make you money.

The telecom example illustrates this perfectly.

The companies building fibre optic networks were not wrong.

Internet traffic did explode.

Streaming became ubiquitous.

Cloud computing became indispensable.

Artificial intelligence itself relies on that very infrastructure today.

Yet many of the businesses that built it went bankrupt before demand finally caught up.

Timing matters.

Capital allocation matters.

Valuation matters.

Investors often underestimate how painful that gap between technological adoption and financial returns can become.

There are two broad scenarios that investors should consider.

Neither is guaranteed.

Both deserve careful consideration.

AI isn't just changing business—it's creating entirely new income opportunities. The Hustle's guide features 200+ ways to make money with AI, from beginner-friendly gigs to advanced ventures. Each comes with realistic income projections and resource requirements. Join 1.5M professionals getting daily insights on emerging tech and business opportunities.

Get the guide

In this scenario, enterprise demand fails to grow quickly enough.

Returns disappoint.

Technology companies begin reducing capital expenditure.

Hiring slows.

Data centre construction moderates.

The stock market reprices future growth expectations.

For investors, this would likely resemble previous technology corrections.

Equity valuations contract.

Momentum reverses.

Sentiment shifts dramatically.

The effects would not remain confined to Silicon Valley.

Countries deeply integrated into global technology spending—including India—would inevitably feel the impact.

Indian IT services firms derive substantial revenue from global technology budgets.

If Big Tech reduces spending, outsourcing demand could weaken.

Graduate hiring could slow.

Technology consulting projects may be postponed.

The ripple effects would extend well beyond AI companies themselves.

Interestingly, this scenario is not necessarily better.

Suppose the infrastructure spending ultimately proves justified.

Companies will still need to generate meaningful profits.

To recover hundreds of billions of dollars invested into AI infrastructure, providers may eventually increase pricing.

Today's generous free usage tiers may disappear.

Inference costs could rise.

Subscription prices may increase.

Only the largest enterprises may comfortably afford frontier AI models.

Ironically, many smaller AI startups could disappear—not because their technology failed, but because operating costs became unsustainable.

In other words, AI becomes successful...

...but significantly more expensive.

Consumers ultimately bear the cost.

There remains one final possibility.

Technological breakthroughs dramatically reduce inference costs.

Model efficiency improves faster than expected.

Hardware becomes cheaper.

Energy consumption declines.

Enterprise productivity rises sharply.

Demand accelerates sufficiently to justify today's investment.

This is undoubtedly the most optimistic scenario.

It is also the scenario current market prices appear to be discounting.

Whether reality unfolds this way remains uncertain.

Rather than asking whether AI is a bubble, perhaps investors should ask better questions.

Is current spending generating sustainable economic returns?

How much future growth is already reflected in today's valuations?

Which companies actually possess durable competitive advantages?

Who controls pricing power?

Who owns scarce infrastructure?

Who benefits regardless of which AI model ultimately wins?

History suggests that infrastructure providers, semiconductor manufacturers, cloud platforms and essential software layers often prove more durable than application developers chasing temporary trends.

Equally important, investors should separate technological optimism from investment discipline.

It is entirely possible to believe AI will transform the global economy while simultaneously believing certain stocks have become excessively expensive.

Those are not contradictory positions.

They are two sides of intelligent investing.

One of the strongest ideas running throughout this discussion is the concept of the capital cycle.

Every major technological revolution follows a remarkably similar sequence.

High returns attract capital.

Capital creates excess capacity.

Excess capacity destroys returns.

Eventually demand catches up.

The survivors become extraordinarily valuable.

The challenge has never been identifying transformative technologies.

The challenge has always been identifying which participants survive long enough to benefit.

That distinction explains why history rewards patience more often than excitement.

The AI revolution is real.

That much appears beyond dispute.

The world's largest technology companies would not collectively commit nearly three-quarters of a trillion dollars annually if they believed otherwise.

The demand for computing power is genuine.

Data creation continues to accelerate.

Artificial intelligence is already reshaping software, research, healthcare, education and manufacturing.

But history also reminds us that revolutionary technologies often experience periods of excessive optimism.

Markets routinely overestimate short-term adoption while underestimating long-term impact.

That is why bubbles exist.

Not because the technology is fake.

But because expectations temporarily outrun economic reality.

As investors, our responsibility is not to predict precisely when sentiment will change.

It is to recognise where expectations have become stretched, where economics remain uncertain and where capital allocation deserves closer scrutiny.

Perhaps the most honest conclusion is this:

Nobody can say with certainty whether today's AI investment boom represents the greatest business opportunity in history or the largest capital misallocation ever witnessed.

The technology is real.

The revenues are real.

The infrastructure is real.

What remains uncertain is whether the prices being paid today will ultimately prove justified.

And that uncertainty—not certainty—is exactly what makes investing so fascinating.

Every generation believes its technological revolution is different.

In many ways, it usually is.

But capital markets have remarkably long memories.

They have seen railroads, electricity, automobiles, telecom, the internet and smartphones.

Every one of them changed the world.

Every one of them also experienced periods when capital outran economics.

The AI era may eventually become the defining technological transformation of the twenty-first century.

Whether today's valuations reflect that future—or have already priced in decades of perfection—is the question investors must continue asking.

Because in investing, being right about the technology is only half the battle.

The harder question has always been whether you're paying the right price for it.

Read our Part-1 and Part-2 here

Join millions of traders putting their knowledge to work on real-world events—from inflation to elections. Buy “Yes” or “No” shares and earn if you’re right.

No house. Peer-to-peer. Cash out anytime.

Get a free $10 to start. Claim it and start trading now.

Start Predicting Now

_{Trade responsibly.}

Disclaimer: The information provided on this website is for educational and informational purposes only and does not constitute financial, investment, or trading advice. Investing in securities involves risk, including the potential loss of principal; always conduct your own research and consult a qualified financial professional before making investment decisions.