Every major market cycle leaves behind a lesson. The challenge is recognising when history is offering a valuable warning and when investors are allowing the past to cloud their view of the future.

For many investors, the technology crash of the early 2000s remains the ultimate warning about what happens when innovation, excitement, and speculation move too far ahead of reality. The dot-com boom was built on a powerful idea that the internet was going to transform the global economy.

That prediction turned out to be right.

The internet changed how people communicate, shop, advertise, consume entertainment, and do business. The mistake investors made was not believing in the technology. The mistake was assuming every company attached to that technology would become a long-term winner.

Many internet companies reached extreme valuations despite having little revenue, limited profitability, and business models that were still unproven. Eventually, expectations ran too far ahead of what those companies could actually deliver. The Nasdaq collapsed, and many businesses that were expected to define the future disappeared.

More than two decades later, investors are asking a familiar question, is artificial intelligence simply the dot-com bubble happening again?

The comparison is understandable. AI enthusiasm has helped drive significant market gains, technology companies dominate major indices, and a relatively small group of businesses has contributed heavily to overall returns. Once again, investors are looking at a breakthrough technology that could reshape the economy.

But similar starting points do not always lead to similar endings.

Assuming that AI must follow the same path as the dot-com era simply because parts of the story feel familiar may be an example of one of the most common mistakes in investing: representativeness bias.

When Investors See the Past in the Present

Representativeness bias is the tendency to judge something based on how closely it resembles a previous experience.

In financial markets, it often appears through simple comparisons.

Technology boom. High valuations. Investor excitement. This must be another 2000.

At first glance, the argument seems reasonable because there are genuine similarities.

Both periods were built around technologies with the potential to transform the world. In the late 1990s, it was the internet. Today, it is artificial intelligence. Both created excitement around productivity improvements, new business models, and the possibility that the economy was entering a new era.

Both periods also benefited from a powerful story.

During the dot-com boom, companies could attract investor attention simply by being associated with the internet. Today, businesses connected to AI infrastructure, semiconductors, cloud computing, and automation have attracted significant interest.

But markets are rarely that simple.

Representativeness bias makes similarities easy to see and differences easier to ignore.

Someone who experienced a devastating storm may become convinced that every dark cloud signals another disaster. Investors often do something similar. They remember the painful ending of previous market cycles but sometimes overlook the unique conditions that created them.

History provides lessons.

It does not provide a script.

Why the AI Boom Is Not a Perfect Repeat of the Dot-Com Era

The biggest difference between today's AI rally and the dot-com period is the quality of many companies leading the market.

During the late 1990s, investors were often paying enormous prices for what companies might become. The vision was exciting, but in many cases the profits simply were not there yet.

Today's AI leaders are generally starting from a very different position.

Many of the companies driving the AI boom, including semiconductor businesses, cloud providers, and large technology platforms, are already some of the most profitable companies in the world. They have established customers, global reach, strong balance sheets, and significant cash generation.

The AI investment cycle is also creating real economic activity. Demand for advanced chips, data centres, networking equipment, and cloud infrastructure has translated into meaningful revenue growth for companies positioned at the centre of this trend.

That difference matters.

However, it does not remove risk.

The dot-com era taught investors that a revolutionary technology can still produce disappointing investments. The internet was not overhyped. If anything, its long-term impact was underestimated. The problem was believing every internet company would succeed and that valuation no longer mattered.

AI investors face a similar challenge today.

Artificial intelligence may prove to be one of the defining technologies of this era, but that does not mean every company connected to AI will justify its valuation.

A great story still needs great numbers.

Ultimately, investors need to determine whether the enormous investment flowing into AI infrastructure will generate returns that support today's expectations.

The Real Risk Is Not Another 2000. It Is Forgetting the Lessons of 2000

While the current environment is different from the dot-com bubble, investors should not ignore the similarities completely.

One of the biggest concerns is concentration.

A small number of technology companies now represent a significant portion of major equity indices. This means overall market performance has become increasingly dependent on a handful of companies continuing to deliver strong earnings growth.

That alone does not mean the market is in a bubble, but it does increase vulnerability. When expectations are high, even strong companies can disappoint.

There are also important questions around the scale of AI investment. Companies are spending enormous amounts on infrastructure, and eventually investors will want evidence that this spending can translate into attractive long-term returns.

Some areas of the AI market may ultimately discover that expectations moved faster than profits.

The danger sits at both extremes.

Investors can become too optimistic and assume AI changes everything, making valuation irrelevant. They can also become too cautious and assume every technology boom must end like 2000.

Neither approach is particularly useful.

The better question is not, "is this the dot-com bubble again?"

The better question is, "are today's expectations supported by tomorrow's earnigs?"

The dot-com era showed that revolutionary technologies can create extraordinary winners while still producing painful losses along the way. Some companies disappeared, while others survived and eventually became some of the most valuable businesses in history.

AI may follow a similar path.

The lesson from 2000 is not to avoid innovation. It is to separate innovation from speculation.

Markets in 2026 share some similarities with the technology boom of the late 1990s, and those similarities deserve attention. But similarity does not mean destiny.

Representativeness bias encourages investors to believe that because something looks familiar, the outcome must also be familiar.

The more valuable lesson is that technological revolutions can change the world, but even the greatest innovations must eventually be supported by earnings, cash flows, and realistic expectations.