Highlights

  • Artificial intelligence may reshape industries, but technological success does not guarantee investment success.
  • AI exposure extends beyond software companies to chips, data centres, energy, cybersecurity and other supporting industries.
  • High valuations and concentrated market leadership can make AI-related shares particularly sensitive to disappointing earnings or changing expectations.
  • Understanding where a company sits within the AI ecosystem can be as important as believing in the technology itself.

Artificial intelligence has become one of the defining investment themes of the current market cycle.

AI is moving into workplaces, software platforms, data centres, cloud infrastructure and everyday consumer products. At the same time, enormous amounts of capital are being directed towards the chips, computing infrastructure, energy systems and digital services needed to support its expansion.

That combination has created a compelling market story.

But there is an important distinction between believing that AI will transform the global economy and assuming that every company associated with AI will deliver attractive investment returns.

History offers plenty of examples where a technology changed the world while many businesses built around that technology disappeared along the way.

For anyone looking at AI-related investments, the real challenge may therefore be less about deciding whether AI matters and more about understanding who could ultimately capture the economic value.

Why Has AI Become Such a Powerful Market Theme?

AI is no longer simply a story about experimental software.

Its expansion requires physical infrastructure on an enormous scale. Advanced semiconductors need to be produced. Servers need to be installed. Data centres need to be constructed and cooled. Cloud platforms must provide computing capacity, while electricity networks have to support increasingly power-intensive facilities.

That creates potential economic activity across multiple industries.

Companies supplying the infrastructure behind AI have consequently attracted significant market attention because many are already generating revenue from the current investment cycle.

This is an important difference between some AI-related businesses and purely speculative technology themes.

Parts of the AI ecosystem are supported by existing customers, real capital expenditure and current cash flows.

Yet fundamentals are only one reason prices can rise.

Market psychology also matters.

When a technology is widely described as potentially transformative, investors can become concerned about missing the next major market trend. That fear of missing out can push valuations higher than fundamentals alone might justify.

Separating genuine earnings growth from market enthusiasm becomes particularly important when expectations are already high.

AI Is Not a Single Investment Theme

One of the biggest misconceptions around artificial intelligence is treating AI exposure as though it were one single market segment.

In reality, AI operates through several interconnected layers.

Each layer has different business models, competitive conditions, capital requirements and risks.

An investor buying shares in a semiconductor producer may technically be investing in AI, but that exposure looks very different from investing in an enterprise software company using generative AI.

Understanding those differences can help clarify what is actually driving an investment.

AI Infrastructure and Hardware

The hardware layer sits at the foundation of the AI ecosystem.

AI models require significant computing power, which has increased demand for specialised chips, servers, networking equipment and cooling systems.

These businesses can benefit directly from infrastructure spending because customers need physical equipment before many AI applications can operate at scale.

However, infrastructure companies can also be highly exposed to the pace of investment.

Building advanced semiconductor facilities, manufacturing equipment or data centres requires substantial capital. If spending slows or customers decide they have built enough capacity, revenue expectations can change quickly.

The same investment boom that supports growth can therefore create risk if future demand does not match the amount of capacity being developed.

Data Centres, Energy and the AI Supply Chain

Artificial intelligence does not operate in a digital vacuum.

Behind every AI query sits physical infrastructure consuming electricity, processing data and transferring information across networks.

As demand for AI computing increases, attention has widened towards industries that provide the supporting infrastructure.

Data-centre operators are an obvious example, but power suppliers, grid infrastructure businesses and property-related companies can also form part of the wider AI investment theme.

These businesses may not appear as technologically exciting as companies creating large language models or generative AI applications.

Yet they can still participate economically in the expansion of AI because the technology cannot scale without physical capacity and reliable energy.

This illustrates why understanding the broader supply chain matters.

Sometimes the companies supplying infrastructure to a technological boom can generate significant revenue regardless of which individual platform eventually dominates.

Software Platforms and AI Applications

At the other end of the AI ecosystem are the businesses attempting to embed artificial intelligence directly into software products and services.

The potential market is enormous.

AI can potentially change how companies analyse information, create content, automate tasks, interact with customers and develop software.

But the application layer may also carry considerable uncertainty.

A company can successfully develop an AI feature without necessarily creating a profitable long-term business around it.

Competition can be intense, technology can evolve quickly and customers may not always be willing to pay substantially more for AI capabilities.

That leaves an important question: who captures the profit?

The technology provider? The cloud platform? The software company? The customer using AI to lower costs? Or a combination of all four?

Until business models mature, answering that question may remain difficult.

The Less Obvious AI Beneficiaries

Some companies may participate in AI growth without building models or manufacturing chips.

Cybersecurity is one example.

As organisations adopt more digital systems and AI applications, the number of systems, connections and data flows that require protection can also increase.

Data management represents another supporting area.

AI systems depend heavily on access to information. Businesses therefore need ways to organise, store, secure and process the data used by their applications.

These supporting industries illustrate why investing in AI can mean very different things.

A portfolio can be exposed to the theme through infrastructure, energy, software, data, security or combinations of these areas.

Before considering an AI-related investment, understanding exactly where the company sits within this ecosystem can provide more context than the AI label alone.

Why Can AI Stocks Be So Volatile?

High expectations can be powerful when companies meet them.

They can be equally powerful in the opposite direction.

When shares are priced on assumptions of rapid future growth, the market may leave little room for disappointment.

A company does not necessarily need to report poor results for its share price to fall. Sometimes growth simply needs to be slightly weaker than investors expected.

A change in demand forecasts, slower customer spending or weaker guidance can force the market to reassess how much future growth is already reflected in the share price.

That can lead to sharp movements.

The higher the expectations embedded in a valuation, the more sensitive an investment may become to changes in those expectations.

AI-related stocks can therefore experience significant volatility even when the underlying industry continues to grow.

Valuation Risk: A Great Technology Can Still Be Expensive

One of the most important lessons from previous technology cycles is that being right about the technology does not automatically mean being right about the investment.

A company can participate in a rapidly expanding industry and still generate disappointing returns for shareholders if its shares were purchased at an excessively high valuation.

The price paid matters.

When a valuation assumes near-perfect execution, there may be little room for delays, weaker margins, increased competition or slower adoption.

This creates an important distinction between company quality and investment value.

A high-quality business can still be a risky investment at the wrong price.

Similarly, a transformative industry can contain businesses whose valuations ultimately prove unsustainable.

Concentration Risk Can Hide Inside a Portfolio

Another risk surrounding AI is concentration.

A relatively small group of very large companies has played an outsized role in the development and commercialisation of artificial intelligence.

That can create an unexpected issue for investors who already own broad market investments.

Someone may believe they have separate exposures through technology shares, an index fund and an AI-themed investment while several of those holdings ultimately contain many of the same companies.

The portfolio may therefore be more concentrated in AI than it appears.

Diversification is not determined only by the number of investments owned.

What matters is what sits underneath them.

Examining the largest holdings within funds and understanding overlaps between investments can help reveal whether a portfolio is disproportionately dependent on a small number of AI-related businesses.

The Right Trend Can Still Produce the Wrong Companies

Technology booms tend to create winners and losers.

A company can correctly identify a major technological shift and still fail because of excessive debt, weak management, competition, poor execution or an unsustainable business model.

The internet provides a useful historical lesson.

The technology itself transformed commerce, communication and entertainment. Yet many companies created during the internet boom did not survive.

AI could produce a similar distinction between the success of the technology and the success of individual businesses.

Picking a structural trend can therefore be easier than identifying which companies will dominate it over the long term.

That uncertainty is especially relevant in areas where competitive advantages are still developing.

When Will AI Spending Generate Returns?

Another question facing the market concerns the relationship between spending and future revenue.

AI infrastructure requires substantial investment today.

Data centres, advanced computing equipment, energy infrastructure and software development all involve significant upfront expenditure.

The investment case frequently assumes that these costs will eventually produce greater revenue, productivity or efficiency.

But the timing remains uncertain.

If companies spend heavily on AI infrastructure while the commercial benefits arrive more slowly than expected, the market may become less willing to support elevated valuations.

That does not necessarily mean the technology has failed.

It may simply mean expectations were too aggressive.

This gap between technological development and financial returns can become an important driver of market volatility.

Regulation Adds Another Layer of Uncertainty

Artificial intelligence is also developing within a rapidly changing regulatory environment.

Governments are considering questions around employment, privacy, intellectual property, cybersecurity, competition and the broader social consequences of AI.

Future regulation could affect how AI products are developed and deployed.

Rules that increase compliance requirements could raise costs for businesses. Restrictions on data use or certain AI applications could also affect commercial opportunities.

At the same time, clear regulation may provide greater certainty for businesses and consumers.

The direction and impact of policy remain difficult to predict, making regulatory developments another factor to watch when assessing AI-related companies.

Can AI Still Be a Long-Term Structural Trend?

The risks surrounding AI do not eliminate its potential significance.

If artificial intelligence produces sustained improvements in productivity, the impact could extend far beyond the technology industry.

Businesses may use AI to automate processes, analyse information more efficiently, develop products faster and improve customer experiences.

Infrastructure built during the current expansion could also remain economically useful for years.

The source material compares this possibility with previous infrastructure cycles, where assets created during periods of intense investment continued to support economic activity long after some of the original companies disappeared.

This is why the debate around AI is unlikely to be as simple as boom versus bust.

A technology can simultaneously create genuine economic value and experience periods of market excess.

What Should Someone Consider Before Investing in AI?

The first question should not simply be whether artificial intelligence has a future.

A more useful set of questions may include what part of the AI ecosystem the company operates in, how it makes money, how much growth is already reflected in its valuation and what could disrupt that growth.

Portfolio exposure also matters.

An investor may already hold significant AI-related companies through broad market funds without deliberately buying an AI-specific investment.

Timeframe and risk tolerance remain important as well.

Emerging technology themes can experience steep corrections as markets reassess growth expectations.

Anyone considering exposure therefore needs to recognise that a long-term technology story can still produce considerable short-term volatility.

Ultimately, AI investing involves separating three different questions.

Is the technology important?

Can the company benefit from it?

And is the current share price reasonable relative to that opportunity?

Those questions may produce very different answers.