Artificial intelligence has become one of the biggest investment stories in recent years. From the chips powering AI models to the data centres running them, companies across the technology ecosystem are attracting enormous investor attention.
But there’s an important question worth asking: Is AI creating sustainable business growth, or are stock markets becoming too dependent on expectations about its future?
To understand the debate, let’s look at both sides.
The positive side: Why AI is driving market optimism
1. AI is creating new revenue opportunities
AI is no longer just a technology experiment. Companies are integrating it into cloud computing, enterprise software, advertising, customer service, and other business operations.
For companies such as Microsoft, Alphabet, and NVIDIA, AI creates opportunities to sell new products, expand existing services, and serve growing demand for computing power.
If these investments translate into higher revenue and profits, today’s enthusiasm could have a meaningful business foundation.
2. The opportunity extends beyond technology companies
The AI opportunity isn’t limited to chipmakers and software developers.
Banks can use AI to improve fraud detection and automate processes. Healthcare companies can apply it to research and diagnostics. Manufacturers can use it to improve efficiency, while energy companies may benefit from the additional electricity and infrastructure required by data centres.
This means AI could influence productivity and profitability across several industries, potentially broadening its economic impact.
3. Investors have seen what AI demand can do for earnings expectations
NVIDIA offers a striking example.
In May 2024, its shares rose more than 9% in a single trading session after a strong revenue forecast reinforced expectations of sustained demand for AI chips. The company added approximately $218 billion in market value that day.
The rally illustrated how quickly investors can revalue a company when they believe demand for its products will exceed expectations.
For investors, the attraction is straightforward: if AI spending continues to grow and companies convert that spending into profits, the businesses supplying essential infrastructure could benefit significantly.
The negative side: Where does the risk begin?
1. Expectations can rise faster than actual profits
A promising technology doesn’t automatically make every company associated with it a good investment.
Investors may price in years of future growth before that growth appears in financial results. If revenue disappoints, margins weaken, or competition increases, a stock can fall even when its underlying business is still growing.
This is particularly important for companies whose valuations depend heavily on expectations of future AI demand.
2. The cost of building AI infrastructure is enormous
Developing advanced AI systems requires substantial investment in chips, data centres, electricity, networking, and specialised infrastructure.
Technology companies may continue spending heavily to remain competitive. The challenge is whether the additional revenue and productivity gains eventually justify those costs.
If companies invest more than they can earn back, the financial returns may disappoint, even if the technology itself proves useful.
3. Competition can change the investment story overnight
A major example came on 27 January 2025, when the emergence of Chinese AI startup DeepSeek’s lower-cost models unsettled investors.
NVIDIA shares fell nearly 17%, wiping approximately $593 billion off its market value in one session, a record single-day market capitalisation loss for a US company at the time. The Nasdaq fell around 3.1%, while semiconductor stocks suffered a particularly sharp sell-off.
Why such a dramatic reaction?
Investors began questioning whether AI systems could be developed and operated more cheaply than previously expected. If less expensive models could deliver competitive performance with fewer computing resources, demand for some high-cost chips and data-centre investments might not grow as strongly as anticipated.
The episode didn’t prove that AI was overvalued. It showed how quickly expectations can change when a new development challenges the assumptions behind an investment boom.
The sharp recovery in technology stocks over the following sessions also illustrated how volatile sentiment around AI had become.
So, is the stock market becoming too dependent on AI?
There are reasonable arguments on both sides.
The case for continued optimism:
- AI could create new revenue streams and improve productivity.
- Demand for computing infrastructure may continue to expand.
- Companies that successfully commercialise AI could generate substantial long-term returns.
The case for caution:
- Some valuations may already assume exceptional growth.
- Heavy infrastructure spending doesn’t guarantee attractive returns.
- Competition, cheaper models, and slower adoption could challenge existing expectations.
- When investor enthusiasm is concentrated in a limited group of companies, disappointment in those businesses can affect broader market indices.
The last point matters because large technology companies account for a substantial share of major US stock-market indices. When these companies rise, they can lift the broader market. When they fall together, the impact can spread to investors who hold diversified index funds as well.
However, AI’s influence on the market doesn’t automatically mean the market is too dependent on it. That depends on how much investors are paying for future growth, how broadly earnings growth is spreading across industries, and whether companies can generate returns that justify their investments.
The bigger question for investors
AI may genuinely transform industries, improve productivity, and create valuable businesses. At the same time, even a technology with enormous potential can become the focus of excessive market expectations.
The technology and the investment opportunity are two different things.
A company can be an AI leader and still be an unattractive investment if its share price already reflects unrealistic expectations. Equally, a volatile market reaction doesn’t necessarily mean the underlying business opportunity has disappeared.
The real test will be whether the revenue, profits, and productivity gains generated by AI can keep pace with the money being invested in it.
What do you think: Is the stock market becoming too dependent on AI, or are investors recognising a genuine long-term transformation?