The artificial intelligence boom may seem unstoppable, but a growing number of investors and observers are increasingly worried that this surge resembles a bubble on the verge of bursting.

After the Nasdaq technology index rose more than 50% from its April lows, it has fallen nearly 5% this month. Investors fear it may take longer than expected to generate the major profits they had hoped for after pouring trillions of dollars into the next wave of technology.

Those who witnessed the dot-com bubble and its collapse in the early 2000s say parts of todays enthusiasm feel familiar. Optimists, however, believe the situation is different this time.

Nvidia, the AI-focused chipmaker, has led the stock-market rally, becoming the worlds most valuable company on the back of investor excitement surrounding artificial intelligence. The Santa Clarabased firm produces advanced chips used by tech companies to train AI models, power data centers, robotics, and more.

Both bullish and bearish investors were waiting to see what Nvidia would reveal about the state of its business in Wednesdays earnings report. The company kept optimism alive after reporting quarterly earnings and guidance that exceeded analyst forecasts. Its shares jumped more than 4% in early after-hours trading.

Nvidias chief executive, Jensen Huang, said during a call after the results: There has been a lot of talk about an AI bubble. From our point of view, we are seeing something entirely different and just to remind you, Nvidia is not like any other accelerator. We outperform at every stage of AI.

From social media to self-driving cars, Huang stressed that AI capable of generating content and performing tasks without human intervention will affect every industry.

Nvidias results may help revive AI-linked market momentum. Still, investors and analysts remain concerned about whether current stock valuations are justified for all companies entering the AI race. After the dot-com bubble, many companies disappeared, but those that survived are now among the worlds largest and most profitable firms.

The extremely high valuations of Silicon Valleys tech giants and other major AI players have pushed investors to question when and how their bets on the future of technology will pay off. Tech firms have become more interconnected as they invest hundreds of billions of dollars into one another, as well as into data centers, AI research, and generous employee compensation packages.

In September, Nvidia said it plans to invest up to 100 billion dollars in OpenAI the maker of ChatGPT to fund extensive construction of data centers hosting equipment used to store and process the enormous volumes of information required to run AI systems. OpenAI has also committed to purchasing at least ten gigawatts of Nvidias AI chips for its data centers.

According to an October research note from New Street Research, the capital expenditure needed to meet OpenAIs computing requirements could reach 130 billion dollars by 2027, meaning OpenAI alone could spend 52 billion dollars on Nvidia technology.

Despite its massive valuation of around 500 billion dollars, OpenAI continues to lose billions as it spends heavily on infrastructure, computing capacity, and other expenses.

OpenAI chief executive Sam Altman said in a talk at Stanford University last year: Whether we lose 500 million dollars a year or 5 billion or 50 billion, it doesnt matter. I truly dont care. Its going to be expensive but its absolutely worth it.

But as the losses pile up, investor concerns have grown.

Around 45% of global fund managers surveyed by Bank of America said there is an AI bubble that could negatively affect the economy and markets.

Debate will continue over whether a bubble truly exists.

Samuel Hammond, chief economist at the American Innovation Foundation, said he does not believe AI investments are in a bubble, though he expects winners and losers: Companies getting massive valuations only because they added the word AI to their pitch but fail to execute could see their value collapse to zero. But most stock-market gains are being driven by large-cap tech companies like Nvidia and Google.

Hammond noted that tech firms are financing these huge data-center projects through equity rather than debt, reducing the likelihood of a bubble ready to burst.

Strategists at Goldman Sachs wrote in an October research paper that although there are risks of overinvestment, technology companies have delivered earnings growth and maintain strong balance sheets: While the success of dominant tech firms is evident, this does not necessarily imply a market bubble about to burst.

However, Gary Smith, an economics professor at Pomona College and author, warned of an AI bubble, pointing to OpenAIs losses, circular funding among tech firms, and limits to AI capabilities.

In an opinion piece for MarketWatch co-written with Jeffrey Funk, he wrote: OpenAI is in a very fragile position and when the AI bubble bursts, it will be among the first casualties.

Some analysts have compared the current data-center boom to the 1990s telecom boom, when companies invested 500 billion dollars to lay fiber-optic cables to meet the rapid growth of internet usage, resulting in a massive surplus of unused dark fiber that sat idle for years.

Google CEO Sundar Pichai told the BBC that the tech sector has gone through periods of excess: We can look back at the internet. Clearly, there was a lot of overinvestment but nobody doubts now that the internet was a profound transformation.