A growing chorus of voices is warning there could be an artificial intelligence (AI) bubble, as companies whose fortunes are closely tied to the technology see their valuations skyrocket.
High-profile figures, from OpenAI CEO Sam Altman to Amazon founder Jeff Bezos, have suggested in recent months that investors have become overexcited about AI, as companies bet big on the technology with multibillion-dollar investments.
This has been compounded by concerns about the increasingly circular nature of AI spending, as the likes of Nvidia, OpenAI and AMD announce new deals that seem to feed themselves.
"The question is -- are we in an AI bubble?" said James Angel, an associate professor at Georgetown University's McDonough School of Business. "You never really know until afterwards whether today's prices were justified by the future cash flows of these companies or whether investors were overly exuberant."
Since the advent of OpenAI's ChatGPT in late 2022, AI has become a major draw for investors. Nvidia, once a relatively obscure company focused on producing chips for video games, has become the most valuable company in the world.
In July, the chipmaker became the first public company to surpass a market capitalization of $4 trillion, as its chips remain the lifeblood of the AI boom. It currently sits at a massive $4.5 trillion.
Tech giants, like Microsoft, Apple, Amazon, Google and Meta, have also seen expansive growth over the past three years as they seek to cash in on the AI craze, promising vast investments in AI.
Other firms have also felt the power of investors' excitement in AI. Oracle saw its stock surge 40 percent on a single day in September, after projecting massive revenue from several multibillion-dollar cloud computing contracts.
These huge gains have increasingly spurred concerns that stock prices are getting away from their underlying value, the dynamic that results in a bubble.
"When bubbles happen, smart people get overexcited about a kernel of truth," OpenAI's Altman told reporters in August, according to The Verge.
"Are we in a phase where investors as a whole are overexcited about AI? My opinion is yes," he said. "Is AI the most important thing to happen in a very long time? My opinion is also yes."
Bezos offered a similar assessment earlier this month, suggesting that this excitement might be clouding investors' judgment.
"[The] thing that happens when people get very excited, as they are today about artificial intelligence for example, is every experiment gets funded, every company gets funded," he said, adding, "Investors have a hard time in the middle of this excitement distinguishing between the good ideas and the bad ideas."
Bubble fears have continued to grow, particularly as observers question the seemingly circular nature of AI financing in the wake of a series of high-profile deals between key players.
Nvidia announced last month that it planned to invest $100 billion in OpenAI. The ChatGPT maker, in turn, intends to build 10 gigawatts' worth of data centers on the company's systems, which Nvidia CEO Jensen Huang suggested was equivalent to between 4 million and 5 million chips.
Just two weeks later, OpenAI announced it would purchase 6 gigawatts worth of chips from AMD. As part of the deal, the AI firm will have the option to take up to a 10 percent stake in the chipmaker.
The deals have raised concerns about vendor financing, in which a seller helps a buyer finance a purchase of its own product. This was a feature of the dot-com bubble in the late 1990s -- one of several reasons the current AI boom has drawn comparisons to the boom and bust that accompanied the rise of the internet age.
Bank of America analyst Vivek Arya acknowledged in a research note last week that the deals "bring back some unpleasant memories" from the period.
However, he argued that concerns about vendor financing are "highly overstated," suggesting such deals will likely represent a small portion of the expansive spending in AI over the coming years.
Goldman Sachs analysts similarly noted that there are "elements of investor behaviour and market pricing currently that rhyme with previous bubbles," pointing to vendor financing, rising valuations, increasing market concentration and heightened spending.
But they aren't ready to declare an AI bubble quite yet. The rise in stock prices has so far been accompanied by solid underlying growth rather than pure speculation, while spending has largely relied on free cash flow instead of debt, the Goldman analysts noted.
In the case of a bubble, these differences could also make markets more forgiving, Angel suggested.
"When you have a story stock, everything is based on the story, and the moment a crack appears in the story, the markets can respond violently," he told The Hill.
"But when you have a company that has a track record ... even if there's a slight glitch, I suspect the markets may be a bit more forgiving, and that it'll take longer for them to say, all right, where's the cash flow?"
Some have also sought to distinguish the potential AI bubble from others. Bezos suggested there is an "industrial bubble," which could "even be good" for society, given the inventions it produces.
San Francisco Federal Reserve president Mary Daly seemed to be of a similar mind in a recent interview with Axios, emphasizing that not all bubbles are financial.
"Research and economics call it more like a good bubble, where you're getting a ton of investment," she said.
"Even if the investors don't get all the returns that the early enthusiasts think when they invest, it doesn't leave us with nothing. It leaves us with something productive," Daly added.
However, the Bank of England in a meeting earlier this month warned of the potential for a "sharp market correction," noting that markets are "particularly exposed should expectations around the impact of AI become less optimistic."
Even amid all the excitement and fears surrounding AI, Callie Cox, chief market strategist at Ritholtz Wealth Management, emphasized that the technology's impact has yet to flow into the economy in a "significant way." Its impact is not really showing up in jobs or productivity data, she noted.
"Right now, a lot of hopes and dreams are being priced into the stock market around AI," Cox said. "And we just don’t have a lot of proof that all of these assumptions will be true. So, AI is a much bigger driver for the stock market than it is for the economy."
While a dip in the stock market could have some impact on the economy, Cox underscored that the job market is much more relevant.
"Overall, you want to watch the job market if you are trying to form an opinion about the economy," she added. "And right now, AI doesn't have much impact on the job market."