Concerns about an artificial intelligence bubble are growing
CFO/Sipa USA/Alamy
Hundreds of billions of dollars spent on AI appear to have inflated a global financial bubble that is now ready to burst, leaving companies and investors at risk of huge debts that cannot be serviced by the meager revenue generated by current AI services. But what does this mean for the future of the technology behind this financial feeding frenzy?
In recent weeks, warnings about a potential artificial intelligence bubble have come from International Monetary FundBank of England, head of the largest US bankand even OpenAI CEO Sam Altman. “It wasn't just a stock market bubble, it was an investment bubble, it was a public policy bubble,” he says. David Edgerton at King's College London.
The circular nature of some of the deals between major AI players is also surprising. For example, Nvidia which creates the GPU chips that are fueling the artificial intelligence boomrecently invested up to $100 billion in OpenAI to help the company could build a new data center full of Nvidia's own chips. OpenAI, in turn, agreed to a deal that could ultimately lead to 10% stake in rival chip maker Nvidia, AMD..
Concerns about the bursting of the artificial intelligence bubble become even more apparent when one realizes the scale: at least $400 billion is spent on data centers every year, according to Morgan Stanley Wealth Management. And although the US GDP grew by 3.8 percent in the second quarter of the year, according to Jason Furman of Harvard University, if data centers are taken out of the equation, it will lead to in the first half of the year it grew by barely 0.1%..
Karl-Benedict Frey Oxford University says this kind of aggressive deal-making is not unusual in the history of technology – in fact, it would be unusual if the global economy managed to invest in infrastructure for a new technology at exactly the right pace to meet demand. “It's quite common: the same thing happened with the railroad boom, the same thing happened with the dot-com bubble,” he says.
The question is whether the fallout from the AI bubble will only harm the companies involved or could have wider implications. Frey notes that many of these extremely expensive data centers are actually being built “behind the balance sheet.” This involves the creation of new companies, backed by outside investors or banks, who create and own the assets, taking on both the risks and potential rewards.
As a result, we don't know enough about who is at risk. A data center could be financed by a dozen tech billionaires or big banks – and if their losses are big enough, a banking crisis could send shockwaves through the entire economy. “That doesn't mean a financial crisis is imminent, but it's a little opaque. And when things are opaque, there's usually some risk,” Frey says.
Benjamin Arold Cambridge University says the giveaway is the ratio of profits to company valuations, which shows how disconnected public opinion is from the actual money a business makes. He says these numbers are a wake-up call for tech companies today.
“The last time it was this low was 25 years ago, and if you remember, 25 years ago we had the dot-com bubble,” Arold says. “It might go well, but I wouldn’t put my money on it.”
James Poskett from the University of Warwick, UK, believes we're heading towards a fix in the AI industry that could spell the end for many companies, but he says it's definitely not the end for the technology itself. “It's important not to confuse this with the idea that the technology is imperfect or will soon disappear,” Poskett says. “There may be an AI collapse, but that doesn’t mean we won’t have AI.”
Just as the consolidation of numerous railroad companies after the crash left us with a rail network, and the collapse of technology companies during the dot-com crash left us with a legacy of vast fiber-optic networks, we will still have useful technologies, Poskett says.
For consumers, the bursting of the AI bubble will likely mean slightly less choice, perhaps higher access fees, perhaps slower upgrade rates. This may force us to face the reality that using an extremely expensive tool like GPT-5 to write an email is like using a sledgehammer to crack a nut, and that the true cost of using it has previously been hidden by the frantic AI arms race. “There are a lot of free lunches at the moment, but at some point these companies will have to make a profit,” Poskett says.
Topics:
