Nicole Taionescu | iStock | Getty Images
This is the biggest question in the tech industry, and there are a wide range of views on all sides.
Everyone seems to be considering the possibility of an AI boom or bust, as artificial intelligence-powered spending is leading to record deals and valuations.
Economic bubbles result from a rapid rise in market or asset prices in a particular region, often fueled by speculation, followed by a crash that causes a rapid outflow of capital.
More than 1,300 AI startups are currently valued at more than $100 million, and 498 of them are AI “unicorns,” or companies valued at $1 billion or more, according to CB Insights.
Giants like Amazon, Meta, and Microsoft are spending billions of dollars to beef up their data centers, and big deals from OpenAI are being announced all over the place. Nvidia And others.
The deal and spending euphoria resembles the dot-com bubble of the late 1990s and even the 2008 financial crisis.
Some see innovation as having strong demand and money being “well spent.”
Here’s what CEOs and experts have been saying in recent days.
AI bubble or not, we continue to see opportunities because the end result is real dollars being spent on real capital investment with a very long runway for funding.
Anneka Torreon
Global Head of Private Banking, Wealth Management and Investments, ING
Treon said companies are spending a staggering half of their operating cash flow on AI efforts, but they have enough capital to fund it.
“I don’t think we need to be concerned about the source of that funding or the flow of that funding,” Torreon told CNBC’s “Power Lunch” last week.
Total global AI spending is expected to reach $375 billion this year and $500 billion by 2026, according to a UBS report. It won’t be clear until at least a year or more from now whether these huge outlays will actually produce the expected return on capital, Torreon said.
However, Treon believes the current market backdrop is optimistic.
“We’re seeing easing monetary policy, easing fiscal policy, strong earnings growth and a boom in capital spending,” he said.
We point out that the share of the economy devoted to AI investment is nearly one-third greater than the share of the economy dedicated to internet-related investments during the dot-com bubble. Therefore, we think there are enough similarities to make a judgment.
Jared Bernstein
Biden former CEA president
Bernstein told CNBC’s “Squawk Box” last week that soaring asset prices and extreme valuations indicate an AI bubble is a “likely outcome.”
He said bubbles are defined by large gaps between investment levels and actual “credible expectations” of future profits.
Bernstein pointed out that even though OpenAI only plans to generate $13 billion in revenue, it has already done about $1 trillion in AI deals, including $500 billion in data center construction projects.
“You have to consider future revenue,” Bernstein said. “But to us and many others, the disconnect between reliable, plausible expected future returns and this level of investment certainly looks booming.”
Despite The Magnificent Seven spending heavily on AI, Bernstein said most of its profits come from other areas such as advertising and cloud services.
“If you actually look at investments in AI, they tend to be a small percentage of (returns),” Bernstein said. “So it actually contributes to the bubble hypothesis.”
I think some of the investments we’ve seen so far are not in AI, but in the cloud and the power of the cloud. So I don’t think this is a bubble, but I do believe that this is capital well spent most of the time.
Speaking on CNBC’s “Squawk on the Street” last week, Fink said there is no doubt that “a surge of capital” is being poured into AI, and that he doesn’t see this as a sign of a bubble, but rather as a necessary investment for the U.S. to remain the world leader in AI technology.
“Investing in AI doesn’t just mean investing in GPUs and chips, it means investing in HVAC and IT, it means investing in power grids and power supplies,” Fink said.
BlackRock’s CEO says these huge bets will ultimately lead to some failures, but that major hyperscalers meta, microsoftand alphabet They are in a “really good position” to be a winner.
“That’s capitalism,” Fink said. “We’re going to have some big winners, we’re going to have some big losers…but if you have a diversified portfolio, you’ll be fine.”
Are we in an AI bubble? of course! …of course it is. So we’re hyped, we’re accelerating, we’re having a huge impact on the system.
Pat Gelsinger
Former Intel CEO
Gelsinger said the AI-driven market is already in a bubble, but it will be “a few years” before we see an end to it.
“As you can imagine today, the entire Internet and service provider industry is about to be replaced by us,” Gelsinger said on “Squawk Box” last week. “We still have a long way to go.”
The former Intel CEO believes that major disruptive technologies could be developed in the second half of this century and companies could begin to see substantial benefits from them, giving the bubble enough time to burst.
“That being said, things will change,” Gelsinger said. “These are fundamental improvements in AI efficiency that have happened this year.”
For me, the main element of a bubble is psychological excess. The price is not too high. And since I don’t feel that much of a manic state at this point, I didn’t label this incident as a bubble.
howard marks
Co-founder of Oaktree Capital Management
In an interview with CNBC’s Sarah Eisen last week, Marks said that while there is undoubted excitement about AI, he has not seen it reach a “critical population” that would indicate an obvious bubble.
“My reaction so far has been that valuations are not abnormal,” Marks said. “High but not crazy. And if things are high or low but not crazy, you can’t make observations that are likely to be correct.”
The veteran investor likened the current AI frenzy to the internet boom of the late 1990s. The Internet boom rightly promised to change the world, even though it ultimately rendered many companies worthless.
Marks warned of a common behavior seen in “bubble psychology,” where investors support any company with even a small chance of making huge profits. Marks sees signs of this pattern of excessive risk-taking in today’s AI boom, but said a true bubble phenomenon has yet to occur.
We certainly see a lot of evidence of bubble-like behavior in the AI space. We’re seeing things like cyclical revenue trading, and we’re seeing a lot of very aggressive price action.
ben inker
GMO Asset Allocation Co-Head
Inker told CNBC’s “Money Movers” last week that companies are shifting from free cash flow financing to relying on debt and a large stake from Nvidia.
“Whether Microsoft or Meta wants to invest a lot of money into a data center that has cash flow is another question,” Inker said. “OpenAI is not like that, and neither is xAI. Even Oracle is issuing a lot of debt to do this.”
Last year, OpenAI expected an operating loss of about $5 billion on revenue of $3.7 billion, and it remains in the red. Inker said this raises concerns about Nvidia’s $100 billion investment in OpenAI aimed at building data centers powered by Nvidia chips.
“What this whole ecosystem means is that the capital from the cash flow of the hyperscalers that have been funding them has dried up and now they have to be financed by debt, some very strange deals between Nvidia and AMD, and some money-losing companies that have huge capital needs,” Inker said.
Featured: $1 trillion worth of AI trades in one graph
