It wasn't long after the launch of ChatGPT in November 2022 that talk of an artificial intelligence (AI) bubble started to form.
The introduction of the breakthrough AI chatbot set off an unprecedented arms race in the new AI technology, driving demand for Nvidia (NVDA +2.55%) graphics processing units (GPUs) through the roof and giving it a market value that's now above $3 trillion, rivaling Apple and Microsoft for the most valuable company in the world.
Plenty of naysayers have pointed out that for the billions of dollars spent on AI infrastructure, there's relatively little to show for it in end-user demand for AI products and services. There's no "killer app" in AI yet, and businesses and everyday users seem to be struggling to find ways to take advantage of the new technology.
One such skeptic is David Cahn, a partner at Sequoia Capital, a well-known Silicon venture capital firm and a backer of at least 70 AI companies. Cahn wrote a paper in June saying that the AI bubble was reaching a tipping point.
Image source: Getty Images.
A $600 billion AI shortfall?
As Cahn sees it, the tens of billions of dollars that are being spent on AI infrastructure need to generate an even greater amount in revenue for its owners to justify the expense.
For example, based on some back-of-the-envelope math, he estimates that if Nvidia's data center run-rate revenue reaches $150 billion by the end of the year, then the companies spending $150 billion on those Nvidia components would have to generate $600 billion in revenue to make back their investment.
That calculation is based on data center chips accounting for half of the cost to run those data centers and a 50% margin for the software that's run on those cloud infrastructure services buying Nvidia hardware like Amazon Web Services, Microsoft Azure, and Google Cloud.
He also argues that the supply shortage from late 2023 has subsided, GPU stockpiles are growing, and pure-play AI companies are only making minimal revenue, with the vast majority accruing to OpenAI. OpenAI recently reached $3.4 billion in annual run-rate revenue, but it's also on track for a $5 billion loss this year.
While Microsoft has talked up spending on Azure OpenAI, for example, Cahn makes a fair point that direct revenue from generative AI is relatively negligible thus far.
There's more to the story
Despite his belief that AI is in a bubble, Cahn characterized the bubble as a short-term one, saying he believes that "a huge amount of economic value is going to be created by AI."
However, he also called the current environment in AI, both among the major cloud infrastructure players and investors, a "speculative frenzy," implying that there will be a substantial pullback in valuations of stocks like Nvidia and its peers.
Big tech companies like Alphabet and Microsoft are already getting pushback from investors about their sharp increases in capital expenditures in AI as both stocks sold off after their earnings reports, and that resistance is likely to pick up speed until investors are convinced that they can generate the kind of revenue sums that Cahn spells out above.
NASDAQ: NVDA
Key Data Points
Is it time to sell AI stocks?
It's likely that there will be some kind of pullback in AI stocks like Nvidia at some point, as the first sign that demand isn't quite as strong as expected or that prices for GPUs are starting to fall will send investors heading for the exits.
However, there's a reason why so many tech titans believe generative AI is as transformative as the internet and why the bets are so big. It's still very early in the ramp of the new technology, and it will take years for its utility to grow and make today's spending look reasonable.
When you consider innovations on the horizon like self-driving cars, artificial general intelligence (AGI), and the creative potential of generative AI, the technology will be disruptive if it works, and in order to work, it requires the massive infrastructure being built now.
While a pullback in AI stocks is likely, over the long term, the future still looks bright for the AI sector as the technology looks poised to be as transformative as its champions hope it will be. A temporary pullback shouldn't deter long-term AI investors.
