Michael Burry – the investor known for predicting the pre-2008 housing crisis – has turned his attention to one of the market's most popular topics: artificial intelligence.
Burry recently deregistered his hedge fund firm Scion Asset Management, removing it from routine regulatory disclosure requirements. But he continues to invest actively and is redoubling his efforts in what he believes will be the next big mispricing in the markets.
At the heart of this view is Phil Clifton, the former associate portfolio manager at Scion, whose research supports the skepticism. Clifton argues that while the adoption of generative AI is accelerating, the costs of the industry's massive infrastructure buildout are not yet justified.
In his farewell letter to Scion investors in late October, Burry called Clifton “the most amazing thinker” he had ever met. CNBC obtained several of Clifton's research notes from earlier this year, written before he founded his own firm, Pomerium Capital, that help outline Scion's pessimistic thesis on AI.
The investment world “expects this technology to have far greater economic impact than it is likely to provide,” Clifton wrote. “Just because a technology is good for society or revolutionizes the world doesn’t mean it’s a good business proposition.”
Low margins
On the surface, the use of AI appears to be ubiquitous. According to the Pew Research Center, more than 60% of U.S. adults say they interact with AI at least several times a week. Still, Clifton said the economy was “surprisingly weak” on the demand side.
OpenAI – market leader and cultural phenomenon – is expected to reach over $20 billion in annual revenue this year, but that number is minuscule compared to the scale of AI expansion. According to Man Group, hyperscalers have quadrupled their capital spending in recent years to nearly $400 billion per year and expect to reach $3 trillion in the next five years.
“We expect that other generative AI services will not be sufficient overall to justify the amounts spent on infrastructure,” Clifton wrote.
History's warnings
Scion sees a clear historical parallel to the telecommunications boom of the early 2000s, when heavy investments in fiber optic networks far exceeded actual usage. U.S. capacity utilization has fallen to about 5% and wholesale telecom prices have plunged about 70% in a single year, Scion noted.
Clifton argues that cloud giants are currently in a similar race, building out AI infrastructure on the assumption that future demand will eventually catch up. However, if mass adoption of AI takes longer than expected, the economics of these massive data center deals could become unsustainable.
Some Big Tech companies are already starting to follow through on their commitments, he noted. Microsoft has canceled data center projects consuming 2 gigawatts of electricity in the US and Europe, citing oversupply. Alibaba's CEO has warned that a bubble is forming in AI infrastructure.
The Nvidia exposure
No company has benefited more from AI spending than Nvidia. The stock has surged along with unprecedented GPU orders from cloud providers. But Scion questions whether these customers will ever see an economic return on this investment.
Stock chart iconStock chart icon
Nvidia one year
A key element here is the depreciation policy. Tech giants have extended the lifespan of their servers to six years. Still, Nvidia's product cycles now expire every year, causing older chips to become functionally obsolete and less power efficient long before they are written off, Scion claims.
Nvidia has rejected that claim, saying its hardware stays productive for much longer than critics claim, thanks to the efficiencies achieved by the company's CUDA software system.
Still, Burry and other critics raise a contradiction. Nvidia says the latest chips are superior in terms of performance, efficiency and performance, but at the same time promises that older chips will remain economical. One of these defenses, it is said, must give way.
Burry has launched a new Substack newsletter to lay out his pessimistic thesis on AI. Whether generative AI ultimately proves to be a bubble remains to be seen, but for now, Burry is once again positioning itself on the cautious side of a rapidly evolving story.