Three years in the past, Sequoia accomplice David Cahn was one of many first folks to do the mathematics and put a quantity on on the implications of Silicon Valley’s titanic spend on AI infrastructure.
In 2023, he was reacting to Nvidia’s reported annual GPU income of $50 billion. Beginning with that determine, and including within the implied prices of working the information facilities and the margins for his or her operators, he deduced that $200 billion in income could be required to pay again the up-front funding.
He took it as a problem, asking entrepreneurs to give you AI services and products to utilize, and generate income from, all that infrastructure. Quick ahead to immediately, including up three years of hyperscaling, and Cahn’s acquired a new quantity on AI infrastructure spending for 2026: $1.5 trillion.
All informed, he calculates that the AI business should earn $3 trillion to justify all these chips and different information heart expenditures. And that’s most likely an underestimate—the rising prices of reminiscence and the growing use of unique or inference-specific chips will drive that quantity up. “Just lately,” he writes, “the required income per GW of CapEx has sharply elevated because of these bottleneck dynamics and rising prices of development.”
On the opposite aspect of the ledger, Anthropic is assumed to have hit $60 billion in ARR, whereas OpenAI reportedly earned $13 billion in 2025 (though in November 2025, it stated it was at $20 billion ARR) and is presumably making extra this yr. However there’s clearly a big hole to be closed.
Somebody minding that hole is Torsten Slok, the chief economist at Apollo, the enormous asset supervisor. In a current be aware, he factors out that the hyperscalers — Google, Meta, Microsoft and Amazon — are all predicting huge accelerations of their free-cash circulation in 2028. That’s, they anticipate to see the pay-back from all these chips they purchased.
What in the event that they don’t? Slok notes a threat we’re presently seeing throughout AI utilization: Extra organizations turning to cheaper open weight fashions, typically Chinese language, not these constructed by the frontier labs, and total token costs falling. OpenAI’s newest mannequin, per CEO Sam Altman, is 54% extra token environment friendly on coding duties. That’s good for customers fretting about the price of their AI brokers, however it could be unhealthy for firms constructing token factories ought to customers not wildly improve their total token utilization with them.
Slok worries that if hyperscalers don’t meet their money circulation objectives, the market response might be extreme—
“with a lot driving on so few names,” he writes, “a slower payoff wouldn’t simply be a sector downside, it could threat tipping the economic system into recession and the S&P 500 right into a correction.”
Simply one thing to remember have in mind as you’re herding your AI brokers towards cheaper tokens.
While you buy by hyperlinks in our articles, we might earn a small fee. This doesn’t have an effect on our editorial independence.
