Whale Rock Capital founder and veteran hedge fund manager Alex Sacerdote laid out one of the clearest bull cases for the AI buildout on Episode 477 of the Invest Like the Best podcast. His argument is that AI-powered coding is the proven unlock of the entire AI wave, and he believes the market is barely getting started. Sacerdote backed that view with capital, investing in Anthropic at a $180 billion valuation when the underlying coding technology was, by his account, only 7-9 months old.
Why Sacerdote Believes Coding Is AI’s First Breakout Use Case
Sacerdote’s core argument is that AI-powered coding has become the first large-scale AI application with clear evidence of real-world adoption. As he put it, “Just coding alone has completely taken off.” To illustrate how quickly the technology has improved, he pointed to two respected technologists who, in his telling, “completely flip-flopped” on AI coding tools: Andrej Karpathy and Linus Torvalds.
Sacerdote cited Karpathy’s framing that last year’s tools wrote 20% of code while humans wrote 80%, a ratio Sacerdote says has now fully reversed. As he put it, Karpathy “hasn’t written a line of code except in English.”
That arc lines up with what Anthropic CEO Dario Amodei has described on the Dwarkesh Podcast, where Amodei noted that coding adoption inside Anthropic itself has been startling. “It’s seen such fast adoption within Anthropic, like, you know, coding is a lot of what we do,” Amodei said on the show. Readers can listen to the full conversation on the Dwarkesh Podcast.
The Half-Trillion-Dollar Market Math
Sacerdote built his market-size case from two inputs he supplied himself. He observed that some Anthropic employees were spending $100 a day on tokens, which he said extrapolates to $20,000 to $30,000 per year per heavy user. Layered against roughly 20 million coders worldwide, his conclusion was “a half a trillion-dollar market just from coding alone.”
That $500 billion figure is Sacerdote’s own extrapolation from his token-spend and coder-count assumptions. It is a bull-case projection, not an established or third-party-validated market size.
Early on the S-Curve
Sacerdote believes we are in the very early innings. He noted Anthropic currently has 14-15 million daily active users, yet believes the enterprise market is “less than 1% penetrated.” He cited Sundar Pichai’s estimate that true agentic AI users represent just “10 basis points of the knowledge workers of the world.”
His analogy was that in 1998, businesses knew they needed a website but struggled to build one. The inflection, in his words, is happening now: “A light switch this year went off in the enterprise.”
Pushing Back on the Commodity Narrative
A common bear argument is that frontier AI models are interchangeable commodities racing to zero margin. Sacerdote sees real differentiation, arguing that “Anthropic, they’re very good for anything that has to do with private equity and finance.” If model strength varies by domain, that supports pricing power and defensibility, weakening the thesis that AI models are simply commodities.
Skeptics, including The Lean Startup author, Eric Ries, have called AI valuations crazy and warned that productivity gains are exaggerated. Sacerdote counters that coding productivity has demonstrably taken off. Our own coverage on The AI Investor Podcast has highlighted the key difference between today’s AI boom and prior technology bubbles: much of the infrastructure buildout is being funded by real cash flows and substantial revenue generated by mega-cap technology companies.
At the same time, investors continue to debate whether parts of the ecosystem are developing the kind of circular investment dynamics that were seen during the dot-com era.
Key Takeaways
Sacerdote’s pitch is unusually concrete: coding is the use case with the clearest traction, the runway is long, and differentiation is real. He believes the market could reach a half-trillion dollars. Anthropic has confidentially filed for an IPO but is not yet trading, so retail investors cannot directly own the thesis Sacerdote is articulating. The size of the coding market is debatable, but the traction AI is seeing in coding is not.