The Essay Still Waiting to Be Written

Packy McCormick's "Power in the Age of Intelligence" is the best thing I've read on AI strategy this year. His argument: stop asking which software companies have moats against AI. Start asking which companies are positioned so that better AI makes them stronger. The winners won't sell tools to incumbents — they'll use AI to break the core constraint in an industry, seize the high ground, integrate vertically, and dominate. Rockefeller did it with refining. Carnegie with steel. Ford with the assembly line. Now it's SpaceX with launch costs, and a new generation of startups going after energy, aviation, and internet infrastructure.

The framework is good. The historical pattern-matching earns its keep. And his sharpest line lands clean: "If your technology is so good, why aren't you using it to compete?"

Here's the thing, though. McCormick writes from the venture capital chair. He's asking which companies to fund. The 99% of professionals reading about AI strategy aren't founding the next Standard Oil. They're sitting inside the organisations these companies intend to displace, trying to figure out what to do on Monday morning.

For them, "own the scarce asset in your industry" is about as actionable as "be Rockefeller." The concentration thesis is real — the academic evidence he cites is solid — but the essay treats it as a natural law when it's at least partly a function of a specific regulatory environment. Antitrust is not a footnote. And presenting SpaceX's plan for 10,000 annual Starship launches and lunar mass drivers alongside Carnegie's actual steel output is doing a lot of work that the word "plan" shouldn't have to carry.

The insight that survives all of this: when AI makes a capability abundant, value migrates to whatever remains scarce. That's worth internalising whether you're a founder, an investor, or a marketing director trying to justify next quarter's budget. But the essay about what that means for the people inside the incumbents — the ones who need to figure out whether they're the landowner or the labourer in McCormick's analogy — hasn't been written yet.

Somebody should write it.