AI’s got a supply-side problem
Federal Reserve Governor Lisa Cook said the AI buildout could keep inflation sticky into 2027 if shortages of materials and skilled workers don’t ease up. Translation: the data-center arms race isn’t just about chips and cloud contracts — it also needs concrete, copper, electricians, and enough people who can actually build the thing.
Why investors should care
That matters because inflation isn’t just a macro headline; it’s the stuff that shapes rates, valuation multiples, and how expensive it is for companies to borrow. If AI infrastructure demand keeps pulling on the same limited pool of labor and materials, the Fed could end up staring at a familiar villain: persistent price pressure that refuses to chill out.
The weird part
AI is often pitched as the ultimate productivity booster, the kind of tech that makes everything smoother and cheaper. But in the near term, the buildout itself can look a lot like the opposite — a giant construction project with a appetite for scarce stuff and human skill. Big picture: the AI revolution may be deflationary in theory, but the bill for getting there could be inflationary in practice.