The Fed may be making the economy more concentrated by fighting the part of the economy it can actually hurt.

The Federal Reserve may be creating the economic concentration it is trying to fight.

That sounds backwards until you look at where the growth is coming from.

The U.S. economy is being pulled in two directions. AI is generating an extraordinary wave of spending on data centers, electricity, chips, construction and related infrastructure. At the same time, many of the parts of the economy that normally respond to higher interest rates are already slowing.

The Fed sees the first part.

The interest-rate-sensitive economy feels the second.

And that creates a strange feedback loop.

Higher rates make housing, commercial projects, small-business borrowing and ordinary investment harder to finance. But the companies driving the AI buildout have enormous cash flows, access to capital markets and a strategic reason to keep spending.

So the Fed’s policy can weaken the parts of the economy most exposed to rates without necessarily stopping the spending that is keeping aggregate growth elevated.

The Federal Reserve’s own July report already showed the split. Business fixed investment was running at an 11% annual rate in the first quarter, with much of that strength tied to infrastructure for AI services. Outside those areas, particularly offices and manufacturing structures, investment was weak.

The September Beige Book showed the same thing at the ground level. Residential construction was declining while nonresidential construction was increasing, with several districts specifically reporting a high concentration of activity around data-center projects.

That distinction matters.

Suppose a hyperscaler decides to spend billions on another data center.

A 25-basis-point increase in the federal funds rate does not necessarily stop it.

But a small developer deciding whether to build apartments has a completely different financing equation.

A small manufacturer considering a new facility has a different equation.

A homeowner has a different equation.

A business dependent on a bank loan has a different equation.

The Fed therefore has a tool that hits the economy unevenly.

And the unevenness may be getting worse.

This is where the Fed’s own internal debate gets interesting.

On September 28, Governor Lisa Cook acknowledged that AI-driven demand is pushing up prices for chips, computers and software. But she also made a distinction that is crucial to this entire debate.

She said some of those price increases reflect demand shifting toward AI-related sectors rather than an increase in economy-wide demand. She warned that trying to fight sector-specific inflation with monetary policy could be a mistake because the Fed’s tools are too blunt to target narrow sectors.

Then she voted for the September rate hike anyway.

Her reason was that the pressure may be spreading.

That is the actual argument the Fed needs to prove.

Not that AI spending is enormous.

Everyone knows it is.

Not that some AI-related inputs are getting more expensive.

That is happening too.

The question is whether th