At Jackson Hole, the Fed Chair Called AI a Possible New Factor of Production. Why That Matters for Interest Rates

Behind the technological language was a major economic question.

What happens to inflation and interest rates if AI allows the economy to produce much more with the same resources?

Start with productivity

Productivity measures how much output can be produced from a given amount of input.

Imagine two accounting firms with identical numbers of employees.

If new software allows one firm to complete twice as much work during the same number of labor hours, its productivity has increased.

At the scale of an entire economy, persistent productivity growth can raise living standards because more goods and services can be produced without requiring proportional increases in labor or capital.

This matters enormously to the Federal Reserve.

One challenge for central banks is estimating how quickly the economy can grow without producing excessive inflation.

If AI significantly increases productivity, the economy may be able to grow faster before running into capacity constraints.

The difficult part is determining whether that improvement is already occurring.

Warsh explicitly framed this as an unresolved question, asking whether AI would generate a sustained productivity increase and when that increase might arrive.

Why AI could lower some inflation pressures

Suppose a company previously needed ten employees to process a certain volume of routine administrative work.

AI allows the company to handle the same workload with fewer labor hours.

The company's cost per unit of output may decline.

If similar productivity gains occur across many industries, businesses could increase production without costs rising at the same speed.

That creates room for stronger economic growth with less inflationary pressure.

The effect would not necessarily appear immediately.

Companies first have to buy new technology and redesign workflows. Employees need training. Some AI applications may fail to generate meaningful savings.

Productivity revolutions often take time to appear in national economic statistics.

That delay creates a policy challenge. A central bank cannot confidently assume that future efficiency gains have already increased the economy's current productive capacity.

The investment boom can push inflation the other way first

AI also requires enormous physical investment.

Data centers need semiconductors and electricity. They require construction and specialized equipment.

This creates demand in the economy before all of the promised productivity benefits have arrived.

Warsh pointed to the extraordinary amounts of capital flowing into AI infrastructure.

That creates a fascinating sequence for monetary policy.

AI could eventually make the economy less inflationary by improving productivity.

During the buildout, the investment required to create that future could increase demand for scarce resources and contribute to price pressure.

A central bank therefore has to distinguish between today's construction boom and tomorrow's productivity gains.

The timing determines the policy implications.

AI might also change the labor market

Warsh raised another unresolved issue: whether AI usage will complement workers or compete with them.

Those outcomes have different consequences.

If AI allows employees to produce more while companies continue hiring, productivity could rise alongside wages and economic output.

If AI substantially replaces certain categories of labor, employment could shift between industries or occupations.

The Federal Reserve cares because maximum employment is part of its mandate.

Technology has displaced some forms of work throughout economic history while creating others. What makes AI particularly interesting is the range of cognitive tasks it may affect.

The adjustment may therefore reach industries that historically viewed automation as primarily a manufacturing issue.

Warsh also changed how the Fed communicates

The speech contained another important message for investors.

Warsh criticized excessive reliance on forward guidance, the practice through which central bankers signal likely future policy decisions.

He argued that regular forward guidance can cause markets to rely too heavily on the Fed's own signals. The central bank then watches market prices that were themselves influenced by Fed communication, creating what he described as a hall-of-mirrors problem.

This matters because investors have spent years trying to interpret every phrase from Federal Reserve officials as a clue about future interest rates.

Warsh's approach suggests that markets may receive fewer explicit hints.

That could make economic data more influential.

If investors cannot depend on the Chair to telegraph the next meeting, each employment or inflation report can create a larger adjustment in expectations.

Why uncertainty can increase market volatility

Financial markets prefer clarity because asset prices depend heavily on future interest rates.

A bond investor wants to know whether short-term rates are likely to rise.

A business considering a loan needs to understand financing costs.

Stock investors use interest rates when valuing future corporate profits.

Central banks cannot remove uncertainty from the economy.

Forward guidance can reduce some uncertainty by explaining the likely direction of policy.

Warsh's argument is that excessive guidance creates a different problem by encouraging financial markets to anchor too closely to central-bank forecasts that may later prove wrong.

The tradeoff is subtle.

Less guidance could make markets more responsive to genuine economic information. It could also cause larger swings when investors disagree about what the Fed will do.

AI turns an old Fed problem into a new one

Central banking has always involved estimating an economy that cannot be observed perfectly in real time.

Economists do not know the exact sustainable growth rate of the economy at any moment. They estimate it from incomplete data.

AI makes that estimation more difficult because a structural transformation may be occurring at the same time policymakers are trying to control inflation.

If AI raises productivity faster than expected, the Fed could keep interest rates too high because it underestimates how quickly the economy can grow.

If policymakers assume AI has already improved productivity when the gains remain years away, they could allow demand to grow too quickly and worsen inflation.

That is why an AI discussion belonged in a monetary-policy speech.

The Federal Reserve is ultimately trying to answer a question about the economy's capacity.

Artificial intelligence may change that capacity.

Until economists understand by how much, every interest-rate decision has another source of uncertainty built into it.