Federal Reserve Chair Kevin Warsh used the Jackson Hole economic policy conference on August 28 to make an argument his institution has long resisted saying out loud: artificial intelligence may need to be treated not as a technology story, but as a fundamental input to economic growth — alongside labor, capital, and land.
The remarks, delivered at the annual symposium that central bankers and economists treat as the Fed's most important platform for policy signaling, put the central bank squarely into the AI debate. Coverage from 24/7 Wall St. via Yahoo Finance, The Washington Post, NPR, and Decrypt indicates the speech will shape how economists and investors think about AI's macroeconomic footprint heading into the fall policy calendar. For more context on this story, see our ongoing AI news.
'A Hinge Point in History'
Warsh framed the moment in unusually sweeping terms for a central banker.
"Times sure have changed. We've come to a hinge point in history," he said, according to remarks quoted by multiple outlets.
The core of the argument was that the Fed now recognizes AI as something categorically new in its models of the economy. "AI is a new variable, potentially a new factor of production that will have consequences both for the economy and for the conduct of monetary policy," Warsh said.
In introductory economics, the classical factors of production are labor, capital, and land. Adding a fourth is more than a rhetorical flourish. The Fed's forecasts of potential output — the amount the economy can produce before wages and prices overheat — rest on how those inputs combine. If AI genuinely functions as a fourth factor, the economy's non-inflationary growth ceiling rises with it, and a policy stance that looks restrictive today could look neutral tomorrow.
The $100 Billion Token Economy
To anchor the claim in evidence rather than hype, Warsh cited a figure that has circulated in industry reports but rarely in a Fed chair's remarks.
"Reports put annualised token sales for the two leading labs alone at more than $100 billion, an increase of 500% from just 12 months ago," he said.
Tokens — the fragments of text billed whenever a model reads a prompt or writes a response — are a rough proxy for actual AI usage by paying customers. That distinction matters. Data-center capital expenditure tracks capacity built; token revenue tracks capacity consumed. A 500% year-over-year jump in consumption, if the reports Warsh cited are accurate, is the kind of adoption curve that eventually shows up in productivity statistics.
And that, in part, was Warsh's point: it has not yet shown up in the official data. The gap between what industry reports suggest and what government statistics capture is exactly the measurement problem the Fed now faces.
What It Means for Interest Rates
Warsh was careful to note that these considerations will not immediately shape current policy decisions. The fed funds target upper bound has held at 3.75% since the easing cycle that ran through late 2025, and NPR reported that Warsh paired his AI commentary with a warning that inflation remains too high — a combination that has markets pricing the possibility of rate hikes rather than cuts.
The tension is deliberate. The productivity case for AI argues the economy can grow faster without inflation; the inflation data argue the opposite is true right now. Real GDP growth came in at 1.5% for the quarter ending April 1, 2026, below the 2-to-3% range the Fed's own guidance identifies as healthy, while bond markets have been leaning skeptical — the 10-year Treasury yield sat at 4.67% on August 27, near the top of its trailing-year range.
The Complement-or-Compete Question
Warsh distilled the debate into a single question: "Will the application of AI cause a significant, sustained rise in productivity across the economy? If so, when will token usage be complementary or competitive to labor?"
Both outcomes can be true at once, in different jobs. The wage data so far suggest AI has not yet moved the aggregate picture — real average hourly earnings in July 2026 stood close to where they were a year earlier, according to figures cited in the 24/7 Wall St. analysis.
The Fed Builds an AI Task Force
Perhaps the most concrete institutional signal in the speech was procedural. The Fed has established an internal task force on productivity and jobs to study the implications of AI — a sign that the reframing Warsh described is happening inside the building, not just at the podium.
Central banks have historically treated technology as background noise: important eventually, but rarely urgent. A standing task force, a Jackson Hole platform, and the factor-of-production framing together suggest the Fed now regards AI as a monetary policy problem to solve rather than a technology story to observe.
Why It Matters Beyond the Fed
For the AI industry, the speech is a double-edged validation. On one hand, a Fed chair citing $100 billion in annualized token sales is the strongest mainstream economic endorsement yet that AI demand is real consumption, not just capital spending. On the other, the same data will now feed into decisions about interest rates that shape how cheaply AI labs and data-center builders can finance the infrastructure buildout the industry depends on.
Investors got the message quickly. Analysis of the remarks across financial media over the weekend focused on what Warsh's framing means for discount rates and long-duration assets — growth equities and long bonds whose valuations are most sensitive to assumptions about future productivity.
The honest answer to Warsh's central question is that nobody knows yet — including the Fed. What changed at Jackson Hole is that the institution responsible for the answer has officially started trying to find one.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
