Photo: Photo: Steve A Johnson / Pexels / Pexels
Kevin Warsh called the current moment a 'hinge point in history' as business capital spending on AI infrastructure surges and the Fed scrambles to understand the economic fallout.
The Federal Reserve’s top official just admitted that artificial intelligence caught everyone off guard, including the people who were already bullish on it.
Fed Chair Kevin Warsh used his keynote at the Jackson Hole Economic Policy Symposium on August 28 to deliver a message that landed somewhere between awe and alarm: AI progress has outrun even the most optimistic forecasts from prior years. He described the current moment as a “hinge point in history,” a phrase that carries considerably more weight when it comes from the person steering US monetary policy.
The numbers behind the warning #
Business capital expenditure growth hit roughly 9% over the trailing four quarters. More than half of that spending went directly into AI infrastructure, meaning data centers, chips, and the computing backbone that powers large language models.
Token sales for leading AI labs reportedly topped $100B, a 500% increase from the prior year.
The Fed has been paying attention. AI references have appeared at least 18 times in recent FOMC minutes, with policymakers weighing productivity gains on one side of the ledger and financial stability risks on the other. Earlier in 2026, the central bank appointed a dedicated task force focused on productivity and jobs, specifically to assess how AI is reshaping the labor market and output potential of the US economy.
Why the Fed cares about large language models #
Central bankers don’t typically spend their summers talking about neural networks. But AI creates a genuinely novel problem for monetary policy.
If AI delivers the productivity gains its proponents promise, the economy can grow faster without generating inflation. In that scenario, the Fed could afford to keep rates lower for longer, because growth wouldn’t automatically translate into rising prices. But there’s a darker scenario that clearly occupies space in Warsh’s thinking. If the massive capital flowing into AI creates asset bubbles, concentrates market power in a handful of companies, or displaces workers faster than the economy can absorb them, the Fed faces a much messier situation. You’d potentially get inflation from supply chain disruptions in the labor market, combined with financial instability from overvalued AI assets.
The productivity question #
Warsh’s framing as a “hinge point” echoes a debate economists have been having since the mid-2010s. Productivity growth in the US has been sluggish for decades, and there’s a long history of promising technologies that were supposed to fix that problem but didn’t, at least not on the timeline their advocates predicted.
But productivity is notoriously difficult to measure in real time, and even harder to attribute to a single cause. The Fed’s new task force exists precisely because the central bank doesn’t want to be caught flat-footed. Either direction of error, overestimating AI’s impact or underestimating it, leads to a policy mistake.
Investors watching the Fed for rate signals should note the tension embedded in the FOMC minutes. Eighteen references to AI in recent transcripts means this isn’t a side conversation anymore. It’s becoming central to how the Fed thinks about the trajectory of the economy, which means it’s becoming central to how the Fed sets interest rates.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our