AI Productivity Won't Clear the Fed's Path

The New York Times piece sells a clean story: an “A.I. productivity boom” lowers inflation pressure and clears the runway for more rate cuts, a view it pins to Trump’s Fed pick. Sounds tidy on paper. In practice, the transmission from software efficiency to easier monetary policy is where the story breaks.

Let’s start with what the article gets directionally right. Productivity does matter for inflation and rates. Higher output per worker can, in theory, ease cost pressure, support real wages, and give the Fed more room to cut without overheating demand. Economists have been chasing that holy grail for decades, and AI is the latest vessel for that hope.

But turning that theoretical channel into a near-term policy guide is where the wheels start to wobble.

Productivity is not a magic wand

Yes, efficiency gains should help lower unit labor costs and, all else equal, take some pressure off prices. The piece leans heavily on that textbook logic. Frankly, the math doesn't lie in principle: more output for the same labor input should, eventually, be disinflationary.

“In principle,” however, is doing a lot of work. The article slides from “AI could boost productivity” to “the Fed could then cut rates” without wrestling with three hard constraints: what we can actually measure, when we measure it, and how gains get passed through.

Official productivity data are backward-looking and noisy. By the time a “boom” shows up cleanly in those series, the Fed will already have been reacting for months to realized inflation and labor-market data. A smattering of AI pilot projects, tech earnings calls full of buzzwords, and a few efficiency anecdotes at big firms do not equal broad-based, economy-wide productivity strong enough to anchor policy.

Even when companies do capture real savings from AI, the paths those savings follow are not policy-friendly or even consumer-friendly by default. Some firms lower prices. Others plow gains into product expansion, new hires in adjacent roles, or simply higher margins. Some, like the largest platforms, have the market power to keep prices steady and let shareholders enjoy the spread. The Fed doesn’t set rates based on hypothetical price cuts managers might deliver if they feel generous.

Right now, treating a potential AI-driven productivity bump as a near-term lever for rate cuts is closer to storytelling than risk management.

Narratives, politics, and the Fed’s blind spot

Here’s the part the article tiptoes around: when a Fed nominee publicly connects a technology narrative to future rate cuts, they’re not just opining on economics — they’re guiding expectations. Markets trade that story. Politicians repeat it. And suddenly, what should be a conditional, data-dependent scenario starts to look like a soft commitment.

I spent years on trading floors watching how clean narratives get turned into blunt positioning: “AI means higher supply, so lower inflation, so more cuts.” Then every data release is forced into that frame. When the numbers don’t instantly validate the story, pressure lands on the Fed to “look through” sticky inflation or a tight labor market because the model says productivity is coming.

Central banks need distance from hype cycles, especially ones with clear political fingerprints. The article’s framing — productivity as justification for loosening — flips the order of operations. First you see confirmed disinflation; then you cut. You don’t cut because you expect disinflation from a technology adoption curve that hasn’t fully started.

The optimistic case — and its problems

The strongest counterargument is actually quite reasonable: if AI really does cut costs across logistics, software, and professional services, and if that cost compression bleeds into both headline and core inflation while employment holds up, then the Fed will have textbook cover to ease.

That’s logically coherent. It’s also doing a lot of conditional work.

You still have the distribution problem: narrow productivity gains won’t move aggregate inflation much. Broad gains can create offsetting frictions — labor reallocation, wage premiums for scarce technical skills, investment waves that juice demand. Those second-round effects can keep core inflation stickier than the first-round efficiency shock would suggest.

We’ve seen a version of this movie before. The late-1990s tech-driven productivity surge coincided with strong growth and, for a while, benign inflation. The Fed did allow real rates to drift lower than previous cycles would have implied. But that was after years of observed data, not a bet placed on the promise of networking hardware and the early internet. And even then, the landing wasn’t exactly gentle.

More recently, look at large platforms aggressively rolling out AI tools. Some are cutting costs and headcount. Others are using AI to build new products and ad formats that expand revenue opportunities rather than undercut pricing. The “productivity dividend” is showing up more clearly in earnings calls than in grocery bills.

What the AI-cut story leaves out

Three practical risks get short shrift in the article’s framing.

First, timing: productivity that actually matters for monetary policy arrives with lags, both in implementation and in measurement. Betting on it now is betting blind.

Second, measurement: AI’s biggest impact may initially show up in product quality or entirely new services, not in lower sticker prices. Our metrics are not great at distinguishing “same price, better product” from “inflation stayed high.”

Third, distributional spillovers: if AI concentrates gains among dominant firms and highly skilled workers, you can end up with pockets of intense wage pressure and strong pricing power, which is not exactly a recipe for clean, low, stable core inflation.

Let’s be real: central banking is risk management under uncertainty, not a technology conference keynote.

If Trump’s Fed pick wants to hinge a case for cuts on an AI productivity boom, the real test isn’t the story — it’s the guardrails: which indicators they would track, what evidence would actually trigger a change in stance, and how they’d respond if wage and price data refuse to cooperate with the model.

The article’s headline promises a path cleared by AI; the more likely outcome is a Fed that still has to bushwhack through the same old inflation and labor data, with an extra layer of politically charged tech optimism fogging the route.