Fed Chair Kevin Warsh Outlined 3 Possible Futures for AI Yesterday. One Could Be Devastating for Stocks

AI & Compute

Warsh’s three AI scenarios test a trillion-dollar investment bet

Nvidia lost roughly $1 trillion in market value in less than a week as Warsh mapped three AI paths: stronger productivity, worker displacement, or weaker returns and stock pressure.

By
Rebecca Stone
Filed
Channel
AI & Compute
Read
4 min read

By Rebecca Stone

Nvidia’s shares lost approximately $1 trillion in market value in less than a week after CEO Jensen Huang’s departure, as Federal Reserve Chair Kevin Warsh set out three AI outcomes with sharply different consequences for growth, employment and stock prices.

That reversal lands amid trillions of dollars invested on the belief that AI will transform the global economy. Technology companies are building enormous data centers, semiconductor manufacturers are expanding production, and investors are paying premium valuations for expected beneficiaries. The premise: AI’s economic gains must justify the spending.

What did Warsh’s three AI scenarios say?

On November 16, 2026, Warsh spoke at the San Francisco Fed’s Productivity and Jobs Task Force Conference and mapped three paths for economic growth, employment and financial markets.

  • AI delivers: Productivity rises, GDP accelerates and employment remains healthy; stronger earnings could support valuations.
  • AI displaces workers: Productivity improves, but unemployment rises; corporate gains could come at consumers’ expense.
  • AI disappoints: Investment returns fall short, growth slows and unemployment increases; expensive technology stocks face valuation pressure.

The third path carries the sharpest market consequence. After years of capital investment and stock-market gains, disappointing AI returns could trigger declining equity valuations, slower economic growth and rising unemployment. The danger does not require AI to fail technically.

In Warsh’s framing, the risk reaches further: “the investment boom itself could become a source of economic weakness.” That is the commercial problem.

How does infrastructure spending transmit to chips?

The stakes extend well beyond semiconductor manufacturers. The article put the aggregate commitment at trillions of dollars without supplying a company-by-company total; Microsoft, Alphabet, Amazon and Meta Platforms are committing billions to computing infrastructure.

Their investment supports demand for Nvidia processors, networking equipment, electrical systems and data-center construction. Spending is not earning. An AI data center can operate successfully while generating less return than investors anticipated, and billions committed upfront can magnify even modest disappointments. Returns still matter.

The Nvidia reversal shows how quickly expectations can change before the industry’s long-term prospects resolve. The report linked growing scrutiny of the company to Huang’s departure and put the market-value loss at approximately $1 trillion in less than a week.

OpenAI CEO Sam Altman raised a related supply concern: multiple chipmakers may be assuming “they’ll capture the same future customer demand.” If suppliers collectively plan for more purchases than customers can make, excess capacity becomes a genuine risk. That would add pressure to AI investments whose economics depend on years of future growth.

Can productivity growth rescue investor returns?

Warsh was not predicting an AI collapse. He highlighted AI’s potential to reverse a long-running research problem: research efforts increased more than 20-fold since the 1930s, yet research productivity declined. AI could help scientists and businesses discover ideas faster, supporting continued investment.

Productivity does not guarantee that every AI company prospers. The internet transformed commerce without making every internet stock purchased in 1999 a winning investment. The same distinction applies to AI: broad economic gains can coexist with poor returns for companies whose valuations require years of growth.

Investors concerned about market concentration face another version of the issue. Broad index funds hold the same technology companies funding the AI expansion, so diversification across a market index does not remove exposure to their capital-spending and valuation risks.

What should investors and chip suppliers watch?

None of Warsh’s three scenarios is a forecast. They separate economic progress from investment returns: AI can transform productivity while disappointing shareholders who paid too much for anticipated growth. The practical response is not to abandon technology stocks, but to examine free cash flow, capital spending and valuations.

For chip suppliers, the hinge is whether customer demand matches the capacity suppliers plan to build. The greatest risk may not be that AI fails; it may be that AI succeeds without delivering returns sufficient to justify what Wall Street has already priced in.

The next test is whether actual orders can absorb planned chip purchases without creating excess capacity. If they cannot, infrastructure spending, supplier returns and stock valuations could weaken even as AI delivers broader productivity gains.

Source: 247wallst.com

Share this article:

More from Rebecca Stone

Rebecca Stone

Show full bio

Correspondent covering technology and product news in Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.

1,061 articles

Related articles

« Previous article