Warsh Names Tech Visionary Marc Andreessen To Lead New AI Task Force - Forbes

Marc Andreessen has spent thirty years betting private capital on the idea that software rearranges economies. This week the Federal Reserve gave him a role in deciding what that idea means for American interest rates. Chairman Kevin Warsh named the leadership of five new task forces on Thursday, and the one that should hold investors' attention is charged with assessing "the economic impact of new general-purpose technologies, including artificial intelligence, to inform the Federal Reserve's policy judgments." Andreessen co-leads it, and Warsh has asked the groups to deliver recommendations by the end of the year.

Central bank working groups rarely rate a second read. This one is different in kind. The Fed is building formal machinery to treat AI as an input to the price of money, and that machinery now has names, a mandate, and a deadline. Investors spent the week parsing chip stocks for signals about the AI trade. A quieter and larger signal came out of Washington.

What Warsh Announced

The five groups cover communications, balance sheet policy, data, productivity and jobs, and inflation frameworks, staffed by fifteen outside experts asked to rethink how the institution does its job. The rosters read like a central-banking hall of fame. Mervyn King, who ran the Bank of England through the financial crisis, co-leads communications. Raghuram Rajan, who ran India's central bank, joins former Fed governor Jeremy Stein on the balance sheet group. Greg Mankiw and the Nobel laureate Thomas Sargent take the inflation framework.

The productivity and jobs group is where the AI question lives, and it was staffed differently. Alongside Andreessen sit Charles I. Jones, the Stanford economist whose career work is the theory of technology-driven growth and who is currently on leave at Anthropic, and Asha Sharma, the Xbox chief executive who previously ran Microsoft's CoreAI product group. Three people, one shared prior: the technology is economically transformative, and the open questions are magnitude and timing.

The Greenspan Precedent

Warsh has been explicit about the model he intends to follow. He and the administration read AI as an internet-style productivity boom, the kind the American economy last experienced in the 1990s, and the reference matters because of what the Fed did back then.

In 1996 and 1997, with unemployment falling and conventional wisdom demanding preemptive rate increases, Alan Greenspan held rates steady. He had detected, in scattered corporate reports, productivity gains that had not yet surfaced in the official statistics, and he reasoned that if companies were producing more per worker, growth would not translate into inflation. He was right. Productivity growth accelerated from roughly 1.5% in the early 1990s to between 2.5% and 3% from 1996 through 2004, inflation stayed contained, and the expansion ran for years past the point where the old models said it should have been choked off.

The lesson Warsh drew from that episode is the operating theory of his chairmanship. A central bank that recognizes a productivity boom early can let the economy run. A central bank that misses one strangles the boom by fighting inflation that was never coming. The task force is the institutional version of Greenspan's scattered corporate reports, a body designed to detect the productivity shift faster than official statistics will show it.

A Committee Of Believers

The composition drew immediate attention, most of it skeptical. Andreessen's firm has billions riding on AI, Jones is spending his leave inside one of the frontier labs, and Sharma built Microsoft's AI platform business before taking over Xbox. The Washington Post noted that everyone named to the group has recently spoken or written in sharply positive terms about AI's economic effects, and Axios pointed out that Warsh and Andreessen have been friends for decades.

The conflict-of-interest reading writes itself, and it stops one step short of the useful conclusion. Committees telegraph their conclusions through their membership. The Fed convened a group that already believes AI is reshaping the economy and asked it what follows for policy, which means the direction of the recommendations is visible before the first meeting. What remains open is how far they go and how quickly the policy committee absorbs them.

What It Changes For Investors

The Fed's models deserve a word of realism here. They are tools, and the record says they are frequently wrong tools; the inflation call of 2021 is the freshest example. What makes this week matter is that markets price off those tools anyway. Much of the 2022 drawdown in growth stocks was institutions rerunning valuation models with new rate assumptions after the Fed changed its posture. When the reaction function moves, everything priced against it moves too, whether or not the underlying businesses have changed by a dollar.

The old reaction function treated strong growth and tight labor markets as inflation warnings that justified higher rates. A Fed that formally credits AI-driven productivity reads the same data differently: growth becomes evidence the technology is working rather than evidence of overheating. Economists argue about the mechanics, and some contend a genuine productivity boom should push rates up rather than down. The Greenspan episode is the useful referee. The argument over the theoretically correct rate raged through the late 1990s too, and what mattered for markets was the chairman's willingness to let the economy run while the theorists fought it out.

Most portfolios treat AI as an earnings story, weighed company by company. The task force points at the second channel, the one that gets far less attention: AI as a rates story. If the Fed concludes that productivity gains are real and disinflationary, it will tolerate expansions the old playbook would have cut short, and that reaches every asset discounted against the policy rate, including the large majority of stocks that have nothing to do with AI.

The recommendations are due by year-end. Between now and then, the tell is language: whether productivity starts doing real work in FOMC statements, whether Warsh's speeches lean harder on the 1990s, and what the task force actually publishes. If the Fed decides AI lets the economy grow without stoking inflation, rates sit lower than the old playbook allowed, and every valuation in the market gets measured against that lower bar.