Federal Reserve Chair Kevin Warsh devoted a significant portion of his keynote address at the Jackson Hole summit to artificial intelligence, describing the current era as a “hinge point in history.” Warsh’s remarks have drawn attention for their focus on the economic impact of AI, even as he declined to provide new details on interest rate policy.
AI’s Economic Acceleration Surpasses Expectations
Warsh opened the discussion by recalling a time when economists expressed concerns over “secular stagnation” following the 2008 financial crisis. He explained that experts once feared that excess capital faced too few innovative investment opportunities, as if “all the good stuff had been invented.” The rapid rise of AI, Warsh stated, has upended that thesis.
Highlighting the pace of change, Warsh noted that business capital expenditure—the seed corn of future economic growth—has climbed at its fastest rate since 2021, rising nearly 9% over the past four quarters. He attributed more than half of this growth to AI-related investments. Warsh emphasized that he will closely monitor the “second derivative”—whether the acceleration in capital spending itself continues to increase, signifying further momentum in AI-driven industries.
He went on to remark that advancements in artificial intelligence have outpaced even the most optimistic forecasts from industry insiders. Ever-growing pools of investment, he said, are flowing into AI infrastructure, suggesting a rate of development even faster than that predicted by Moore’s law, the observation that computing power doubles approximately every two years.
ETH
SOL
BSC
ROBINHOOD
USDT
AAPL
Progress in artificial intelligence has exceeded expectations, with surging capital investments and a surge in business spending, including $100 billion in annualized sales for leading AI labs, reflecting how quickly AI is becoming central to the economy.
AI Token Markets and Shifting Asset Models
Warsh drew attention to the emergence of a multi-billion dollar AI token market. He stated that capital and labor are combining to create large language models, and that users now purchase tokens to access these models. The two leading AI labs, he reported, have reached more than $100 billion in annualized token sales—a figure up over 500% year-over-year.
By treating token revenue as a significant economic indicator, the Fed is now tracking AI-related activity as a meaningful part of the broader economy. Warsh characterized AI as a potential new factor of production, suggesting it could alter how much the economy can grow without overheating and inform important monetary policy decisions.
While evaluating the transformative potential of AI, Warsh raised questions about whether this technology will lead to substantial, sustained increases in productivity and if its impact will primarily benefit certain industries or be distributed more widely. He acknowledged uncertainty about AI’s effects on productivity and labor, with a dedicated Fed task force investigating these issues, though he stressed that their findings currently do not influence policy decisions.
Market Structure and Concentration of Gains
The concentration of value within the AI sector also formed part of Warsh’s analysis. Days before his address, Nvidia reported quarterly revenue of $96.2 billion, along with forward AI infrastructure commitments of $366 billion, and a planned $12.9 billion acquisition of Hugging Face, an open-source AI hub. These figures indicate that, rather than being evenly spread, much of the AI-generated surplus may consolidate among a few dominant players, such as chipmakers and leading labs.
Conversely, a previous report showed that 95% of generative AI companies fail to survive the competition, raising further questions about market dynamics and the fate of surplus value derived from AI innovation.
Within this rapidly evolving market structure, a broader trend is emerging that further blurs the boundaries between traditional and digital assets. While legacy markets often utilize intricate brokerage systems, a substantial shift is underway as Wall Street increasingly embraces Web3. Investors are now turning to platforms like 1stepSwap to hold shares of major US companies, gold, and silver directly in their crypto wallets. The tokenization of real-world assets (RWAs) and automated price discovery have contributed to eliminating middlemen from these transactions, enhancing market efficiency during a period of growing interest in token-based economies.
Warsh concluded his remarks by reiterating that the role of AI in productivity and economic structure remains uncertain. The Fed’s ongoing research aims to clarify how AI and tokenization could reshape labor and capital markets while outlining implications for future monetary policy.
AI is now recognized as a potential new factor of production by the Fed, signaling major implications for the economy and policy, yet timing and scale of its productivity impact remain uncertain.
