BlockBeats news, November 17 — The US market is confronting a deeper question: whether AI will become the next productivity revolution, or first turn into a large investment cycle that raises the cost of capital and inflation. The governance dispute that recently erupted around NVIDIA reflects the other side of this competition. Employees' concern that a departure by Jensen Huang would drain core talent shows that the competitive advantage of AI companies rests not only on chips and capital, but more on talent, technical decision-making and organizational integration. As companies must keep committing enormous capital to stay ahead, markets have begun to reassess the sustainability of AI investment and whether it can eventually convert into profit and productivity.
The same question pushes the Fed into a policy dilemma. US third-quarter GDP grew at an annualized 3.4%, and the October unemployment rate stood at 4.1%, so the economy has not visibly stalled, yet energy prices and services inflation still limit the room for cuts. More notable is that AI infrastructure is forming a new source of demand: data centers, semiconductors and power investment raise capital expenditure and energy demand in the short run, while potentially lowering services costs and improving productive efficiency over the long run. The problem is that the productivity dividend has not yet shown sufficiently in economic data, while the demand pressure from the investment already exists. Markets have therefore raised their sensitivity to monetary tightening; the 10-year Treasury yield broke above 5% and short-term yields rose in step, indicating investors no longer focus solely on when the Fed will cut but also demand greater term and inflation risk premia.
What Warsh faces is precisely this structural shift. He has begun to examine, through an AI working group, technology's impact on productivity, inflation and monetary policy, because the framework that judged inflation from demand and employment data may not be adequate to explain the changes in supply and demand brought by AI. If AI ultimately does lower operating costs for companies, it will help ease medium- and long-term inflation; but if computing and energy investment keep expanding while productivity gains fail to keep pace in time, the Fed may hold rates high even with the economy still resilient. This also explains why the White House wants lower borrowing costs while Treasury officials are unwilling to over-tighten in the face of an energy shock, and why market disagreement over the December policy direction has widened.
On another front, the debate inside the Fed over its balance sheet is likewise part of this transition. Hammack supports distinguishing temporary bond purchases during financial-market disorder from ordinary monetary policy operations, so that emergency liquidity support is not mistaken by the market for a new round of quantitative easing. Should the idea develop further, it would allow the Fed to keep room for handling liquidity risk in financial markets while maintaining its tightening stance. For crypto markets, what truly warrants attention is not the size of any single rate move, but whether AI investment keeps absorbing market liquidity, whether dollar funding costs stay elevated, and whether the Fed can balance inflation control against financial stability. When the capital demand of AI is realized ahead of its productivity returns, high-valuation assets face a rising discount rate and constrained liquidity at the same time.
