Demand is arriving before the productivity dividend
The infrastructure boom creates immediate orders for scarce equipment, skilled trades, grid connections, and capital. Productivity improvements, by contrast, depend on adoption, complementary process changes, reliable models, and time. Monetary policy has to manage the economy during that gap rather than assuming the payoff has already arrived.
The Federal Reserve's June meeting minutes describe continued AI-related capital spending as support for growth. Officials have also discussed sustained pressure on technology-product and electricity prices. Both can be true: AI can raise long-run capacity while straining near-term supply.
One forecast now supports many balance sheets
AI investment is no longer funded only from the cash reserves of large technology companies. Debt markets, utilities, developers, equipment suppliers, and local economies are increasingly tied to the buildout. That spreads opportunity and distributes the downside if utilization, pricing, or model economics disappoint.
Regulators and investors need scenario tests that delay the productivity payoff, increase power and construction costs, or reduce demand growth. The question is not whether AI will matter. It is whether the financial system can absorb a slower path than its most expensive projects assume.
Go to the source
Read the evidence behind this analysis. External links open in a new tab.
Reuters — AI investment enters the Federal Reserve's risk calculus Federal Reserve — June 2026 policy meeting minutes Federal Reserve — AI, the economy, and the financial system
