Powell – unlike the dotcom boom, AI spending isn't a bubble (fortune.com)

Federal Reserve Chair Jerome Powell told reporters the AI investment wave is fundamentally different from the dotcom bubble: companies driving the surge have real earnings, business models and are financing build-outs largely through corporate cash flow rather than speculative debt. He said AI spending isn’t primarily a rate-sensitive phenomenon but a structural, productivity-driven capital investment—hundreds of billions into data centers and semiconductors, with firms like Nvidia, Microsoft and Alphabet leading massive capex plans. Goldman Sachs estimates AI-driven productivity could be worth $8 trillion in present value (up to $19 trillion in high scenarios) and notes AI investment is still under 1% of U.S. GDP versus 2–5% in past tech cycles, implying room to grow. Powell also pointed to concrete, real-economy effects: surging demand for power, faster grid expansion, cranes, concrete and other capital goods as evidence the boom isn’t just financial froth. JPMorgan projects AI infrastructure could add roughly 0.2 percentage points to near-term GDP. He tempered optimism with caution—outcomes remain uncertain, gains are concentrated among a few capital-intensive firms, and automation may slow hiring (recent layoffs cite AI). Powell warned it’s too early to call a permanent productivity revolution, noting job growth is “pretty close to zero,” so policymakers and firms must monitor distributional and labor-market risks even as investment scales up.