
The AI boom's massive infrastructure spending, particularly on data centers, is warping the U.S. economy, potentially leading to a bubble. Venture capitalist Paul Kedrosky argues that unprecedented levels of capital are being deployed into a narrow set of recipients and geographies, impacting GDP and diverting resources from sectors like manufacturing. This AI build-out differs from historical infrastructure booms like railroads due to the short lifespan of GPUs, creating a need for rapid recoupment of investment. The trend is also causing energy inflation and NIMBY-esque pushback, potentially leading to offshoring of data centers. Kedrosky suggests watching for increasing use of special purpose vehicles and delays in ancillary equipment as signs of a bubble.
Sign in to continue reading, translating and more.
Continue