Goldman economist warns AI's productivity payoff may take 15 years

A Goldman Sachs economist argues AI boosters may be badly miscalibrated on timing, noting the computing revolution took 15 years to show up in macro productivity data. The analysis found productivity actually got measurably worse for four years after the PC's 1981 commercialization, then flatlined, becoming statistically significant only in year eight. The drag, the research finds, has less to do with the technology than with humans reluctant to adopt it, sometimes even sabotaging rollouts. Goldman still expects AI to "meaningfully boost productivity" over the next decade, but the findings suggest investors pricing in fast gains could be disappointed.

View full digest for July 15, 2026