AI and Productivity: Can Technology Rescue Stagnant Labor Markets?
Productivity growth is the single most important variable for long-run living standards. Is AI finally delivering, or is it another hype cycle?
Every so often, a technology arrives with the promise of transforming productivity — the output produced per hour of labor. Productivity is the quiet engine of prosperity: it determines whether wages can rise without inflation and whether an aging society can sustain its social contract. The question is whether AI is the next general-purpose technology, or another overhyped disappointment.
The Productivity Problem
Productivity growth in advanced economies has been weak for two decades. The gains of the late 1990s and early 2000s — driven by the internet and enterprise software — faded after the global financial crisis. Since then, output per hour has averaged barely over 1% annually in the United States, and even less in Europe and Japan. With labor forces shrinking across the developed world, that stagnation has become a macroeconomic emergency.
Is AI Different This Time?
The optimistic case is that generative AI is a true general-purpose technology, one that diffuses across every sector. Early evidence is encouraging in specific domains: software development, customer support, and legal document review have all shown measurable efficiency gains of 20-40% in some knowledge-work tasks.
The skeptical case is that these gains remain too narrow to move aggregate numbers. The productivity paradox — the gap between technological capability and measured economic impact — has fooled forecasters before. Diffusion takes time: the full dividend from electricity and the internet took decades, because it required complementary investment in skills and organizational redesign.
The Labor Market Question
The deeper uncertainty is distributional. Unlike previous technologies, AI directly competes with cognitive labor — the kind of work long insulated from automation. The transition could displace middle-skill white-collar jobs even as it creates new roles demanding different capabilities.
Our view is that AI will eventually show up in the productivity statistics, but with a lag measured in years, not quarters. The economies that adapt fastest — through retraining and capital investment — will capture the gains. For investors, the productivity boom, when it arrives, will be the single most important macro tailwind of the decade.