Technology and Jobs (paulkrugman.substack.com)

This primer revisits a long-running debate—dating back at least to David Ricardo’s 1821 worries about “machinery” and echoed in dystopian fiction like Kurt Vonnegut’s Player Piano—about whether technological progress helps or hurts workers. The author emphasizes that anxiety over automation is not new, but argues we should be careful about extrapolating from high-profile examples of sectoral job loss to predictions about overall employment. The piece sets out to clear common confusions and give historical and theoretical context rather than settle the controversy. Technically, the primer will examine how technology destroys jobs in specific industries (substitution) versus how it creates new tasks or raises demand elsewhere (complementarity), why sectoral displacement doesn’t automatically imply lower aggregate employment, and the conditions under which productivity gains can depress wages or employment (e.g., weak demand, falling labor share, slow institutional adjustment). It will then ask whether current AI advances are likely to tip the balance toward persistent job loss, and highlight a crucial “fork in the road” around energy: AI’s compute and infrastructure choices could amplify carbon and resource use or, if steered differently, help decarbonize systems. The upshot is a nuanced, policy-relevant framing: outcomes depend on technological detail, market responses, and collective choices about distribution and energy.