AI, productivity, and economic growth

Abstract

This chapter examines how artificial intelligence (AI) may affect productivity and long-term economic growth. It begins by clarifying what economists mean by productivity, total factor productivity (TFP), and growth accounting, then explains why AI can raise task-level efficiency without immediately producing a visible aggregate productivity boom. The chapter develops the productivity paradox, the role of intangible complementary investment, the AI productivity J-curve, and the distinction between genuine productivity-enhancing innovation and so-so automation. It then analyzes the main channels through which AI may increase growth: cheaper prediction, process innovation, better allocation, scientific discovery, organizational redesign, and new products. The chapter also considers the limits of growth optimism by contrasting singularity-style scenarios with Baumol-style sectoral bottlenecks. The central claim is that AI’s growth impact will depend less on raw model capability than on diffusion, reorganization, competition, institutions, and the distribution of complementary capital.