Artificial Intelligence, Industrial Structure Transformation and Escaping the Middle-Income Trap

Abstract

Abstract This paper constructs a multi-sector general equilibrium model to analyze the impact of artificial intelligence (AI) on structural transformation, by examining the complementary and substitutive relationships between AI and high- and low-skilled labor, as well as the differences in labor skill intensity across industries. From the dual perspectives of productivity and income inequality, the paper explores policy pathways for China to overcome the “middle-income trap.” The findings indicate that AI accelerates the rising share of the service sector through both capital-structure and skill-structure effects, thereby enhancing aggregate labor productivity, while simultaneously widening the income inequality between high-skilled and low-skilled workers. On this basis, the paper conducts counterfactual analyses under three scenarios: labor skill-structure transition, AI-biased technological change, and deepening AI penetration. In all three scenarios, AI raises aggregate labor productivity. However, the income inequality between high- and low-skilled labor narrows only under the labor skill-structure transition scenario. Further analysis reveals that the development of generative AI increases the risk of job polarization, and excessive reliance on technological substitution may constrain long-term productivity growth potential. Only when AI technology achieves an optimal match with the labor skill structure can efficiency and equity be jointly attained; thus, fostering a new human-machine collaborative production model represents a strategic direction for the future.