THE GROWTH EFFECT OF ARTIFICIAL INTELLIGENCE, R&D, AND SUBSIDIES.

Abstract

This thesis characterizes Artificial Intelligence as a capital embodied technology, improving the productivity of capital while reducing reliance on labor. It presents a framework that endogenizes households’ saving decisions and firms R&D investment decisions to adopt A.I., compares the dynamics between the social and decentralized equilibrium, and examines some policy implications. Three results emerge: Firstly, the model has two stable equilibria, a primitive economy with no technology adoption and a modern economy that attains perpetual growth. Secondly, compared to the social optimum, the decentralized economy invests and consumes less in the long run, and has a lower likelihood of adopting advanced technology. Thirdly, both investment and production subsidies effectively reduce the gap between the social and decentralized equilibrium, with an optimal production subsidy (the first-best policy) allowing the decentralized economy to completely achieve the socially optimal outcome.