March 2021

IZA DP No. 14171: Modelling Artificial Intelligence in Economics

Economists' two main theoretical approaches to understanding Artificial Intelligence (AI) impacts have been the task-approach to labor markets and endogenous growth theory. Therefore, the recent integration of the task-approach into an endogenous growth model by Acemoglu and Restrepo (AR) is a useful advance. However, it is subject to the shortcoming that it does not explicitly model AI and its technological feasibility. The AR model focuses on tasks and skills but not on abilities, while abilities better characterize AI services' nature. This paper addresses this shortcoming by elaborating the task-approach with AI abilities for use within endogenous growth models. This more ability-sensitive specification of the task-approach allows for more nuanced and realistic impacts of progress in artificial intelligence (AI) on the economy to be captured.