Abstract
This paper theoretically and empirically analyzes the interaction between economic policy and the artificial intelligence (AI) revolution. To this end, we construct a general equilibrium model where R&D investments in automation increase the probability of AI development. If AI is successfully developed, then competitive firms substitute AI for labor. The model exhibits multiple market equilibria, all of which are Pareto-inefficient. Using AI investment data from 80 countries collected between 2009 and 2018, we also find strong empirical support for multiple equilibria—the primary prediction of our model. Hence, the theory and the available evidence suggest that decentralized markets can cause coordination failures in AI development due to multiple equilibria. We show that subsidizing investments in AI development can restore the efficiency of the markets and also act as an equilibrium selection device.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
