Are Automating Jobs Subject to Moral Hazard in a Dynamic Principal-Agent Model?
-
SeriesPhD Lunch Seminars
-
Speaker(s)Vincent Tena (Toulouse School of Management, France)
-
LocationErasmus University Rotterdam, Tinbergen Building, Room H09-02
Rotterdam -
Date and time
November 20, 2018
13:00 - 14:00
The speed and depth of current advancements in artificial intelligence provoke numerous and passionate debates in our society. In this paper, we explore the relationship between automation and the inherent agency issue of task delegation. We present a continuous-time agency model where the principal can automate at any date the production at a sunk cost. Thanks to the advantages offered by continuous-time models and results in optimal stopping theory, we solve for the optimal contract. We show that giving to the principal an option to automate does not change how the contract is terminated, but hasten the arrival of payments and of termination.