Can I read Agent Incentives: A Causal Perspective on EtoBox?

Agent Incentives: A Causal Perspective by Everitt, Tom; Carey, Ryan; Langlois, Eric; Ortega, Pedro A; Legg, Shane is a scholarly article available to read on EtoBox.

What is Agent Incentives: A Causal Perspective about?

We present a framework for analysing agent incentives using causal influence diagrams. We establish that a well-known criterion for value of information is complete. We propose a new graphical criterion for value of control, establishing its soundness and completeness. We also introduce two new concepts for incentive analysis: response incentives indicate which changes in the environment affect an optimal decision, while instrumental control incentives establish whether an agent can influence its utility via a variable X. For both new concepts, we provide sound and complete graphical criteria. We show by example how these results can help with evaluating the safety and fairness of an AI system.

Author

Everitt, Tom; Carey, Ryan; Langlois, Eric; Ortega, Pedro A; Legg, Shane

Published

2021

Language

EN