Incomplete Contracting and AI Alignment

USC CLASS Research Papers Series No. CLASS18-10

USC Legal Studies Research Papers Series No. 18-10

17 Pages Posted: 19 Apr 2018

See all articles by Dylan Hadfield-Menell

Dylan Hadfield-Menell

University of California, Berkeley - Department of Electrical Engineering & Computer Sciences (EECS)

Gillian K. Hadfield

University of Toronto; Vector Institute for Artificial Intelligence; OpenAI; Center for Human-Compatible AI

Date Written: April 19, 2018

Abstract

We suggest that the analysis of incomplete contracting developed by law and economics researchers can provide a useful framework for understanding the AI alignment problem and help to generate a systematic approach to finding solutions. We first provide an overview of the incomplete contracting literature and explore parallels between this work and the problem of AI alignment. As we emphasize, misalignment between principal and agent is a core focus of economic analysis. We highlight some technical results from the economics literature on incomplete contracts that may provide insights for AI alignment researchers. Our core contribution, however, is to bring to bear an insight that economists have been urged to absorb from legal scholars and other behavioral scientists: the fact that human contracting is supported by substantial amounts of external structure, such as generally available institutions (culture, law) that can supply implied terms to fill the gaps in incomplete contracts. We propose a research agenda for AI alignment work that focuses on the problem of how to build AI that can replicate the human cognitive processes that connect individual incomplete contracts with this supporting external structure.

Suggested Citation

Hadfield-Menell, Dylan and Hadfield, Gillian K., Incomplete Contracting and AI Alignment (April 19, 2018). USC CLASS Research Papers Series No. CLASS18-10. Available at SSRN: https://ssrn.com/abstract=3165793 or http://dx.doi.org/10.2139/ssrn.3165793

Dylan Hadfield-Menell (Contact Author)

University of California, Berkeley - Department of Electrical Engineering & Computer Sciences (EECS) ( email )

Berkeley, CA 94720-1712
United States

Gillian K. Hadfield

University of Toronto ( email )

78 Queen's Park
Toronto, Ontario M5S 2C5
Canada
4169784214 (Phone)

Vector Institute for Artificial Intelligence ( email )

OpenAI ( email )

Center for Human-Compatible AI ( email )

310 Barrows Hall
Berkeley, CA 94720
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
112
Abstract Views
481
rank
242,037
PlumX Metrics