Improving Confidence in the Estimation of Values and Norms

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

7 Downloads (Pure)

Abstract

Autonomous agents (AA) will increasingly be interacting with us in our daily lives. While we want the benefits attached to AAs, it is essential that their behavior is aligned with our values and norms. Hence, an AA will need to estimate the values and norms of the humans it interacts with, which is not a straightforward task when solely observing an agent's behavior. This paper analyses to what extent an AA is able to estimate the values and norms of a simulated human agent (SHA) based on its actions in the ultimatum game. We present two methods to reduce ambiguity in profiling the SHAs: one based on search space exploration and another based on counterfactual analysis. We found that both methods are able to increase the confidence in estimating human values and norms, but differ in their applicability, the latter being more efficient when the number of interactions with the agent is to be minimized. These insights are useful to improve the alignment of AAs with human values and norms.
Original languageEnglish
Title of host publicationCoordination, Organizations, Institutions, Norms, and Ethics for Governance of Multi-Agent Systems XIII - International Workshops COIN 2017 and COINE 2020, Revised Selected Papers
Subtitle of host publicationInternational Workshops COIN 2017 and COINE 2020 Sao Paulo, Brazil, May 8–9, 2017 and Virtual Event, May 9, 2020 Revised Selected Papers
EditorsAndrea Aler Tubella, Stephen Cranefield, Christopher Frantz, Felipe Meneguzzi, Wamberto Vasconcelos
PublisherCornell University Library - arXiv.org
Pages98-113
Number of pages16
ISBN (Electronic)978-3-030-72376-7
ISBN (Print)9783030723750
DOIs
Publication statusPublished - 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12298 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Keywords

  • Autonomous agents
  • Norms
  • Ultimatum game
  • Values

Fingerprint

Dive into the research topics of 'Improving Confidence in the Estimation of Values and Norms'. Together they form a unique fingerprint.

Cite this