Human adaptation to adaptive machines converges to game-theoretic equilibria

Abstract Here we test three learning algorithms for machines playing general-sum games with human subjects. The algorithms enable the machine to select the outcome of the co-adaptive interaction from a constellation of game-theoretic equilibria in action and policy spaces. Importantly, the machine l...

Full description

Saved in:
Bibliographic Details
Main Authors: Benjamin J. Chasnov, Lillian J. Ratliff, Samuel A. Burden
Format: Article
Language:English
Published: Nature Portfolio 2025-08-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-12998-1
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849237764034265088
author Benjamin J. Chasnov
Lillian J. Ratliff
Samuel A. Burden
author_facet Benjamin J. Chasnov
Lillian J. Ratliff
Samuel A. Burden
author_sort Benjamin J. Chasnov
collection DOAJ
description Abstract Here we test three learning algorithms for machines playing general-sum games with human subjects. The algorithms enable the machine to select the outcome of the co-adaptive interaction from a constellation of game-theoretic equilibria in action and policy spaces. Importantly, the machine learning algorithms work directly from observations of human actions without solving an inverse problem to estimate the human’s utility function as in prior work. Surprisingly, one algorithm can steer the human-machine interaction to the machine’s optimum, effectively controlling the human’s actions even while the human responds optimally to their perceived cost landscape. Our results show that game theory can be used to predict and design outcomes of co-adaptive interactions between intelligent humans and machines.
format Article
id doaj-art-4914a2fdf23a404f8472c7684a5fdf36
institution Kabale University
issn 2045-2322
language English
publishDate 2025-08-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-4914a2fdf23a404f8472c7684a5fdf362025-08-20T04:01:52ZengNature PortfolioScientific Reports2045-23222025-08-0115111010.1038/s41598-025-12998-1Human adaptation to adaptive machines converges to game-theoretic equilibriaBenjamin J. Chasnov0Lillian J. Ratliff1Samuel A. Burden2Department of Electrical & Computer Engineering, University of WashingtonDepartment of Electrical & Computer Engineering, University of WashingtonDepartment of Electrical & Computer Engineering, University of WashingtonAbstract Here we test three learning algorithms for machines playing general-sum games with human subjects. The algorithms enable the machine to select the outcome of the co-adaptive interaction from a constellation of game-theoretic equilibria in action and policy spaces. Importantly, the machine learning algorithms work directly from observations of human actions without solving an inverse problem to estimate the human’s utility function as in prior work. Surprisingly, one algorithm can steer the human-machine interaction to the machine’s optimum, effectively controlling the human’s actions even while the human responds optimally to their perceived cost landscape. Our results show that game theory can be used to predict and design outcomes of co-adaptive interactions between intelligent humans and machines.https://doi.org/10.1038/s41598-025-12998-1human-AI interactionmulti-agent systemsgame theory
spellingShingle Benjamin J. Chasnov
Lillian J. Ratliff
Samuel A. Burden
Human adaptation to adaptive machines converges to game-theoretic equilibria
Scientific Reports
human-AI interaction
multi-agent systems
game theory
title Human adaptation to adaptive machines converges to game-theoretic equilibria
title_full Human adaptation to adaptive machines converges to game-theoretic equilibria
title_fullStr Human adaptation to adaptive machines converges to game-theoretic equilibria
title_full_unstemmed Human adaptation to adaptive machines converges to game-theoretic equilibria
title_short Human adaptation to adaptive machines converges to game-theoretic equilibria
title_sort human adaptation to adaptive machines converges to game theoretic equilibria
topic human-AI interaction
multi-agent systems
game theory
url https://doi.org/10.1038/s41598-025-12998-1
work_keys_str_mv AT benjaminjchasnov humanadaptationtoadaptivemachinesconvergestogametheoreticequilibria
AT lillianjratliff humanadaptationtoadaptivemachinesconvergestogametheoreticequilibria
AT samuelaburden humanadaptationtoadaptivemachinesconvergestogametheoreticequilibria