Cooperating with machines
Name
s41467-017-02597-8.pdf
Size
1.33 MB
Format
Adobe PDF
Checksum (MD5)
9a75ba6dc8d24d6503d109530134164a
Author(s)
Crandall, Jacob W.
Oudah, Mayada
Tennom, Mayada
Ishowo-Oloko, Fatimah
Abdallah, Sherief
Bonnefon, Jean-François
Cebrian, Manuel
Shariff, Azim
Goodrich, Michael A.
Rahwan, Iyad
Date Issued
January 2018
Journal
Nature Communications
Publisher
Nature Publishing Group
Citation
Crandall, Jacob W. et al. “Cooperating with Machines.” Nature Communications 9, 1 (January 2018): 233 © 2018 The Author(s)
Version
Final published version
Abstract
Since Alan Turing envisioned artificial intelligence, technical progress has often been measured by the ability to defeat humans in zero-sum encounters (e.g., Chess, Poker, or Go). Less attention has been given to scenarios in which human-machine cooperation is beneficial but non-trivial, such as scenarios in which human and machine preferences are neither fully aligned nor fully in conflict. Cooperation does not require sheer computational power, but instead is facilitated by intuition, cultural norms, emotions, signals, and pre-evolved dispositions. Here, we develop an algorithm that combines a state-of-the-art reinforcement-learning algorithm with mechanisms for signaling. We show that this algorithm can cooperate with people and other algorithms at levels that rival human cooperation in a variety of two-player repeated stochastic games. These results indicate that general human-machine cooperation is achievable using a non-trivial, but ultimately simple, set of algorithmic mechanisms.
MIT Department
Massachusetts Institute of Technology. Media Laboratory
Terms of Use
Attribution 4.0 International (CC BY 4.0)
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1038/s41467-017-02597-8
