Designing ethical environments using multi-agent reinforcement learning

Arnau Mayoral-Macau, Manel Rodriguez-Soto, Enrico Marchesini, Maite López-Sánchez, Marti Sanchez-Fibla, Alessandro Farinelli, Juan Antonio Rodriguez Aguilar

Keywords: Ethical values, learning for value alignment, multi-agent reinforcement learning

TL;DR: Presenting algorithm to create ethical environments where agents learn to be ethical

Abstract:

This paper introduces the Approximate Ethical Embedding Process,
an algorithm for automating the design of ethical environments for
learning agents. Our algorithm helps build environments wherein
multiple agents learn policies that align with an ethical (moral) value
while simultaneously pursuing their individual objectives. Therefore,
we contribute to endowing environment designers with algorithmic
tools for building ethical environments. Moreover, we demonstrate
the ethical design process for two different settings of a gathering
environment, where agents have to adhere to beneficence to promote
the collective survival of the population. Our experiments show
that our approximate embedding process successfully generates
environments that incentivise the learning of value-aligned policies.

Type Of Paper: Full paper (max page 8)

Anonymous Submission: Anonymized submission.

Submission Number: 1