Towards Machines that Trust: AI Agents Learn to Trust in the Trust Game
Artificial Intelligence
2023-12-21 v1 Neurons and Cognition
Abstract
Widely considered a cornerstone of human morality, trust shapes many aspects of human social interactions. In this work, we present a theoretical analysis of the , the canonical task for studying trust in behavioral and brain sciences, along with simulation results supporting our analysis. Specifically, leveraging reinforcement learning (RL) to train our AI agents, we systematically investigate learning trust under various parameterizations of this task. Our theoretical analysis, corroborated by the simulations results presented, provides a mathematical basis for the emergence of trust in the trust game.
Cite
@article{arxiv.2312.12868,
title = {Towards Machines that Trust: AI Agents Learn to Trust in the Trust Game},
author = {Ardavan S. Nobandegani and Irina Rish and Thomas R. Shultz},
journal= {arXiv preprint arXiv:2312.12868},
year = {2023}
}