Interactive AI with a Theory of Mind

TL;DR

This paper advocates for developing interactive AI systems with a computational theory of mind, enabling AI to understand and anticipate human users by modeling them as agents in a multi-agent framework, demonstrated through multi-armed bandit experiments.

Contribution

It introduces a novel categorization of user modeling approaches based on agency levels and applies nested multi-agent modeling to enhance AI's understanding of users.

Findings

Effective user models for multi-armed bandit systems

Demonstrated potential of nested multi-agent modeling

Proof-of-concept user study supports approach

Abstract

Understanding each other is the key to success in collaboration. For humans, attributing mental states to others, the theory of mind, provides the crucial advantage. We argue for formulating human--AI interaction as a multi-agent problem, endowing AI with a computational theory of mind to understand and anticipate the user. To differentiate the approach from previous work, we introduce a categorisation of user modelling approaches based on the level of agency learnt in the interaction. We describe our recent work in using nested multi-agent modelling to formulate user models for multi-armed bandit based interactive AI systems, including a proof-of-concept user study.