Emergent Multi-Agent Communication in the Deep Learning Era

TL;DR

This paper surveys recent research on how deep learning agents develop shared language for cooperation, aiming to understand language evolution and improve AI communication capabilities.

Contribution

It provides a comprehensive overview of recent studies on emergent multi-agent communication in deep learning, highlighting scientific insights and practical applications.

Findings

Language emerges under specific conditions in deep agent communities

Emergent communication features resemble aspects of human language

Enhanced communication improves AI cooperation and problem-solving

Abstract

The ability to cooperate through language is a defining feature of humans. As the perceptual, motory and planning capabilities of deep artificial networks increase, researchers are studying whether they also can develop a shared language to interact. From a scientific perspective, understanding the conditions under which language evolves in communities of deep agents and its emergent features can shed light on human language evolution. From an applied perspective, endowing deep networks with the ability to solve problems interactively by communicating with each other and with us should make them more flexible and useful in everyday life. This article surveys representative recent language emergence studies from both of these two angles.