Emergent Complexity via Multi-Agent Competition

TL;DR

This paper demonstrates that competitive multi-agent environments with self-play can induce agents to learn highly complex behaviors in relatively simple 3D worlds, revealing a natural curriculum for skill development.

Contribution

It introduces new competitive multi-agent environments where agents develop complex skills through self-play, surpassing the environment's inherent complexity.

Findings

Agents learn diverse complex behaviors such as running, blocking, and tackling.

Self-play creates a natural curriculum for skill progression.

Simple environments can produce highly sophisticated agent behaviors.

Abstract

Reinforcement learning algorithms can train agents that solve problems in complex, interesting environments. Normally, the complexity of the trained agent is closely related to the complexity of the environment. This suggests that a highly capable agent requires a complex environment for training. In this paper, we point out that a competitive multi-agent environment trained with self-play can produce behaviors that are far more complex than the environment itself. We also point out that such environments come with a natural curriculum, because for any skill level, an environment full of agents of this level will have the right level of difficulty. This work introduces several competitive multi-agent environments where agents compete in a 3D world with simulated physics. The trained agents learn a wide variety of complex and interesting skills, even though the environment themselves are relatively simple. The skills include behaviors such as running, blocking, ducking, tackling, fooling opponents, kicking, and defending using both arms and legs. A highlight of the learned behaviors can be found here: https://goo.gl/eR7fbX