Google’s DeepMind Artificial Intelligence has achieved human-like performance levels in Quake III Arena.
As described in a fascinating blogpost, the pioneering AI company run by games industry veteran Demis Hassabis has recently unleashed its AI agents on Quake III Arena’s Capture The Flag mode. DeepMind said: “Mastering the strategy, tactical understanding and team play involved in multiplayer video games represents a critical challenge for AI research. Now, through new developments in reinforcement learning, our agents have achieved human-level performance in Quake III Arena Capture the Flag, a complex multi-agent environment and one of the canonical 3D first-person multiplayer games. These agents demonstrate the ability to team up with both artificial agents and human players.”
As the blogpost explains, DeepMind homed in on Quake II Arena partly in a bid to teach its AI agents to co-operate: “This is a setting we call multi-agent learning: many individual agents must act independently, yet learn to interact and cooperate with other agents. This is an immensely difficult problem — because with co-adapting agents the world is constantly changing.”
DeepMind continued: “We ran a tournament including 40 human players, in which humans and agents are randomly matched up in games — both as opponents and as teammates. The FTW [For The Win] agents learn to become much stronger than the strong baseline methods, and exceed the win-rate of the human players. In fact, in a survey among participants they were rated more collaborative than human participants. Through unsupervised learning we established the prototypical behaviours of agents and humans to discover that agents in fact learn human-like behaviours, such as following teammates and camping in the opponent’s base.”
E-sports pros need not worry: the bots aren’t coming to get you just yet. DeepMind is essentially training its AI for ever more complex tasks, using ever more complex videogames. If you want to keep abreast of what DeepMind is up to, keep an eye on its website.