DeepMind, an AI that was able to dominate scores of human players in Quake III Capture The Flag, has started to learn how to play the rest of the game, with promising results.

DeepMind, the company behind the AI of the same name, got its start in London in 2010 and was acquired by Google in 2014. Through the direction of Doctor Demis Hassabis, Doctor Shane Legg, and Mustafa Suleyman it became one of the first AI programs on the market and helped gaming companies create more complex and human-like NPCs within multiplayer environments. DeepMind hopes that its AI system can better understand human nature by working with human teams in an ever-changing real-world environment. The AI that dominated CTF in Quake III has now managed to learn more complex game modes, so it seems like that goal is a real possibility.

Since IMB's Deep Blue beat Grandmasters at chess and IBM's Watson achieved worldwide fame for beating Jeopardy Mega champion Ken Jennings, this is certainly not something that is new in the world of artificial intelligence. What makes DeepMind different, however, is its ability to complete multiple complex tasks within an environment that it has no prior knowledge of. Other AI relies on interactions within the same environment with little variations, whereas DeepBlue is able to learn from its mistakes as they come up.

RELATED: Quake Champions Is Going Full Fortnite And Introducing A Battle Pass, Dropping Loot Boxes

via engadget.com

The DeepMind staff shared a blog post update on the AI's progress in which they detailed how their AI controlled players or "FTW Agents" and was able to quickly understand other game modes within Quake III. The results indicated that "agents can play multiple game modes and multiple maps competitively and are starting to challenge the skills of our human researchers in test matches," which suggests that that the AI may soon give them an idea of its full potential. Furthermore, the company hopes that its FTW agents will one day be able to interact with human players in a way that mimics human-to-human interactions.

Given that AI players or bots struggle to assist human players currently within a multiplayer environment, DeepMind's success could mean that bots will become more difficult to play against. Additionally, this advancement might also help professional players improve their skills and allow Esports teams to run simulations multiple times.

DeepMind still has a way to go before it can be deployed alongside human teams. Cooperation is one thing, but being able to judge emotions and feelings is another and DeepMind will have to learn how to do this before it can truly achieve the goals set by its creators.

READ NEXT: Known Cheater Still Qualifies For Fortnite World Cup, Despite Being Previously Banned