We have in the past seen AI take on human players in a variety of games. However for the most part these are one-on-one matches. However in recent times we’re starting to see how researchers are trying to train AI to have teamwork, such is the case with Elon Musk’s OpenAI which has recently been training to take on professional DotA 2 teams.
However it appears that OpenAI is not alone on this front as Google’s DeepMind AI has also been training its teamwork skills by playing Capture The Flag (CTF) on id Software’s Quake III Arena (via VentureBeat). The AI has played nearly 450,000 games in a bid to establish its understanding on how to work with other machines and humans.
According to DeepMind’s blog, “We train agents that learn and act as individuals, but which must be able to play on teams with and against any other agents, artificial or human. From a multi-agent perspective, [Capture the Flag] requires players to both successfully cooperate with their teammates as well as compete with the opposing team, while remaining robust to any playing style they might encounter.”
This is actually not the first time that Google’s DeepMind AI has been taught to play games. In the past Google and Blizzard have worked together to see if its AI was capable of learning strategic games such as StarCraft 2.
Filed in . Read more about AI (Artificial Intelligence), Deepmind, Google and Quake.