The University at Buffalo, New York, Artificial Intelligence Institute, plans on creating an AI by studying brain waves of gamers. This AI is to be used DARPA, the U.S. Defense Advanced Research Projects Agency, to train military robots.
For the research, engineers of the University at Buffalo received a funding of $316,000 from DARPA.
Reportedly, the researchers at university are studying the brain waves and eye movements of 25 gamers with “strong gaming experience.”
“We’re trying to recruit participants who have strong gaming experience,” said Souma Chowdhury, one of the lead researchers in the project to Digital Trends.
Studying the Gamers By Developing A New Game
It’s not like the gamers play any game they like. The researchers have created their own real-life strategy game for the study, which is currently unnamed. In the game, the players have to defeat enemies while building units.
While the gamers play this unique strategy game, researchers hook them up to electroencephalogram (EEG) technology. It helps the game designer to study the brain activity of players.
Moreover, according to Digital trends, there are ultra-high-speed cameras equipped to study the fast eye movements of players.
All the collected information will then be used by machine learning algorithms to develop new algorithms, which will guide a fleet of 250 robots on the ground and in the air. So, the robots will be able to complete difficult tasks in complex, uncertain environments.
“Humans can come up with very unique strategies that an AI might not ever learn. A lot of the hype we see in AI are in applications that are relatively deterministic environments. But in terms of contextual reasoning in a real environment to get stuff done? That’s still at a nascent stage.” said Souma Chowdhury to Digital Trends.
According to Chowdhury, developing Artificial Intelligence will not take much time, as the project is already moving at an “aggressive pace.” They only require each participant to play around six to seven games to get all the required data.
However, it’s uncertain how long it will take to train a whole fleet of robots using the collected data.