If you believe that a Skynet-like future is destined to happen as artificial intelligence (AI) becomes more and more advanced, there's another reason for you to start warning the people you meet on the street. A robot developed by students at Brigham Young University students in the U.S. can be taught how to play foosball and is becoming skilled enough to beat human players.
"It's becoming a challenge for us to beat the artificial intelligence," Nathan Warner, one of the six students who created the AI, told SiliconRepublic. "You think, 'Oh yeah, humans ultimately should always be better than the computer,' but we're actually struggling to keep up."
Foosball is a table-top recreation of football — or soccer, as it is referred to in the U.S — where two or more players control players mounted on rotating rods. The rods can be moved sideways, but not back and forth, and requires players to essentially pass the ball along with the aim of scoring a goal. Much like football, the players are arranged to take the place of strikers, midfielders, defenders and goalkeepers.
The AI-powered robot relies on a camera to judge the match as it's played and controls the players fast enough to take on human players. The AI is reportedly getting better, anticipating its opponent's moves. This is apparently possible as the students have programmed what they consider are commonly used tactics. In comparison, both AlphaGo, which beat world champion Lee Sedol in a Go tournament in March as well as IBM's Deep Blue, which defeated chess grand master Gary Kasparov in 1997, began with a bank of moves.
