Hot on the heels of Chinese technology companies reaping the rewards of their battle with their U.S. counterparts, comes the news in the China Daily that a Chinese-made computer has won a Chinese Chess or ‘Go’ game against another Artificial Intelligence (AI) machine. The computer, known as ‘FineArt’, beat its Japanese counterpart at the annual Worldwide Computer Go Tournament at the University of Electrocommunications, in Tokyo, which has been held for the past ten years.
Around 30 teams from around the world competed in the tournament and the next challenge to face the Chinese machine, built by the tech company Tencent, will be a match against a Japanese Go Master Ryo Ichiriki, which it will hopefully win. The machine won all eleven games that it played and entered the tournament for the first time this year. This win has come after another, American designed, AI machine successfully fought off a human poker player for the first time in an important victory for the technology. This is due to poker being an ‘imperfect’ game, meaning that contestants are unable to openly see their opponents’ hand and so must make judgments with information that they do not have. This, obviously, moves into the realm of human nature and behavior and was once considered impervious to computing. The ability to undertake strategic reasoning utilising imperfect information, has now surpassed the best that humans can master and unlocks the door to AI being used for many medical, military and scientific purposes.
FineArt is similar to other AI programs in that it uses deep learning to intelligently go about its business, constantly playing against itself to improve and refine its learning of whatever task it has been set. It goes without saying that any game undertaken against a human counterpart also becomes part of the learning process and this is what has enabled such programs to become so successful in beating human opponents’ – by using their flaws against them.
Obviously, the scope for such technology is huge, however, the worrying part for me is what will happen to us? Once computers have taken all of our jobs, what are we going to do? How long we have until humans are effectively obsolete in relation to their daily lives is anyone’s guess and maybe we will go full circle and pay more for human service and production, as opposed to that handled by a computer. Whatever the answer, it appears that C3PO and R2D2 are not that far off anymore and AI could indeed be the technology that propels us into the next century!