Who’s the victor anyway?
The historic match between Korean Go 9-dan-ranked world champion Lee Se-dol and the strongest-yet computer Go player AlphaGo developed by Google is set to kick off. Koreans are glued to the game that is due to start on Wednesday in Seoul, and Go aficionados and scientists around the world are closely watching an event that could possibly signal a new chapter for artificial intelligence and the game of Go.
Spectators will see how much of a reality sci-fi-like artificial intelligence has become. Bets behind each player are tight with a 50:50 chance. But while the media and the Internet community are abuzz about who will win, we have to probe the real motive behind Google DeepMind’s latest prized software development. Is defeating Lee really the goal?
The concept of artificial intelligence first emerged in 1956 with idealistic designs of developing a machine capable of mimicking human intelligence. Scientists believed intelligence was the crucial factor that made humans superior to all other beings. They believed natural instincts and emotions that even animals have should come under the control of the logical and reasoning capacity of the brain.
They aspired to develop a machine that could be faster and more efficient in rational and logical information processing than the human mind. Machines already excel human capacity in some limited capacity. Computers can work out a mind-boggling mathematical problem in seconds and process multiple navigational inputs to take us to our destination by the fastest route. Thanks to these machines, our lives have become so much easier and more convenient.
But we are comfortable using them with the firm belief that they do not pose a threat to humans - at least for the moment - and that they are under human control. Machines do not have a will of their own. They do not have their own desires, complain or protest about given work, nor do they do anything beyond their command. The thought that machines one day could have a mind of their own is in itself too frightening. It is scary to imagine a computer refusing to log off and continuing with work on its own.
But if we look closely, they are already lurking around us. We often find ourselves shouting “Stop!” when our car navigation won’t stop beeping and insist we follow its orders when we make a turn on our own. A computer infected with a virus frantically goes its own way. Someone could be behind all this, but the person at the machine cannot feel comfortable. It is why many warn of the dangers of the evolution and proliferation of artificial intelligence.
AlphaGo receives the spotlight because of its fast adaptation and development. It has already studied and processed all the games played by professional human counterparts, and its training is said to be equivalent to what a human player can do in 1,000 years. Its skills have progressed enormously, playing one million games in just weeks.
But AlphaGo is designed to dominate as much space as possible on the 19-by-19 grid. It does not know why nor does it think about what the game means. It does not get tense or enjoy the game. It does not feel the thrill of winning. It just moves on immediately to analyzing its moves and other possibilities that could have made it win a bigger score.
AlphaGo, which trains day and night without knowing the purpose, reminds us of Korean youths. Our children spend every day moving from one after-school class to another, solving problems, memorizing and testing. The hours, amount of learning materials and competition they go through are as substantial as a machine that plays a millions games a month. But how many students are doing it with a purpose and know the meaning of all of their efforts? Korea’s education field looks like the early scientists who believed human intelligence was the highest value.
Parents wouldn’t want their children’s minds working like a computer. So why are we trying to mimic a computer? Humans seek artificial intelligence because they know what they are after.
No matter who wins the game between Lee and AlphaGo, the real winner will be Google. The technology giant has been investing in the software and game because it has a specific purpose. AlphaGo is just the instrument for its goal.
So what are we trying to make our children into - super machines or Google-like visionaries?
Translation by the Korea JoongAng Daily staff JoongAng Ilbo, March 7, Page 35
*The author is a professor of psychology at Korea University.
by Hur Tae-kyun