當年今日﹕Chess computer beats human
【明報專訊】11 May 1997 marked a milestone in the progress of artificial intelligence (AI). That day Deep Blue, a chess-playing computer developed by IBM, defeated Garry Kasparov (卡斯帕羅夫), the greatest chess player ever born.
◆1. The games
Garry Kasparov is a Russian chess grandmaster (國際象棋大師). He is regarded as the greatest chess player of all time. From 1986 to his retirement in 2005, he ranked first in the world for 225 out of 228 months.
Kasparov first took on the chess computer Deep Thought in 1989, winning both games. In 1996, he faced IBM's Deep Blue. He lost the first game, but went on to win three and draw two of the following five, defeating Deep Blue by 4-2.
Then IBM programmers upgraded the supercomputer and made it twice as powerful and capable of scanning 300 million positions a second and analyzing 74 moves ahead. It eventually got the better of Kasparov, whose IQ is estimated to be 190. The last game took place on 11 May 1997 (two wins for Deep Blue, one for Kasparov, three draws).
◆2. The debates over artificial intelligence
Artificial intelligence (AI) is an area of computer science about creating intelligent machines that work and react like humans. AI research dates back to the 1950s. Since then a debate about its power has gone on.
1．Can computers think, learn, observe and innovate like humans?
Few would doubt that, in areas like calculation, computers leave humans in the dust. But the human mind is capable of learning new ideas, making observations and inventing new things. Will computers be able to do them one day?
Though some argue that they never will, computer scientists are trying to give computers at least some of the skills listed above. For example, Iamus, a piece of AI software created by Spanish programmers, writes melodies.
2．Will computers become cleverer than humans?
Some computer-science experts think this is likely. Last year, Ray Kurzweil, Google's director of engineering, argued at an interview that, by 2029, computers would be cleverer than humans. He believes they will be "able to understand what we say, learn from experience, make jokes, tell stories and even flirt (調情)."
Given his track record, Kurzweil's predictions carry much weight. In 1990 he predicted a computer would defeat a world chess champion by 1998. That happened between Deep Blue and Kasparov on 11 May 1997. He predicted, also correctly, the prominence (顯著地位) of the World Wide Web when it was used by a few academics.
3．Will computers threaten humans' existence?
If computers can become so intelligent, will they threaten our existence? Prominent theoretical physicist Stephen Hawking (霍金) believes that the full development of AI could spell the end of the human race, as we will not be evolving (進化) fast enough to match the fast development of computers. Bill Gates, the founder of Microsoft, also thinks computers could grow too strong for humans to control.
Some computer scientists are less pessimistic (悲觀的), however. Rollo Carpenter, who created the AI learning software Cleverbot (which is designed to talk as a human does), thinks that programmers are still a long way from having the computing power to achieve full artificial intelligence, and that humans will remain in control of computers "for a decently long time".
