The AlphaGo shock
The historic five-match Go tournament between Korean champ Lee Se-dol and Google DeepMind’s AlphaGo ended with a 4-1 victory Tuesday for the artificial intelligence (AI) program. Despite Lee’s persistent attempts to shake the machine, AlphaGo remained strong. The contest between man and supercomputer defied previous expectations that a machine would hardly be able to surpass the complexity of human intelligence, at least in the world of Go, the most complex strategy game ever invented.
Lee’s defeat is almost immeasurable. AlphaGo clearly demonstrated its idiosyncratic computing ability throughout the tournament. But what really grabbed public attention was AlphaGo’s ability to play the game just as humans do. Endlessly judging the status of the matches, it sometimes provoked Lee and sometimes retreated. In fact, the predictions that AlphaGo would be incapable of seeing the whole picture - despite its strength in partial battles - were completely off track. Its intellectual prowess was nearly reminiscent of the styles of both Lee Se-dol and Lee Chang-ho, another legendary Korean Go player.
We cannot help but recognize AI’s potential: AlphaGo’s triumph translates into a dictum that artificial intelligence will soon exceed human ability, at least in the gaming realm. And the day when it helps - or outright replaces - people in other areas is likely coming faster than we expect. DeepMind CEO Demis Hassabis said his goal is to develop AI that can be applied everywhere, remarks that herald the era of a fourth industrial revolution characterized by a stunning transition from hardware to software and from separate products to total integration.
Yet, Korea lags behind advanced countries in this area due to its blind adherence to a hardware-focused industrial structure largely based on memory chips like DRAM. The government and industrial sector still tend to dismiss the amazing potential of software, as seen in its prejudice and restrictions on the gaming industry.
The AlphaGo shock could serve as an opportunity to turn the tide. We need a long-term and comprehensive plan to integrate the industrial, social and cultural sectors. Without sticking to short-term profits, the government must revise the law to expedite investments in research and development. At the same time, it must educate younger generations about the imminent arrival of the artificial intelligence era. All parties involved must address the many challenges AlphaGo has presented us.
JoongAng Ilbo, March 16, Page 30