BEIJING — Since Google’s computer program AlphaGo won four out of five matches against South Korea’s champion Go player, Japanese governmental officials are seriously wondering whether artificial intelligence (AI) is the way to rewrite Japan“s blueprint for the future.
There is precedent for programs beating humans. IBM’s Deep Blue beat chess champion Gary Kasparov in 1997, and in 2012, computer programs beat professionals of Shogi, also known as Japanese chess. But the Japanese government has never been as shaken as this time. That’s because Go, which is played on a grid of 19 horizontal lines and 19 vertical lines, is considered “the last bastion of human intelligence.” It’s a game of greater depth than chess or Shogi, and the player depends mostly on intuition and experiences.
Google has said the program uses four combined technologies — big data, algorithms, super computing and deep learning. In other words, AlphaGo simulates the human nervous system and has conducted a thorough study of 30 million tournaments played by professionals.
As one Japanese player remarked after watching the five AlphaGo matches against Lee Se-dol, “AlphaGo can do what humans are incapable of doing, because in the middle phase of the game a player’s moves can only rely on intuition. Yet AlphaGo is moving successively with confidence. Right from the beginning AlphaGo already grasps the overall situation, something that’s impossible for a human.”
Committing to AI
Over the past few years, Japanese firms have competed in investing huge amounts of funds in the research and development of artificial intelligence. Toyota, the largest Japanese automobile maker, is among them. In a recent press conference, company president Akio Toyoda announced that the firm will spend $1 billion over the next five years in a joint AI research project with Preferred Networks, a small company located near Tokyo University.
At CES 2016, the world’s largest consumer electronics show, which was held in Las Vegas earlier this year, Toyota showcased the results of its cooperation with Preferred Networks — “vehicles that will not collide with each other.” During the demonstration, five Toyota smart cars freely glided around without hitting each other.
“Toyota is a large enterprise that led Japan’s rapid economic development in the latter half of the 20th century,” one Japanese official said. But he added that it wants shift focus to avoid losing competitiveness, given that the future is electric cars without engines. “Other major manufacturers like Sony and Matsushita Electric will join the automotive industry, a move that could lead to the end of Toyota’s empire. In order to prevent such an outcome, Toyota’s ambition is to be at the forefront of autopilot technology.”
On the other hand, Japan’s Ministry of Health, Labor and Welfare is also nursing an eager expectation. In the summer of 2013, Japan underwent a huge “immigration debate.” At the time, Prime Minister Shinzo Abe advocated his economic policy, nicknamed Abenomics, by setting the country’s GDP growth target at 600 trillion Japanese yen. But no matter how hard the Japanese work, they know this is an impossible mission given the fact that Japan is an aging nation with a low birth rate. Over 25% of its population is composed of people over 65 years old.
For a sustainable Japanese economy, the Abe government proposed accepting immigrants from neighboring countries. But as a mono-ethnic island state, the Japanese are adamantly opposed this idea. And Japan also began to see Europe’s migration crisis as a warning, ultimately abandoning the proposed immigration policy.
Instead, the Ministry of Health, Labor and Welfare has begun envisioning AI development as a way to avoid relying on immigrants. “Whether it’s about taking care of the elderly, children or most kinds of factory and household labor, robots equipped with AI are the solution.”
Japan’s doomed future lit up
“What AI development requires is cutting-age technology and a spirit of persistent pursuit,” says Toru Nishikawa, the 33-year-old founder and CEO of Preferred Networks. Japan excels in these areas.
One example is Onogawa Naoki, a 25-year-old Japanese artist who has suddenly become highly recognized lately. His work involves folding miniscule origami cranes from paper sized at 12mmx12mm. He can make 500 of them per day. Then he glues these tiny colorful cranes, in the tens of thousands, to tree branches.
At Onogawa’s recent New York exhibition entitled “Art on Paper,” viewers were very generous in their praise. Such delicate techniques and persistence is deeply rooted in Japanese DNA, and passed from generation to generation. And that’s what developing AI requires.
But can Japan really resurrect itself through developing the AI? Experts seem to have diverging opinions.
Takuya Matsuda, a Kobe University emeritus professor and former president of Japan’s Astronomical Society, is an expert in the AI field. His book, Issues in the Year 2045, is a bestseller in Japan and has sounded the alarm for mankind’s future. According to Matsuda, AI will surpass the intelligence of mankind collectively by 2045. If that happens, our future is unpredictable, because not only can AI bring us convenience but also abuse.
Take war as an example. In 2045, country A’s AI may battle with that of country B. But if during the war, one country’s AI realizes that the initiator of the war was actually the humankind of both countries, the AI could decide to team up and simply eliminate both countries.
A similar situation has indeed occurred in Japanese history. During the Heian era, a particularly extravagant period about a thousand years ago, aristocrats used to hand down the tasks of laboring and fighting in wars to their samurai. These warriors worked like slaves to defend their lords, who only knew how to give orders. In the end, these samurai turned around their swords and eliminated the lords to become Japan’s rulers. The Japanese society established by samurai lasted nearly 700 years.
“In the 21st century, humans created artificial intelligence,” Matsuda has warned. “This can be said to be a creation of God, or the devil. How can we make sure that the AI we create is a god, instead of a devil? America and the EU have embarked on the very issue, while in Asia this is left completely unattended.”
Lee Se-dol has a chance of continuing the challenges even after his tournament setbacks. Were a war between artificial intelligence and mankind really to happen, the humans probably won’t have a second chance. China and Japan are Asia’s locomotive for the 21st century. They should make it a priority to team up and envisage how to deal with the “issues in the year 2045.”
