+ A

But is it AI?

Dec 26,2019
이미지뷰
이미지뷰
Lee In-ah
The author is a professor of brain/cognitive science at Seoul National University.

Artificial intelligence (AI) is a much-heard buzzword. The government, corporations, universities and research sectors are all investing heavily in AI-related projects. People worry if their jobs will one day be replaced by smart machines, and students wonder what they have to study to live in the AI future.

Computer intelligence inspires awe and fear and offers a challenge. To brain scientists, AI is an enigma because they cannot understand machines mimicking the human brain, which remains a mystery.

The mind-related studies like philosophy and psychology run deep in countries like the United States, where AI researches and applications are active. The historical work on the mind when converged with scientific and biological brain research has built up brain and cognitive sciences. Doctors devoted to the mystery of the brain — and computer scientists and engineers trying to expand the spectrum of artificial intelligence with the help of brain sciences — ended up giving birth to AlphaGo, a marvelous Go-playing computer program. The product of Google’s DeepMind software demonstrated its superfast learning skills when it beat human Go champion Lee Se-dol in March 2016 to become the world’s No. 1 Go player. Lee Se-dol, who was recently defeated 2-1 by HanDol — a Korean equivalent of AlphaGo — has retired from the world of Go, also known as baduk in Korea.

AlphaGo’s capabilities stunned the world by showing how far artificial intelligence can go. But the voice-activated assistants on mobile devices and speakers, which are used by many consumers, remain frustratingly simple. AI research powerhouses like Google may want to do away with “artificial” in the compound word of AI. Demis Hassabis, founder and CEO of DeepMind, returned to academia after working on AI programming to pursue more study on cognitive neuroscience to seek inspiration from the human brain for a new AI algorithm. His Ph.D. study included a paper on patients with damage to their hippocampus — which is known to cause amnesia — to examine the connection between the constructive process of imagination and the reconstructive process of episodic memory recall. His focus on the essence of the cognitive power has been wondrous.

IT multinationals like Google and Facebook have worked closely with universities and research centers because of the cognitive science behind AI development.

The ultimate goal of investment in brain science is to achieve a physical understanding of the brain’s mechanism to reconstruct perfect virtual reality, combat amnesia, seek solutions to various mental illnesses, augment cognitive powers, and build future AI capabilities. Such a future would drive human happiness, grow enterprises and improve the society and country.

But such a future may be far-fetched for Korea. Companies have their eyes on immediate profits through AI applications to their technologies. Universities are unwilling to groom new convergence studies beyond what is currently available.

Neuroscience and cognitive science studies and majors are rare in Korea. Korea’s AI focus is entirely commercial. For the country’s future, the government must set higher goals and far-sighted vision and investment.

There is an ancient saying that one is a fool if he looks at his finger when he is pointing to the moon. If AlphaGo had a human feature, it would have pointed to our brain with envy. Despite multibillion-dollar projects on AI the government has announced, the money still may be going to the finger, not the moon.

Translation by the Korea JoongAng Daily staff.

JoongAng Ilbo, Dec. 24, Page 29