NATIONAL

[EYE] ‘Social diversity key to facing AI world’

By Korea Herald

Renowned neuroscientist stresses need of ‘scarcity’ for humans to be creative

  • Published : May 20, 2016 - 20:37
  • Updated : May 23, 2016 - 11:39
As apprehension spreads among ordinary mortals over the rise of artificial intelligence to replace humans, social diversity would be the key for people to brace for the next AI generation, Korea’s renowned neuroscientist said.

“Current rote-learning method that unilaterally crams knowledge into people’s brains is nothing but an attempt to train the ability that can be replaced by AI. People should come to think of the opposite question: ‘How can we have different experiences and viewpoints?’ Because this is the core of creativity that allows to control and utilize AI,” computational neuroscientist Jeong Jae-seung at the Korea Advanced Institute of Science and Technology said in an interview with The Korea Herald.
 
Jeong Jae-seung, professor at the Korea Advanced Institute of Science and Technology

Jeong, a bio and brain engineering associate professor who originally majored in physics at KAIST, received a Ph.D. from the alma mater with a thesis on computational modeling of Alzheimer’s brains. The 44-year-old expanded his study to psychiatry when he worked as a postdoctoral fellow at Yale University School of Medicine and an assistant professor at Columbia University.

His main research topics include human decision-making, brain-robot interface and brain dynamics.

Having been recognized for his work, Jeong was picked as one of “Young Global Leaders” at the World Economic Forum held in Davos, Switzerland, in 2009.

He became a public figure with a series of best-seller books on science and psychology that fluently elaborated on the otherwise perplexing concepts of the fields.

Witnessing the landmark Go matches between Google’s AI named AlpahGo and South Korea’s Go champion Lee Se-dol in March, he emphasized that the game was a very safe chance for the public to realize how powerful AI could be if it goes through reinforcement learning or deep learning.

In the end, the future will be greatly influenced by the combination of AI, Internet of Things and big data, the professor added.

“The uncut versions of human behavior will be embodied in the IoT and social media in the future, and AI will learn and analyze that behavior, figure out what people desire and come up with customized services beforehand,” he said.

He, however, pointed out that people’s concerns over AI replacing people’s jobs is exaggerated.

“AI has yet to have motivations or desire. There is no need to personify it. Just like other scientific technology, AI will be of very useful means.”

Yet, he stressed that human control is essential.

“The use of AI will be most effective in the medical or health care system. But it cannot serve as a tool once the procedure gets complex. The scientific technology can be used as a tool only when humans have a control over it,” he stressed.

“The more complicated AI becomes, the more often humans will follow its decisions without being aware of how such decisions are made, and also face difficulties in finding its faults.

“We may be in the same situation as Aja Huang playing Go as directed by AlphGo,” he said, referring to the Google DeepMind programmer and amateur 6-dan Go player who was placing stones on behalf of AlphaGo.

Some have raised concerns that robots will ultimately turn into a new “governing class” that controls humans. Jeong, however, underscored that a more serious concern in the AI world will be the digital gap among humans.

“It’s critical to think of how to solve the gap between those who are able to utilize AI and those who are not. The ‘digital divide’ is already visible in society as the classes are formed by the difference in technology adaptability.”

Right after the AlphaGo match sparked an AI wave here, the Korean government vowed in March that it would launch an AI technology research center with support from conglomerates.

Establishing a separate AI team within the Ministry of Science, the government said it would inject 1 trillion won ($840 billion) in the industry for the next five years. It also said a 15-year long-term measures will be drawn up later this year.

Jeong hailed the government’s large-scale investment plans but remained cautious over the way they are state-driven.

“The government’s aggressive investment must be carried out for sure. But once the government attempts to lead the way, reaching the goals will be difficult. No scientific research can be completed with a single vision. It is important that that government maintains a distance by making the investment but not interfering with the industry,” he stressed.

While continuing his studies on brain-machine interface and brain-inspired AI, Jeong participated last year in the government’s project of analyzing problems and issues that the country will face in the future.

The project covered what kind of lifestyle the country has to seek and how scientific technology will play a role.

As one of the project members, the professor predicted that more attention will be paid to the quality of life.

“The Korean society severely lacks a lifestyle that values the quality of life, which will lead to the rise of desire and demand for it. The most important value called happiness has become the most difficult thing to achieve in Korea. ‘Scarcity’ of happiness will make the society to desire the happiness,” said Jeong.

“Scarcity” is what the neuroscientist has emphasized whenever he holds a public lecture.

“Achievement is initiated from voluntary motivation. These days, many parents provide their children in advance with what they may need, giving them no chance to truly desire something. They may go to good schools and get good jobs. But such a life without desire can hardly make great achievements,” he said.

“One of the things that AI does not have but humans do is ‘voluntary motivation.’ Only people with enough voluntary motivations can make great achievements, and scarcity is what drives such motivation.”

As part of his studies analyzing the human decision-making process, Jeong is now working on a dissertation on how patients with depression or anxiety make decisions to commit suicide.

“Not all depression patients kill themselves. If it is possible to predict their suicidal risk, (the society) will be able to provide due care for them. Curbing the suicide rate is one of my academic goals. I would like to contribute to solving problems associated with serious decision-making such as suicide,” Jeong said.

By Lee Hyun-jeong (rene@heraldcorp.com)