AI is still in its early stages... More R&D is needed and access needs to be expanded

“There should be mechanisms in place to control development, but I think we should not try to stop or ban the development of artificial intelligence (AI), because that is absolutely impossible.”

In an interview with Yonhap News on the 4th, Daniel Lee, a professor at Cornell University and a world-renowned scholar in the field of artificial intelligence (AI), emphasized that there is no need to slow down the pace of development, saying, “I think we are now in the early stages of finding out what AI can actually do.”

Professor Lee is a master in AI fields such as robotics and machine learning, and served as vice president of Samsung Electronics [005930] Global AI Center for five years since 2018. Since last year, he has also been serving as a chair professor at the Institute for Advanced Study.

◇ Optimistic about AI ‘existential crisis’… Access needs to be increased further

He compared the introduction of AI technology to the industrial revolution and evaluated it as still at the beginning stage.

Professor Lee said, “The Industrial Revolution started with a steam engine, but it took time to develop into an internal combustion engine,” and added, “To fully develop AI and make it as useful as possible, much more research and development (R&D) is needed in the future.”

He emphasized that there is no need to stop development, saying he is optimistic about the so-called 'existential crisis' in which the recent rapid development of AI could escape humanity's control and threaten its survival.

“Even if one country stops development, there is no way to actually stop development as other countries will do,” he said. “There is no need to stop, but it is necessary to regulate and control development more.”

He continued, "Historically, no matter what kind of technological development you look at, whether it's the Industrial Revolution or the Internet, there was chaos at first and people were always very worried that this technology would completely change or destroy the world," he said. "But we may not have done it in the best way, but sometimes we found our way through the situation somehow."

He emphasized, "When it comes to AI, we will find ways to mitigate risks and at the same time figure out how to utilize AI."

Professor Lee cited access as the biggest problem with AI today.

“Access to the best AI is controlled by a very small number of companies and people,” he said. “I think in the future it needs to be more democratized and much more widely available, and I think it will be.”

He continued, “Increasing accessibility to AI will not be a limitation (in improving performance, etc.) because new forms of AI will continue to emerge.”

◇ “Korea can act as an intermediary between AI between the U.S. and China… It can also serve as a technology test bed.”

Professor Lee said that Korea is in an interesting position in the race for AI hegemony, and evaluated that it could become a so-called 'intermediary' between the United States and China.

“Korea can become a kind of global AI broker because it is connected to both sides,” he said. “It might be possible to integrate systems or ways of thinking from both sides, so it will be interesting to see whether Korea can leverage this unique global position it already has.”

For example, he said that even in the case of large AI conferences held in both the United States and China, it is difficult to exchange researchers, so many conferences are held in Korea.

Professor Lee predicted that Korea could serve as a test bed for new AI, saying that since Korea cannot compete on scale, it must find a niche.

“Korea has always had a strength in integrating technology or adopting technology faster than other places,” he said. “I don’t know if we can create the best foundation model, but we can try, experiment, and integrate new things faster.”

He continued, "Although winner-takes-all may appear in the AI ​​industry in the short term, the situation will eventually change. Even in the early days of the Industrial Revolution, such as the oil market, a few companies dominated, but over time, other markets and countries were able to do the same."

◇ “Next-generation AI will be inspired by basic science”

Professor Lee highly praised the collaboration with the Korea Advanced Institute of Science, saying he believes the next-generation AI technology will still be inspired by basic science ideas.

He said, “The beginning of the industrial revolution was a steam engine, but now internal combustion engines are used and battery technology is used. There was science behind this development.”

He continued, "New AI will not simply be based on the current large-scale Transformer model or Large Language Model (LLM) method," and added, "People who understand deep science need to think about how they can create 'next-generation AI'."

Regarding the reason for continuing cooperation with Korea, he said, "The reason I like Korea is because there are a lot of talented people, we know each other well, and it is good to cooperate."

Professor Lee said, “My role will be to convey a little of the American perspective,” adding, “Introducing the idea in a way that the United States is thinking about this, but in Korea, we can think about whether it is appropriate or not.”