"Korea can become a robot leader."

Kim Sang-bae, a world-renowned scholar in robotics and a professor in the Massachusetts Institute of Technology (MIT) Department of Mechanical Engineering, emphasized on the 1st at the "1st Seoul AI Robot Show 2025" held at COEX in Gangnam District, Seoul, "It is not easy to match China's hardware prices, but Korea has many students who are equipped with hardware and artificial intelligence (AI) capabilities," adding, "Korea should focus on high value-added processes and move forward by leaving manufacturing to China."

"Seoul AI Robot Show" is an AI and robotics event co-hosted by the Seoul Metropolitan Government, the Seoul City Economic Policy Division, and Seoul Business Agency (SBA) from Sept. 30 to Oct. 2.

At the forum that day, under the theme "A step toward a robot-friendly city, Seoul," world-renowned scholars such as Kong Kyung-chul, a professor at KAIST; Choi Ri-gun, an executive director at Hyundai Motor Robotics Lab; Dennis Hong, a professor in the University of California, Los Angeles (UCLA) Department of Mechanical Engineering; Kim Sang-bae, an MIT professor; and Kim Ik-jae, head of the AI and Robotics Research Institute at the Korea Institute of Science and Technology (KIST), attended and discussed robot policy and technological challenges.

On the 1st at COEX in Gangnam-gu, Seoul, Kim Ik-jae, director of the KIST AI·Robotics Research Institute; Gong Kyung-chul, KAIST professor; Choi Ri-gun, executive director of Hyundai Motor Robotics Lab; Dennis Hong, UCLA professor; and Kim Sang-bae, MIT professor, hold a panel discussion. /Courtesy of Kim Jeong-eun

The scholars warned at the forum about the risks of developing robots solely with a data-driven approach. Professor Dennis Hong explained, "For the past 100 years, to make robots move, we had to understand the laws of physics and control them with mathematical models, but recently, even without knowing math and science, if you feed in a lot of AI data, robots operate—a data-driven approach has emerged as the mainstream."

He warned, "If you approach (robot development) with only data without understanding model-based methods, when a problem occurs you cannot identify the cause, which is very dangerous."

Professor Kim also said, "Seventy percent of people ignore existing disciplines and look only at data," and argued, "We need to use the data approach together with the model-based approach."

In a special lecture on "physical intelligence," Professor Kim stressed the importance of action data. To aid understanding, he asked the audience to put their hands in their pockets and take out an item. When the audience took items out of their pockets, he asked, "Is there anyone among you who understood how the second joint of your index finger moved?" He then said, "I have made 5,000 people do the same, but no one knew."

Professor Kim explained the reason as "because action data belong to the realm of the unconscious." He pointed out, "Currently, IT corporations leading AI, such as Nvidia, Google, and Meta, develop AI only with high-level data (conscious data) such as language," and emphasized, "To conquer physical intelligence, we need an action language in the unconscious realm that is not on the internet."

The subsequent panel discussion addressed "ways to make robots acceptable and trustworthy in everyday life." Executive Director Choi emphasized, "It is important to have clear use cases in everyday life and be offered at a price point people can use," while Professor Dennis Hong said, "As much as people feel fear toward new technology, we need a friendly and acceptable UI and UX."

They also discussed ways to reshape the robot industry ecosystem amid the rush toward medical schools. Professor Kong, CTO of Angel Robotics, said, "It is important to accept the current situation and make efforts to find what can be done under the given conditions," and proposed, "We can use the already established medical infrastructure for R&D and apply it to rehabilitation and assistive robots." Executive Director Choi emphasized, "We need a science-and-engineering-centered ecosystem," adding, "We must instill the idea that if you study hard in the sciences, you can make decisions proactively and lead society."

※ This article has been translated by AI. Share your feedback here.