[[{“value”:”
Jan. 27 (Asia Today) — South Korea’s government has rolled out a “Manufacturing AI Transformation” strategy, promising to combine data and artificial intelligence to raise productivity and help companies achieve self-sustained growth.
But many firms, even as they scramble to buy Nvidia graphics processors and plan large data centers, still cannot answer the basic question: What will they actually do with AI?
The United States has taken a global lead in large language models such as ChatGPT. China has rapidly scaled AI services at home using its capital base and market size. For South Korea, the challenge is building infrastructure while creating demand at the same time. One possible answer lies in an area often treated as old-fashioned: manufacturing.
The emerging concept is “physical AI” – artificial intelligence embodied in machines such as robots and autonomous vehicles. If large language models learned from online text and images, the next phase is teaching machines how to perform real-world tasks. That requires different data: physical skills and movement.
Related
The internet is rich with decades of text and image data. It has far less of what matters on factory floors, such as the manual technique of an experienced welder or the precise feel needed to assemble semiconductor equipment.
Even a simple robot task, like picking up a water bottle, can be mathematically complex. Earlier approaches tried to solve such problems primarily through formulas. Recent development has shifted toward learning by demonstration, with robots trained by observing and imitating human movements.
Examples cited in the industry include teleoperation-style training in humanoid robotics, where human workers perform tasks while wearing sensor equipment, allowing the robot to learn movement patterns through repetition. In that approach, the key input is not an equation but “skilled movement data.”
That is where South Korea could hold an edge. The country remains among the world’s top manufacturing nations. While some global competitors convert the movements of general workers into training data, South Korea could digitize the know-how of veteran skilled workers with decades of experience.
Industrial cities such as Ulsan and Changwon still have seasoned workers whose tacit knowledge is difficult to describe on paper. If that expertise can be captured through teleoperation and converted into training data before retirement, it could become a national asset that is hard for competitors to replicate quickly.
South Korea has often treated its manufacturing-heavy structure as a constraint, arguing it lacks high value-added software design capacity. But as AI begins to take physical form, factories could become “AI learning laboratories,” turning traditional strengths into a new advantage.
The argument is not that South Korea should abandon large-scale computing investment. It is that the country should also pursue a national effort to convert industrial know-how into data – a form of “data sovereignty” grounded in skilled work, not just servers.
— Reported by Asia Today; translated by UPI
© Asia Today. Unauthorized reproduction or redistribution prohibited.
Original Korean report: https://www.asiatoday.co.kr/kn/view.php?key=20260127010012632
“}]]