"What are the children of CEO Jensen Huang doing? They are working on Physical AI."
Jensen Huang, CEO of Nvidia, the world's leading artificial intelligence (AI) chipmaker, visits Korea on the 4th. The first news to emerge was that Huang would meet with Koo Kwang-mo, Chairman of LG Group (003550). When the meeting was reported, shares of major affiliates including LG Electronics (066570) surged so sharply that observers described the move as a "pillar of fire." Why? The industry is focusing on the possibility of collaboration in Physical AI and digital twins.
Jensen Huang: "AI's Final Destination Is Physical AI"
Both Eldest Son and Second Daughter Work in Robotics
Madison Huang Held Prior Meeting with President Ryu Jae-cheol
The AI market has reacted sensitively to wherever Huang is looking. After Huang visited Korea last October and met Samsung Electronics Chairman Jay Y. Lee in Seoul and SK Group Chairman Tae-won Choi in Gyeongju, AI infrastructure investment competition among global big tech firms intensified. Prices of memory semiconductors, an essential component of AI chips, also surged.
But the focus of this visit appears to lie elsewhere. It is Physical AI. Huang has forecast that the AI industry will move from infrastructure investment to agentic AI services that perform tasks on behalf of humans, and ultimately to Physical AI. In fact, in his keynote speech at "Nvidia GTC Taiwan 2026" held on the 1st, Huang presented Physical AI as the next stage of the AI industry.
The fact that both of Huang's children work in robotics shows that this remark is not mere rhetoric. His eldest son, Spencer Huang, joined Nvidia in 2022 and oversees projects in the robotics division. His second daughter, Madison Huang, is Senior Director of Omniverse and Robotics Product Marketing.
The meeting between Chairman Koo and CEO Huang was, to some extent, foreseeable. Director Madison Huang visited Korea in April and discussed Physical AI collaboration with Ryu Jae-cheol, President and CEO of LG Electronics. What did the two heavyweights discuss? While the specifics are unknown, the chances that the talks proceeded positively appear not small.
LG Electronics Unveils Home Robot at 'CES 2026'
Chairman Koo Holds Physical AI Meeting in Silicon Valley
LG Electronics Ranks No. 1 in Global Robot-Related Patents
LG Holds Robot, Auto Parts, HVAC and Battery Technologies
The industry views that Jensen Huang pursued the meeting with Chairman Koo because he highly values LG Group's Physical AI capabilities.
How strong are LG's robotics capabilities? A clue can be found in last year's announcement by the Korean Intellectual Property Office (KIPO). KIPO recently released the results of its analysis of AI-based robot patents over the past 10 years (2012-2021) in five major countries: Korea, the U.S., China, the European Union (EU), and Japan.
In this survey, LG Electronics ranked first in the world with 1,038 patents (18.8%). LG Electronics filed an overwhelming number of patents in core areas of Physical AI, including not only cleaning robots but also service robots, logistics robots, and object and voice recognition technologies. Considering that Samsung Electronics, its rival in the home appliance sector, filed only 41 patents, LG Electronics' Physical AI technology prowess can be gauged.
LG Electronics has begun showcasing its Physical AI capabilities in earnest this year. At "CES 2026," the world's largest information technology (IT) and home appliance trade show held in Las Vegas in January, the company put a humanoid robot, "LG CLOiD," at the forefront rather than home appliances. CLOiD demonstrated folding laundry using robotic fingers, considered the most challenging area of humanoid technology.
Three months later, LG Group released footage of Chairman Koo visiting Skild AI, a robotics intelligence development company in Silicon Valley, where he conversed with co-founder Abhinav Gupta in front of a humanoid robot. Some analysts saw this as Chairman Koo publicly throwing his weight behind LG's Physical AI business.
Beyond this, LG possesses a full lineup of core technologies needed to realize Physical AI, including batteries (LG Energy Solution (373220)), cooling systems (LG Electronics), robot eyes (LG Innotek (011070)), faces (LG Display (034220)), joints (LG Electronics), and data centers (LG CNS). LG AI Research also has its own AI model, "EXAONE."
Nvidia services its robot learning platform "Isaac" based on Omniverse, a virtual world or digital twin platform. Also, at "Computex 2026" held in Taiwan on the 2nd, Nvidia unveiled "Nvidia Cosmos 3," an open-world foundation model for Physical AI.
Nvidia Needs Real-World Data
LG Operates Home Appliance, Display and Chemical Plants
Sharing Manufacturing Data Could Forge a 'Win-Win AI Alliance'
There is also a view that interprets this meeting through the lens of last year's so-called "best-friend meeting" between Nvidia CEO Jensen Huang and Hyundai Motor Group Chairman Euisun Chung. Through Omniverse, Nvidia provides a virtual world where autonomous vehicles and robots can learn.
The problem is that for such digital twins to become precise, "real-world" data is essential. No matter how outstanding a virtual world Nvidia builds, without on-site data, it is difficult to implement a digital twin at the same level as the real world.
There are observations that since the meeting between the two leaders last year, Nvidia has shared driving data and automobile factory production data through its collaboration with Hyundai Motor. Considering that Nvidia unveiled Alpamayo, an open-source-based next-generation autonomous driving integrated platform and AI model, this view carries a degree of credibility.
But LG Group also boasts world-class competitiveness in terms of manufacturing site data. Through affiliates such as LG Electronics, LG Display, LG Energy Solution, LG Innotek, and LG Chem, LG arguably possesses on-site data from an even wider range of industries than Hyundai Motor.
Physical AI must be trained to fit the actual environments in which humans work. This requires training in digital twins that reflect human work environments and process data. LG Group, in the end, holds the core data that will determine the competitiveness of Physical AI.
What will Chairman Koo Kwang-mo of LG Group and CEO Jensen Huang discuss at this meeting? While the specifics are unknown, after this meeting, it appears difficult to describe LG Group without mentioning Physical AI.
