While the rest of the AI ​​industry is quick to label its work as “AGI” or “superintelligence,” Alexandre LeBrun, CEO of AMI Labs, Yann LeCun’s global model startup, avoids the terms altogether. Lebrun said in an interview with TechCrunch that the company doesn’t use terms like “AGI” or “superintelligence” at all.

“We never used the word AGI. And I realized that no one uses it anymore; they changed to superintelligence,” he said. “Next time we’ll change to something else.” He is also not convinced by the new label. “There’s no good definition. What is superintelligence? I don’t know. It’s not a very useful word.”

It’s a strong stance from a founder at the center of AI’s latest race.

TechCrunch spoke with LeBrun while he was in Seoul last week for the International Conference on Machine Learning, where he was looking for local industry partners, global companies and researchers. AMI Labs is still early product, but it’s already courting players in robotics, manufacturing, and electronics. A global model, which incorporates physics to predict and work with the real world, needs to prove its effectiveness outside the laboratory, LeBrun explained.

One area where global models are expected to have a big impact is robotics. For now, robots simply run fixed, “completely static” routines, and AI is still “really dumb in the physical world,” LeBrun said.

Even when AI can simply make robots “context aware,” that would make “a very big difference to the world.” This context-sensitive AI would have been useful, for example, in preventing a robot that was dancing and doing kung fu at a public event from walking up and kicking a child. “The hardware is very advanced; the hardware progress in recent months is incredible, but there is no brain.”

A large language model (LLM) predicts the next word or text and a world model predicts the next state. If you push a glass off the table, you already know that it will tip over and spill; That’s the intuition a world model should capture: predicting the next state of the world, LeBrun explained.

It does not claim that world models are better than LLMs, which are “complementary, not replaceable” when it comes to AI systems that understand the physical world, LeBrun said. Drawing a parallel to the various language and reasoning functions of the human brain, he added that LLMs will continue to be the most efficient tools for processing language, while world models will provide context and understanding of the real world.

Almost all industries that “touch the real world” could eventually make use of robotics based on world models, LeBrun said, arguing that physical environments are still where LLMs are weakest.

A factory robot that repeats the same motion works quite well today, he said. The challenge begins when you “take your robot into a more open environment, in your home or on the street,” where it must understand its environment and operate safely. “Robots are not safe right now,” he said. “Today there is no solution for that.”

Healthcare offers a more personal example for LeBrun, whose previous company was Nabla, an artificial intelligence-based health startup. He compared current artificial intelligence systems to a doctor trained solely with textbooks and no residency. LLMs can be useful in medicine, he said, but they cover “only 1% of healthcare.” The rest depends on real world experience.

But a world model, LeBrun said, cannot be built inside a laboratory. To train on reality, AMI needs real environments and close partners, according to the CEO. “We need access to the real world” and it is “easier for us to do it with partners.” That’s part of what draws him to Asia, where the robots and chips and factories actually are.

LeBrun will not yet spell out a full strategy for Asia. “It’s too early,” he said. But the attraction to South Korea comes down to two things. First, Korea has advanced industries in robotics, semiconductors, and manufacturing; the hardware-heavy sectors that the first wave of AI barely touched.

The second attraction is speed. LeBrun pointed to Korea’s national plan to invest money in AI and its track record as an early adopter. “Korea was the fastest country to adopt the Internet 25 years ago,” he said. It’s that combination, a deep industrial base plus a willingness to quickly adopt AI, that he calls “unique” and why “we want to be here from day one.”

“I’ve been telling Alex and the team to come to Korea,” JP Lee, CEO of SBVA and one of AMI’s sponsors in Asia, told TechCrunch.

The government has done “a tremendous job” funding local sovereign LLM models, Lee said, and those already work “pretty well” for general-purpose tasks, but it is pushing for Korea to continue investing in physical AI as well. He points to Seoul’s June plan to mobilize some $880 billion for chips, AI data centers and physical AI as one of its three stated pillars: “They must coexist.”

Korea’s value to foreign companies, Lee argued, is not just in the hardware. Local developers quickly adopt and adapt new tools, a pattern that has produced local Internet players like Naver and Kakao.

Despite all the star power and billion-dollar check, AMI has nothing to sell yet. The startup, co-founded by Turing Award winner Yann LeCun after he left Meta, raised $1.03 billion in March at a pre-money valuation of $3.5 billion. There is no product or timeline yet to commit to. “We’ll do a surprise when we’re ready,” LeBrun said.