Why AMI Labs’ Alexandre LeBrun doesn’t call his AI “AGI” or “super intelligence”

While the rest of the AI ​​industry races to call its work “AGI” or “superintelligence,” Alexandre Lebrun, CEO of… Yan to be Universal model start, AMI Laboratories, It avoids the terms completely. LeBron said in an interview with TechCrunch that the company does not use terms like “AGI” or “super intelligence” at all.

“We never used the word artificial general intelligence,” he said. “And I noticed that no one uses it anymore; they’ve switched to superintelligence.” “Next time we’ll switch to something else.” It is not sold on the new label either. “There’s no good definition. What is superintelligence? I don’t know. It’s not a very useful word.”

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

TechCrunch spoke to LeBrun while he was in Seoul last week for the International Machine Learning Conference, where he was looking for local industrial partners, global companies, and researchers. AMI Labs is still in the pre-product phase, but is already courting robotics, manufacturing and electronics companies. LeBrun explained that the global model, which involves physics to predict and work with the real world, needs to prove itself outside the laboratory.

One area where global models are expected to have a major impact is robotics. Right now, robots only run fixed routines, “quite static,” and AI remains “really stupid in the physical world,” Lebrun said.

Even when AI can make robots “context aware,” this would make “a very big difference to the world.” Such context-aware AI could have been useful, for example, in prevention A robot was dancing and practicing kung fu at a public event From approaching the child and kicking him. “The hardware is very advanced; the progress that has been made in hardware in the last few months is incredible, but there is no brain.”

The Large Language Model (LLM) predicts the next word or text, and the Global Model predicts the next state. Push the cup away from the table, and you already know it’s going to tip over and spill; Lebrun explained that this is the intuition that the global model aims to capture: predicting the next state of the world.

LeBrun said he does not claim that universal models are better than MBAs, which are considered “complementary and non-substitutable” when it comes to AI systems that understand the physical world. In parallel with the distinct language and reasoning functions of the human mind, he added that MAs would remain the most efficient tools for language processing while global models would provide context and understanding of the real world.

Almost every industry that “touches the real world” could eventually benefit from robotics based on global models, Lebrun said, arguing that physical environments remain where MBAs are weakest.

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

Healthcare provides a more personal example for LeBrun, whose previous company was Nabla, an AI health startup. He likened today’s artificial intelligence systems to a doctor who was trained only on textbooks and without a residency. Master’s degrees in law may be useful in medicine, but they cover “only 1% of health care,” he said. The rest depends on real world experience.

But Lebrun said the global model cannot be built in a laboratory. For real-life training, AMI needs real environments and close partners, according to the CEO. We need access to the real world, and it’s easier for us to do that with partners. That’s part of what draws him to Asia, where robots, chips and factories already exist.

LeBron won’t spell out his full Asia strategy yet. “It’s too early,” he said. But the attraction to South Korea is due to two things. First, Korea has advanced industries in robotics, semiconductors, and manufacturing; Hardware-heavy sectors that were barely touched by the first wave of AI.

The second attraction is speed. LeBron pointed to Korea’s national plan to pump money into artificial intelligence and its track record as one of the first countries to adopt it. “Korea was the fastest adopter of the Internet in 25 years,” he said. It’s that combination, the deep industrial base plus the desire to embrace AI quickly, that he describes as “unique,” and why “we want to be here from day one.”

“I asked Alex and the team to come to Korea,” JB Lee, CEO of SBVA and a backer of AMI in Asia, told TechCrunch.

Lee said the government has done a “tremendous job” in funding domestic sovereign LLM models, and these models already work “well enough” for general-purpose missions, but he is pressing Korea to continue investing in physical AI as well. He points to Seoul’s June plan Mobilizing about $880 billion for chips, AI data centers and physical AI, as one of its three stated pillars. “They must coexist.”

Lee said that Korea’s value to foreign companies is not limited to hardware. Local developers are quick to adopt and adapt new tools, a pattern that has spawned local online players like Naver and Kakao.

Despite all the star power and the billion-dollar check, AMI has nothing to sell yet. The startup, co-founded by a Turing Award winner Yan to be After he left dead, It raised $1.03 billion in March At a pre-cash valuation of $3.5 billion. There’s no product yet, and no timeline he’ll stick to. “We’ll surprise you when we’re ready,” LeBron said.