AMI Labs, co-founded by Turing Award winner Yann LeCun, is currently scouting for industrial partners in Seoul to bridge the gap between AI and the physical world. While current large language models excel at processing text, LeBrun describes them as akin to doctors who have studied only textbooks, lacking the practical residency required for real-world application. His vision centers on world models that predict physical states, providing the context-awareness necessary to make robotics safe and functional outside of static, controlled environments.
Why AMI Labs CEO Alexandre LeBrun Rejects the AGI Label
While the broader AI industry pivots from the term AGI to superintelligence, AMI Labs CEO Alexandre LeBrun is discarding both. LeBrun contends that these labels lack meaningful definitions, preferring to focus his startup’s efforts on building world models capable of navigating the complexities of physical environments rather than chasing abstract hype.
LeBrun identifies a critical disconnect between the rapid advancement of hardware and the current lack of a digital brain to guide it. By integrating physics into AI, he hopes to move beyond robots that merely repeat fixed routines. The company is actively courting partners in manufacturing and electronics to train these models in real-world settings, viewing South Korea’s industrial base and aggressive investment in AI infrastructure as a vital testing ground. Despite a $3.5 billion valuation and a $1.03 billion funding round, AMI Labs remains pre-product, with LeBrun opting to withhold any specific launch timelines until his team is ready to deliver a tangible, functional breakthrough.
Comments (0)
No comments yet. Be the first!