Yann LeCun predicts next AI breakthrough will focus on understanding the physical world

On October 12, 2026, AI pioneer Yann LeCun, founder and chairman of AMI Labs, stated in Seoul that the next major breakthrough in artificial intelligence will not come from large language models, but from systems capable of understanding and predicting the physical world. He emphasized that while large language models excel at processing language, they fall short in comprehending the complexities of the real world. LeCun noted that even a teenager can learn to drive in a few hours, yet no AI system has achieved L5-level autonomous driving or created fully autonomous robots without training. He argued that AI must bridge the gap between understanding language and grasping the physical world, as the latter is far more unpredictable. LeCun advocates for the development of 'world models' that can learn the physical laws and causal relationships governing reality, enabling AI to predict outcomes and interact with its environment. LeCun, a co-inventor of convolutional neural networks (CNNs), has significantly advanced modern computer vision and is widely recognized as one of the 'fathers of AI.' He was awarded the Turing Award in 2018. The article highlights his belief that current AI systems lack the ability to comprehend the physical world, unlike humans, and that future progress depends on overcoming this limitation.