Jin Rong Jie·2026-01-22 09:22Core Insights - The introduction of the VLA (Visual-Language-Action) driver model by Li Auto in 2025 marks a significant leap in AI's understanding of the physical world, moving beyond traditional assisted driving to a more advanced form of intelligent interaction [1][9][14] Group 1: Evolution of Assisted Driving - Li Auto's evolution in assisted driving reflects the technological advancements and cognitive deepening within China's smart automotive industry, transitioning from rule-based systems to AI-driven models [3][6] - The shift from "imitation" to "thinking" in driving models highlights the limitations of earlier systems, which relied heavily on human-engineered rules and lacked deep reasoning capabilities [3][6] - The introduction of the VLA model aims to overcome the communication barriers and slow reasoning speeds that hindered previous models, enhancing the overall performance of assisted driving [6][10] Group 2: VLA Technology and Its Implications - The VLA model is positioned as a foundational technology for the next generation of AI systems, transforming vehicles from passive tools into proactive service robots [9][12] - Li Auto's commitment to the VLA model is supported by a vast dataset from over 1.5 million vehicles, enabling a data loop that enhances the model's training and performance [6][10] - The VLA model's ability to exhibit "human-like" behaviors signifies a shift towards a more intuitive interaction between humans and vehicles, allowing for voice commands and complex decision-making [7][10] Group 3: Future of the Automotive Industry - The ongoing development of the VLA model is expected to lead to the emergence of "automobile robots," capable of achieving higher levels of automation and intelligence [12][14] - Li Auto's exploration of VLA technology is influencing other leading companies in the industry, indicating a broader trend towards integrating large-scale data and advanced algorithms into automotive systems [12][14] - The ultimate goal of the assisted driving industry may evolve from merely achieving "autonomous driving" to creating mature embodied intelligent products, marking a new era in the automotive and AI sectors [12][14]