Artificial intelligence (AI) is accelerating its expansion from the digital world to the physical world. During the 2026 Intel Connection event, Gao Song, Corporate Vice President of Intel and General Manager of the Edge Computing and Physical AI Business Unit in China, shared Intel's perspectives in a media interview on trends in physical AI development, robot architecture, chip development roadmaps, and standardization. Gao Song pointed out that with the integration of technologies such as multimodal models and reinforcement learning into the robotics field, robots are evolving from performing preset actions to possessing perception and cognitive capabilities, requiring computing platforms to simultaneously run AI models and handle motion control tasks. Leveraging its XPU architecture, which combines CPU, GPU, and NPU, Intel is building a computational foundation and has introduced an Embodied AI Development Kit, integrating the 3rd Gen Core Ultra processor to lower the barrier to robot development. Regarding chip selection, Intel believes that the current market has a clear and sizable demand for general-purpose chips due to their stable supply and mature ecosystem; however, if clear customization demands emerge in the market, Intel will also consider launching dedicated chips. Additionally, Intel is driving standardization efforts in the field of physical AI to reduce the learning curve and foster industry consensus.