Yao Qizhi Academician: Artificial intelligence needs to move from imitation to reasoning, from local skills to whole-body coordination.

On December 13, 2025, the opening ceremony of the Global Developers Pioneer Conference and the International Embodied Intelligence Skills Competition was held. Turing Award winner, Chinese Academy of Sciences academician, and Chairman of the Competition Guidance Committee Yao Qizhi articulated the key directions for the development of embodied intelligence in his speech at the opening ceremony, and shared four forward-thinking insights: First, we must move from imitation to reasoning. Current methods rely heavily on imitation and lack interpretable world models and physical causal reasoning. In the future, reasoning, planning, and control should be placed in the same closed-loop framework. Secondly, we must move from data scarcity to data leap. The cost of embodied data collection is high, and new data collection technologies need to be explored to build scalable data factories and verify the path to scale embodied intelligence. Third, we must move from local skills to whole-body collaboration. The current movement and operation capabilities of robots are still two separate systems, and in the future, full-body control and fine hand operation planning need to be unified to support multi-step general skills. Finally, we must move from individual combat to unified evaluation. Establishing open benchmarks, safety standards, encouraging open-source reproduction, and promoting excellent algorithms through challenges that are repeatable, verifiable, and industrializable. Academician Yao Qizhi emphasized that industry and academia working together will accelerate intelligent robots entering thousands of industries and households, truly serving humanity.