When will embodied intelligence reach its breakthrough? industry leaders discuss challenges and future prospects

On September 11, 2026, during the Inclusion · Bund Conference, industry leaders, investors, and young entrepreneurs in embodied intelligence discussed the industry's current state, challenges, and future potential. Many believe the sector is still in its early stages, with key capabilities emerging but not yet forming large-scale productivity. Jia Peng of Zhijian Dynamics noted that the 'bubble' feeling stems from a mismatch between valuation and commercial value, with the industry potentially evolving into a larger scale than mobile phones or cars. Han Zheng of Sudo Technology compared the current state of embodied intelligence to 2018-2019 for large language models, emphasizing the need for integrating technologies into a commercially viable system. Xiao Wanggang of Daxiao Robotics highlighted data as the biggest bottleneck, stressing the importance of zero-shot capabilities for large-scale adoption. Young entrepreneurs like Li Yiming and Ding Ning emphasized the importance of world models and physical-world integration, with Ding Ning noting China's potential in high-precision data and hardware. Chen Boyuan of Nix Matrix Technology argued that physical AI requires understanding causality and is a complex system involving models, data, and real-world interaction. He stressed the importance of first principles and small-scale validation over individual experience. The discussion highlighted the industry's challenges, including system engineering, computing power, and data, while also pointing to opportunities in vertical applications and global markets.