At the World Robot Conference, Yushu CEO Wang Xingxing addressed the misunderstandings surrounding embodied ontology, questioning why it hasn't seen widespread adoption. He questioned whether the hardware is inadequate or the cost is high. The reality is that current hardware is adequate, but not great enough. He stated that there's currently relatively little focus on models, while there's a bit too much emphasis on data. The industry's biggest challenge right now is the sheer inadequacy of embodied AI. Large-scale robotic models are similar to the situation 1-3 years before ChatGPT's release: everyone has identified the direction and technical path, but no one has yet successfully implemented it.