周鸿祎罗永浩对谈大模型，称模型能力不与大小成正比

Core Viewpoint - The discussion between Luo Yonghao and Zhou Hongyi emphasizes that larger models in AI are not necessarily better, and the importance of specialized knowledge remains crucial in various fields [1] Group 1: AI Model Capabilities - Zhou Hongyi agrees with Luo Yonghao that the definition of "full-time and all-capable" models is critical, suggesting that until such models surpass human capabilities in all aspects, roles like product managers will continue to be relevant [1] - Zhou provides examples indicating that large models may struggle with highly specialized tasks, such as designing an aircraft engine or a battleship, due to the necessity of expert knowledge that is not readily available on the internet [1] - The realization that "big" does not equate to "capable" has emerged, with many smaller models demonstrating significant effectiveness [1]