At the EVOLVE 2025 Zhongguancun Kejin Model and Smart Device Industry Innovation Summit held today, Tan Jianrong, an Academician of the Chinese Academy of Engineering, shared that “we should focus on big models, but we must never ignore small models. There are no small models, only big models, and it is also very difficult for artificial intelligence to be implemented.” Tan Jianrong pointed out that in addition to the three elements of artificial intelligence models, computing power, and algorithms, knowledge engineering is also one of the core key technologies for implementing artificial intelligence. Among them, knowledge can be divided into two categories: qualitative and quantitative, and models are quantitative knowledge. Large models require computational power to train different data to eventually generate knowledge. Therefore, behind big data and big models, large computing power also needs to be supported.

At the EVOLVE 2025 Zhongguancun Kejin Model and Smart Device Industry Innovation Summit held today, Tan Jianrong, an Academician of the Chinese Academy of Engineering, shared that “we should focus on big models, but we must never ignore small models. There are no small models, only big models, and it is also very difficult for artificial intelligence to be implemented.” Tan Jianrong pointed out that in addition to the three elements of artificial intelligence models, computing power, and algorithms, knowledge engineering is also one of the core key technologies for implementing artificial intelligence. Among them, knowledge can be divided into two categories: qualitative and quantitative, and models are quantitative knowledge. Large models require computational power to train different data to eventually generate knowledge. Therefore, behind big data and big models, large computing power also needs to be supported.