Liu Zhiyuan wrote in The Guangming Daily: Establishing an artificial intelligence evaluation system based on practice.

Liu Zhiyuan, professor at Tsinghua University and co-founder and chief scientist of ModelScope, published a signed article in The Guangming Daily proposing an AI evaluation system grounded in practice. The piece advocates improving the system across three dimensions: evaluating specific applications, assessing industry impact, and measuring overall contribution, while incorporating efficiency and cost to guide AI from scale competition toward efficiency and value competition.