Zhang Peng, CEO of Zhipu AI: The commercial logic of "price war" for large models is not normal and will not last long

At today's World Artificial Intelligence Conference (WAIC 2024), under the theme of "Constructing a New Value Chain Driven by Large Models," a clash of ideas unfolded between Yan Junjie, founder and CEO of MiniMax, Zhang Peng, CEO of Zhipu AI, Xu Bin, General Manager of the National Local Co-Building Humanoid Robot Innovation Center, and Xie Ling, founder and CEO of Yufeng Future.

Zhang Peng expressed that the essential feature of large models is the ability to provide generalized capabilities based on a single model, capable of addressing a series of scenarios and application requirements. From this perspective, Zhang Peng believes that to empower the real economy with generative AI, it is necessary to build more universal and fundamental capabilities.

In the future, Zhang Peng believes that besides accuracy, another breakthrough point for AI models is multimodality.

Why is multimodality important? Zhang Peng explained that solving problems in the real world involves multiple modes: in addition to natural language, there are also visual, auditory, tactile, and common sense aspects. Therefore, if large models can make a breakthrough in this area, it will make AI more inclusive.

Zhang Peng also believes that the phenomenon of "price wars" among various AI large models will not continue, as this is not a normal business logic. In his view, the reason why major models are lowering prices is mainly due to improving technology and decreasing costs. However, going too far is not beneficial, and true value empowerment should be achieved step by step.

Zhang Peng stated that they aim to provide better quality services to everyone, hoping that users will use these services to create greater value, which will then be reciprocated back. This is a normal and reasonable market value logic.

Therefore, Zhang Peng believes that while the current price war does benefit users, it will not last long.

Finally, Zhang Peng concluded that in the future, the new value chain brought by AI will exhibit the characteristics of "single-point outbreak, comprehensive breakthrough," meaning that the changes triggered by AI will be comprehensive, involving various industries and fields, leading to an "exponential" value creation expectation