The thesis announced at Baidu World
In his speech at Baidu World 2026, Baidu founder Li Yanhong identified the so-called intelligent organization as the next stage in the evolution of artificial intelligence. According to IT之家's report, Li recalled that in 2016 he had predicted the rise of AI and that today that reading has become a fixed point. The novelty, therefore, does not concern technology itself but the way companies should structure themselves to use it.
The proposal is presented as a programmatic statement by the founder, not as the result of documented research or experimentation. The value of the news lies in the fact that Baidu, one of China's leading operators in the sector, places the discussion of AI adoption on an organizational and not merely technical plane.
From the Technology Stack to the Corporate Dimension
Li described the architecture of AI as a stack different from the traditional computing one: instead of the chip-operating system-software scheme, it places the GPU at the center, above which sit cloud infrastructures, then the model layer and finally the layer of agent-based applications. He also recalled the five-layer theory attributed to Nvidia founder Jensen Huang, which adds an energy layer.
Hence the question Li posed to the audience: if energy is the upstream of AI, what constitutes its downstream? His answer is organization, understood as an AI-native system and an alternative to traditional corporate hierarchy. The image he used is that of electricity: if AI is the new energy, the intelligent organization is the new factory, the place where productivity is concretely produced.
The Three Directions and the Limits to Be Verified
According to IT之家, Li stated that organizational intelligence represents the deepest moat for a company and outlined three principles: overcoming role boundaries to unleash internal energies, embedding tacit knowledge to build organizational intelligence, and designing closed-loop processes to foster the organization's evolution. He also urged companies not to treat AI as a mere tool for individual productivity, but to integrate digital workers into sales, customer service, research and development, and operations.
However, several aspects remain to be clarified. The material does not indicate company cases, metrics, or experiments demonstrating the effectiveness of the proposed model, nor does it clarify how the three principles translate into management practices. The scope of the thesis, therefore, should be read as a strategic vision declared by Baidu's leadership, useful for orienting the debate on AI adoption in companies, but not yet as empirical evidence.
