Zhou Hongyi Discusses Large‑Model AI Trends and Their Role in New‑Quality Productivity at the 6th National Youth Entrepreneurs Conference

On May 16, the 6th National Youth Entrepreneurs Conference jointly hosted by the All‑China Federation of Industry and Commerce and the Central Committee of the Communist Youth League opened in Wuhan, Hubei. Deputy Minister of the United Front Work Department and Party Secretary of the All‑China Federation of Industry and Commerce Xu Lejiang, Hubei Provincial Vice Governor Wang Zhonglin, and other senior officials attended. 360 Group founder Zhou Hongyi gave a speech titled “Large Models Empower New‑Quality Productivity and High‑Quality Development.”

Zhou Hongyi believes that the nation is vigorously developing new‑quality productivity, with digitalization and artificial intelligence as key drivers. He stresses that AI must be combined with traditional industries, especially manufacturing, to effectively build new‑quality productivity.

He argues that in the AI era, people should establish an “AI faith,” believing that large models are truly intelligent, that AGI is accelerating, and that large models will reshape all products and services. With this belief, large models are seen not as a bubble but as the catalyst for a new industrial revolution.

Regarding how entrepreneurs should engage with the large‑model wave, Zhou advises young entrepreneurs not to chase super‑general large models, which require algorithmic, computational, cost, talent, and security strengths that only giant firms possess. Instead, entrepreneurs should “think different,” focusing on specialized, smaller models.

Zhou outlines three current trends in large‑model development: (1) Mixture‑of‑Experts (MoE) architectures that combine multiple expert small models, (2) the terminalization of large models, and (3) the explosion of open‑source large models. He believes that algorithm, model, and compute are no longer bottlenecks; the critical challenge is identifying “star scenarios” where large models add value.

He emphasizes that large models only possess capabilities such as drawing, writing, and analysis, not finished user products. The notion that “model equals product” is a false premise; the capabilities must be demonstrated through concrete scenario applications that solve real problems. Accordingly, 360 leverages the 360AI browser to provide scenarios for rapid processing of video, audio, text, and images, benefiting teachers, students, and content creators.

Facing the opportunities presented by the rapid rise of large models, Zhou suggests deeper integration with enterprise scenarios to boost productivity. He hopes to establish a 360 large‑model factory in Hubei to offer tools and deployment services for entrepreneurs, enabling low‑cost, efficient model construction and accelerating the development of new‑quality productivity.