On January 10, 2026, Tsinghua University and Zhipu hosted the AGI-Next summit in Beijing, featuring key leaders in Chinese AI, including Tang Jie from Zhipu, Yang Zhilin from Moonshot AI, Lin Junyang from Alibaba's Qwen, and Yao Shunyu from Tencent. They engaged in discussions about the US-China AI rivalry, the dynamics of open-source versus closed systems, and business models shaping the future of AI. A translated transcript reveals insights into China’s technology landscape, emphasizing challenges in lithography, compute capacity, and the need for a more robust enterprise market. Yao Shunyu highlighted that China has advantages in electricity and infrastructure but faces significant hurdles in production capacity and software development. He noted that the willingness of companies to take risks is hampered by the current economic and cultural environment, leading to a preference for safer projects. Lin Junyang pointed out that while US companies can invest heavily in innovative research, Chinese firms often focus on immediate delivery needs, limiting their capacity for groundbreaking work. Tang Jie shared his journey from academia to founding Zhipu, driven by the ambition to make machines think like humans. His philosophy, inspired by a coffee addiction metaphor, underscores the importance of focus and perseverance in research. The summit indicated that while China excels in certain areas, it still needs to foster a culture of risk-taking and innovation to truly compete on the global stage.