China's AI Landscape Reveals Challenges in Achieving General Intelligence

The recent Conference on General Artificial Intelligence in Chongqing, held from October 21 to 23, 2026, highlighted significant advancements and ongoing challenges in the pursuit of artificial general intelligence (AGI). This event brought together researchers and industry leaders to discuss how AI can transition from performing specific tasks to applying learned knowledge in unfamiliar and complex real-world situations. As China solidifies its position in the global AI landscape—with over 6,600 AI companies representing about 15% of the global total—these discussions are critical for understanding the future trajectory of AI development.

Current AI systems excel at narrow tasks but struggle with generalization and autonomous decision-making. Fields Medalist David Mumford emphasized that achieving agency—AI's ability to act independently—remains a formidable hurdle. He pointed out that for AI to operate autonomously in society, it must be equipped with appropriate values. This need for ethical frameworks was echoed by Zhu Songchun from the Beijing Institute for General Artificial Intelligence, who introduced the CUV framework. This model integrates cognitive architecture, capability systems, and value systems, aiming to create adaptable intelligence that can function across various domains.

Robotics was identified as a practical area for exploring these AGI challenges. Experts discussed the potential of world models to enhance robots' capabilities, enabling them to navigate complex environments and tasks. Feng Liang from Chongqing University noted that the data generated by these models must accurately reflect real-world conditions to be effective. This highlights a critical requirement for AI development: not only improved algorithms but also robust computing systems that can support complex tasks.

Li Zicheng of Hygon, a semiconductor company, pointed out that achieving cross-domain reasoning in AI requires more than just increased computational power. The interplay between CPUs and accelerators, along with high-speed interconnects, is essential for facilitating advanced AI functions, including task planning and workflow orchestration. This underscores a growing trend in the industry toward developing integrated hardware and software solutions that can handle the demands of sophisticated AI applications.

In practical applications, companies like Alibaba are already leveraging AI to enhance marketing workflows. Li Mu, head of Alibaba's Alimama Marketing and Experience Center, noted that while AI can generate valuable insights, human expertise remains crucial for strategic decision-making. This reflects a broader reality in the AI landscape: while technology can augment human capabilities, it cannot yet replace the nuanced understanding that human judgment provides.

The conference also addressed the need for AI to comprehend the complexities of human society, where outcomes are influenced by social interactions and information dynamics. Research presented by Wang Hui from the Communication University of China focused on simulating public opinion evolution through AI interactions. This work illustrates the potential for AI to model and predict social phenomena, but translating these insights into actionable applications remains a challenge.

The Mingjing platform, launched at the conference, aims to bridge the gap between research and practical application. It facilitates the development and testing of AI systems in real-world contexts, with a focus on urban governance and public services. Such initiatives are essential for assessing AI performance and identifying limitations, guiding future research efforts.

While these advancements signal progress toward AGI, the consensus remains that achieving true general intelligence is a long-term goal without a clear timeline. The initiatives emerging from China's AI sector, particularly in Chongqing, represent a fertile ground for experimentation and innovation. As companies and researchers collaborate to refine AI capabilities, the implications for industries ranging from marketing to public policy are profound. The continued integration of AI into diverse sectors will likely accelerate the pace of innovation, shaping how businesses operate and interact with consumers in an increasingly complex digital landscape.