CSET

China’s Artificial General Intelligence

We’re Still Getting It Wrong

William Hannas

and Huey-Meei Chang

August 29, 2025

Recent op-eds comparing the United States’ and China’s artificial intelligence (AI) programs fault the former for its focus on artificial general intelligence (AGI) while praising China for its success in applying AI throughout the whole of society. These op-eds overlook an important point: although China is outpacing the United States in diffusing AI across its society, China has by no means de-emphasized its state-sponsored pursuit of AGI.

Recent op-eds by Eric Schmidt and Jack Shanahan comparing the United States’ and China’s artificial intelligence (AI) programs fault the former for its narrow focus on artificial general intelligence (AGI)—a hypothetical state in which AI exceeds human cognitive skills in all domains—while praising China for its success in applying AI throughout the whole of society.

Their main point is that the United States’ singular focus on large language models (LLMs) in pursuit of AGI is distracting companies from developing practical AI applications and exploring alternative options. Our research supports this position. But at the same time, these op-eds overlook an important point: although China is outpacing the United States in diffusing AI across its society, China has by no means de-emphasized its state-sponsored pursuit of AGI.

Instead, China pursues both projects simultaneously. Moreover, China invests in not one but multiple paths to AGI. This includes the large statistics-based “frontier” models favored in the West (think DeepSeek) as well as wholly different approaches, such as brain-inspired (类脑) AI, hybrid models using brain-computer interfaces, and “embodied” (具身) AI embedded in applications that interact with the real world and learn from it continuously.

In fact, a chorus of top AI scientists (such as Ya-Qin Zhang, Tan Tieniu, and Wang Yequan and Sun Aixin) believe embodied AI is the approach most likely to achieve some early form of AGI. Immersing AI in physical environments grounds the AI in reality, exposes it to limitless data, facilitates autonomous learning, and enables sharing between agents—conditions Chinese AI researchers believe are ideal for AGI to emerge.

China’s scientists are aware of the salutary effects of embodiment, see no contradiction between practical AI and AGI development, and have already taken concrete steps to realize this merger on a city-wide scale. So, there are two lessons to learn here from China: the first is not to take all eggs from one basket and put them into another but rather to adopt multiple approaches to AI development. The second lesson is to back these approaches with national resources.