China leads US in everyday AI apps but firms are overvalued, experts say
Executives say China is ahead in getting AI apps to users, even as its model capabilities still trail the US and AI firms may be overvalued.
Intelligence analysis by GPT-5.4 Mini
At an HKEX tech summit, Alibaba Cloud's Chi Zhang and Infinigence AI's Lixue Xia said China's edge is in putting AI into everyday use, not in raw compute. They also warned that Chinese AI firms may be getting overvalued even as the US-China model race tightens.
China is getting AI helpers into phones and apps people use every day faster than the US, even if its smartest computer brains are still a bit behind. It is like making toy cars that people can drive now while the engine contest is still going on.
Analysis
What the executives argued
At the 2026 HKEX Future Tech Summit in Shenzhen, Alibaba Cloud Intelligence Group general manager Chi Zhang said China is still behind the US in computing power, but only about “100 days behind” in frontier model capabilities. The more important advantage, he said, is in AI applications: China has a large base of entrepreneurs and engineers, and its stage of economic development gives it room to deploy AI widely in everyday products and services.
Lixue Xia, co-founder and CEO of Infinigence AI, made a similar case from the infrastructure side. She said China’s growth in a possible AGI race would be driven by sectors such as energy, infrastructure buildout, and its open-source ecosystem. That framing puts less emphasis on raw model prestige and more on where AI gets used in the real economy.
The article also notes that Chinese labs have recently released proprietary and open-source models that are close behind US peers such as OpenAI and Anthropic. Lower costs have already pushed some US firms to use Chinese vendors, including DeepSeek, which signals that price and deployment can matter as much as top-end capability.
The cautionary part of the story is valuation. Even as Chinese companies gain practical traction, the executives warned that AI firms may be getting ahead of themselves financially. In other words, the technology race looks closer, but market pricing may already be assuming too much success too soon.
Key points
- Chi Zhang said China is still behind the US in compute, but only about 100 days behind in frontier model capabilities.
- The article argues that China's biggest edge may be in getting AI into everyday applications.
- Lixue Xia said energy, infrastructure buildout, and open-source development could drive China’s AI growth.
- The story says Chinese models are closing the gap with US peers like OpenAI and Anthropic.
- Executives warned that Chinese AI firms may now look overvalued.
If China keeps turning AI into useful apps, adoption could spread faster across energy, infrastructure, and other sectors. Lower-cost models and an open-source ecosystem could also make AI tools easier for more firms to use.
If valuations keep rising faster than real usage, investors may overpay for AI firms. China still lags in computing power, so a gap in frontier capability could limit how far applications can scale.
