Explainer: How China’s AI Surge Could Outpace U.S. By Next Year

The former president of Google China and current founder of 01.AI and Sinovation Ventures, Kai-Fu Lee, has stated that Chinese artificial intelligence models currently lag approximately six months behind their American counterparts.

AI (Artificial Intelligence) letters and robot hand are placed on computer motherboard in this illustration taken, June 23, 2023. REUTERS/Dado Ruvic/Illustration

However, as Chinese companies develop large language models (LLMs), the gap has narrowed significantly, positioning them now just six to nine months behind in technological capability. Lee shared these insights during the AVCJ Private Equity Forum in China, as reported by CNBC on September 11.

Lee posits that Chinese AI applications may soon surpass those available in the United States, largely due to substantial reductions in the cost of training AI models witnessed in recent months. He predicts that by early next year, the proliferation of AI applications in China will occur at a pace much quicker than that in the United States. However, he acknowledged that it remains uncertain whether these applications will emerge from smaller startups or larger enterprises.

Furthermore, he speculated that achieving consumer-ready generative AI functionality comparable to a “super app”—an application capable of performing a multitude of tasks—may take between five to eight years. Lee asserted that the advancement of AI technology will necessitate novel hardware devices, as current smartphones may be insufficient to meet the increasing demands of AI applications. He articulated a vision for AI devices that are perpetually operational and attentive, a notion echoed by the release of the “Friend” necklace in July, an AI-powered wearable designed to serve as a virtual companion.

In parallel with developments in the United States, the Chinese AI sector has garnered significant interest from major tech corporations such as Alibaba and Tencent. These firms have not only introduced multiple iterations of proprietary AI models but have also invested billions in funding smaller AI startups. Tencent, for instance, launched its own LLM known as “Hunyuan” in September of the previous year, positioning it as a direct competitor to OpenAI’s ChatGPT. Since its inception, Hunyuan has been integrated into Tencent’s extensive ecosystem, impacting its cloud computing, marketing, and gaming divisions.

While Chinese AI models currently demonstrate a temporal lag behind those in the United States, the rapid advancements in application development paired with aggressive investment strategies suggest a looming opportunity for China to establish prominence in the global AI landscape in the near future.