China Celebrates DeepSeek's AI Success, But The Race Is Far From Over

The logo of DeepSeek was showcased at the Global Developer Conference in Shanghai, highlighting the company's recent achievements in artificial intelligence.

DeepSeek's recent advancements in AI represent a significant milestone for China, yet experts caution that it does not alter the United States' dominant position in this crucial field. Zheng Yongnian, who serves as the dean of the School of Public Policy at the Chinese University of Hong Kong's Shenzhen campus, noted that while the rise of DeepSeek has stirred strong nationalist feelings, it is important for China to recognize its current technological shortcomings.

Zheng voiced his concerns on a social media platform associated with CUHK-Shenzhen, advising against excessive national pride in the face of global technological competition. He explained that the gap between China and the U.S. remains vast, particularly in terms of technology and the quality of data.

Breakthrough Innovations by DeepSeek

DeepSeek, located in Hangzhou, recently astonished observers by introducing two significant AI models: the V3 large-scale language model in December and the R1 inference model in recent weeks. These open-source models demonstrate capabilities comparable to leading American chatbots from firms such as OpenAI and Google, yet they are less expensive to develop.

Chinese state media quickly hailed DeepSeek as a national treasure in the race for AI supremacy. The People’s Daily, the official newspaper of the Communist Party, lauded the company as "a testament to China's rapid tech adoption and relentless innovation drive," especially amidst ongoing chip export restrictions from the U.S. and its allies.

Challenges Ahead in Tech Competition

Despite these successes, Zheng warned that embracing a populist mentality could hurt China’s future in AI development. He pointed out that significant gaps still exist between China and the U.S. regarding advanced semiconductor production, which is vital for many industries including manufacturing and national security.

According to Zheng, while the U.S. has access to extensive global data, China's data remains largely isolated due to limited internet openness, making data integration difficult among businesses, government entities, and different regions.

To enhance its competitive edge in AI, Zheng advises the Chinese government to ease regulatory constraints. He emphasized that China should consider adopting a more open regulatory framework, comparable to that in the U.S., to prevent losing AI talent to other countries.

Regulatory Framework and Future Directions

Currently, all generative AI technologies in China must pass security evaluations conducted by the Cyberspace Administration of China before being launched publicly. The Chinese government is also in the process of developing a unified AI law, with a draft submitted for review to China’s top legislative body. However, this initiative has encountered significant resistance from both academic and industry sectors.

Zheng advocates for a national policy that supports open-source AI models, explaining that many Chinese companies initially based their products on Western open-source frameworks. However, in the end, many of these products have turned out to be closed source, hindering technological advancement and leading to a potential decline in global competitiveness.

Adopting a strategy of looser regulations and promoting open-source development will be critical for China as it navigates pressures from the U.S. and works to catch up with advancements from Europe, Japan, India, and other nations.

China, AI, DeepSeek, Technology, Competition