Experts praise country's vision of AI development

China is fostering its own vision for the development of artificial intelligence through reliable and efficient applications of AI tools across a wide range of industries, a wise path that differs from that of the United States, said industry experts and company executives at the ongoing Zhongguancun Forum.

They made the comments as several AI-related panel discussions saw packed houses over the weekend, with attendees filling every seat and some even standing along the walls. The overwhelming interest mirrored the prominence that AI has commanded across the nation, which was also seen at the recently concluded Boao Forum for Asia Annual Conference and the China Development Forum.

Turing Award winner Joseph Sifakis said at the Zhongguancun Forum that China is crafting its own AI vision, distinct from that of the US. "China has a solid and extensive industrial base and a unified domestic market. This enables the country to develop more reliable AI solutions that better align with the needs of the real economy, especially in the long-awaited transition to autonomous driving," he said.

Sifakis noted that China's strong industrial foundation, in particular, gives it an edge while industries such as self-driving vehicles, smart cities, smart factories and intelligent farms present more opportunities. "If developed well, (all these) will give China a dominant position in industrial AI," he said.

Kai-Fu Lee, a prominent AI expert and chairman and CEO of investment company Sinovation Ventures, said that China has reached its "DeepSeek moment". He predicted that 2025 would mark a breakout year for AI applications and large-scale model deployment in China.

Lee recalled that about nine months ago, he had expressed frustration over China's lack of a "ChatGPT moment", as promising Chinese AI models at the time failed to stand out and spark nationwide adoption.

"However, DeepSeek has changed that landscape. Its success has awakened the Chinese market, ushering in a new AI era for the country," he said.

According to Lee, DeepSeek's rise proves that "closed-source AI has no future", and only open-source development will drive greater progress.

"As AI scaling laws shift from the pretraining stage to the inference stage, AI applications will accelerate exponentially this year," he said.

Gao Xuefeng, founder and CEO of AI data infrastructure company Fabarta, said that after DeepSeek's breakthrough, many Chinese companies no longer hesitate to invest in AI.

"Every few months, new technological advancements emerge. Companies can't afford to wait; they must integrate AI into their business and industry now," Gao said, adding that in China, every sector can be reimagined through AI in the years to come.

Wang Zhongyuan, head of the Beijing Academy of Artificial Intelligence, a leading nonprofit AI institute, said that DeepSeek has indeed delivered impressive results, proving that China can train large-scale AI models parallel to ChatGPT 4 despite limited computing power.

"Such technologies could also be adopted by international institutions and other nations, accelerating the global development of AI," he said.

Regarding artificial general intelligence, in which an AI system can match or exceed the cognitive abilities of humans in any real-world task, Wang said the industry is still far from achieving such capabilities. "I think we are still at least five to 10 years, or even longer, from reaching that level," he added.