Collaborative, innovative AI ecosystem highlighted

Chinese tech companies' intensified efforts in achieving breakthroughs in artificial intelligence will bolster the application of this cutting-edge technology in a diverse range of fields, drive industrial upgrading and inject fresh momentum into global economic growth, despite the tightened export controls of the United States on AI chips, said experts and business executives.

Noting that Chinese AI startup DeepSeek has recently caused a global sensation with its cost-effective and open-source large language model, they added that the U.S.' attempt to contain China's rise in the high-tech sector will only motivate Chinese enterprises to double down efforts on independent innovation, in order to make progress in crucial technologies.

Industry insiders also called for strengthening international cooperation to build an open, collaborative and innovative AI industrial ecosystem and to gain an upper hand in the increasingly fierce global AI chatbot race.

Leading Chinese tech heavyweights, including Alibaba Group, Baidu Inc and ByteDance, are scrambling to bolster technological advancements in AI models and propel the use of generative AI across various sectors such as healthcare, manufacturing and finance.

Alibaba Cloud, the cloud computing arm of Alibaba Group, recently unveiled its latest AI model, Qwen2.5-Max, which boasts enhanced math and coding capabilities and has outperformed other leading AI models such as U.S.-based OpenAI's GPT-4o and DeepSeek's V3.

Wu Yongming, CEO of Alibaba Group, emphasized the importance of an open-source approach regarding LLMs, saying it will lower the threshold for the development and application of AI, greatly reduce the cost of computing power and boost the popularization of the state-of-the-art technology in a wide array of industries.

Last week, Baidu announced that it has launched China's first third-generation Wanka cluster based on its self-developed Kunlun chips, marking a milestone in the AI computing power field and significantly improving training efficiency of AI models.

The Wanka cluster is a high-performance computing system composed of 10,000 or even more GPU computing accelerator chips that are mainly used to train and fine-tune AI models.

The Ministry of Industry and Information Technology said on Saturday that China's three largest telecom operators have all integrated DeepSeek's LLM into their networks and provided dedicated computing power solutions and infrastructure to bolster the performance of DeepSeek's model.

These technological developments followed the recent launch of DeepSeek-R1, the latest AI model developed by DeepSeek, the performance of which is on a par with leading models from OpenAI in tasks such as mathematics, coding and natural language reasoning, but at only a fraction of the cost and computing power of its rivals.

Zhou Mi, a senior researcher at the Chinese Academy of International Trade and Economic Cooperation in Beijing, said the AI industry's development has long relied on huge amounts of computing capacity and capital input, while DeepSeek's emergence has overthrown the existing paradigm and redefined the global AI landscape amid Washington's heightened export restrictions on cutting-edge semiconductor technology.

"The open-source approach adopted by Chinese companies will be conducive to propelling international cooperation in science and technology, further driving the globalization of the world economy," Zhou said.

Pan Helin, a member of the ministry's Expert Committee for Information and Communication Economy, said that despite facing restrictions in terms of advanced chips, homegrown AI models have made significant breakthroughs with low costs, high performance and open-source features.

The meteoric rise of DeepSeek has not only demonstrated Chinese companies' growing independent innovation capacity in the AI sector but also underscored that AI innovations can thrive under restrictions, he said, adding that China has the ability to take the lead in technological advancements.

Pan also underlined the need to foster technological collaboration and innovation through open-source principles.

Ouyang Rihui, assistant dean of the China Center for Internet Economy Research at the Central University of Finance and Economics, said: "The U.S. has sought to curb China's technological rise through export controls and investment restrictions. However, Chinese tech companies have developed alternative solutions to train AI models with less high-end chips by optimizing algorithms, model architectures and training procedures."

Ouyang noted that the U.S.' technological blockade would not impede China's innovation progress, but instead it would prompt Chinese enterprises to seek more breakthroughs.