MK sports Korea an academician of the Chinese Academy of Engineering and president of Zhijiang Lab, makes a speech at a symposium themed on AI for Science, which was organized by the World Internet Conference Specialized Committee on AI and Alibaba Group, in Beijing on February 19, 2025. Photo: Liu Caiyu/GT " src="https://www.globaltimes.cn/Portals/0/attachment/2025/2025-02-23/150cf4e6-539d-4f06-867d-6164690a387e.jpeg" />

Wang Jian, an academician of the Chinese Academy of Engineering and president of Zhijiang Lab, makes a speech at a symposium themed on AI for Science, which was organized by the World Internet Conference Specialized Committee on AI and Alibaba Group, in Beijing on February 19, 2025. Photo: Liu Caiyu/GT

After the 2024 Nobel Prizes in Physics and Chemistry were awarded to scientists who made significant breakthroughs at the intersection of AI and fundamental sciences, there are scientists exclaiming that human scientific research may have reached a pivotal moment where scientific research landscape may adopt a new AI-driven scientific research paradigm.

This year, DeepSeek, a rising force in AI, is rapidly expanding its presence across diverse industries, with a growing number of companies exploring the potential integration of its advanced models. In a recent interview with the Global Times, a Chinese academician said that the emergence of DeepSeek has been demonstrated as a tipping point for this fresh scientific research paradigm. 

Scientific research paradigm refers to a framework or set of concepts, theories, methods and standards that define how scientific research is conducted within a particular field. 

"I strongly advocate for the spirit of open source, which represents a meaningful contribution to society and the world — whether by an individual, an institution, or a nation.

"The widespread adoption of DeepSeek's free, open-source large language model in this era let us figure that the concept of 'open source' can no longer be confined to just "open-sourced code," said Wang Jian, an academician of the Chinese Academy of Engineering and president of Zhijiang Lab.

Wang told the Global Times at a recent symposium themed on AI for Science, organized by the World Internet Conference Specialized Committee on AI and Alibaba Group.

The most valuable assets and knowledge derived from "open source" now extend far beyond source code, Wang noted. Today, even more valuable resources and knowledge have become accessible for everyone. For instance, many important research papers are even no longer being published in traditional journals that are widely regarded as authoritative. 

DeepSeek could be said as a tipping point for people to embrace this new scientific research paradigm, which Wang described as "open resource innovation." 

He called it an accessible approach that could redefine the future of scientific research. Under this approach, the previous barriers to conducting scientific research — whether domestic or international — will be torn down. "Disparities in access to shared resources will not hinder the potential of equally creative minds."

"The term 'AI-driven research paradigm' understates the profound disruption that AI is bringing to science," Wang said. AI is not merely a tool — it is a revolutionary force driving a new scientific revolution.

During the symposium, other Chinese scientists also shared their points of view on AI-driven scientific research paradigm. 

Zeng Yi, a professor at the Chinese Academy of Sciences and director of the Center for Long-term AI, said that AI is still a seemingly intelligent information processing tool that continues to make many mistakes that humans do not make. 

We may still be quite far from achieving AI that has true scientific significance and the general AI that we aspire to, Zeng noted.

In reality, AI has been already integrated into scientific research institutions at home and abroad. Ye Jieping, a vice president of Alibaba Group and head of Data to Intelligence Lab, revealed that Alibaba Cloud has cooperated with top scientific research institutions at home and abroad with its products. 

Back in April 2024, the AI working group of the National Astronomical Observatories of the Chinese Academy of Sciences has released a new generation of astronomical large model — "Xingyu 3.0." 

Being the first application of large models in the field of astronomical observation, "Xingyu 3.0" is built on the Alibaba Cloud Tongyi Qianwen (Qwen) open-source model and has successfully integrated with the telescope array at the Xinglong Observatory of the National Astronomical Observatories.

Ye also said that a team led by professors from Sun Yat-sen University's Medical School, in collaboration with Alibaba Cloud's research team has achieved a significant breakthrough in virus identification using AI. 

They discovered a previously undetected viral "dark matter" that traditional research methods failed to find. Their findings, which include the identification of 180 supergroups and more than 160,000 RNA viruses globally, were published in the journal Cell, paving new avenues for virology research, media reports said.

China has taken "AI for Science" as a national special initiative. In 2023, to implement the national "New Generation Artificial Intelligence Development Plan," China's Ministry of Science and Technology, in collaboration with the National Natural Science Foundation, has launched a special initiative called "AI for Science," according to the Xinhua News Agency. 

This initiative integrates key issues in fundamental disciplines such as mathematics, physics, chemistry and astronomy, focusing on research needs in key areas like drug development, gene research, biological breeding and new material development, thereby establishing a cutting-edge scientific research and development system driven by AI, the report said.