These past seven years have been a period of rapid advancement in artificial intelligence. From the introduction of Transformer to the emergence of ChatGPT and then to the birth of Sora, the evolution of AI technology has accelerated. What gaps exist between us and the international leading level? How can we catch up with and surpass the international leading level in terms of computational power and models? How can we seize the opportunities? These questions are right in front of us, unavoidable and should not be ignored.

On May 9, 2024, Zhejiang Lab (ZJ Lab) hosted the "Zhijiang Roundtable", gathering nearly 100 scientists, entrepreneurs, and investors to address problems, engage in exchanges, have dialogues, and contemplate the future.

"In the rising tide of AI, we must not be absent and must seize the opportunity, striving to be pioneers and leaders in the field of artificial intelligence." As TONG Guili, Secretary of the Leading Party Members' Group of the Department of Science and Technology of Zhejiang Province and Secretary of the Party Committee of ZJ Lab, said in her speech that, in the era of big science, major technological innovation increasingly relies on industry-university-research collaboration, as well as cross-disciplinary innovation. ZJ Lab sincerely hopes to use the roundtable as a platform to explore the establishment of a new mechanism for innovation through collaboration among "scientists, entrepreneurs, and investors", working together to accelerate the "integration of four chains" (i.e., innovation, industry, capital, and talent chains) and the seamless integration of "education, science and technology, and human resources" to achieve extensive consultation, joint contribution, shared benefits, and common development.

The first session of the Zhijiang Roundtable under the theme of "AI Competition - Opportunities and Challenges" commenced. Here are a few key highlights extracted from the roundtable to inspire further contemplation.

"System Innovation" and "Value-Driven"

XU Yinghui, a research fellow from the Artificial Intelligence Innovation and Incubation (AI³) Institute of Fudan University, stated that the development of artificial intelligence should be systematic innovation. In other words, it should encompass advancements in high-performance computing as well as low-performance computing. Breakthroughs are also needed in technologies such as high-speed storage and mass storage. Systematic architectural thinking is necessary to propel China's AI development further.

"Catch-up and Leapfrog Development" and "Full of Confidence"

"We should approach the technical roadmap with a pragmatic attitude, taking into account both the present and the future." ZHANG Peng, Senior Vice President of H3C, suggests that innovation cannot be planned; trial and error should be encouraged in rapid iterations to achieve catch-up and leapfrog development. Innovation should not be narrow-minded; it should embrace differences while being inclusive. China has a rich industrial ecosystem and application scenarios, offering great potential for AI development.

ZHANG Ruyun, a research fellow of ZJ Lab's Intelligent Computing Infrastructure Innovation Center, explained that ZJ Lab focuses on intelligent computing, not only in advancing related cutting-edge basic research but also in establishing a solid foundation for computational power. By empowering computational power with software, the goal is to make it truly user-friendly, encouraging people to use it confidently and willingly. HAN Dingyi, a distinguished expert of ZJ Lab's Research Center for Data Hub and Security, mentioned, "We still need a lot of talent, followed by data and then computational power." He added, "China has the advantage of massive data. I hope to unite all forces, seize the historical opportunity, and consolidate our country's high-quality data. I believe we could build excellent foundation models."

"With the efforts of the Hangzhou Industrial Alliance for Artificial Intelligence, a good mechanism has been formed to unite institutions and enterprises that work toward shared goals. I believe we will be able to firmly follow this path," said ZHU Shenghong, President of CloudNine Information Technology Co., Ltd.

Ecosystem! Ecosystem! Ecosystem!

Several experts at the meeting believe that ecosystem conservation is crucial for the development of the artificial intelligence industry. LIU Shufeng, Chairman of Hundsun Technologies Inc., stated that in the future, in addition to further enhancing the capabilities of foundation models, the industry can also work together on industrialization and business model exploration to promote the development of the foundation model ecosystem. "The combination of innovative talent and innovation capital is crucial for the success of the AI industry," said CHEN Bin, Chairman of Zhejiang Saizhi Bole Equity Investment Management Co., Ltd.

CHEN Xiangming, President of Ginkgo Valley Capital, stated, "We have not yet seen a surge in applications due to the sharp decrease in costs for foundation models. This is an area where concerted efforts are needed. We look forward to the Hangzhou Industrial Alliance for Artificial Intelligence taking the lead on this issue."

At the Roundtable, SHI Dongcai, Deputy Director-General of the Department of Science and Technology of Zhejiang Province, and ZHANG Jilin, Director of the Data Resources, Department of the Data Bureau of Zhejiang Province, exchanged views on AI innovation layout and data collection, governance, and circulation. CAO Guoxiong, Chairman of Puhua Group, CHEN Xiaofeng, Chairman of ZJKF Capital Management Co., Ltd., and other heads of investment institutions analyzed the investment opportunities for AI uses in industrial production, life and health, low-altitude economy, and other fields.

ZJ Lab focuses on intelligent computing. The understanding of intelligent computing can be summarized in two sentences. First, computing relies on intelligence, that is, intelligent technology is used to achieve high-performance computing. Second, computing aims at intelligence, indicating that the purpose of computing is to enable artificial intelligence. WANG Jian, Academician of the Chinese Academy of Engineering and President of ZJ Lab, believes that artificial intelligence is not simply a revolution in technological tools; it is a tool to revolutionize scientific fields. The development of artificial intelligence gives us the opportunity to independently evaluate the technological, social and commercial value of innovations. "To advance artificial intelligence, human creativity is paramount. Reassessing innovation logic and mechanisms is also essential. I hope that we could achieve joint contribution, shared benefits, and common development through extensive consultation, and become creative executors to jointly create something new."

At the Roundtable, sincere and candid discussions were conducted on many issues regarding how to make breakthroughs in the future and when the future will arrive. In the future, Zhijiang Roundtable will be held regularly to build itself into a platform for open communication and collaboration, promote innovations through government-industry-academia-research-application-finance collaboration, and inject new vitality into the development of new quality productive forces.