Interview with President Jin Li of Fudan University: Both scientific research and education in universities should embrace AI

Tech 2023-07-14 12:27:39 Source: Network

Reporter Sun Bing from China Economic Weekly | Beijing ReportWhile the public is still actively discussing how the AI big model will change university education and scientific research, Fudan University has already arranged a "magic tool" for all teachers and students in the school.On June 27th, the largest cloud based research and intelligent computing platform (CFFF) for domestic universities was officially launched at Fudan University

Reporter Sun Bing from China Economic Weekly | Beijing Report

While the public is still actively discussing how the AI big model will change university education and scientific research, Fudan University has already arranged a "magic tool" for all teachers and students in the school.

On June 27th, the largest cloud based research and intelligent computing platform (CFFF) for domestic universities was officially launched at Fudan University.

It is reported that this scientific research "supercomputer" built to discover and solve complex scientific problems was jointly built by Fudan University, Alibaba Cloud, China Telecom, etc. It provides ultra thousand card parallel intelligent computing in an advanced public cloud model, and can support the training of large models with hundreds of billions of parameters. This is not only the first among domestic universities, but also ahead of internationally renowned universities such as Stanford University.

Although the big model and other AI technologies are still in dispute, Jin Li, president of Fudan University and academician of the CAS Member, said in an interview with the China Economic Weekly that human mastering AI technology is like human learning to generate electricity. AI is also a technology and tool at this level, which will profoundly change human lifestyle and social structure.

Therefore, the positioning of Fudan is very clear. With the advent of AI technology, we need to embrace it. Although its negative effects also need to be considered, we believe that both teachers and students need to have the ability to understand AI and use AI tools well, "Jin Li said.

Jin Li, President of Fudan University and academician of the CAS Member (photographed by Sun Bing, reporter of China Economic Weekly)

How does AI change research?

To understand the value of AI from the perspective of paradigm change in scientific research

If you want to do good work, you must first sharpen your tools. For researchers, in the current AI era, a supercomputer with powerful computing power is undoubtedly a "big scientific device" that can enhance scientific research work and a new type of "infrastructure" for scientific innovation.

It is reported that the CFFF platform of Fudan University consists of two parts: the AI for Science intelligent computing cluster "Qienyi" for multidisciplinary integration and innovation and the dedicated HPCC "Jinsi No.1" for advanced research. Since the first day of its construction, it has received a variety of research needs from different Fudan colleges and departments, not only covering life science, Atmospheric science, materials science, but also social science research such as financial system analysis.

At present, the first scientific research achievement on the CFFF platform has been born. Li Hao's team from the Institute of Artificial Intelligence Innovation and Industry Research of Fudan University recently released a large model of medium and short term weather forecast with 4.5 billion parameters. The prediction effect reached the industry recognized collective average level of ECMWF (European Centre for Medium-Range Weather Forecasts) for the first time in the public data set, and the prediction speed was shortened from the original hourly scale to 3 seconds.

Based on the CFFF platform, a large-scale parallel intelligent computing model with thousands of cards can be trained in just one day. It is difficult to achieve this using traditional computing platforms, "said Li Hao.

And this is also the first big model nurtured on the CFFF platform. Jin Li revealed that in the future, Fudan hopes to build a number of world-class scientific models based on the CFFF platform, such as life science model, material science model, Atmospheric science model, integrated circuit model, etc.

The intelligent computing platform represented by the CFFF platform, as an emerging research supercomputing architecture, will become an important support force for scientific research, greatly improving research efficiency, reducing research costs, accelerating scientific principle discovery and technological breakthroughs, and effectively promoting the implementation of scientific models.

Li Hao's team from the Artificial Intelligence Innovation and Industry Research Institute of Fudan University has developed the first large-scale meteorological model based on CFFF (provided by respondents)

In addition to allowing teachers and students at Fudan University to use "supercomputing", Jin Li further believes that the position of artificial intelligence in scientific research needs to be viewed from the perspective of paradigm change in scientific research.

According to Jin Li, the traditional research paradigm has undergone four stages of evolution and development. Firstly, it is the "empirical paradigm" that describes natural phenomena through experiments; The second is a "theoretical paradigm" for research through models or induction; The third is the "computational paradigm" of applying computer simulation to solve disciplinary problems; The fourth is the "data paradigm" of studying the internal relationship of things through Big data analysis.

However, with the continuous growth of Big data resources and the increasing complexity of scientific problems to be solved, the exploration of the "Fifth normal form" of scientific research has been triggered, that is, on the basis of the data paradigm, the introduction of intelligent technology, the emphasis on the integration of human decision-making mechanism and data analysis, and the effective combination of data science and computing intelligence, which is the arrival of the "AI for Science" era.

Facing the widespread discussion triggered by the AI model, how to win the initiative and achieve innovative breakthroughs in key areas in the rapidly changing technological innovation environment is a new proposition given to education by the times. This is not only related to talent cultivation, but also to future international competition. "Jin Li emphasized.

Wang Jian, an academician of the CAE Member and the founder of Alibaba Cloud, also agrees with this view. Wang Jian comes from the industry, but he was also a college educator. This time, Alibaba Cloud and Fudan University jointly built the CFFF platform, which also stems from an important consensus of both sides: the inclusive value of cloud computing plays an important role in the collaborative innovation of industry, university and research.

Wang Jian told China Economic Weekly that giving universities the same computing power as technology giants will help promote collaborative innovation between industry, academia, and research. Currently, many studies rely on computing, but even in the United States, universities do not have the research and computing platforms that many companies have today. Many internationally renowned universities are still in the stage of using mainframes or personal computers for research, "he said.

Zhao Dongyuan, a professor in the Department of Chemistry of Fudan University and an academician of the CAS Member, is guiding students in experiments and using AI tools to process experimental data (respondents provide pictures)

How does AI change education?

Breaking free from the confusion of the "Tiankeng" profession through the integration of industry and education

According to data from the Ministry of Education, the number of applicants for the national college entrance examination in 2023 was 12.91 million, an increase of 980000 compared to last year, setting a new historical high. In 2023, the number of college graduates in China reached 11.58 million, an increase of 820000 compared to the same period last year, also reaching a new historical high.

12.91 million young people who are about to enter university are facing the choice of voluntary majors. They are afraid to enter the "Tiankeng" major, and they study hard but find employment difficult; 11.58 million graduates who are about to enter society are facing the choice of employment career, fearing to become "waste" of talent and unable to match social needs with their knowledge and abilities.

Their confusion and confusion are undoubtedly issues that educators need to consider. Jin Li stated that this is also the focus that Fudan University has been exploring: how to help universities break free from the confusion of "sinkhole" majors? How can the people produced by universities create greater value for society?

Colleges and universities have established a system for cultivating students that has been in place for many years, but with the rapid development of technology, our existing system has encountered great challenges: firstly, the continuous generation of new technologies, and secondly, the cross integration of disciplines.

Jin Li takes Fudan's innovative exploration as an example. At present, the "2+X" undergraduate training system of Fudan University has been basically improved. Under the flexible educational system, students can freely take various courses and obtain various learning resources while completing Liberal education and major basic education.

In Jin Li's view, the students cultivated in a school are like the "product" of the school. If it is difficult to contribute to social development, it is a very practical problem.

I believe that an important and effective way for universities to avoid the confusion of 'Tiankeng' majors is the integration of industry and education. This integration of industry and education is not a simple technology transfer, but is reflected in two aspects: close interaction and cooperation in innovation; and close interaction and cooperation in talent cultivation. There should be a two-way empowering relationship between schools and enterprises in innovation and talent cultivation, "Jin Li said.

Enterprises can see more about social needs, so they can help universities set goals and directions, thereby helping universities have clearer goals in talent cultivation. At the same time, we often say that enterprises are the main body of innovation, but there are not many enterprises in China that can truly take on the responsibility of innovation as the main body. Talents are the key factor in enterprise innovation, and enterprises can also obtain more innovation needs through close cooperation with universities Talent Jin Li said.

Jin Li also emphasized that AI, as an important technology and tool, will also change the culture of innovation. Because AI can make difficult and cumbersome tasks easier and more efficient. As people who previously did not understand and did not know how to use AI learn to use AI and are able to use the best AI tools, there will be an explosion of innovation.

Wang Jian also stated that joint innovation between universities and enterprises can create many wonderful things, especially in the fields of computing and communication. For example, the collaboration between IBM and Columbia University gave birth to the computer science major; The cooperation between MIT and Bell Labs has enabled almost every university to have a communication major. The cooperation between enterprises and universities will also bring more possibilities for collaborative innovation between industry, academia, and research.

In the 'big science era' driven by data and intelligent technology, how to win the initiative and achieve innovative breakthroughs in key areas in the rapidly changing technological innovation environment is a proposition given by the times, and requires the cooperation of all parties in politics, industry, academia, and research to answer.

Professor Liu Zhipan from the Department of Chemistry at Fudan University explores chemical research and develops the LASP platform using new methods such as AI (provided by interviewees)

Zhou Yang, a teacher from Big data Research Institute of Fudan University, uses computing to explore social sciences (the picture is provided by the interviewees)

Editor in Chief: Guo Jiyao

(The copyright belongs to the China Economic Weekly magazine, and no media, website, or individual is allowed to reprint, extract, link, repost, or use it in any other way without authorization.)

Disclaimer: The content of this article is sourced from the internet. The copyright of the text, images, and other materials belongs to the original author. The platform reprints the materials for the purpose of conveying more information. The content of the article is for reference and learning only, and should not be used for commercial purposes. If it infringes on your legitimate rights and interests, please contact us promptly and we will handle it as soon as possible! We respect copyright and are committed to protecting it. Thank you for sharing.(Email:[email protected])

Mobile advertising space rental

Tag: Interview with President Jin Li of Fudan University Both