The "cold" thinking of artificial intelligence in colleges and universities

When setting up relevant majors, colleges and universities should organize professor committees and academic committees. In combination with the national talent policy, the demand for talents in industrial development, and the training of similar professionals in other universities at home and abroad, there are no realistic conditions for the establishment of this major in the university. How to carry out teacher construction, curriculum construction, professional characteristics, etc., and conduct full argumentation.

Speaking of artificial intelligence (AI), perhaps the public is not unfamiliar. Because in recent years, whether AlphaGo defeated human champions Li Shishi and Ke Jie in the Go game, or the driverless car was issued a test license, similar news events are labeled with the same label: artificial intelligence.

In July 2017, the State Council issued the “New Generation Artificial Intelligence Development Plan”, which opened a new journey for the rapid innovation and development of artificial intelligence in China. In April this year, the Ministry of Education issued the “Innovative Action Plan for Artificial Intelligence in Colleges and Universities”, which clearly stated that it will support the establishment of artificial intelligence disciplines in computer science and technology disciplines and promote the construction of first-level disciplines in the field of artificial intelligence.

It is reported that in the first batch of 612 "new engineering" research and practice projects identified in 2018, 57 human intelligence projects were laid out. As of December 2017, a total of 71 universities across the country have set up 86 secondary disciplines or interdisciplinary subjects around the field of artificial intelligence. In May of this year alone, there were four colleges and universities in Nankai University, Tianjin University, Nanjing University and Jilin University.

But in this context, should we give artificial intelligence a "cool down" and carry out some "cold" thinking?

A variety of factors to promote the boom

In fact, the artificial intelligence of colleges and universities is not without reason, but a chain reaction that conforms to the social development situation and actual needs.

As Tan Tieniu, an academician of the Chinese Academy of Sciences and vice chairman of the Chinese Society of Artificial Intelligence, said at the academic conference of the two academies this year, in the past decade, the external forces in the development of information technology such as big data, cloud computing, the Internet, and the Internet of Things have been promoted and widely Driven by demand, artificial intelligence technology has successfully crossed the “technical gap” between science and application, breaking through the technological inflection point of “cannot be used, not used” to “useable”, before entering the explosive growth period. .

In this regard, Fang Yongchun, dean of the Institute of Artificial Intelligence at Nankai University, is deeply impressed.

"When the network was started at the beginning of this century, everyone was worried that there would be no content on the network. Later, after years of informatization construction and development, a large amount of data was accumulated. After the big data technology, artificial intelligence technology was needed to extract useful information. Information serves humanity." Fang Yongchun said in an interview with the Journal of the Chinese Academy of Sciences that from the networking to the informationization to the big data and artificial intelligence, this development has a distinctive feature, that is, the application promotes technological development, technology. Development and application are intertwined and promote each other.

At present, the world's scientific and technological development trends and national economic construction have put forward very strong and tangible needs for intelligent talents, but the reality is not optimistic. Goldman Sachs’ “Global Artificial Intelligence Industry Distribution” report shows that in 2017, China’s emerging artificial intelligence projects accounted for 51% of China’s total, and the number has surpassed that of the United States. However, the global artificial intelligence talent reserve is only about 5%, and the talent gap exceeds 500. Ten thousand people.

"The biggest bottleneck in the development of artificial intelligence industry is talent. Now it has entered the era of global competition for artificial intelligence talents. The 'hematopoietic function' of high-level talent training will directly affect the core competitiveness of the artificial intelligence industry." Institute of Artificial Intelligence, Nanjing University Chang Zhou Zhihua said in an interview with the media.

Therefore, the "New Generation Artificial Intelligence Development Plan" particularly emphasizes that "the construction of high-end talent team is the most important part of the development of artificial intelligence", "improving the discipline layout of artificial intelligence field" and "establishing an artificial intelligence college in pilot schools as soon as possible". In April this year, the Ministry of Education issued the "Innovative Action Plan for Artificial Intelligence in Colleges and Universities", which also requires that by 2030, colleges and universities should become the core force for building the world's major artificial intelligence innovation centers and the talented highlands that lead the development of a new generation of artificial intelligence. It is among the forefront of innovative countries to provide technology support and talent protection.

It can be said that whether it is the trend of world science and technology development or the needs of national economic construction, it provides a very good environment for the establishment and construction of artificial intelligence colleges or majors in colleges and universities.

Course construction is especially important

At present, the situation of artificial intelligence colleges or majors established by many universities in China is slightly different: some recruit undergraduates, some are only scientific research and postgraduate training; some schools focus on artificial intelligence methods, and some schools have more Focus on intelligent systems that are tightly integrated with applications.

"This situation is in full compliance with the characteristics and development laws of artificial intelligence." In Fang Yongchun's view, artificial intelligence itself is a highly interdisciplinary subject. It is very natural for schools to choose different entry points for construction. On the other hand, The rapid development of a new generation of artificial intelligence technology and the construction of artificial intelligence colleges in various universities are recent events. The situation in all aspects is not yet mature and is in a period of rapid development. Therefore, it is also bound to have no problems in curriculum setting and teaching planning. Less space.

For example, Nankai University's Institute of Artificial Intelligence is based on intelligent science and technology. The major was established in 2004. After nearly two years of careful preparation, the focus was on meticulous construction in terms of curriculum and teaching practice. Only in 2006, the first undergraduate students were recruited. “Although after more than ten years of development, it has been relatively stable, but even so, in the past few years, Nankai has been repetitively optimizing curriculum construction based on the development of artificial intelligence disciplines. Curriculum development is particularly important if artificial intelligence colleges are to develop well. ."

For this point of view, Zhou Zhihua coincides with him.

“Currently, the number of courses that can be dedicated to artificial intelligence is far from meeting the demand, which leads to the concentrating of artificial intelligence professional courses to the level of ‘high-level science.’” Zhou Zhihua said.

Taking computer science and technology as an example, undergraduate graduates need to complete 150 credits, of which about 60 credits are general education courses, 15 credits are graduation design and employment and entrepreneurship courses, in the remaining 75 credits, subjects Platform courses and professional core courses account for about 55 credits. At present, there are only about 20 credits left, but there is no special course for artificial intelligence. These credits still need to consider the "wide-caliber" talent training in computer science, and balance the needs of multiple professional directions.

“In fact, even if you don’t consider the number of courses, you only have a lot of distance from the content of the courses you have already set up.” Zhou Zhihua said, such as the linear algebra + matrix theory required for artificial intelligence, The linear algebra content of computer science is very shallow. Usually, matrix theory is not established. Many students have not even touched matrix derivation, which poses a great obstacle to the learning of artificial intelligence core courses such as machine learning.

Therefore, in his view, instead of tinkering under the existing discipline training system, it is better to build a new curriculum system based on the characteristics of the artificial intelligence discipline itself.

Clarifying the connotation is the key

Speaking of course construction, Wu Fei, director of the Institute of Artificial Intelligence of Zhejiang University, said that in fact, the core courses required for artificial intelligence are now in colleges and universities, but they are scattered in different disciplines. "The key is to clarify the meaning of artificial intelligence." On this basis, determine which are compulsory courses and which are elective courses."

For the connotation of artificial intelligence, it is possible to refer to the curriculum set up by the undergraduate major of artificial intelligence that Carnegie Mellon University will open this fall.

“Their compulsory courses include matrix calculations, optimization analysis and integration and approximation in mathematics and statistics courses, computer programming, computer systems, data structures and algorithms in computer science courses, and in artificial intelligence courses. Machine learning, knowledge representation and problem solving, and natural language/computer vision, etc.” Wu Fei introduced that the school divided the required courses of artificial intelligence into 16 categories, involving mathematics, computers and artificial intelligence itself. "Artificial intelligence has certain discipline independence, but it also intersects with other disciplines, such as computer science and control disciplines."

In fact, according to the 2012 Undergraduate Professional Catalogue of the Ministry of Education, there are four of the 506 undergraduate majors involved in “intelligence”, namely, intelligent science and technology in computers, and building electrical and civil engineering. Smart grid information engineering and electrical engineering and intelligent control in intelligent and electrical categories.

However, whether it is connotation or extension, there is a certain difference between these professions and artificial intelligence.

“The relationship between 'artificial intelligence' and 'smart' is similar to the relationship between 'aircraft (artificial bird)' and 'bird'. In this regard, Zhou Zhihua said that the research aircraft is obviously different from the study of bird science. Bird science itself is important, but it is not the scientific knowledge that must be mastered by the aircraft manufacturer. The lack of clarity on the bird does not hinder the construction of the aircraft. The flight of the aircraft does not have to be the same as that of the bird.

In Wu Fei's view, colleges and universities offer artificial intelligence colleges to train three types of talents, namely, frontier researchers who master the frontier artificial intelligence theory, implementers of artificial intelligence technology, and AI+X users who intersect with artificial intelligence and industry, respectively. Frontier theory crossover, technology crossover and industry crossover. The first two types of talents need to be cultivated by the Artificial Intelligence Institute. The third type of talents need to be cultivated by the Artificial Intelligence Institute and other colleges.

“First of all, we must clarify the difference between the artificial intelligence college and the existing colleges, and clarify the difference between the entrance and exit of the artificial intelligence college students and other college students.” Wu Fei said, otherwise, if it is related to computer science or intelligent science and technology If the content and training objectives of the students are the same, the artificial intelligence college will not have a separate existence, and it may also form a development bubble.

Don't blindly follow the trend

How to avoid the artificial intelligence college to form a development bubble? In Fang Yongchun's view, colleges and universities should first identify their own characteristics and avoid blind follow-up.

“To solve the practical problems in the typical industry, this is a development trend in the field of artificial intelligence. Therefore, when building artificial intelligence colleges, colleges and universities must consider this development trend and use application-driven construction of artificial intelligence.” Fang Yongchun said Artificial intelligence is a cross-disciplinary subject involving many fields. Therefore, in order to build an artificial intelligence major, universities must fully consider their own characteristics, foundations and advantages, and combine these advantages with the development trend of artificial intelligence disciplines. In order to form their own characteristics.

At the beginning, it was the same as the wind engineering professional, but the development of the ordinary car, such as software engineering. In 2000, the State Council issued the “Several Policies to Encourage the Development of Software Industry and Integrated Circuit Industry”; the following year, 35 demonstration software colleges were established nationwide; by 2016, the number of software engineering majors nationwide reached 563. It is undeniable that the software engineering profession has trained a large number of high-quality talents for the country and the society, and promoted the development of the industry. At the same time, however, many colleges and universities are low-level repeat settings, lacking features and quality guarantees. Therefore, the “2016 Dynamic Adjustment of Cancelled and Added List of Degree Authorization Points” published by the Academic Degrees Committee of the State Council shows that 175 colleges and universities have cancelled 576 degree points, of which 35 software engineering degree points have been revoked.

For the current artificial intelligence boom in colleges and universities, Xiong Bingqi, vice president of the 21st Century Education Research Institute, bluntly said that it is necessary to prevent colleges and universities from chasing hot spots and rushing to open artificial intelligence colleges or majors, resulting in excessive professional offerings, lack of talent training, and quality of personnel training. It is difficult to meet the needs of society.

“When colleges and universities offer related majors, they should organize professor committees and academic committees. In combination with the national talent policy, the demand for talents in industrial development, and the training of similar professionals in other universities at home and abroad, there is no reality on whether to open a major in this field. Conditions, how to carry out teacher construction, curriculum construction, the formation of professional features, etc., to fully demonstrate." Xiong Bingqi stressed.

In this regard, Wu Fei also expressed similar views.

Not long ago, Lei Chaozi, director of the Science and Technology Department of the Ministry of Education, said at a press conference that the Ministry of Education will further demonstrate the connotation of artificial intelligence and promote the construction of the first-level discipline in the field of artificial intelligence. "Before the formal organization of artificial intelligence disciplines at the national level, universities should still base their original foundations, combine their own actual conditions and existing characteristics, and train artificial intelligence talents in a down-to-earth manner." Wu Fei said.