Peking University, June 10, 2021: In April of 2021, President Xi Jinping delivered a speech at the 110th anniversary of Tsinghua University, highlighting the importance of unleashing the magnificent potential that lies in every student, "To build a first-class university, the key is to encourage and propel students to strive for the best. The direction of personal development should gravitate towards the urgent needs of our country."
With the recent rapid development of AI, namely artificial intelligence, China has been cultivating and gathering AI professionals to develop further technological advancements for the country. In order to achieve these goals, Professor Zhu Song-Chun, dean of the Institute for Artificial Intelligence at PKU and Chair Professor at Tsinghua University, is now promoting the undergraduate program focusing on Artificial Intelligence across PKU and THU, which offers a unique curriculum unlike any other in this country. To explore this topic further, PKU News talked with Prof. Zhu regarding his views on AI.
General Unified AI: New Height in Technology
"AI" the term was first coined in 1956, describing a novel type of technology which can simulate human functioning and reasoning. With the concept of AI presented to the global stage, it gathered much intrigue and excitement from the public, especially when AI signaled an impending future with endless potential and maximized efficiency. Nonetheless, a sudden turn of events quickly cooled down the rising optimism, when the "two dark clouds" was unveiled during AI development.
"Two dark clouds" was a metaphor describing the two critical issues that existed in the formation of AI. The first dark cloud, also called "Symbol Grounding", was pinpointed the five senses - sight, hearing, touch, smell, taste - and converted signals into symbols which could be interpreted by logic reasoning techniques; the second cloud was called "Commonsense Reasoning", which included human insights, thought processes, values and intentions that we possess everyday when engaging in activities and decision-makings. Both of these two components defined the core of our human experience. By encrypting alone, it was impossible to convert the very essence of the human experience into systematic computer coding and apply them accurately to real-life applications.
Soon after the "two dark clouds" was brought up, the development of AI quickly reached its halt and entered a depression phase. During the depression, the research of AI was divided into six sub fields: computer vision, natural language processing, cognitive reasoning, robotics, machine learning and multi-agent systems. "The real intelligent agent must contain all six sub fields, all of which act as an integral component of artificial intelligence," said by Prof. Zhu, "It was the main-stream trend since 2015. Nowadays general artificial intelligence has become the most advanced form of scientific development, incorporating cutting-edge technology." Prof. Zhu, compelled to contribute to the cause, established the Institute for Artificial Intelligence which is now supported by PKU, with the hopes of creating an undergraduate AI program that can cultivate more talents in the field.
Concepts：General Liberal Art Knowledge, Intelligence and Applications
"At present, the AI related undergraduate programs do not match the exact social needs in China," said Prof. Zhu, "Taking computer science programs as an example, generally speaking, the portion of AI courses in these programs take up less than 10% due to the unique disciplinary characteristics that define the subject."
It became clear to Prof. Zhu that redesigning the existing AI courses was much needed. When it came to summarizing the main principles of the AI program, Prof. Zhu pointed out three important factors: general liberal art knowledge, general intelligence and general applications. General intelligence encompassed six sub fields of AI and Prof. Zhu incorporated them into the course design. General knowledge denoted the integration, overlapping and unification between AI and humanities including music, art, ethic, social governance, etc. General utilization referred to the applications of AI in numerous fields such as medicine, education and finance. Although the newly established program emphasized the importance of generalizing the usage of AI in multiple fields, Prof. Zhu made it clear that, "Besides the general curriculum, students can enroll in more in-depth professional courses that are offered in graduate school as well as entering labs whenever they prefer, with the choice of working on research projects if they wish to start early. The ultimate goal for this program is to offer a clear path for students to become AI professionals."
Practice：Systematization of Professional Training
Until now, admissions to the undergraduate program of AGI was through PKU's Yuanpei College. In terms of the qualities that are looked for in the program's candidates, Prof. Zhu said, "They have to express a deep interest in AI and possess the courage to face major challenges in research." As part of its teaching, the program will adhere to the systematic teachings of AI and apply them to real-life scenarios. Prof. Zhu cherished the hope that graduates from the program would create AI innovations applicable for the Chinese society, and pay tribute to the knowledge they have learned while they continue on the path towards success in the field of AI. Prof. Zhu noted, "We will introduce talents from aboard continuously and build a new platform for AI research. I believe that more excellent students will be encouraged to stay in China and make their own contributions when world-level experts of AI are around them."
Cooperation：Mutual Strengths between PKU and THU
The program was launched in PKU and THU at the same time. Facing the realistic needs of the country, Prof. Zhu realized that bilateral cooperation between the two schools, which combined the essence of both PKU's science and humanities and THU's engineering, was vital and perfectly suited for the creation of the first program dedicated to AI studies. "The curriculum is similar in the two universities," said by Prof. Zhu, "And the qualifications for degree completion are underway. Our teaching staff members are pioneers in AI research as well. With most taken care of, we are now left with supporting work that is still in the process of completion."
Future：Deep integration of civilization and AI
Prof. Zhu said, "We cultivate talents in order to create a better life. This is our ultimate goal." As AI advancements propel forward, the significant impacts AI would bring to the society as a whole and our own lifestyles would be substantial, therefore universities should be prepared to meet the challenges of an upcoming technological revolution. "The goal for this undergraduate program is to offer a rich and vibrant playground for students to engage their interests in AI", said Prof. Zhu. Regarding the integration of PKU's humanities with AI, Prof. Zhu noted that "We may act as the Confucius of robots. Traditional scriptures, representing our culture and language, can form as pieces of an intricate puzzle, with the other half being the advanced mathematical models in AI. Together they form as one, creating a new interpretation of humanity." With the propelling advancements in Information Technology, the deep integration of human civilization and AI can work together to provide a better future for us all.
Zhu Song-chun
Professor Zhu Song-Chun is dean of the PKU Institute of Artificial Intelligence and Chair Professor at both Peking University and Tsinghua University. In 2020, Prof. Zhu returned to China from the United States to join Peking University and Tsinghua University in advancing artificial intelligence. As a world-renowned and leading computer vision expert, statistical and applied mathematician, and artificial intelligence expert, he has won many awards in recognition of his achievements, notably the Marr prize, which is the highest international award in the field of computer vision. He also served as the general chair of the International Conference on Computer Vision and Pattern Recognition and the director of the Center for Computer Vision, Cognition, Learning and Autonomy at the University of California, Los Angeles.
Reported by: Fu Jiaqi, June Tan Rui Min
Edited by: Rose Li
Photo credit to: Zhu Song-Chun, VCG,Song Mengli, Cai Xiangyu
