What can AI bring to education? At the World Artificial Intelligence Conference (WAIC), the development of education in the future has become a major focal point.
The 2023 WAIC Forum on Intelligent Education was convened at the World Expo Center on July 8th with the theme of “Leading the Learning Revolution and Creating the Future of Education”. Over 46,000 people participated in the forum both online and offline.
This forum was hosted jointly by ECNU and Shanghai AI Laboratory.
How can AI empower intelligent education?
“In face of the rapid iteration of new AI technologies, how can education respond more quickly and scientifically?” Mei Bing, Secretary of CPC ECNU Committee, pointed out in her speech that in the context of accelerated technological advancement, it is imperative to promote the construction and talent cultivation in intelligent education disciplines more scientifically and rapidly. This is crucial for ensuring the high-quality development of education.
Wang Hao, Deputy Director of Shanghai Municipal Education Commission, introduced Shanghai’s efforts in promoting AI to empower digital transformation in education and accelerating the construction of a modern education system in the era of intelligence. He stated that Shanghai would expedite the deployment of digital infrastructure in schools to create interconnected and intelligent educational facilities. The focus would be on expanding new applications of intelligent education and exploring personalized learning approaches based on AI. Additionally, efforts would be made to construct an intelligent education framework and accelerate the development of a digital platform for sharing educational resources.
Happy Sithole, representative of the BRICS countries and Head of the South African delegation, emphasized in his speech the significance of utilizing advanced AI to develop education while advancing AI technology. He also highlighted the necessity of strengthened collaboration in the development of intelligent education, particularly through enhanced exchange and cooperation between South Africa and China.
Where does the machine’s knowledge come from?
Zheng Qinghua, President of Tongji University, raised a profound scientific question in his report: ‘Where does the machine’s knowledge come from?’ He proposed that the two challenges confronted by the “big data knowledge engineering” are cracking the information explosion and eliminating data silos.
On this basis, Zheng Qinghua shared his team’s research on big data knowledge engineering and proposed the solution of “knowledge forest”. He summarized the concept of “knowledge forest” as the process of data knowledgeization, knowledge systematization, and knowledge inferability. He also introduced the typical applications of knowledge forest in intelligent education and national golden tax project etc.
He believed that the future of intelligent education, both large-scale models and knowledge forests, would be timely and promising.
Indispensable role of humans in intelligence
During the forum, Prof. Wang Huaimin, an academician of the Chinese Academy of Sciences and a professor of National University of Defense Technology, delivered a keynote speech on “Collective Intelligence and Intelligent Education”. He emphasized that humans are indispensable in intelligence, and technology is merely a tool. The enhancement of human educational capabilities stems from the improvement of means and levels in stimulating and aggregating collective intelligence. Large language models represent the latest technological achievement in stimulating and aggregating collective intelligence, and high-quality large language models acceptable to humans stem from the collective intelligence of human beings.
Wang Huaimin also expressed that large language models may serve as intelligent assistants for both excellent teachers and students, fostering a mutually beneficial learning environment. Large language models represent the latest technological means to support intelligent education, and intelligent education can provide high-quality human feedback and corrective data to further improve large language models.
To cultivate smarter people
Guo Yike, Provost of Hongkong University of Science and Technology and a Fellow of the Royal Academy of Engineering in the UK, pointed out that AI would exert a fundamental impact on education. Traditional education emphasizes knowledge accumulation, while in the AI era, AI serves as a tool for traditional education. The changing learning environment necessitates a reevaluation of fundamental educational questions, such as "What is knowledge?" "What is learning?" and "What is teaching?"
In the world empowered by AI, the three most important directions for teachers and learners are "human-centeredness", "critical thinking and creativity," and "the ability of communicating between humans and machines".
Guo Yike proposed that smart humans should use intelligent machines created through creativity to cultivate even smarter individuals, thereby creating even more intelligent machines.
Yuan Zhengguo, a lifelong professor of ECNU, presided over the keynote report section of the forum. In his concluding remarks, he emphasized that in the context of AI as the new generation of scientific and technological revolution, improving human-machine collaboration has become a crucial topic in the era of intelligence. Throughout, humans will always remain the masters of technology.
This forum also featured two roundtable discussions on “Large-scale Models and Educational Transformation” and “Cultivation of Normal University Candidates for the New Generation of AI”. During the discussions, experts and scholars further shared their understanding and insights on such topics as “Applicability of Large-scale Models in Education”, “Relationship between Large-scale Models and Teachers”, “Large-scale Models and Educational Transformation”, and “Cultivating Innovative Talents in the Era of AGI".
Release of Standards for Educational General AI Large-scale Models
On that day, Prof. Wu Yonghe and other guests from the Faculty of Education, ECNU jointly released the research outcomes regarding the Standards for Educational General AI Large-scale Models, including research reports and a set of standards.
The research report provides a detailed introduction to 24 general AI large-scale models both at home and abroad. On this basis, the report proposes the establishment of a standard system for Educational General AI Large-scale Models, which involves six stages including preliminary research, system planning, standard development, trial verification, training implementation, and application promotion. The aim is to construct trustworthy, secure, efficient, and user-friendly educational general AI large-scale models. This initiative seeks to create a human-centered AI that empowers and enhances education, improves the digital education standard system, promotes educational digitization, and facilitates educational innovation, transformation, and high-quality development.
This research report is scheduled to be officially published by Tsinghua University Press this August.
The series of Standards for Educational General AI Large-scale Models aims to regulate the overall framework, constituent elements, functional requirements, principles for constructing datasets, evaluation framework, security and privacy considerations, as well as teaching application requirements within the field of education.
These standards will be subsequently included in the development of industry standards and national standards, and the construction of a testing platform for Educational General AI Large-scale Models will also be initiated.
Please scan the code for review the live broadcast.
Copy editor: Zhang Yan, Philip Nash
Editor: Yvaine Lyu