Exploring New Research Results in Science of Learning and Creating a New Teaching Paradigm of Human-AI Collaboration —Forum on “Science of Learning and Human-AI Collaboration” Held in Beijing
date：2025-09-01 15:30author：adminsource：Smart Learning Instituteviews：
On the afternoon of August 18, the Forum on “Science of Learning and Human-AI Collaboration” under the Global Smart Education Conference 2025 (GSE2025) was successfully held at Beijing Normal University. The forum brought together leading experts and scholars from psychology, AI, neuroscience, and other fields to jointly explore how to promote the digital transformation of education and build a future-oriented educational ecosystem featuring human-AI collaboration through the deep integration of technology and education.
Group photo of attendees
Opening Remarks Session: Discussions on the Future of Education by Experts from Home and Abroad
Prof. Mohamed Jemni, Director of ICT Department, Arab League Educational, Cultural and Scientific Organization (ALECSO), underscored in his speech the vital importance of gaining a deep understanding of the learning process and improving learning outcomes. He highlighted the need to strengthen the dialogue and cooperation among experts, scholars, policymakers, and relevant institutions. He also noted that human-AI collaboration has emerged as a key characteristic of the digital age, and Arab countries are advancing the integration of AI and education through various projects and initiatives to develop a more inclusive future education system.
Prof. Mohamed Jemni, Director of ICT Department, Arab League Educational, Cultural and Scientific Organization (ALECSO)
Prof. MA Ning, Dean of the School of Educational Technology, Faculty of Education, Beijing Normal University, China, pointed out that AI is profoundly reshaping education. She stressed the need to combine the science of learning and technology to explore a new paradigm of human-AI collaboration. She also underscored that AI should act as a “super assistant” to teachers, unleashing its potential in areas such as personalized tutoring, thereby enabling teachers to focus on their irreplaceable roles in providing emotional support and fostering creativity.
Prof. MA Ning, Dean of the School of Educational Technology, Faculty of Education, Beijing Normal University, China
Keynote Speech Session: Research on the Science of Learning with AI Empowerment
Mr. Gwang Chol Chang, Chief of Section of Education Policy, UNESCO Division for Policies and Lifelong Learning Systems, mentioned in his keynote speech that the world is still facing systemic educational crises, including 272 million out-of-school children. He stressed that technology should serve educational equity, and policy-making should be based on empirical data. He also noted that UNESCO's “Happy Schools” initiative advocates the integration of technology in multiple dimensions (including who, how, and where) to address educational inequality.
Mr. Gwang Chol Chang, Chief of Section of Education Policy, UNESCO Division for Policies and Lifelong Learning Systems
Prof. Fred Paas, Professor of Erasmus University Rotterdam, the Netherlands, and Editor-in-Chief of Educational Psychology Review, emphasized in his keynote speech that the design of smart education should be grounded in an understanding of the limitations of human cognition. Drawing on cognitive load theory, he discussed several effects and highlighted strategies such as “multimodal effects” (visual, auditory, etc.) and “embodied effects” (physical participation) to optimize cognitive resource allocation, reduce irrelevant information load, and enhance learning efficiency.
Prof. Fred Paas, Editor-in-Chief of Educational Psychology Review
Prof. LIU Sannyuya, Vice President of Central China Normal University, proposed in his speech that AI-driven educational research should target the fundamental proposition of “understanding learning.” He put forward three perspectives for approaching this issue: individual level (AI analysis of large-scale real-world learning data), social level (Agent simulation and interaction), and system level (the construction of a general intelligent tutoring architecture). This approach aims to break through the limitations of traditional small-sample research.
Prof. LIU Sannyuya, Vice President of Central China Normal University
H.H. Prince Fahad Faisal T. Jalowi Al Saud, Kingdom of Saudi Arabia, noted that AI has introduced innovative approaches to the sharing of knowledge and wisdom, and human-AI collaboration can help us better disseminate culture. He shared Saudi Arabia's exploration in smart classrooms and AI education tools, and emphasized the importance of international cooperation in driving educational innovation.
H.H. Prince Fahad Faisal T. Jalowi Al Saud, Kingdom of Saudi Arabia
Prof. Wayne Holmes, UNESCO Chair in AI Ethics & Governance and Professor of University College London (UCL), UK, critically examined the impact of AI on education. He emphasized the need to remain vigilant about its potential downsides, such as energy consumption and privacy risks. He cautioned that using AI tools without assessing their effectiveness is tantamount to conducting “experiments on children.” Therefore, he advocated for stronger regulation and ethical research to prevent technology from blindly replacing the indispensable role of teachers.
Prof. Wayne Holmes, UNESCO Chair in AI Ethics & Governance
Prof. DONG Yuqi from Shanghai Normal University, proposed that achieving “personalized learning at scale” is the key to building a leading education system. He presented empirical research to validate the feasibility of AI learning companions in supporting personalized and collaborative learning, and underscored the necessity for deeper research into the underlying aspects of cognition, emotions, and thinking. Meanwhile, he called for collaborative efforts among government, industry, universities, and research institutions to address bottlenecks in teacher training and system development.
Prof. DONG Yuqi from Shanghai Normal University
Panel Discussion Session: Promoting Human-AI Collaboration Through Interdisciplinary Collaboration
During the panel discussion, experts and scholars from various countries and fields engaged in discussions on the science of learning and educational practices.
Panelists
Prof. Nancy Law, Founding Director of the Centre for Information Technology in Education (CITE), University of Hong Kong, proposed that technology design needs to be based on the science of learning. With the support of AI technology, she and her team designed the “Integration and Learning Platform,” harnessing big data from learning interactions to better understand the human learning process.
Prof. Nancy Law, Founding Director of the Centre for Information Technology in Education (CITE), University of Hong Kong
Prof. John Shawe-Taylor from the Centre for Computational Statistics and Machine Learning, University College London (UCL), stressed the importance of the science of learning. He emphasized the need to explore fundamental educational issues like how people learn and create knowledge. These insights, he said, should guide our thinking on what kind of AI education we need.
Prof. John Shawe-Taylor from the Centre for Computational Statistics and Machine Learning, University College London (UCL)
Prof. Diana Laurillard, Emeritus Professor of University College London, argued that teachers need to quickly master AI teaching skills through collaborative efforts. She has been instrumental in developing large-scale online collaboration initiatives based on the MOOC platform for teachers worldwide, facilitating cooperation among teachers, experts, and scholars to seek solutions to commonly concerned issues.
Prof. Diana Laurillard, Emeritus Professor of University College London, UK
Prof. Roberto Lent, Emeritus Professor of the Federal University of Rio de Janeiro, Brazil, introduced that interaction mechanisms between two brains (one responsible for learning and the other for teaching) can be revealed through research on brainwaves. He pointed out that face-to-face classroom instruction, online face-to-face learning, and video-based lessons may yield different effects for different types of learners.
Prof. Roberto Lent, Emeritus Professor of the Federal University of Rio de Janeiro, Brazil
Ms. Sonia Guerriero, Education Specialist of UNESCO Division for Policies and Lifelong Learning Systems, mentioned leveraging the role of the Global Alliance on the Science of Learning for Education to foster collaboration among various stakeholders. This approach aims to strengthen global cooperation while maintaining academic expertise, transforming scientific research into feasible educational policies and practices.Release Session: Co-building of a Regional Community of Practice on Science of Learning in Asia-Pacific
Release of the “Regional Community of Practice on Science of Learning in Asia-Pacific” Initiative
This forum was jointly organized by the UNESCO Global Alliance on the Science of Learning for Education, UNESCO Division for Policies and Lifelong Learning Systems, UNESCO Institute for Information Technologies in Education (UNESCO IITE), UNESCO Chair on Artificial Intelligence in Education, Arab League Educational, Cultural and Scientific Organization (ALECSO), and School of Educational Technology, Faculty of Education, Beijing Normal University. Its co-moderators were Prof. Mohamed Jemni, Director of ICT Department, ALECSO, and Prof. Ahmed Tlili, Associate Professor of Beijing Normal University, and Deputy Editor-in-Chief of Smart Learning Environments.
This page by Smart Learning Institute of Beijing Normal University is licensed under a Creative Commons Attribution 4.0 International License.