Recently, the 11th International Conference on Distance Education and Learning (ICDEL 2026) was held in Hangzhou. During the conference, faculty and students from the School of Education at Shanghai Jiao Tong University (SJTU) focused on cutting-edge research topics, including educational innovation empowered by the synergy of "AI + HI" (Artificial Intelligence + Human Intelligence) and the design of Large Language Model (LLM)-driven multi-agent educational games. They delivered a series of academic reports and held in-depth discussions, fully showcasing the educational research practices and international influence of the SJTU School of Education in the Artificial Intelligence era.
CHIANG Feng-Kuang, Vice Dean of the School of Education and Director of the Center for Future Education, was invited to deliver a keynote speech titled "Reshaping Education for the AI Era: Empowering Educational Innovation through AI + HI" at the conference.
In his talk, CHIANG Feng-Kuang outlined SJTU’s three-year (2024–2026) strategic framework for AI+HI enabled teaching, alongside more than 100 AI-infused innovative instructional practices developed by the university’s Center for Teaching and Learning Development. He unpacked five core paradigms for embedding AI into classroom practice: AI as a learning assistant, AI as a disciplinary research tool, and AI for professional competency cultivation, among others. Drawing on medical education case studies, he systematically analyzed the instructional impacts of generative AI prompts and applied Epistemic Network Analysis (ENA) to unpack GenAI’s role in supporting self-regulated learning. He also showcased SJTU’s in-house multi-agent ecosystem — LearnMaster, TeachMaster and SciMaster — a multi-dimensional collaborative platform built to advance personalized student learning, intelligent teacher instruction and AI-assisted research.
LIN Xihuang, Master of Education Student, 2025 Cohort, delivered a conference report titled "Zero Carbon Tycoon: Design and Development of a LLM-Driven Multi-Agent Educational Game For Fostering Carbon Neutrality Literacy".
This research innovatively utilized LLMs to drive four types of AI agents—government, enterprises, green investors, and international negotiators—to build an educational game named Zero Carbon Tycoon based on a dynamic carbon market simulation environment. Evaluation analysis grounded in Self-Determination Theory (SDT) revealed that the multi-agent design and chain-of-thought (CoT) visualization granted Zero Carbon Tycoon a high degree of operational freedom. The game not only helps learners sort out complicated causal links between carbon emissions and economic investments, but also greatly satisfies learners’ psychological demands for autonomy and competence. It offers an original empirical design solution to tackle the common defects of existing climate change educational games, namely oversimplified plotlines or excessive cognitive burden for players.
ICDEL is a flagship international academic conference in the global field of distance education and learning. Hosted by Hangzhou Normal University, the conference gathers experts, scholars and industry practitioners worldwide to collaboratively explore the innovation and reconstruction of educational models, teacher roles and learning mechanisms in the digital age.
