• With the rapid advancement of large language models (LLMs), AI-empowered educational agents are transforming traditional pedagogy by serving as intelligent collaborators for both teachers and learners. In this talk, I will share our recent studies exploring AI agents as multidimensional role in education: (1) as instructional co-designer, assisting teachers in developing optimized lesson plans; (2) in multi-agent classroom simulation, where teacher-student interactions create feedback loops for continuous lesson refinement; (3) as cognitive modeling agent that simulate authentic student learning process. Together, these studies demonstrate AI’s potential as multiple partners in education, driving innovation in both teaching practices and research methodologies.

  • Digital environments include platforms, online sources and content, as well as how people interact with each other online. Hence, digital environments are different for people from different parts of the world. Whether it is laws and regulations, communication styles, values or aesthetics, countries of the world have their own preferred digital environments. This presentation highlights the impact of these environments on students who travel across cultures and countries. It highlights concepts of digital journeys and transitions in light these diverse digital environments. In addition, information may be perceived differently across cultures and in the digital environment, information is accessed and used by people from diverse perspectives. Therefore, the presentation points to the need for increased research in how this diversity might impact on research into teaching and learning.

  • This study explores how Large Language Models can transform language education through three key techniques, including prompt engineering, retrieval-augmented generation, and parameter-efficient fine-tuning. Drawing on real-world cases in question generation, domain-specific chatbots, and automated essay evaluation, the talk will showcase how these methods enhance teaching efficiency, learner engagement, and assessment quality, while addressing challenges of data scarcity, transparency, and reliability.

  • Generative AI (GenAI) offers opportunities to rethink how we teach, assess, and support learning, but its integration into education raises profound questions that require careful reflection. This talk will synthesize the emerging evidence, drawing on recent empirical studies and technological developments, including adaptive feedback systems and AI-driven assessment tools. It examines how GenAI can foster learning and enable new forms of assessment while also confronting risks such as algorithmic opacity, learner overreliance, and metacognitive laziness. Building on this evidence, the talk outlines priorities for future research, policy, and practice to ensure AI serves as a catalyst for innovation while safeguarding human expertise and agency.