Revisiting Laurillard’s Conversational Framework in the GenAI Era

Abstract

Diana Laurillard’s Conversational Framework has long provided a powerful model for designing technology-enhanced learning, emphasising dialogue, feedback, and the iterative exchange between conceptual understanding and practical application. Historically, building educational tools that supported these rich learning interactions required significant technical expertise and development resources. However, the emergence of Generative AI chatbots capable of generating, previewing, and sharing web application code, such as Anthropic Claude Artefacts as well as Open AI’s ChatGPT and Google Gemini’s Canvas functionality now allow educators and learning designers to describe an idea in natural language and receive working, interactive web applications in return. In this paper we explore how these tools enable the rapid creation of learning experiences that embody core principles of Laurillard’s Conversational framework. Drawing on real examples built using Anthropic Claude’s AI enabled Artefacts, we illustrate how AI-generated applications can support perspective-taking, problem-solving, feedback-driven learning, and structured reflection. From a Linear Regression explainer to a Multi-Perspective Elaboration Tool, these examples show how AI can now scaffold not just content delivery but interactive learning conversations. We argue that this shift lowers the barrier to educational innovation and empowers a broader community of practitioners to design learning tools. We conclude by discussing current limitations and outlining directions for future research and practice in AI-assisted learning design.