A Pedagogy of Effort: Learning as AI Capability Grows

Effort has long been understood in education as constitutive of learning: the struggle is where much of the learning happens. As AI capability improves rapidly, effort that was once unavoidable can now be declined, and this raises questions about how much effort learning should demand, where that effort is best directed, and whose effort is involved.

Five concerns stand out as we examine the pedagogical implications of AI. First, cognitive offloading: all learning involves deciding what to think through and what to hand over, so which parts of a task should stay effortful, and how do we design learning situations that make that effort worth undertaking? Second, the AI supply chain: how can educators and learners engage with AI critically in ways that make visible — and respond to — the socio-material and environmental conditions of these systems? Third, self-efficacy: how do we protect a student's confidence in their own thinking when the tools appear to outperform them at almost everything? Fourth, intrinsic motivation: where do students locate the value and purpose of human learning and personal effort when AI can seemingly "do everything"? Fifth, memory and retrieval: when information is instantly retrievable, what remains worth committing to memory — and what does a person need to already know in order to use AI well?

Taken together, the three questions about effort connect concerns that are usually addressed separately: whose effort, in particular, links the pedagogical to the socio-material. This talk will consider these concerns in relation to our developing understanding of the interrelationships between AI literacy, pedagogical decision-making and pedagogical content knowledge. The talk will draw on research and discussions with several collaborators and communities including EDUSUMMIT, whose 2025 focus was AI literacy and pedagogy; Informatics for All, a European coalition arguing that all students should study informatics, which matters more rather than less as AI capability grows; and the International Federation of Information Processing (IFIP) Working Group 3.9 on AI in Education.

Meet the speaker

Mary Webb is Professor of AI in Education at King's College London and is internationally recognised for her research on pedagogy across both Computer Science as a subject and the uses of new technologies for learning, resulting in over 190 publications. Her AI in education research began in the 1990s with her PhD on children building expert systems. More recently her research has examined computing curricula, AI literacy, programming pedagogy, digital technologies in science education, immersive learning environments, haptics, chatbots, and formative assessment.

Mary is chair of the International Federation of Information Processing (IFIP) WORKING GROUP on AI in Education and led the working group on AI Literacy and Pedagogy at EDUsummit 2025. She is on the steering committee of Informatics for All (I4ALL) in Europe and is passionate about the importance of students learning Informatics/computer science. Mary teaches on Computing PGCE and masters programmes, supervises PhD students, and previously taught science and computing in primary and secondary schools.