How do we learn in ways that truly last? Neuroscience and Psychology has spent decades uncovering the biological mechanisms behind memory consolidation, skill acquisition, and durable knowledge — and the answers are both surprising and deeply practical. In this talk, I draw on findings from cognitive and systems neuroscience to illuminate why some learning sticks while other knowledge fades, and what this means for how we design educational experiences.
From the role of spaced retrieval and interleaving to the neuroscience of prediction error and emotional salience, the brain offers a remarkably clear blueprint for effective pedagogy. But we are now at an inflection point: artificial intelligence tools are reshaping how students learn, how teachers teach, and how institutions think about educational technology. These tools hold genuine promise - enabling personalised learning pathways, adaptive feedback, and scalable support - yet they also carry risks that are not always visible at first glance.
When AI systems do cognitive work for the learner rather than with them, they may short-circuit the very struggles that drive durable learning. Difficulty, desirable or otherwise, is not a bug in the learning process - it is often the feature. This talk explores how educators and technologists can harness AI in ways that are neurologically grounded, pedagogically sound, and alert to the pitfalls of cognitive offloading, over-reliance, and the illusion of competence.
The goal is not to be cautious about AI, nor uncritically enthusiastic - but to be intelligent about it, in every sense of the word.