Less Overload, More Learning: A Research-Informed Guide to AI in Classrooms

It has been a joy—and a genuine professional privilege—to work alongside giants in educational research to explore a question that’s rather timely: how do we introduce generative AI to students in a way that supports, rather than overwhelms, their learning?

In our recently published paper, we bring together the worlds of educational neuroscience and classroom practice. It’s somewhat ironic: the very tools designed to mimic neural networks can, if introduced poorly, overload the human ones we’re trying to nurture.

As a practicing high-school teacher and educational neuroscience researcher, I’m in the middle of this tempest right now—like many of you reading this! Whatever we think about generative AI, being a school teacher means we have a vital role to play. GenAI will be a big part of our students’ world and their future, so it must be part of ours too.

Our students are nervous, curious, despondent, and excited about generative AI. Most of mine want to learn how to use AI to improve their learning and—especially for tasks they find engaging—they’re keen not to outsource the learning journey. We often talk about using AI more like a pushbike than an Uber: amplify your own thinking to go further and faster, rather than bowing out and letting the machine rob you of an education.

Our recent research brings Cognitive Load Theory—a robust, evidence-based understanding of how our brain’s neural networks learn—to bear on the challenge of helping students master artificial ones.

Load Reduction Instruction (LRI) springs from Cognitive Load Theory. It’s an approach that works with how our brains learn, and we explain how LRI provides important guidance for schools and teachers seeking to implement generative AI in a learner-friendly way.

This EducationHQ summary of our research is a great starting point, and for greater depth - head to our source journal article here.

Head to the AI classroom for more :)

Roger Kennett

Thanks to the author team: Andrew J. Martin, Rebecca J. Collie, Roger Kennett, Danny Liu, Paul Ginns, Lala B. Sudimantara, Ema W. Dewi, Lilith G. Rüschenpöhler.