Research School Network: An introduction to Predictive Processing Theory Adam Wray, Associate Leader at ACE Tiverton School, shares his interest in Predictive Processing Theory

Initially, Adam drew on the familiar tools of mainstream educational cognitive science: working memory models, cognitive load theory, dual coding, and the growing body of research around executive functioning and episodic memory. Yet these frameworks still failed to fully explain the patterns he observed in his autistic and ADHD students. It was only when he encountered Peter Vermeulen’s Autism and the Predictive Brain that things shifted. The book provided a clear, accessible introduction to Predictive Processing Theory — a model of cognition that finally resonated with Adam’s real-world experience of neurodiverse learning. What it lacked, however, was a detailed exploration of how these ideas translate into education. That gap became the focus of Adam’s work.

An active prediction engine

At the centre of Predictive Processing is the idea that the brain is not a passive receiver of information but an active prediction engine. Rather than waiting for sensory input to arrive before interpreting it, the brain continually generates models of what it expects to happen next. Incoming sensory data is compared to these predictions, and any mismatch — known as ​“prediction error” — drives learning and attention. Perception, therefore, becomes a process of inference: we see what we expect to see until the world forces us to revise our expectations. This is a dramatic shift away from traditional information-processing views, which assume that cognition begins with sensory input and moves inward. As cognitive scientist Andy Clark famously describes it, our experience of the world is a ​“controlled hallucination” — the brain’s best guess shaped and reshaped by prediction.

Acting to minimise prediction error

Active Inference builds on this theory by proposing that humans do not merely perceive in order to minimise prediction error — we act to minimise it as well. Movements, decisions, and even emotional expressions are attempts to bring the world into alignment with what we expect or to explore the environment in ways that refine our internal models. Grounded in Karl Friston’s mathematical frameworks, Active Inference positions prediction as the organising force behind behaviour, regulation and learning. This makes it one of the most integrative theories of mind currently available.

Recognising that these ideas were almost entirely absent from mainstream education discourse, Adam created his Substack, Predictably Correct. Despite being an avid follower of research, podcasts and professional discussion in cognitive science for education, he was struck by the silence surrounding predictive processing. If this was the most contemporary scientific understanding of how the brain works, why wasn’t it shaping classroom practice? His blog aims to bridge the gap between the theoretical world of neuroscience and the practical world of schools and CPD. Through conference presentations, writing, and community-building, he hopes to bring teachers into the conversation and explore how predictive processing can transform teaching and learning.

Reactions to his work have been enthusiastic, if sometimes overwhelmed. Many educators find the concepts initially unfamiliar or complex, but soon recognise that the underlying ideas are remarkably intuitive. The challenge is not the theory itself but the shift in perspective it demands. Advocates of cognitive load theory and related models have shown interest — Karl Hendrick has already described predictive processing as ​“the next frontier in learning theory” — while others anticipate, in Becky Allen’s words, a ​“messy revolution” as long-held assumptions begin to move. For Adam, the transition is unavoidable; the only question is how quickly the education sector embraces it.

Where to start?

Engaging with predictive processing, he argues, begins with understanding the basics: how prediction, priors, and prediction error shape perception and learning. From there, teachers can start reframing practice through this new lens — thinking less about working memory and load, and more about controlling predictability, designing for desirable surprise, and managing the rate of prediction error learners experience. He encourages educators to share their experiences, contribute to emerging discussions, and help shape the developing field. He also hopes research organisations, teacher-training providers, and MAT leaders will begin incorporating predictive processing into CPD and initial teacher education, ensuring that new teachers understand the evolution of cognitive science beyond the traditional models of Baddeley and Hitch.

Looking ahead, Adam is turning his attention to an area he sees as the missing link: the precise neuroscience of how the brain updates its generative model — how learning changes neural wiring so that predictions improve over time. This underpins key educational concepts such as forgetting, spacing and retrieval, and offers promising implications for curriculum sequencing and timetable design. Alongside this work, he is collaborating with Nathan Burns (@Mrmetacogntiion) to develop an explicitly taught, TA-facilitated ​“Thinking Skills” metacognition programme.

For teachers encountering Predictive Processing for the first time, Adam offers reassurance. Despite the technical language, at its core the theory is beautifully simple: every brain strives to make better predictions. Everything educators do — from routines and modelling to feedback and relationships — supports that predictive process. Once seen through this lens, the classroom becomes a radically different space.

To explore these ideas further, you can visit Adam’s Substack, Predictably Correct, where he continues to unpack topics such as ​“Move over CLT: Predictive Processing Theory is here.”

Additional reading

Clark, A. (2023) The Experience Machine: How Our Minds Predict and Shape Reality. New York: Pantheon Books.

Vermeulen, P. (2022) Autism and the Predictive Brain: Absolute Thinking in a Relative World. London: Routledge, Taylor & Francis Ltd. ISBN: 9781032358970.