A New Era of Overlearning?
How AI *could* enable anyone to master anything.
During the 25 years or so that I’ve been exploring how humans learn, one of the concepts I’ve been most interested in is “Overlearning”.
Whether we know it or not, for people who design and deliver learning experiences of any kind, Overlearning is the north star which we all aspire to achieve. It refers to a state in which a learner continues to study or practice well past the point of initial understanding to full mastery of concepts and skills.
Overlearning is closely associated with the concept of Flow State, which was popularised by Hungarian-American psychologist Mihaly Csikszentmihalyi. Csikszentmihalyi describes flow as, “A state of complete immersion and focus in a particular activity”. It’s characterized by:
Challenge-Skill Balance: The task at hand is challenging yet matches the individual's skill level.
Clear Goals and Immediate Feedback: Individuals have clear objectives and receive immediate feedback on their performance.
Concentration and Focus: There's a high level of concentration on the task.
Loss of Self-Consciousness and Time Distortion: Individuals may lose awareness of themselves and time may seem to pass faster.
A third concept associated with both Overlearning and Flow State is Automaticity.
Automaticity refers to a human’s ability to perform tasks without occupying the mind with the low-level details required, allowing it to become an automatic response pattern or habit. It’s usually the result of ongoing repetition, retrieval and practice and usually results in knowledge and//or skills being retrieved and performed almost unconsciously. Automaticity, often referred to Mastery, helps to drive learning and growth by freeing up up cognitive resources for other, more complex or higher-order, tasks.
Each of these concepts significantly contributes to the efficiency, effectiveness and enjoyment of the learning process. When combined effectively, they can create a powerful, engaging and intrinsically-motivating learning experience.
TLDR: Through Overlearning, learners can achieve automaticity in certain knowledge and skills, and by designing challenges that match the learners' skill levels, a flow state can be induced, optimising focus, creativity, and overall levels of knowledge and skill acquisition learning.
Why Have We Under-Delivered on Overlearning?
So the big question, of course, is why don’t we do a better job of designing and delivering for Overlearning? The research suggests there are two causal factors:
Lack of Individualized Instruction:learning experiences are often designed to cater to a broad audience, which limits the challenge-skill balance necessary for overlearning and flow. The existing one-size-fits-many model doesn’t not provide the right level of challenge for all learners, making it hard for most learner achieve a state of overlearning.
Lack of Immediate, Individualised & Actionable Feedback: In traditional classroom & online setting, immediate , individualised and actionable feedback is impossible to deliver at scale. Feedback is typically delayed, generic and in some cases entirely absent, making it impossible for learners to attain flow state.
AI: A New Era of Overlearning?
One question I’ve been exploring this week is: can AI help us to do a better job of designing and delivering experiences that are optimised for Overlearning?
Here are my findings so far on how AI can be a powerful tool in the context of learning design and delivery:
AI for Learner Profiling: AI’s ability to aggregate and analyse data can help us to profile our learners, define their ZPD and differentiate their experiences in a way that optimises for “flow state”. DuoLingo already do a pretty good job of this.
AI for Dynamic Learning Path Generation: Based off AI-generated learner profiles, Generative AI could be used to create more personalised learning journeys and adapting those experiences on the fly to maintain the challenge-skill balance necessary for Overlearning and flow.
AI for Feedback: Generative AI could also be used to automate assessments and provide real-time responsive feedback, helping learners to stay in the flow and continue their path towards Overlearning.
The Barrier to Overlearning: Humans
There’s no doubt that AI could help us to crack one of the most wicked problems of education and empower us to design and deliver learning experiences which are optimised to drive the intrinsic motivation and achievement of mastery of every learner. The bigger and more challenging question is: how do we implement it?
Designing and delivering effective education requires a multifaceted approach that goes beyond mere technological advancements. TLDR: The technology is ready, but are we?
For any widespread progress to be made, critical questions and considerations need to be addressed. These include:
Standardisation: Establishing standards for AI in education can help ensure that technologies are being used effectively and ethically. It’s vital to have a regulatory framework that guides the use of AI in educational settings.
Privacy & Ethics: With the use of AI, concerns about data privacy and ethical considerations invariably arise. Policies need to be formulated to ensure the protection of students' privacy and data.
Technological Infrastructure: Adequate technological infrastructure is fundamental to support AI-driven education for all learners. This includes reliable internet access, hardware, and software.
Investment & Interdisciplinary Collaboration: Significant investment is required not only in technology but also in training educators and developing high-quality, AI-powered learning technologies. Collaboration between educators, technologists, psychologists, and policymakers is vital to ensure that AI in education is being developed and used in a way that truly benefits learners.
As we explore what AI can offer the world of education, the key question must transition from a technological 'can we?' to a social, cultural, economic and political 'how do we?'.
Happy innovating!
Phil 👋
PS: If you design learning experiences and want to get hands on and experiment with AI supported by me, you can apply for a place on an upcoming cohort of my AI-Powered Learning Science Bootcamp here.
PPS: If you work in a large company ahead of the AI curve, you can apply to take part in my & Gianluca Mauro’s AI research (or just sign up to learn more about it) here.