Most learning technology starts from the content and asks: how do we deliver this efficiently? Wayfinder starts from the opposite end. It asks: who is this person, how do they think, what are they ready for, and what kind of conversation will help them grow?
People approach learning in fundamentally different ways. Some work step by step. Others scan for patterns. When the teaching approach mismatches how someone thinks, learning doesn't just slow down — it fails.
Recognising the right answer on a test is not the same as being able to derive, explain, and apply it independently. Most systems assess what people can remember. Wayfinder assesses what they can do with what they know.
Current adaptive systems adjust the difficulty or sequence of content. They don't adapt to how the learner thinks, what developmental stage they've reached, or what kind of conversation will help them grow. The adaptation is shallow.
A vehicle for driving through knowledge
Wayfinder is not a speculative approach to AI-assisted learning. It is the practical implementation of three complementary frameworks, each developed over decades of independent research. The theory is complete. The technology to implement it has only recently become possible.
The mechanism by which understanding is constructed through dialogue, derivation, and teachback. Not content delivery — genuine conversation. Not testing — reconstruction.
The structural stages through which a person's capacity to handle complexity grows. Four orders of information complexity. The developmental map that tells us what a learner is ready for.
How to see where a person currently stands without reducing them to a test score. Appreciation, not evaluation. Tasking, trusting, tending — the three obligations of any system that supports growth.
Domains are built as entailment meshes by the Domain Authoring Workbench — cyclic knowledge structures where every idea can be reached from every other. Wayfinder then meets each learner within that structure, adapting its conversation to how they think and what they're ready for.
The domain is structured as a cyclic entailment mesh. Multiple entry points, multiple paths, multiple perspectives. The learner chooses the destination and the route.
Through the pattern of exploration and engagement with domain-adapted phrase cards, the system detects learning strategy and current processing mode. No tests. Appreciation, not evaluation.
Step-by-step derivation for those who work sequentially. Overviews and analogies for pattern-seekers. Complexity calibrated to the learner's developmental stage. Not one path for everyone.
Can the learner explain, derive, and teach the knowledge back? That is understanding. Everything else is recall. The learner does not feel tested. They feel engaged.
A visualisation of the entailment mesh with their progress overlaid. What they've understood, where they're heading, where the domain extends beyond what they've explored. The learner becomes their own teacher.
Which topics are understood across your organisation, where learners struggle, strategy profiles, development trajectories. Insight into how your people learn, not just what they've been shown.
Clinical reasoning, diagnostic knowledge
Regulatory knowledge, risk assessment
Systems knowledge, expert capture
Leadership, strategic thinking
Ocean literacy, science education
If your organisation's success depends on people genuinely understanding complex knowledge — not just being exposed to it — we'd like to hear from you.
peter.tuddenham@coexplorer.com