Contract work
AI Education & Literacy
Interaction patterns and learning flows that build AI literacy, helping users interpret, question and calibrate their trust in AI output.
Role
Product Designer
Domain
AI literacy / educational UX
Focus
Trust calibration, interpretability, learning flows
01 The Setup
Context & Problem
AI now runs through everyday tools, yet most users cannot tell when a system is confident, guessing, or wrong. Without a shared vocabulary for uncertainty, people either trust AI blindly or reject it entirely.
Designing literacy patterns that meet users where they are, turning abstract model behaviour into legible, teachable moments inside real product flows, without turning the product into a tutorial.
02 Design Decisions
What We Tried & Why We Pivoted
Confidence as a readable signal
Replaced opaque AI outputs with tiered confidence states paired with plain-language explanations, so users learn to read what the model knows and what it does not.
Learning inside the flow
Embedded short, contextual explainers at the moment a user meets an AI decision, teaching trust calibration through the product itself rather than a separate onboarding.
A shared literacy vocabulary
Defined consistent terms for confidence, uncertainty, and override across surfaces, so the same mental model carries from one AI feature to the next.
03 Outcomes
Results
3
interaction patterns prototyped
Pattern library
delivered for product teams to adopt
04 Reflection
What I'd Do Differently
AI literacy is not a course you complete. It is a product teaching you, quietly, every time you use it.