Fifteen working prototypes built on the same twelve sub-scores from three assessments. Type 1 produces a single number and works hard to explain what is inside it. Type 2 argues that the pieces do not cohere well enough to be one number, and shows what to publish instead. Type 3 takes a third position: sort the scores by how they are measured and how fast they change, which yields three buckets that behave differently by design. Every prototype is interactive — switch the learner in the top right to see how each model behaves on different profiles.
Shortlist — three options, cross-referenced against the AI Fluency Framework v0.2 →