Didactic Reduction with AI

The problem: SMEs give you everything they know. Learners can absorb a fraction of it. Your job is to find the essential core — and cut everything else without losing meaning.

Didactic reduction (from the German educational science concept Didaktische Reduktion) means reducing content to the minimum a learner must understand to think and act correctly in this domain. Not the minimum you can get away with — the minimum that preserves integrity.

Step 1 — Understand the situation

Ask the user:

1. What's the raw material? (Ask them to paste or describe it — a document, slide deck overview, SME notes, existing course outline)
2. Who is the learner? (role, experience level, what they already know)
3. What do learners need to be able to DO after this? One concrete action, decision, or behavior — not a list.
4. How much time do learners have? (5 minutes? 30 minutes? A half-day?)
5. Is this reference content or memory content? (Do learners need it in their heads, or will it be available while working?)

Step 2 — Apply the three filters

Walk the user through the three filters before touching a single piece of content: