Feynman Technique Practice

The problem

Many learning sessions get stuck in “I think I get it, but I can’t explain it.” Concepts may be memorized, yet the explanation lacks structure. Follow-up questions can expose terminology stacking, logic jumps, and weak analogies. This skill turns the Feynman technique into an executable check: the user must first explain in plain language, then the AI asks curious layperson follow-up questions to locate gaps.

How it works

The skill is built around active output rather than passive AI lecturing. Core capabilities include:
- Layperson framing: for example a 10-year-old or cross-domain friend, with constraints on jargon
- Follow-up probing: identifying missing middle steps and vague phrases like “sort of”
- Memory vs. understanding: checking whether the user can explain causes, boundaries, and counterexamples
- Reference explanations: giving a concise comparison after iteration instead of replacing the user’s reasoning

The key flow is: confirm one concept, ask the user to explain it, run 2 to 3 targeted follow-up rounds, then evaluate what is clear and what needs more work.

Limits

It fits deep checks on previously learned material, not first exposure to a brand-new topic. The user must actively produce explanations, so it is not an answer generator. For pure factual recall, standard definition lookup, or validating many concepts at once, the benefit is weaker.