Methods · Support & cognitive load
A26
Instant register shift
Demand the same explanation in several registers, expert, peer, child, hostile skeptic, and compare what changes, so you see which load-bearing pieces survive translation and which were vibes.
When to use it
- You can recite the definition, but you freeze when someone asks why.
- A seminar or tutorial is coming and you need your reasoning under pressure.
What to ask AI
Type something like: “Explain … four times, labelled:”.
Ask it to generate material for you to attempt, problems, cues, or a schedule, not finished answers.
Keep the hard part on your side: you retrieve, attempt, explain, or decide before you ask for help.
How it helps
A single explanation can sound clear because of tone, not because of content.
When you force the same idea through several registers, only the structure that is actually understood tends to survive.
What disappears was decoration, borrowed phrasing, or a hedge you never owned.
The hostile-skeptic register is especially sharp: it strips pedagogical kindness and asks what would still be defensible.
The child register is dangerous if it becomes false (Minimal-sufficient explanation); used carefully it reveals which distinctions you thought were essential and cannot yet motivate.
How to run it
- Ask for 3-4 registers at once, same idea, labelled.
- Read for differences, not comfort. Make a three-column note: kept / dropped / distorted.
- Restate in a register the AI did not use (e.g. "explain to my lab partner who thinks this is trivial").
- Next: mark distortions. Child versions that lie get discarded or repaired.
- Keep the densest true version you can now say, often a hybrid.
Example exchange
Explain natural selection four ways: expert to expert, peer in psychology, bright twelve-year-old without becoming false, and a hostile skeptic. Then tell me what survived all four.
I'll give the twelve-year-old version first so we don't drown in jargon: living things vary; some variants leave more offspring; those variants become common. What's one thing that version risks distorting?
It might sound like individuals try to adapt on purpose.
Good catch, that distortion is why the skeptic register matters. Ask the skeptic's objection next, in your own words.
Copyable prompt
Explain [IDEA] four times, labelled:
(1) expert to expert,
(2) strong peer in an adjacent field,
(3) bright 12-year-old WITHOUT becoming false, if you can't, say so,
(4) hostile skeptic who doubts the idea is coherent.
Afterward: list what claim survived all four, what only appeared in (1), and
what (3) risked distorting.The Tell
Here is how you know this method has flipped: explanations in different registers feel like different topics, and you can't say what the shared claim was.
In that moment the AI (or the schedule, or the story) did the thinking, and you only recognised a finished product.
Tighten the prompt for instant register shift so you attempt, decide, or retrieve before anything is handed to you.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
The underlying learning idea is rated weak. Register / audience shifting has roots in rhetoric and language education more than in a single learning-science effect size. Comparing registers can surface conceptual invariants (related to Multi-level explanation/Simplify, then restore) but direct retention evidence is thin. Principle weak-to-moderate; AI register performance highly variable, speculative.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
The learning principle may be weak, but that does not prove an AI session delivers it well. Treat AI delivery as speculative unless a study of tutoring with this method is named, and none is claimed here.