Methods · Making it fun, memorable & personally alive

J16

Personal character insertion

Put yourself, or someone you find interesting, inside the examples, so the material is happening to a person rather than to nobody.

When to use this
First time through the material
Learning impact
Clear help, modest gain
Learning evidence
Credible with caveats
Where you are with the topic
Meeting the idea
How often to use it
Worth repeating often
What AI is doingCreate practice, examples, or schedules

When to use it

  • Textbook examples are about nobody, and nothing feels at stake.
  • Abstract material will not stick until it happens to a specific person.
  • You can solve the personalised version and stall on the anonymous train problem.

What to ask AI

Type something like: “Rewrite these examples so I’m the person in them. I’m [low-stakes details]. Keep every number and principle exactly the same; only change who it’s happening to.”

Ask it not to simplify difficulty while personalising.

Keep the hard part on your side: run the specific constraints, then abstract the principle with personal detail removed.

How it helps

Anonymous examples signal generality, and they often cost care and memory.

Material encoded in relation to yourself is often remembered better than equivalent material about strangers, the self-reference effect.

Inserting yourself also forces specificity: vague “a consumer faces a budget constraint” becomes a real trade-off.

Boundary: do not insert yourself into illness, bereavement, or trauma you are carrying. Use a liked persona instead, it works nearly as well.

How to run it

  1. Supply real but low-stakes details: course, city, hobbies, nothing you would mind in a chat log.
  2. Insist numbers, mechanism, and difficulty stay identical.
  3. Use a persona when you would rather not use yourself.
  4. Let the specificity bite: run real constraints rather than idealising.
  5. Abstract back: restate the principle with personal detail stripped, then try one unrelated example.

Example exchange

You

Can you rewrite these budget-constraint problems so I’m the person in them? I study in Manchester, have £140 a month after rent for food and phone, and my laptop just died. Keep every number and the difficulty the same.

AI

You’re in Manchester with £140 left this month. Rent is already paid. Your laptop has failed, ninety pounds to repair. Food and phone must still fit. Same trade-off structure as the textbook widget firm. Solve for the feasible set if you repair versus if you delay.

You

If I repair, I have £50 for everything else. Too tight. I’ll delay repair and keep £140 for food and phone. Can you check the principle without me naming it?

AI

Abstract: opportunity cost and a hard budget line. Now do one widget-firm version with no personal details to prove you have the concept, not only your story.

Copyable prompt

Rewrite these examples so I'm the person in them. I'm [a few true, low-stakes
details: what I study, where I live, what I do at weekends]. Keep every number
and every principle exactly the same, only change who it's happening to.

The Tell

The Tell is solving every version about you and stalling on the version about a train.

In that moment you bound to the surface story, not the principle.

Always finish by abstracting out the personal detail and checking a depersonalised twin.

Principle evidence

Strength of the underlying learning idea, not a claim about AI products.

The underlying learning idea is rated moderate. The self-reference effect, better memory for material encoded in relation to oneself, is a relatively robust memory finding. Personalised examples also force specificity. Engagement benefits beyond memory are thinner; this is an encoding and engagement layer, not a full study system.

AI delivery evidence

Whether an AI tutor delivers this method well is a separate question.

AI personalisation quality and safety boundaries are delivery concerns. Difficulty drift (quietly easier numbers) is common. If the substance changes, delivery failed. Prefer personas for emotionally loaded topics.