Methods · Making it fun, memorable & personally alive · Short guide
J4
Difficulty normalisation
Learn that the confusion you feel is expected at this stage, so productive struggle isn't misread as proof you should quit.
When to use it
- Confusion feels like proof you should quit, and you cannot tell if that is normal for this stage.
- You are stuck after a real attempt and need to separate “expected struggle” from “missing prerequisite.”
- You want one next move with a time box, not endless reassurance.
What to ask AI
Type something like: “I’m confused about [topic] after trying [what I did]. Normalise vs diagnose: is this expected-stage confusion or a missing prerequisite? Ask me two questions to tell. Then prescribe one next move.”
Ask it to distinguish comfort talk from diagnosis.
Keep the hard part on your side: answer the diagnostic questions honestly, then do the one next move for a set time.
How it helps
Useful struggle often feels like incompetence from the inside.
Difficulty normalisation names where you are on the learning curve, and what confusion usually means at that stage.
It also has to separate stage-typical struggle from a missing earlier skill.
Normalisation without a plan is just soothing. Pair the verdict with one concrete next action and a time box.
How to run it
- State the stuck feeling and the exact task.
- Ask what confusion typically signals at this stage for this skill.
- Separate “need a prerequisite” from “expected struggle.”
- Choose one next move: fix the earlier gap, or keep going with support.
- Time-box the struggle so normalisation does not become endless flailing.
Example exchange
I’ve tried two textbook induction proofs and I keep stalling on the inductive step. I’m not sure if this is normal early confusion or if I’m missing a prerequisite. Can you ask me a couple of questions and then suggest one next move?
First: can you write a clear base case and say in words what “P(k) implies P(k+1)” means for a simple sum, without looking? Second: where exactly do you stall, the algebra, or not knowing what you’re allowed to assume?
Base case is fine. I freeze because I don’t know what I’m allowed to use from P(k).
That pattern is common early in induction: expected struggle around the assumption, not a missing prerequisite like algebra. Next move: twenty minutes on three proofs where you underline the exact line that uses P(k). Stop after three, don’t binge a whole chapter.
Copyable prompt
I'm confused about [ ] after trying [ ].
Normalise vs diagnose: is this expected-stage confusion or a missing
prerequisite? Ask me two questions to tell. Then prescribe one next move.The Tell
The Tell is feeling reassured after the chat and still having no next action on the page.
In that moment the AI soothed you, and the struggle never became a plan.
Leave with one diagnosis and one time-boxed move, or you only bought comfort.
Principle evidence
Strength of the underlying learning idea, not a claim about AI products.
The underlying learning idea is rated moderate. Desirable-difficulty and productive-struggle traditions hold that confusion and effortful processing are often expected features of durable learning, not proof you should stop, when the struggle is well targeted. Misreading normal hardness as failure kills persistence. Pure cheerleading without diagnosis is weaker.
AI delivery evidence
Whether an AI tutor delivers this method well is a separate question.
AI “stage calibration” is speculative. The useful delivery is a short diagnose-vs-normalise fork plus one next move. If the chat only says struggle is good and never checks for a missing prerequisite, treat it as comfort, not method.