AI Guides › Claude Mastery
The Difference Between Asking For An Answer And Asking For A Method
By Nigel Guy · 2 min read
"What should I do here?" is the most natural question in the world to ask,
and it's often the wrong one, because it hands the entire judgement call
over and gets back a single confident-sounding recommendation with no
visible working. It reads decisively. Whether it's actually right for your
specific situation is much harder to tell, precisely because you can't see
how it got there.
The rule: asking for an answer gets you a conclusion you have to take on
faith; asking for a method gets you a way of deciding that you can check,
reuse, and disagree with on the merits.
The mechanism: reframing the request
Before sending a decision question as-is, check whether it's asking for an
answer or a method, and reframe it deliberately if it's the former:
- Instead of "what should I do," ask "what are the factors that should
decide this, and how would each push the decision one way or the
other." This surfaces the actual reasoning as something you can weigh
yourself, rather than a single verdict you either accept or reject
wholesale.
- Instead of "which option is better," ask for the criteria a good
decision here would need to satisfy, then apply them to your specific
options yourself, or ask it to. This separates the (genuinely useful)
judgement about what matters from the (much more fallible) judgement
about your particular case.
- Instead of "write the plan," ask for the planning approach — what
order to tackle things in and why, what the biggest risks are — and then
have it produce the specific plan from that approach. The approach is
the reusable part; the specific plan is disposable and situation-bound.
Why the method version holds up better
A method exposes its own weak points. If a piece of reasoning is thin,
asking for the method surfaces that thinness directly — you can see which
factor is doing the most work and question it. A bare answer hides the same
thinness behind confident phrasing, and confident phrasing is not evidence
of a sound method underneath it.
What to skip
Skip asking for a method on questions that are genuinely simple factual
lookups — "what time zone is Tokyo in" doesn't need a framework, it needs an
answer. Reserve the method-first approach for judgement calls: decisions
with real trade-offs, ambiguity, or consequences, where the reasoning behind
the answer matters as much as the answer itself.
Guardrails
- Asking for a method doesn't remove the need for judgement — it relocates
where you apply it, from evaluating one bare conclusion to weighing the
criteria yourself. That's a better place to apply it, not a way to avoid
applying it.
- A well-articulated method can still rest on a wrong assumption about your
situation — check that the factors it names actually match your real
constraints before trusting the framework wholesale.
- For decisions with legal, medical, financial or safety consequences, a
method for thinking it through is not a substitute for the relevant
professional's judgement — use it to prepare better questions for them,
not to replace the conversation.
All 751 AI guides · JulieMango plans from £17/mo