Ask why before asking AI to build
https://patrickz.ai/learn/ask-why-before-building/

OUTCOME
Write a one-page problem brief with an explicit finish line.

YOU NEED
A chat assistant and a task that currently annoys you.

1. Describe the friction
Say who has the problem, when it happens and what they do today. Use an example rather than a list of desired features.

2. Invite useful disagreement
Ask the assistant to test your assumptions and offer a non-AI alternative. It should ask a small number of questions before proposing a solution.

3. Choose a narrow outcome
Write what must become easier, what is out of scope and how you will observe success. Include constraints such as time, device or privacy.

4. Turn the brief into acceptance checks
Create three actions a person can perform to verify the result. Use those checks to steer the work and to decide when it is done.

PROMPT
Before solving this, help me clarify what I want. Situation: [example]. Who is affected: [people]. Current workaround: [workaround]. Ask three important questions, challenge one assumption, and propose a simpler non-AI option. Then write an outcome, constraints, exclusions and three acceptance checks.

EXAMPLE / EXPERIMENT
“I need a dashboard” becomes “I need to know which two invoices require a follow-up today.” A sorted list may solve the problem before a dashboard is justified.

CHECK YOUR RESULT
[ ] The problem is described without naming a preferred tool.
[ ] At least one assumption has been examined.
[ ] Three checks define completion.

IF IT FAILS
If the assistant immediately writes code, restate that this is a problem-definition task. If the brief stays vague, provide one real example of the current failure.

MY RESULT / NEXT CHANGE