📘 CHAPTER 5
ASK: Structured Prompting
🎯 The second mistake people make
Even when people think before asking…
They still fail here:
They don’t know how to ask properly
⚠️ The reality
Most prompts are:
Short
Vague
Unstructured
Example:
“Write an email”
“Give ideas”
“Explain this”
These are not prompts.
👉 These are instructions without direction
🧠 What AI needs from you
AI does not read your mind.
It responds to:
Structure + clarity in your request
🔥 Insight
The quality of your prompt determines the quality of your response.
🧩 The ASK layer
A strong prompt has four elements:
🔷 1. ROLE
👉 Who should AI act as?
Examples:
“Act as a project manager”
“Act as a teacher”
“Act as a marketing expert”
🔷 2. CONTEXT
👉 What is the situation?
Examples:
“For beginners”
“For working professionals”
“In India”
🔷 3. OUTPUT
👉 What do you want?
Examples:
“Give 5 bullet points”
“Write a 1-page report”
“Create a step-by-step guide”
🔷 4. CONSTRAINTS
👉 Any limits?
Examples:
“Keep it simple”
“Under 200 words”
“No jargon”
🧠 The universal prompt structure
Use this everywhere:
Act as ___
Help me ___
Give output in ___ format
Constraints: ___
🔁 Before vs After
❌ Weak prompt
“Write about leadership”
Output:
Generic
Predictable
Not useful
✅ Strong prompt
“Act as a leadership coach.
Give 5 practical leadership tips for new managers.
Use bullet points. Keep it simple.”
Output:
Focused
Relevant
Actionable
⚡ What changed?
Not AI.
👉 Your prompt structure
🧠 Why most prompts fail
Because people:
Skip role
Ignore context
Don’t define output
Forget constraints
👉 Result:
AI fills the gaps with assumptions
🧩 Your job vs AI’s job
| Your Job | AI’s Job |
|---|---|
| Define direction | Generate content |
| Provide clarity | Expand ideas |
| Set constraints | Follow structure |
👉 If you don’t do your part, AI cannot do its part well.
🔁 The refinement loop
Even a good prompt is not final.
You improve it by asking:
“Make it more practical”
“Simplify this”
“Add examples”
“Reduce length”
👉 Prompting is not one-shot.
It is iterative.
🧠 Advanced insight
The best users don’t write perfect prompts.
They:
Refine prompts progressively
⚡ Prompt evolution example
Step 1:
“Give business ideas”
Step 2:
“Give 5 business ideas in education”
Step 3:
“Act as a startup advisor.
Give 5 practical education business ideas in India.
Include problem and solution.”
👉 Each step improves clarity and usefulness.
🧠 A powerful habit
After writing a prompt, ask yourself:
Is the role clear?
Is the context defined?
Is the output specific?
Are constraints included?
👉 If not, improve it.
💡 One-line insight
Don’t ask quickly. Ask clearly.
🧠 Points to Ponder
Do I write vague prompts?
Do I define output format?
Do I refine or accept first response?
⚡ Practice (Do this now)
Take a weak prompt you usually use.
Step 1 — Rewrite using:
Role
Context
Output
Constraints
Step 2 — Run both prompts
👉 Compare the difference.
🔚 Chapter Close
Good prompts don’t come from creativity.
👉 They come from structure
🔜 Next Chapter
JUDGE: Don’t Trust AI Blindly
Why thinking after AI matters as much as thinking before
✅ Summary (1 line)
Structured prompts turn AI from generic to useful.
No comments:
Post a Comment