How to Write Better AI Prompts: A Practical Guide

Most disappointing AI answers come from vague requests. Ask “write a blog post about laptops” and you’ll get something generic. Give the model the context a smart colleague would need and the results improve dramatically.

1. Give context

Who is this for? What’s the situation? “I’m a data science student in Lahore with a budget of Rs300,000” is far more useful than “recommend a laptop”.

2. Be specific about the task

Say exactly what you want: compare three options, write a 200 word email, list five risks, explain like I’m new to the topic.

3. Show examples

If you want a certain tone or format, paste an example of what good looks like. Models are excellent at matching examples.

4. Set the format

Ask for a table, bullet points, a numbered plan or plain paragraphs. Specify length.

5. Break big tasks into steps

Instead of “write my thesis chapter”, ask for an outline first, review it, then work section by section.

6. Iterate

Treat the first answer as a draft. Ask for changes: shorter, more formal, add an example, remove jargon.

7. Ask it to check itself

Ask the model to list assumptions or point out what it’s unsure about.

Always verify

AI models can make mistakes, including invented facts and citations. For anything important, such as health, legal, financial or academic work, check against reliable sources.

Protect your data

Don’t paste passwords, ID numbers, private client data or confidential business documents into public AI tools unless your organisation has approved them.

Sources

• Anthropic, Prompt engineering overview
• OpenAI, Prompt engineering guide