Prompting Fundamentals
What is Prompting?
A prompt is the text you send to an AI model to get a response. It's your way of communicating with the AI—telling it what you need, providing context, and guiding its output.
Why Prompting Matters
The quality of your prompt directly affects the quality of the response. The same AI model can give vastly different outputs depending on how you ask:
Vague prompt: "Write about dogs" → Generic, unfocused response
Better prompt: "Write a 200-word guide for first-time dog owners on choosing between a puppy and an adult rescue dog" → Specific, useful, actionable response
The Prompt-Response Loop
Working with AI is iterative:
- You send a prompt → AI generates response
- You evaluate → Is it what you needed?
- You refine → Adjust your prompt based on the output
- Repeat → Until you get the desired result
This back-and-forth is natural. Even experts rarely get perfect results on the first try.
Components of a Good Prompt
Most effective prompts include some combination of:
- Task: What you want the AI to do
- Context: Background information that helps
- Format: How you want the output structured
- Constraints: Limitations or requirements
- Examples: Sample inputs/outputs (optional but powerful)
A Simple Example
Instead of: "Help me with my email"
Try: "I need to write a professional email to decline a meeting invitation. I want to be polite but firm, explain I have a conflict, and suggest rescheduling. Keep it under 100 words."
The second prompt gives the AI everything it needs to produce exactly what you want.
Key Insight
Think of prompting as giving instructions to a capable but literal assistant. Be clear, specific, and provide the context they would need to help you effectively.
:::