Prompt Engineering

The Art and Science of Prompting: Your Secret to AI Success

Feeling like your AI tool gives you a different response every time you ask a question? Or maybe the answers are just... not quite what you were hoping for?

The secret isn't in the AI tool itself—whether you're using ChatGPT, Gemini, or Copilot—it's in the prompt.

Think of prompting as the art and science of giving the right ingredients to get the perfect result. Just like a good recipe requires a precise list of components and clear instructions, a good AI prompt needs a few key ingredients to go from a vague query to a brilliant, tailored response.

This practice is known as prompt engineering, and it's the single most important skill you can learn to unlock the true potential of AI in your classroom.

The Ingredients of a Great Prompt

The goal is to move beyond simple, one-line questions like "Write a lesson plan on the water cycle." A more effective prompt provides the AI with a clear purpose and all the necessary information.

Microsoft's framework provides a fantastic model for this, using four key components. Let's break them down and see how they work together.

1. Goal

This is your core objective. What is the specific task you want the AI to accomplish?

• Example: "Create a lesson plan."

• Without this, the AI might not know if you want an activity, a summary, or a full plan.

2. Context

This is the background information and relevant details needed to complete the goal accurately. It helps the AI understand the setting.

• Example: "It should be for a Year 6 class and should be engaging and interactive."

• Without this, the AI might give you a plan suitable for older students or a generic, non-interactive plan.

3. Source

This is the specific information the AI should use for its response. This is a game-changer because it grounds the AI in your own materials, preventing generic or irrelevant answers.

• Example: "Use the provided curriculum document for key learning objectives and refer to my previous lesson on weather patterns."

• This tells the AI to use your specific, trusted information, rather than pulling from its general knowledge.

4. Expectations

This is the desired format, tone, length, and style of the final output. It's how you tell the AI exactly how you want the finished product to look.

• Example: "The plan should include a list of materials, a step-by-step procedure, and a short assessment activity. The tone should be instructional and the format should be a bulleted list."

• This ensures you don't get a long, narrative response when all you wanted was a quick, scannable list.

Putting It All Together: A Masterful Prompt

By combining these four elements, we transform a vague request into a powerful set of instructions.

Simple Prompt: "Write a lesson plan on the water cycle."

Full Prompt (Using the framework above):

"Create a lesson plan on the water cycle for a Year 6 class. It should be engaging and interactive. Use the provided curriculum document for key learning objectives and refer to my previous lesson on weather patterns. The plan should include a list of materials, a step-by-step procedure, and a short assessment activity. The tone should be instructional and the format should be a bulleted list."

Notice the difference? The full prompt is clear, specific, and gives the AI all the information it needs to deliver an accurate, useful, and perfectly formatted response.

Learning to craft a good prompt is truly a form of professional development in this new age of AI. It's the difference between asking for a generic "lesson plan" and getting a personalized, ready-to-use resource tailored to your exact needs.

Take the first step today: the next time you use an AI tool, try including all four of these "ingredients" in your prompt. The results will surprise you!