The Art & Science of Prompting with Gen AI

Effective prompting is the key to unlocking the real potential of generative AI. When students learn to be clear, specific, and purposeful in the way they ask for information, they move beyond simply getting answers and begin shaping meaningful responses.

By providing context, defining roles, setting expectations for format, and refining their prompts through iteration, students develop critical thinking, creativity, and communication skills.

Encouraging strong prompting habits also helps them question outputs, check for accuracy, and use AI as a thoughtful collaborator rather than a shortcut, building the kind of human capabilities that matter most in an AI-rich world.

When students learn how to prompt well, they’re not just “using a tool.” They’re actually building some pretty powerful skills along the way. They’re learning how to get clear on what they mean, how to communicate it precisely, and how to look at what comes back with a critical eye.

That’s where the real learning kicks in.

The Science of Prompting (What Works)

  • Clarity and specificity – Clear instructions produce more accurate outputs
  • Context matters – Background, purpose, and audience shape relevance
  • Defined roles – Assigning a role improves tone and perspective
  • Structured inputs – Specifying format (e.g. list, table, paragraph) guides output
  • Constraints improve focus – Word limits, tone, and scope reduce ambiguity
  • Step-by-step prompting – Breaking tasks down improves complex responses
  • Iteration refines results – Adjusting prompts leads to better outcomes
  • Patterns and predictability – AI responds to recognisable language patterns

The Art of Prompting (What Elevates It)

  • Asking better questions – Curiosity drives deeper, more meaningful responses
  • Framing intent – Shaping how a task is approached, not just what is asked
  • Creative exploration – Using prompts to generate ideas, perspectives, and possibilities
  • Tone and nuance – Crafting prompts to achieve voice, style, and emotion
  • Judgement and discernment – Knowing what to accept, reject, or refine
  • Blending human voice – Using AI as a collaborator, not a replacement
  • Play and experimentation – Trying variations to discover unexpected outcomes
  • Knowing the limits – Recognising when AI falls short and human thinking matters most

Next time you’re using generative AI, it’s worth keeping this in mind: better prompts don’t just get you better answers… they lead to better thinking. And better thinking is what drives real learning.

AI can generate an answer—but you’re the one shaping the question.

11 steps for effective gen AI prompting

  1. Be clear and specific – Vague prompts = vague results. Precision improves quality.
  2. Provide context – Include purpose, audience, and situation to guide the response.
  3. Define the role – Ask the AI to act as a teacher, marketer, historian, etc.
  4. Set the format – Specify output style (dot points, paragraph, table, script, etc.).
  5. Use constraints – Word limits, tone, structure, or level of complexity sharpen responses.
  6. Break tasks into steps – Complex prompts work better when staged or scaffolded.
  7. Iterate and refine – Prompting is a conversation—adjust based on outputs.
  8. Ask for examples – Improves clarity and usefulness of responses.
  9. Encourage reasoning – Ask “why,” “how,” or for explanations to deepen thinking.
  10. Check and verify – AI can be wrong—always review for accuracy and bias.
  11. Blend human creativity – Use AI as a collaborator, not a replacement for thinking.

Activity – Prompt Makeover Challenge

Purpose:
Help students understand how the quality of a prompt shapes the quality of an AI response.

1. Start with a weak prompt
Give students a vague prompt like:

“Write something about climate change.”

2. Generate the AI response
Use a tool like ChatGPT, Copilot, Claude, or Gemini to show the output.
Students quickly realise it’s generic and lacks depth.

3. Analyse the problem
Ask students:

  • What’s missing?
  • Why is the response so broad?
  • What would improve it?

4. Prompt makeover

In pairs or small groups, students rewrite the prompt using:

  • Clear purpose
  • Defined audience
  • Specific format
  • Added detail/context
  • Constraints (e.g. word limit, tone)

Example improved prompt:

“Write a 150-word explanation of climate change for Year 6 students, including one real-world example and a possible solution.”

5. Compare outputs

Run the improved prompts and compare results:

  • Which is clearer?
  • Which is more useful?
  • What changed?

6. Reflect and refine

Students answer:

  • What made the biggest difference?
  • How could you improve your prompt again?

Extension ideas

  • Prompt competition: Best improvement wins
  • Role-based prompting: “Act as a scientist / journalist / teacher”
  • Cross-curricular: Apply in English, Science, Humanities

Learning takeaway

Students see that better prompts = better thinking, and that AI works best when guided by human clarity, creativity, and intent.