Prompt, Critique, Refine

What you will learn

• Understand that a clear starting prompt includes the task, the audience, and the constraints, and that the quality of the first prompt strongly shapes the quality of the first output.
• Learn to critique an AI output by naming specific things that are wrong, missing, or weak, in short concrete feedback that can feed directly into the next prompt.
• Learn to refine a prompt based on critique and run at least one more round, so that the loop closes and the final output is visibly better than the first.
• Combine prompt, critique, and refinement into a Refine Loop template the student can reuse on any future task, documenting three complete rounds on one real task.

What happens in this lesson

1. Chapter 1: Writing A Clear Prompt

   Understand that a clear starting prompt includes the task, the audience, and the constraints, and that the quality of the first prompt strongly shapes the quality of the first output.

   What you make: A saved prompt draft for a real task including the task itself, the audience it is for, and at least two specific constraints.

2. Chapter 2: Critique The Output

   Learn to critique an AI output by naming specific things that are wrong, missing, or weak, in short concrete feedback that can feed directly into the next prompt.

   What you make: A saved critique note identifying at least three specific weaknesses in a real AI output and explaining what each weakness is.

3. Chapter 3: Refine And Iterate

   Learn to refine a prompt based on critique and run at least one more round, so that the loop closes and the final output is visibly better than the first.

   What you make: A saved iteration log showing one prompt, one critique, and one refined prompt that addresses the critique specifically.

4. Chapter 4: Refine Loop Template

   Combine prompt, critique, and refinement into a Refine Loop template the student can reuse on any future task, documenting three complete rounds on one real task.

   What you make: A saved Refine Loop template for one real task showing three rounds of prompt, critique, and refinement, with a short reflection on what each round improved.

5. Chapter 5: Refine For A Real-World Brief

   Apply the prompt-critique-refine loop to a more complex, realistic brief and evaluate which feedback should change the next round and which feedback should be ignored or deferred.

   What you make: A completed three-round iteration log for a realistic brief, plus a short decision note explaining which critique points were acted on, which were not, and why.