Most people stop after one attempt. If the output is not great, they assume the tool is not useful. The skill you are building this week is the opposite: treat every output as a draft, and tell the AI what to change.
Today's task
Take the best output you got from Day 1 or Day 2 and ask for a revision. Use one of these follow-up prompts:
- "This is too formal. Make it more conversational."
- "Shorten this by half. Keep the key points."
- "Add a specific example for each point you made."
- "I disagree with point three. Here is why: [explain]. Revise accordingly."
Notice what happened: your feedback is getting more specific. That specificity is the skill. Each round teaches you what information the AI needed that it didn't get the first time. Over a few rounds, you learn that the first prompt is rarely the best prompt, it is just the starting point.
Specific revision feedback is rarer and more valuable than the ability to write perfect first prompts. Every round of feedback teaches you what information the AI needed that it did not get the first time. This pattern, try-evaluate-specify-retry, is the same cycle that powers every iterative process from scientific research to product design. The AI is just the fastest practice partner you will ever have for building it.