Fix a weak AI video prompt
Rewrites a video prompt that gave poor results into three sharper versions, and names what was missing, vague or conflicting in the original.
Task:
Rewrite the AI video prompt below into three sharper versions and tell me what was wrong with it.
Context:
I used the prompt in [tool name] and the problem was [what went wrong, for example the person's face changed or nothing moved]. I want the result to show [what you wanted to see].
Rules:
- Keep the idea, the subject and the setting of the original.
- Remove conflicts, such as two camera moves or two moods, and replace vague words with something a camera could film.
- Give each version one action, one camera move and at most 60 words.
- Make the three versions differ in how much detail they give: minimal, balanced and detailed.
- Do not claim that a version will fix the problem. Say what the change targets and what I should check in the result.
Result:
First a list of at most five problems found in the original, each with a quote. Then the three versions as a numbered list, each followed by one line on what it tests.
Material:
[paste the prompt that gave poor results here]