Learning Timeline
Key Insights
Prompting as a Continuous Loop
Prompting isn't a one-and-done process (straight line). It is a loop of 'Ask, Check, Adjust, and Ask again' until you achieve the perfect result.
The Power of Constraints
Setting 'Constraints' actually forces the AI to be more creative. Without limits, AI tends to provide generic and boring answers.
Avoid AI Information Overload
AI processes information much like humans do. If you pack too many instructions into a single large paragraph, the AI can become overwhelmed and might miss crucial details.
Prompts
Example of a Simple Prompt Structure
Target:
ChatGPT / Google Gemini
Create a strategy for Q1. Target Gen Z, include a budget, add KPIs.
Example of a Prompt with Constraints
Target:
ChatGPT / Google Gemini
Give me video ideas. It must be under 90 seconds, must focus on one single tip, and must start with a question.
Step by Step
Iterating and Refining Failed Prompts
- Analyze AI outputs that don't meet your criteria to identify weaknesses in the response.
- Review your original prompt framework to ensure 'Context' and 'Persona' elements have been included.
- Identify instruction paragraphs that are too long or complex (run-on sentences).
- Break down those instructions into simpler, clearer sentence structures (e.g., use bullet points).
- Set specific 'Constraints' such as timeframes, target audiences, or specific sentence opening styles.
- Enter the refined prompt into the AI chat box.
- Click the 'Send' or 'Generate' button to see the improved results.
- Repeat this loop until the AI output reaches your desired standard.
