By Large Language Models (LLM's) I am reffering to the likes of ChatGPT, Claude etc...
- Start with the basics:
- Begin by asking simple questions that don't need a lot of background information.
- Add a little at a time:
- Slowly start asking more detailed questions that make the computer think harder and use more of what it knows.
- Learn as you go:
- Use what worked and what didn't from your earlier questions to make your next questions better.
- Ask for the computer's thought process:
- Get the computer to explain how it came up with its answers step by step.
- Keep tweaking:
- Keep changing and improving your questions based on how well the computer's answers turn out.
This information is based off the article: