One single line that drastically improved my ChatGPT and Agentic Worker outputs
“Before you answer, assess the uncertainty of your response. If it’s greater than 0.1, ask me clarifying questions until the uncertainty is 0.1 or lower.”
That's it, this forces the the AI to think harder and prompt ME to fill in the gaps.
I now include this in the personas of all my Agentic Workers and ChatGPT system prompts.
4
2 comments
One single line that drastically improved my ChatGPT and Agentic Worker outputs