GenAI is better at writing prompts than I am

I’ve been trying something recently. Instead of writing detailed prompts from scratch, I give GenAI a rough description of what I want and ask it to write the prompt for me.

A recent example from doing competitor analysis for a client:

“write a prompt to do competitor research analysis about <company name> and AI initiatives”

The output was better than I expected: a fully structured research brief with defined roles, clear objectives, and five detailed sections. It even added example competitors and guardrails for keeping the analysis objective — things I hadn’t mentioned and would probably have forgotten to add. It’s saved me about 10 minutes of thinking and writing per prompt.

I’ve tried it across a few other use cases:

• Creating infographics: “write a prompt to create an infographic based on this text”
• Analysing user research transcripts: “write a prompt that pulls out key themes from these transcripts”
• Extracting writing style: “write a prompt to analyse my blog posts and extract my tone and style into a guide I can use with GenAI”

Across all of these, the pattern is the same: a rough description in, a structured and detailed prompt out.

What I’ve noticed is that GenAI is just better at writing prompts than I am. It knows the structure and applies it automatically — things like role, context, objectives, guardrails, and format — which I’d either forget or have to look up.

It also helps me with something I struggle with: the blank page problem. Instead of staring at an empty prompt wondering what to write, I just describe what I’m trying to do in a few words and let it build from there.

The important skill I’ve learned isn’t writing the perfect prompt — it’s describing the goal I’m trying to achieve.