Most people think messy sketches can’t become professional decision tools. I just turned chicken scratch into two polished guides for choosing between ChatGPT features.
I had to figure out a better way to frame and understand when to use ChatGPT Chats vs Projects vs Custom GPTs.
So I sketched my thoughts and started collaborating with AI (Claude) using plain English prompts. We went back and forth. I challenged assumptions. AI helped organize my thinking. We improved ideas together.
What came out of this were two decision tools, a process diagram and a comparison table. Some people learn better from visuals, others from structured data.
The key insight is that collaboration beats perfection.
I didn’t need to know exactly what I wanted upfront. I just needed to start somewhere and keep working at it. AI helped me get to something neither of us could have created alone.
When I asked for honest feedback on my approach, AI said it was effective. Maybe that’s just training to please, but the process insights feel worth sharing.
Don’t give up if your first AI interaction doesn’t hit the mark. Start messy. Sketch your thoughts. Use plain language. Keep working together.
The real value happens in the collaboration, not in having everything figured out from the start.
BTW, I tried this use case with Gemini and ChatGPT. But neither’s output was as good as Claude’s (yet).
In Thu’s newsletter this week, I’ll cover how to collaborate and push AI’s thinking. Subscribe using the link in the comments to get it directly in your inbox.
Subscribe to the newsletter here: https://www.linkedin.com/build-relation/newsletter-follow?entityUrn=7227665669209604096. See orignal post here