Comment by bottlepalm
13 hours ago
One strategy I use in the planning phase is even when I know how I'd implement the solution, I ask the Claude/Codex how they would solve the problem or implement the feature without giving them any clues - and then compare their solutions to my own. Often I am pleasantly surprised by alternative ways of doing things and ideas that we integrate into the final design.
Same. I've been creating "research" documents where I let it do a freeform survey of possible solutions/have sketch out it's own solution. I'll then sketch out a plan based on what I think is good or what I think it missed, and then I'll have it interrogate me for a final PRD document. It then implements the feature in reviewable chunks, and I'll give it feedback or tweak the PRD doc as needed.
Finally feel like I have a good workflow where I can fully benefit from these things without sacrificing my understanding of what they're doing.
Same here. Step 1 is usually a research doc where I simply describe the task and tell it to research the relevant parts of the codebase. This gets refined to a high-level plan, which gets distilled to a detailed step-by-step implementation plan.
When it comes to the actual implementation I prefer to work through it in small steps, where the AI explains to me exactly what it's about to do and why (and I approve) along the way. This enables me to catch it if it's about to do something I disagree with beforehand. And reduces the time I need to spend reviewing in the end.
I like this, though it does leave me feeling more nervous when I really don't know how I'd solve the problem, still requires trust.