By skyforbes I’ve been catching myself doing something kind of dumb.
I’ll open a fresh chat with a clear intention in my head — I know I want analysis, or criticism, or concrete steps — but what I actually type is average, Vague, Soft.
It’s not that I don’t know better. I’ve written better prompts before. I’ve seen them work. But in a new chat, staring at the empty input box, I default to whatever feels fastest.
Then I blame the output.
At some point I started paying attention to where this goes wrong, and it didn’t feel like a "prompt skill" issue. It felt more like friction.
Every conversation resets. No trace of how I like to think. No reminders. No scaffolding. Just a blank box and my memory doing all the work.
The surprising thing was how much changed when I stopped relying on memory altogether.
Instead of "remembering" how to phrase things, I made a few of my usual thinking nudges something I could just click in when I needed them. Tiny stuff. Not templates, not big personas. Just short prompt starters I reach for all the time.
Things like:
– asking it to be critical instead of agreeable
– forcing it to zoom out before answering
– pushing it to turn abstract ideas into actual steps
What felt weird was how quickly the quality stabilized across chats. Fewer lazy starts. Fewer rewrites. Less irritation at myself for knowing better and not using it.
It made me think that a lot of what we call "prompt engineering" is really about whether the interface helps or fights your habits.
Anyway, curious if this resonates or if I’m just unusually bad at typing what I’m already thinking.