The first time I asked an AI to “fix my code,” it gave me something that compiled… but made absolutely no sense.
The problem wasn’t the AI — it was me. I hadn’t told it what I was actually trying to do, why the code existed, or what success even looked like.
And that’s when it hit me: I’ve seen this exact thing happen on software teams for years.
We do the same thing to each other.
We assume people can read our minds.
When I Left My Team Guessing
I’ve done it plenty of times.
Pushed up half-finished code thinking, “They’ll get it.”
Fixed a bug without explaining why I made a tradeoff.
At the time it felt efficient. But later someone had to stop and retrace my steps, trying to understand what I was thinking. It slowed everyone down.
It’s the same as giving AI a bad prompt — you don’t get what you expect because you didn’t provide enough context.
What AI Taught Me About Context
Working with AI reminded me of a few simple truths that also apply to working with people:
- Be explicit. AI needs details. So do your teammates. Clear commits, comments, and TODOs save hours later.
- Use the right language. AI can’t guess what you mean — neither can humans. Clarity beats cleverness every time.
- Keep it current. Stale prompts confuse AI; stale tickets confuse teams.
- Leave breadcrumbs. If you want continuity, you have to leave a trail someone else (or future you) can follow.
Context Is a Gift
Leaving context isn’t busywork. It’s a gift — to your future self, your team, and even your AI assistants.
It’s how progress keeps moving when you’re out sick, pulled into another meeting, or switching projects.
That’s why I built imdone — to leave context right where we work: in the code.
Because whether you’re collaborating with humans or AI, context isn’t optional.
It’s the difference between chaos and clarity.
Top comments (0)