Happy Friday. Week 0 of doing these roundups publicly. Let's see how it goes.
This week was mostly about figuring out where AI actually saves time in an engineering workflow versus where it just creates a convincing illusion of speed. Here's the honest tally.
✅ What worked
1. Using AI for the first draft of boilerplate, not the logic.
Generating scaffold code — config files, test stubs, README sections — is where I stopped fighting the output and started trusting it. The model doesn't need to understand your domain to write a Dockerfile or a pytest fixture skeleton. Ship the boring stuff fast, think harder on the real problems.
2. Prompt iteration as a debugging loop.
Treating a bad AI output like a failing test — rather than a reason to give up — changed everything. Reframe, constrain, give an example. Three rounds of that usually gets you somewhere useful. The mental model shift is: you're not asking, you're specifying.
3. Writing documentation before the code.
Sounds backwards but it works. Describe the function in plain English, hand it to the model, get a first implementation back. The doc becomes the spec. You catch ambiguity before it's baked into 200 lines of code.
❌ What didn't work
Letting AI "own" a refactor end-to-end.
I handed off a medium-complexity refactor with minimal checkpoints. The output was locally coherent and globally wrong — it made decisions silently that I would have caught immediately if I'd stayed in the loop. The lesson: keep the AI in a copilot seat, not the driver's seat, any time the task spans more than one file or abstraction layer.
The thread underneath all of this
Every win this week came from having a clearer mental model of the task before touching the AI. Every loss came from outsourcing the thinking. That's probably the pattern for a while.
If you want a structured version of these ideas — prompts, workflows, and checklists organized into a repeatable system — I've been packaging what actually works into a playbook: grab it here. No fluff, just the stuff I'd send a teammate.
See you next Friday. 🖤
Top comments (0)