Hello, world.
Why this blog exists, what it'll be about, and a note on writing in a second language.
This is the first post.
I’m writing here because the last few months have been mostly spent on a question I can’t get out of my head: what does engineering actually look like when one human and a capable AI ship things together?
I’m not a developer. I’d describe myself as someone who can hold a complex system in their head, who can spot when a thing is structurally off, who can decide what not to build. The tactical layer — typing, syntax, boilerplate — I delegate. To Claude, mostly. The strategic layer — what problem is this even solving, who pays, what breaks at scale, what does the law require — stays mine.
That split used to be uncommon. With AI as a serious co-pilot, I think it’s becoming a normal mode of operating, especially for solo founders and small teams. This blog will be a build log of that mode: real projects, real diagnoses, real moments where I overruled the AI and was right, and the moments where I should have overruled it sooner.
A few things I expect to write about:
- Build logs from active projects (more on those soon)
- Diagnoses — patterns where AI tools quietly produce subtly wrong things, and how to design around it
- System thinking without code — the part that doesn’t go away when the models get better
- Decisions: when to automate, when to override, when to ship anyway
A note on language. English isn’t my first language. I’m writing here partly to think and partly to get better at it. If something reads slightly off — that’s the price of doing this in public. If you have a correction or a counter-argument, I’d rather hear it than not.
More soon.