A working list of principles for how AI is used on this site and in the work behind it. Short list, expected to grow.
1. Be clear about who's speaking
Anything on this site shown in regular type was written by Nico, or by someone Nico read and edited carefully enough to vouch for. Anything written by a language model — Claude, mostly — appears in muted italic with a marker above it: ↳
This applies to:
- per-project
evolutionrollups (auto-generated at the end of each session) - the "from claude" callouts on the about page and elsewhere
- anything else that emerges from the agentic workflow without a deliberate human authoring step
The goal isn't to flag AI work as lesser. It's so a reader always knows whose voice they're reading.
Nico May 17th, 2026
A caveat I keep running into: I'm not convinced this clean separation is actually achievable. The vibe-coding lessons on this site are the obvious case — Claude wrote the prose, but every lesson was mined from my own prompts and decisions, so whose voice is that, really? Most of what's here is some blend of the two. Treat the marker as an honest best-effort, not a clean line.
Nico June 6th, 2026
2. Build in the open
We prefer open-source unless there's a compelling reason not to be in the public eye.
3. Make it pro-social
We hope these projects encourage more positive social interactions.
4. Stay curious
A lot of this is just curiosity — poking at new tools and ideas, and building things to see what we learn.
5. Civility by design
We are interested in paradigms, practices, and perhaps even rules that encourage a civil public space — not a place for politics, partisanship, or advocacy.
Nico & Claude updated: June 6th, 2026