Imagine I’m the design leader at your org and I present the following guidelines I want us to adopt as a team for doing design work:
- Typography: Use expressive, purposeful fonts and avoid default stacks (Inter, Roboto, Arial, system).
- Motion: Use a few meaningful animations (page-load, staggered reveals) instead of generic micro-motions.
- Background: Don't rely on flat, single-color backgrounds; use gradients, shapes, or subtle patterns to build atmosphere.
- Overall: Avoid boilerplate layouts and interchangeable UI patterns. Vary themes, type families, and visual languages.
How do you think that conversation would go?
I can easily imagine a spirited debate where some folks disagree with any or all of my points, arguing that they should be struck as guidelines from our collective ethos of craft. Perhaps some are boring, or too opinionated, or too reliant on trends. There are lots of valid, defensible reasons.
I can easily see this discussion being an exercise in frustration, where we debate for hours and get nowhere — “I suppose we can all agree to disagree”.
And yet — thanks to a link to Codex’s front-end tool guidelines in Simon Willison’s article about how coding agents work — I see that these are exactly the kind of guidelines that are tucked away inside an LLM that’s generating output for many teams.
It’s like a Trojan Horse of craft: guidelines you might never agree to explicitly are guiding LLM outputs, which means you are agreeing to them implicitly.
It’s a good reminder about the opacity of the instructions baked in to generative tools.
We would debate an open set of guidelines for hours, but if there’re opaquely baked in to a tool without our knowledge does anybody even care?
When you offload your thinking, you might be on-loading someone else’s you’d never agree to — personally or collectively.