Position
We've fully embraced AI as part of the practice — every day, on real client work. The point isn't novelty. The point is that a small, careful studio can now operate at a scope that used to require a much larger one, without trading away the things that make the work worth hiring us for.
Below are the principles that shape how we use it. Short version: AI raises the bar; humans still set it.
Principles
AI is part of the work, every day
Research, drafts, code review, debugging, prose, refactoring, accessibility audits, design exploration. We use it broadly and openly. Pretending otherwise wouldn't serve the work or you.
The bar moves up, not the time down
We don't use AI to ship the same quality faster. We use it to ship a higher quality in the time we already had — more iteration, more polish, more attention to the details that usually get squeezed at the end of a project.
Judgment stays human
AI generates options; we choose, refine, and reject. The taste calls, the structural decisions, the questions about what the work should be — those aren't being outsourced. A studio whose judgment lives in a model isn't a studio worth hiring.
Your work is yours
Client material isn't training data. We don't paste confidential code, content, or strategy into general-purpose AI products, and the tools we do use have terms we've actually read. If a project has unusual confidentiality needs, tell us — we'll match the workflow to them.
Useful, not novel
We don't bolt AI features onto your product because it's the trend. If AI doesn't make the thing measurably better for the person using it, it doesn't ship. "Has AI in it" isn't a feature; solving a real problem is.
Open about how
If you want to know which parts of a project used AI — and how — ask. We'll show you. There's no secret about the tooling, and there shouldn't be.
Talk to us
Want to talk about how this shows up on a specific project, or have a harder question about AI on client work? Write to us.
[email protected]