Our Position
on AI.

AI is a reality. We use it, and we're not going to pretend otherwise.

We use AI tools, but sparingly and only where they genuinely help. When we do, every output is reviewed by a person before it becomes part of anything we ship or publish. We don't use AI to generate code we don't understand, and we don't use it to write things that need a human voice.

We are against vibe coding. Shipping whatever an AI produces because it compiles is not engineering.

Where we do use it:

• ✦ Research. Exploring unfamiliar territory, summarising documentation, understanding tradeoffs.
• ✦ Internal, non-user-facing work. Drafts, notes, and tooling that never reaches a user directly.
• ✦ Porting. Translating logic between languages or platforms where the intent is already well-defined.
• ✦ Peer and code review. A second pass on logic, edge cases, and potential issues.
• ✦ Writing code. With the explicit requirement that every line is read, understood, and owned by the person committing it.
• ✦ Writing documentation. First drafts, structure, and coverage. Always reviewed and rewritten where necessary by the person who built the thing.

In all cases, the output is reviewed by a person before it goes anywhere. We don't ship things we don't understand.

We encourage everyone at Streetwriters to work the same way. Use it where it genuinely helps, review everything it produces, and don't let it substitute for understanding the work.

That's our position. It may change as the technology does, but the underlying principle won't: we are responsible for everything we produce, and that means understanding what we've made.