We test our work. Thoroughly.
A generated answer isn't a finished product. We work to understand what we use, review changes, and thoroughly test what we make. A model sounding confident is not a substitute for verification.
You don't have to.
An open letter about choosing our tools and taking responsibility for our work.
We use artificial intelligence in our work. Not because it solves everything. Not because everyone suddenly has to. Because it can be useful for the things we do.
You can write code by hand, make art without generators, and keep AI out of your workflow. That's your choice. Ours is to use it where we think it makes sense.
AI is part of the process. Not a substitute for thinking. And certainly not a way out of responsibility.
A generated answer isn't a finished product. We work to understand what we use, review changes, and thoroughly test what we make. A model sounding confident is not a substitute for verification.
If we ship it, its quality, mistakes, and fixes are our responsibility. “The AI suggested it” isn't an excuse. Testing doesn't promise that there will be no bugs. When things break, it's on us to fix them.
We don't ask you to use AI, love it, or call it progress. We don't think working without it makes you obsolete. We don't need to choose the same tools to respect each other.
Bugs, data leaks, rights violations, poor quality: specific concerns deserve a real discussion. But using AI isn't, by itself, a reason to insult someone, dismiss everything they've made, or demand they pick the “right” side.
We don't see AI as a universal savior or a universal evil. It has limitations, risks, and useful applications. We don't have to defend every AI product to use some of them.
Found a problem with our work? Let's talk about the problem. Prefer different tools? Great. We don't have to work the same way to coexist peacefully.
We'll keep doing our work, checking the result, and taking responsibility for it. You're free to do things differently.
People and organizations who share these principles.
No signatures yet. You can be the first.
Individual signatures represent the person, not their employer. Organizations sign separately through an authorized representative. This supports the letter's principles; it isn't a certification of project quality.
Share this position? Add yourself or your organization through a pull request.
supporters.jsonc on GitHub and add one entry: a name, a link, and a type — person or organization.Want to make your position clear in your README? Add a badge linking to this letter. It isn't an “AI approved” label or a quality guarantee — just support for these principles.
For the Russian badge, replacebadge.svg with badge-ru.svg and use 248 × 32 for the HTML dimensions. You can also select and copy the text manually.
Open Settings → General → Badges and enter:
https://yes-we-use-ai.geass.su/en/https://yes-we-use-ai.geass.su/badge.svgOn GitHub, use Markdown in your README.