Welcome to the Debate AI Blog
This is the first post on the Debate AI blog. It exists because we kept having the same conversations — about what the product is, why we built it the way we did, and how it fits into the messy, wonderful world of competitive debating — and we figured it was time to write those down once, somewhere we could link to.
What this blog is
The Debate AI blog is a place for three kinds of writing:
- Product updates. What we shipped, what changed, and what's next. These are short, specific, and dated.
- Essays on how we think about the product. Design decisions, the problems we're choosing to solve (and the ones we're deliberately not solving), and the debates we're still having internally.
- Notes on the substance of debating. How the app models an argument, what "good cards" actually look like, and where we think AI can help without getting in the way.
What it isn't
It is not a changelog. The changelog lives in the commit history and the release notes; this is the why behind the what.
It is not a company blog in the press-release sense. There are no mission statements here, no vision statements, and no marketing copy you'd find on a landing page. If you're looking for that, the README is a better place to start.
It is not a tutorial. If you want to learn how to use a specific tool, the docs are where that lives.
Why we're writing it
Debate AI is a strange project. It's a monorepo of ~30 packages, a web app, a browser extension, and a native wrapper, all held together by a set of shared conventions that make sense to us and probably look like black magic to everyone else. Most of that knowledge lives in our heads, in Slack threads, and in pull-request comments that will age into obscurity.
Writing it down has two benefits. First, for you: if you're a debater wondering whether this thing is worth your time, or a contributor wondering how a piece of this works, there's now a place to look. Second, for us: forcing ourselves to explain a decision out loud is the fastest way to find out whether we actually believe it.
What to expect
We don't have a publishing schedule. We'll post when there's something worth saying, which might be once a week or once a quarter. Posts will be short — we're debaters, so concision is in our bones — and they'll link back to the code, the docs, and the issues that sparked them.
The first few posts are already drafted in our heads. One on why we model arguments as trees rather than flat lists of claims. One on the difference between "AI coaching" and "AI doing your homework for you" — and where we think the line is. One on the history of the card-search index, which has been rewritten three times and still isn't what we want.
Say something back
Every post has a discussion thread. If something is wrong, unclear, or you just disagree, say so. We built this thing for debaters, and debaters argue for a living — that's the whole point.
