Every second, hundreds of people are quietly rewriting Wikipedia. This is what that looks like.
Wikipedia is one of the largest collaborative projects in human history, and most of the work that keeps it alive is completely invisible. Every minute, editors around the world are adding citations, correcting facts, reverting vandalism, and updating articles about things happening right now. The scale of it is genuinely hard to comprehend.
Wikimedia publishes a live stream of every edit as it happens, a firehose of raw data: page titles, usernames, timestamps, edit types. It’s fascinating in theory and completely unreadable in practice.
Live Data Stories pipes that stream through an AI model, which watches a rolling window of edits and writes a short narrative about what it sees — what’s being edited, who’s editing it, what patterns are emerging, and what that might tell us about what the world is paying attention to right now.
It updates every 45 seconds. Every story is generated fresh from real data.
This is a personal side project — a way to explore what happens when you point an AI at a live data source and ask it to be a journalist. The stakes aren’t real. The data is.
Wikipedia runs entirely on donations, no ads, no investors. If you find value in it, consider supporting the Wikimedia Foundation.