CorrelateAI pulls real series — interest rates, air quality, phone subscriptions, dozens more — from sources like the Federal Reserve’s FRED and World Bank Open Data, then hunts for pairs that move together despite having no business doing so. It will happily show you Beijing air quality tracking mobile phone subscriptions at a correlation strong enough to put in a slide deck.
That is the point. Real data is full of convincing relationships that mean nothing, and the fastest way to internalise it is to watch a machine mass-produce them from sources you trust. Each result says which series it used and where they came from, so you can check that the absurdity is genuine.
It was also a speed experiment: built end to end in one afternoon with an AI pair. The lesson from that half was less about the tooling and more about scope — an afternoon is enough when the thing only has one job.