Why We’re Building a Dataset of Arab Futures

Today and tomorrow, data decides. What’s collected, indexed, and fed into our machines becomes the future’s source code. And most of what AI currently knows about the Arab world? It’s narrow, extracted, flattened — sometimes absent altogether.

At the Institute for Worldmaking, we are building a dataset that’s not just diverse — it’s deliberate. We’re integrating poetry, speculative fiction, oral histories, mapped materials, fieldwork with craftspeople, and more. Not to make the data colorful, but to make it real. We’re not trying to fix representation — we’re building a different foundation.

Our approach to AI starts with cultural alignment. That means teaching systems to hold complexity, contradiction, story, and context. It means asking not just what machines can do, but what they should preserve, learn from, and imagine alongside.

The Arab world has always been imagining, speculating, and building toward futures. Now we’re building the tools to record and recognize that work — to make it visible, legible, and more powerful.