The future belongs to whoever simulates it first. Today we launched X Research beta. Here's what we built: a simulation world. Upload your research — we construct a cast of AI agents, each a precise model of a real executive, institution, regulator, or market force. They run forward autonomously. They form alliances. They contest positions. Some collapse. You watch from above. God View. Give you the power of predicting everything before you act.

Framer — Launch websites with enterprise needs at startup speeds.

Launch websites with enterprise needs at startup speeds.

Every major decision I've tried to analyze — competitive moves, fundraising timing, market entries — runs into the same wall: you can research the facts, but you can't see how the humans involved will actually react to each other under pressure. I started building X because I kept thinking about two things: Dr. Strange scanning 14 million futures before Infinity War, and Westworld's premise that a world populated by autonomous agents eventually becomes indistinguishable from the real one. Both are fiction — but the underlying idea isn't. If you could simulate the humans in your decision, not just the data, you'd make fundamentally better calls. The hardest part wasn't the AI — it was giving agents a soul. Each one has a hidden agenda, private information, a communication style, and relationships that evolve round by round. They don't just respond to prompts. They scheme, form alliances, and sometimes surprise even us. If you run a simulation and something surprises you, I want to hear about it.

Framer — Launch websites with enterprise needs at startup speeds.

Launch websites with enterprise needs at startup speeds.

X — Simulate the World Before You Act

Run the future before you commit to it

Run the future before you commit to it