AI Meetup #001: Understanding and Simulating Agentic Behaviour

The first edition of a new Paris AI community meetup, and the theme is the one most agent builders are stuck on: how agentic behaviour actually works and how to simulate it, rather than how to wire up another harness. Three short technical talks with live demos — Anis Ayari on the three principles of building a real-world simulation with AI NPCs, Jonathan on what sits behind the Pokemon benchmark used at OpenAI, and a third slot still to be confirmed — then an hour and a half of networking. Approval-gated and already oversubscribed; the exact address in the 2nd arrondissement is sent on acceptance. Free, run in French or English depending on the room. Why go: a first edition draws the people who go looking for rooms like this, and the talks are engineering detail rather than strategy. Best for people building agents and evaluating them.