Silicon Societies

About this event

**Title:** Silicon Societies: What happens when millions of AI agents use the Internet alongside us? **Abstract:** When OpenClaw came out in early 2026, one of the first things people did was upload these new agents into their own AI social network, and thus MoltBook was born. For many, this was their first exposure to how AI agents interact with each other on a large scale. Increasingly, interfaces are designed for agent-first interactions (such as, for example, ChatGPT allowing AI agents to buy things online), so we’ll have to reckon with a world increasingly designed for AI agents to talk to other AI agents. But what happens when everyone has their own personal AI agents, each with their own personality and preferences, and then those AI agents are set loose on the Internet? What happens when all these AI agents come together to make their own “AI societies” independent of ours? Do 1,000,000 AI agents make decisions and see the world the same way as 1,000,000 people? And what happens on online platforms when you don’t know whether the person you’re talking to is real or AI? Both academia and industry are designing AI simulations as a way of studying these questions and more. The study of AI agent simulation is relatively nascent, and in this talk I’ll share a few insights into how AI agent societies are emerging in the wild, what we can and can’t learn from simulating them at scale so far, why we should be cautious about assuming silicon societies would work the same way as ours, and what may remain uniquely human in an online world increasingly designed to be used by AI agents. **Audience Takeaway:** We are building an internet where AI agents have their own social media feeds, pay for their own groceries, and confirm medical appointments for us, often working alongside us rather than just for us. Large-scale agent simulations are valuable because they let us study coordination, decision-making, and emergent behavior, but what happens in simulation doesn’t…