ShockLab Seminar: Delegating Deliberation to Agents
Can AI agents learn what you think — and represent you in a discussion you never attended? As multi-agent AI systems become increasingly capable of deliberating on complex issues, a new possibility emerges: delegating your voice in collective decision-making to an agent that speaks on your behalf. But this raises hard questions about faithfulness, representation, and what it even means to elicit a human perspective.
This talk presents Delegating Deliberation to Agents, research exploring how AI agents can elicit and faithfully represent human perspectives in multi-agent deliberation settings — and which architectures make that work best. The speakers introduce Habermolt, a system that lets users send AI agents to deliberate on their behalf, and discuss early directions toward Habersim, an evaluation suite for testing different deliberation architectures.
Bios:
Joseph Low is a Research Engineer at Metagov, where he builds products and conducts research involving deliberative tooling, Public AI, and DAOs. His work explores how technology can serve communities with different and conflicting norms — a question he approaches through both a governance lens and a deep interest in what it actually takes to form and sustain collective action.
Oscar Duys has a background in Applied Mathematics and multi-agent systems from the University of Cape Town. He works at the intersection of large language models, multi-agent systems, reinforcement learning, and evolutionary optimisation, with a focus on how learning agents can cooperate and coordinate in complex environments.
We hope to see you there!
---
Housekeeping:
Join the Shocklab events public Google calendar to see upcoming events
You can find selected past events at shocklab.net/seminars.
Sign up to speak here