Agentic data science with the Data Agent Kit
cloud.google.com·7h ago
AI Research
arXiv CS.AI·Aug 26·5 min read
Rules Before Oracles: Auditable, User-Configurable Argument Selection for Deliberative Polling
TL;DR
A new framework for deliberative polling allows voters to configure argument selection rules based on transparent criteria, enhancing accountability and user agency. The proposed method outperforms traditional opaque ranking systems in terms of argument coverage and endorsement mass.
✦ Why It Matters
Implement user-configurable argument selection to enhance transparency in your polling or decision-making processes.
Key Takeaways
- Adopt user-configurable rules for argument selection in polling.
- Utilize transparent criteria to enhance voter accountability.
- Implement simulations to evaluate argument effectiveness.
- Consider weight functions to improve completeness in selections.
How It Works
The proposed framework allows voters to define parameters for argument selection based on publicly available evidence. It evaluates argument sets using criteria such as reason coverage, order of arrival, and endorsement mass.
An agentic simulator records every slate at each vote, ensuring transparency and enabling voters to understand how their choices are influenced.