The Empty Chair: Using LLMs to Raise Missing Perspectives in Policy Deliberations
2025-11-17
Summary
The article explores the use of large language models (LLMs) to introduce missing perspectives in policy deliberations by simulating input from absent stakeholders. A tool was developed to transcribe discussions in real time and generate relevant personas, which was tested in a student assembly on sustainability. While participants found the tool useful for sparking discussions and considering new perspectives, concerns were raised about the accuracy of LLMs in representing diverse groups, particularly underrepresented ones.
Why This Matters
Democratic deliberation often suffers from a lack of diverse representation, leading to group polarization and ineffective policymaking. By using LLMs to simulate missing perspectives, this technology holds potential to enrich discussions and broaden understanding in policy settings. However, the tool's effectiveness depends on recognizing its limitations and ensuring it complements rather than replaces genuine participation.
How You Can Use This Info
Professionals involved in policy-making or group discussions can consider integrating AI tools to surface overlooked perspectives, enhancing the depth and quality of deliberations. It's crucial to frame these tools as aids for sparking discussion rather than substitutes for real voices. Awareness of the tool's limitations in accurately representing minority views is essential to mitigate potential biases and ensure responsible use.