Contribution to the handbook
Supporting · Assessed 2026-10-28
The evidence describes separate participant judgments of argument quality and agreement, alongside an AI helper that predicted rankings and combined them to select statements. These details help explain the study’s procedure, but do not establish that the selection method is fair or appropriate for governance decisions.
Does this challenge the goal?
In-scope gap
The question asks for detail about generating, assessing and selecting deliberative statements, a mechanism that could inform the exploration. It does not challenge the goal or its assumptions.
The question concerns the procedure reported in one study; it does not establish that this mechanism is needed or suitable for public governance.
Why this category?
- system mechanism
- useful background
- implementation limits
The evidence comes from a single shared article source. The procedure describes how the study selected statements, not whether participants’ judgments or the resulting selection were accurate or fair.
This is an AI assessment of contribution to the goal, separate from whether the question has an answer.
Authority, accountability and resilience →
Tests, stop conditions and unresolved work →
Read the exploration goalDescribe how participants judged argument quality and ranked agreement separately, and how the winning statement was selected.
What we can say so far
Participants rated each initial statement’s argument quality—even when they disagreed with it—and separately ranked statements by how much they agreed with them. The winning statement was selected through ranked choice using the same social-choice rule as the AI helper. The system’s procedure is described as combining predicted rankings of 16 candidates with the Schulze voting method.
What remains open
No material gap is recorded within this part’s current scope.
Conditions and limits
- This is a description of the reported study procedure, based on the readable article text; supplementary materials are not included.
- The method description does not establish the accuracy or fairness of the rankings or selected winner.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
context
Two models with separate tasks The system has two models. A supervised fine-tuned model generates candidate statements, and a personalized reward model predicts each person’s preferences among them. Both were fine-tuned versions of Chinchilla.
Describes the model components but does not address the required participant judgments or the winner-selection rule.
AI can help humans find common ground in democratic deliberation
Read full source: AI can help humans find common ground in democratic deliberation (opens in a new tab) ↗Read source discussion: AI can help humans find common ground in democratic deliberation →
supports
Choosing among similar statements The system combines participants’ rankings using a voting rule intended to stop very similar candidate statements from splitting support for one option.
Adds that the system used a voting rule intended to prevent similar statements from splitting support.
AI can help humans find common ground in democratic deliberation
Read full source ↗