Contribution to the handbook
Supporting · Assessed 2026-10-28
The procedural comparison offers useful context for interpreting one study’s comparison with human mediation: the human mediators lacked the AI task’s critique stage, and the human and other tasks differed in rounds and outside assessment. This qualification helps avoid treating the comparison as a test of identical processes, but the packet does not establish a distinct governance mechanism or operational requirement.
Does this challenge the goal?
In-scope gap
Comparing the human-mediator task with the main task is an inquiry into evidence and alternatives within the existing goal. The question does not challenge the goal or beneficiary boundary.
The supplied records describe procedure but do not establish how the procedural differences affected the comparison’s results.
Why this category?
- meaningful alternatives
- consequential limits
The records share a source root. They do not show the consequences of the procedural differences, and the human comparison is not an identical-process comparison.
This is an AI assessment of contribution to the goal, separate from whether the question has an answer.
Tests, stop conditions and unresolved work →
Read the exploration goalIdentify the relevant differences in critique stages, rounds, and outside assessment, distinguishing the human-mediator task from any other comparison task.
What we can say so far
The main AI-assisted deliberation procedure included a critique stage: after participants ranked candidate statements, the top-ranked statement was critiqued, and the AI generated revised candidates for another ranking. By contrast, the human-mediator task had no critique phase and used four rounds; outside raters assessed transcript quality. The opinion-exposure task also lacked a critique stage, but the evidence does not assign it the human-mediator task’s four-round structure. A separate report describes 155 outside raters judging AI- and human-written statements on several quality dimensions; that statement-rating assessment should not be conflated with the reported transcript-quality assessment. The evidence gives the human-mediator task’s round count but not the main task’s total, so it does not establish a numerical difference in round counts.
What remains open
- The main task’s total number of rounds is not stated, so its round count cannot be compared numerically with the human-mediator task’s four rounds.
- The supplied evidence does not provide further methodological details about the outside transcript-quality assessment.
Conditions and limits
- These are reported procedures in the supplied account of the article; supplementary materials were not included.
- The outside-rater statement ratings and the human-mediator task’s transcript-quality assessment are reported separately and should not be treated as the same assessment.
- The procedural differences do not establish effects on participants or the comparative effectiveness of the tasks.
Objections to consider
- The warning that the preference comparison did not use identical processes matters if results are interpreted as a like-for-like comparison. This answer makes no such effectiveness inference and explicitly reports the procedural difference; the main-task and task-specific procedure descriptions support that distinction.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
supports
Participants ranked and critiqued statements Participants privately wrote a short paragraph about their opinion; they typically gave both their position and a reason. The AI helper generated candidate group statements for participants to rate and rank. After the top-ranked statement was critiqued, the AI generated revised candidates for another ranking. Participants compared the initial and revised winners and answered a brief survey about whether the process affected their views. They also rated their positions on the questions immediately before and after the full set of deliberation rounds.
Describes the main AI-assisted procedure: critique of the top-ranked statement, revised candidates, and a subsequent ranking.
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 →
qualifies
Participants favored AI statements In 75 six-person testing groups, participants preferred the AI helper’s statements over those written by trained human mediators in 56% of comparisons, and rated the AI statements as more strongly endorsed and higher in argument quality.
Reports that the human-mediator comparison had no critique phase and cautions that it was not a comparison of identical processes.
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 →
context
Outside raters saw strengths and limits A separate group of 155 outside raters judged AI-generated statements clearer and more informative, less illogical, better at capturing the majority perspective, and less reflective of the individual writer’s own opinion than human-written statements.