Contribution to the handbook
Priority · Assessed 2026-10-28
The comparison bears directly on how much authority, if any, citizens may be willing to delegate and whether expert judgment remains preferred. This is material to comparing governance arrangements, while the evidence must remain qualified as stated willingness rather than endorsement or proven legitimacy (q-live-5b99b8b6b2cb419b0fe26939; s-hidalgo-augmented-democracy-2a2c70e42a4dc107; s-hidalgo-augmented-democracy-f5df4161df628d34).
Does this challenge the goal?
In-scope gap
The question examines public willingness across task types and degrees of delegation, an unresolved authority and citizen-service issue within the existing definition; it does not contest the goal.
The supplied study summaries leave task details and sample context unclear, so they do not establish broad public preferences or support for AI control (q-live-5b99b8b6b2cb419b0fe26939; s-hidalgo-augmented-democracy-2a2c70e42a4dc107; s-hidalgo-augmented-democracy-f5df4161df628d34).
Why this category?
- Compares different degrees of AI influence and delegation with expert judgment, informing a consequential authority choice.
- Makes task-specific variation an empirical question rather than assuming a general public preference.
The reported preference for experts and the limited-influence survey result are secondhand study summaries with important context missing. Neither demonstrates actual use, public endorsement of AI rule, or how safeguards affect willingness.
This is an AI assessment of contribution to the goal, separate from whether the question has an answer.
- Does working with AI help people understand and shape decisions, or can it replace their judgment?
- Do people want AI to make technical decisions?
- How does public willingness to give AI influence over technical decisions vary across task types and degrees of delegation, and how does it compare with preference for expert judgment?
- Experts remain preferred
- Bias may differ across tasks
- Some support for limited AI influence
Authority, accountability and resilience →
Tests, stop conditions and unresolved work →
Read the exploration goalDetermine whether willingness differs across specified types of technical decisions, identifying the tasks, participants, measures, and limits of any comparison rather than assuming variation.
What we can say so far
The material reports that a cited study found people more willing to give AI some influence on technical tasks, while still preferring expert judgment. It does not establish whether willingness differs among types of technical decisions: the tasks, participants, measures, and task-specific comparisons are not provided. The evidence therefore supports only a broad comparison between openness to some AI influence on technical tasks and preference for experts, not a comparison across technical task types.
What remains open
- Which technical decisions or task types were studied.
- Who participated and how participants were selected.
- How willingness and degree of AI influence or delegation were measured.
- Whether willingness differed across task types and the basis for any such comparison.
Conditions and limits
- The comparison is attributed to a study cited by the authors, not presented as independently verified evidence.
- The source is based on the full readable text of the retrieved preprint; figures were not independently interpreted.
- The passage does not specify how technical tasks were defined or presented.
Objections to consider
- The passage does not identify the technical tasks or provide participant and measure details, so it cannot resolve whether willingness varies across task types.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
context
Bias may differ across tasks The authors distinguish bias in language tasks, such as assumptions linking jobs with gender, from bias in predicting civic preferences. They caution that a bias found in one task should not simply be assumed to operate in the other.
Distinguishes language-task bias from bias in predicting civic preferences and cautions against assuming findings transfer across tasks. It does not report public willingness by technical task type.
Large Language Models (LLMs) as Agents for Augmented Democracy
Read full source: Large Language Models (LLMs) as Agents for Augmented Democracy (opens in a new tab) ↗Compare willingness for AI assistance or partial influence with willingness to delegate final authority, and compare each with preference for expert judgment. Keep secondhand study reports distinct from direct participant accounts.
What we can say so far
The supplied material offers limited, secondhand evidence about willingness to allow AI partial influence. One cited study reportedly found that more than 60% of participants would accept a process assigning AI 25% of the influence and politicians 75%. Another cited study reportedly found greater willingness to give AI some influence on technical tasks, while people still preferred expert judgment. These reports suggest openness to limited AI influence does not displace preference for experts. They do not establish willingness to delegate final authority to AI, so the requested comparison between partial influence and final delegation cannot be completed from this evidence. Neither item is a direct participant account.
What remains open
- No supplied evidence compares willingness to accept AI assistance or partial influence directly with willingness to delegate final decision authority to AI.
- No direct participant accounts are supplied; the evidence consists of secondhand reports of studies.
- The evidence does not provide enough detail to assess how the technical-task finding varies across task types.
Conditions and limits
- Both findings are reported by the source authors from other studies; they are not findings from the authors’ own experiment or direct participant accounts.
- The 25% AI / 75% politician result concerns stated acceptance of a decision process, not actual use or support for AI control.
- The passage reporting greater openness to AI influence on technical tasks does not specify the tasks or clearly delimit the result’s scope.
- The cited study’s sample and broader applicability are not sufficiently described in the supplied passage.
Objections to consider
- The report does not specify the technical tasks or sufficiently clarify the finding’s scope, limiting the comparison across task types.
- The passage omits sample detail and reports stated acceptance of partial influence only; it does not establish broader applicability or willingness to delegate final authority.
Evidence and related ideas
Supporting evidence, qualifications and context play different roles in this answer.
qualifies
Experts remain preferred The same cited study found that people were more willing to give AI some influence on technical tasks, but still preferred to defer to experts.