Theme of inquiry
How should agents manage information and memory for reliability?
A question within its area, explored through 4 lines of inquiry below — each a family of specific questions the research asks.
48 specific questions
- Can explainability and appropriate trust work against each other?
- How do confidence signals in AI outputs mislead human trust calibration?
- Why do users trust overconfident AI outputs even when accuracy drops?
- Why do AI-generated answers carry unearned authority in decision-making contexts?
- Can we measure appropriate trust levels in human-AI assistant relationships?
- Can trust in AI be formally parameterized and measured?
- Can organized response format trick users into overestimating AI reliability?
67 specific questions
- Can AI systems produce genuinely new validity claims without community participation?
- Why do stakeholders interpret the same explanation differently in practice?
- How can correct explanations coexist with failed applications in AI?
- Why is AI output fundamentally unverifiable against underlying reality?
- Can ethical constraints in AI address the gap between performance and actual understanding?
- Can artificial systems develop the authority to challenge expert claims?
- Do different AI models independently converge on the same social outputs?
66 specific questions
- How do live human evaluations differ from ground-truth benchmarks?
- How do satisfaction scores differ from genuine cognitive improvement?
- How should we evaluate AI systems we cannot directly observe?
- Why do users report satisfaction that diverges from actual cognitive clarity?
- Can cognitive governance help users interpret AI outputs better?
- Should explanation quality be measured by user satisfaction or behavior prediction?
- Should evaluations shift toward open-world messy tasks instead of contests?
47 specific questions
- Can polished presentation authority substitute for actual accuracy in AI outputs?
- Why is confidence a dangerous proxy for accuracy in human-AI interaction?
- Why does AI fluency create false impressions of expert judgment?
- How does AI presentation authority substitute for actual expert judgment?
- Why does polished explanation make wrong AI systems more persuasive than poorly explained ones?
- What structural evidence shows that polished presentation substitutes for actual thinking in AI output?
- Can users learn to discount fluency as a signal of their competence?