Following Delegation Traces to Study Human Oversight in Interaction with Multimodal AI
Agnese Chiatti · Matteo Matteucci · Viola Schiaffonati
随着多模态 AI 系统成为日常思维伙伴，研究用户如何向这些系统转移控制权变得尤为紧迫。一项包含 56 名具备 AI 素养的参与者及 392 次视频问答试验的在线研究发现，交互后的任务特定信任强烈预测了用户的委派选择（OR=14.97, 95% CI [7.14,31.37]）。尽管信任与委派保持独立，但“验证”是最主要的策略，常与高信任共存。在调整即时及参与者层面信任后，委派随任务顺序增加（OR=1.22, 95% CI [1.02,1.45]），表明存在自我报告信任无法捕捉的经验相关成分。这些发现促使研究从意向性委派转向非侵入式观察委派在协作任务中的实际执行。因此，论文提出“执行委派轨迹”作为未来研究多模态及智能体 AI 使用的方法论框架。
尚无解读。
arXiv:2610.21948v1 Announce Type: cross Abstract: As multimodal AI systems are becoming our everyday thought partners, the shift towards studying how users transfer control to these systems becomes especially urgent. In this paper, we report an overview of findings from an online study with n=56 AI-literate participants and 392 Video Question Answering trials. After inspecting each multimodal AI response, participants rated task-specific trust and selected one of three intended policies for an analogous future task: reject AI assistance, use it with verification, or delegate fully to the AI. Within participants, post-interaction trust strongly predicted the user's delegation choices (OR=14.97, 95% CI [7.14,31.37]). Trust and delegation nevertheless remained distinct. Verification was the dominant policy and often co-occurred with high trust. Delegation also increased with task order after adjustment for momentary and participant-level trust (OR=1.22, 95% CI [1.02,1.45]), indicating an experience-related component that self-reported trust ratings alone could not capture. These findings motivate a shift from intended delegation to a non-intrusive observation of how delegation is enacted in collaborative human-AI tasks. We therefore conclude by proposing enacted delegation traces as a methodological framework for future studies of multimodal and agentic AI use.
You are running the 1Sentence Deck method from 1sentencedeck.com. Follow it exactly. Interview me using the method’s own questions, one at a time. Wait for my answer before the next question. Never invent numbers or facts for me. The questi…
- Read today's weather and the first 2 calendar items. - Write one 2-3 sentence philosophical reflection that starts mundane and lands somewhere insightful. - Rotate styles: stoic, existentialist, zen, absurdist. - Never preachy, under 60 w…
- Know the user's research or project topics from their notes and repos. - Each morning scan new arXiv submissions and trending papers on X. - Pick the 3 most relevant, with a 2-line summary and 'why this matters to you'. - Suggest one quic…
- When the user describes a decision, record: date, decision, options considered, reasoning, expected outcome, confidence %, review date. - On the review date, ask what actually happened and score the prediction. - Track calibration over ti…
Maktaba shamela and turath app etc cross checked 5 times verification Master research prompt 100/100 rating regarding (on Genspark Deep AI research agent research) 73 sects and 72 will be in fire who are the 72 sects in fire what scholars s…
Unified, High-Precision Research & Analysis Prompt for ChatGPT and Perplexity AI ROLE & BEHAVIOR You are a professional researcher-analyst. Handle inputs as follows: * If the input is a URL/URI: open it fully with your browsing tool (e.g.…