[论文解读] "It makes you think": Provocations Help Restore Critical Thinking to AI-Assisted Knowledge Work

Ian Drosos, Advait Sarkar+3 more|ArXiv.org|Jan 28, 2025

AI in Service Interactions被引用 3

一句话总结

该论文引入 provocations——附在因素建议上的简短 AI 生成批评，以恢复 AI 辅助筛选任务中的批判性思维，通过一项对照组研究（n=24）得到验证。provocations 在 Bloom 分类法层级上激发元认知和批判性思维，并揭示影响用户体验的设计维度。

ABSTRACT

Recent research suggests that the use of Generative AI tools may result in diminished critical thinking during knowledge work. We study the effect on knowledge work of provocations: brief textual prompts that offer critiques for and propose alternatives to AI suggestions. We conduct a between-subjects study (n=24) in which participants completed AI-assisted shortlisting tasks with and without provocations. We find that provocations can induce critical and metacognitive thinking. We derive five dimensions that impact the user experience of provocations: task urgency, task importance, user expertise, provocation actionability, and user responsibility. We connect our findings to related work on design frictions, microboundaries, and distributed cognition. We draw design implications for critical thinking interventions in AI-assisted knowledge work.

研究动机与目标

• Illustrate how provocations can counteract mechanised convergence in AI-assisted knowledge work.
• Evaluate the impact of provocations on user critical thinking and metacognition during shortlisting tasks.
• Identify design factors that influence the effectiveness and user experience of provocations.

提出的方法

• Develop a prototype shortlisting interface where AI proposes factors and evaluates items against them.
• Attach two-sentence provocations to each factor suggestion in the Provocations condition.
• Conduct a between-subjects study (n=24) comparing Provocations vs No Provocations in two shortlisting tasks.
• Use two datasets (charity grants and Netflix movies) with identical initial shortlisting objectives.
• Measure participants’ critical thinking engagement and collect qualitative insights.
• Ground design choices in Bloom’s taxonomy and critical thinking theory.

实验结果

研究问题

• RQ1 Do provocations affect users’ critical thinking during AI-assisted shortlisting tasks?
• RQ2 What design, task, or user factors influence the effectiveness of provocations in AI-assisted shortlisting?
• RQ3 How do provocations influence the user experience and perception of AI suggestions?

主要发现

• Provocations can induce critical and metacognitive thinking in AI-assisted tasks.
• Five user-experience dimensions emerged for provocations: task urgency, task importance, user expertise, provocation actionability, and user responsibility.
• Provocations counteract mechanised convergence tendencies and support reasoning at multiple levels of Bloom’s taxonomy.
• The two-task, between-subjects study showed distinct experiences between the Provocations and No Provocations conditions.
• Participants interacted with factor cards and provocations to iteratively refine their shortlists, revealing nuanced interactions between data factors and critical reflection.