"ChatGPT, what am I missing?": Designing AI Workflows around Professional Task Structure to Shape Analytic AI Use

📅 2026-09-14
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
研究设计了两种AI工作流程以支持专业任务,通过实验比较发现结构化的用户导向工作流能更广泛地满足分析需求并降低主观努力。
📝 Abstract
General-purpose AI lets users choose what support to request, but leaves them to structure the support a professional task requires. We examine how interactive workflows can embed professional task structure without prescribing how users engage with AI. We designed two scaffolded interfaces around the same negotiation scaffold: one presented a completed AI analysis, while the other supported user-directed, incremental development. A four-condition randomized experiment with 800 participants compared these interfaces with no-AI and an AI chat interface. AI-supported conditions improved preparation coverage over unaided work; the scaffolded workflows further improved coverage over chat. Although the scaffolded workflows produced similar coverage, the user-directed workflow elicited a broader repertoire of analytic requests and lower subjective effort. Professional scaffolding therefore depends not only on displayed structure but on how workflows organize users' engagement with it. Effective professional AI must structure how users and AI build analysis together.
Problem

Research questions and friction points this paper is trying to address.

AI workflows
professional task structure
interactive workflows
scaffolded interfaces
Innovation

Methods, ideas, or system contributions that make the work stand out.

interactive workflows
professional task structure
user-directed development
scaffolded interfaces
🔎 Similar Papers
No similar papers found.
Zilin Ma
Zilin Ma
Harvard University
Human-computer interactionnegotiationmental healthlarge language model
S
Suzi Jazmati
Frontline Associates, Canada
M
Marco Chimenton
Independent Researcher, Italy
Yiyang Mei
Yiyang Mei
Emory University School of Law
J
Jacqueline Lane
Laboratory for Innovation Science at Harvard, United States
K
Krzysztof Z. Gajos
Harvard John A. Paulson School of Engineering and Applied Sciences, United States
Finale Doshi-Velez
Finale Doshi-Velez
Professor, Harvard
Machine LearningHealth