提出人机协作的三层统一框架，让AI能动态调整工作流程。
Interaction, Process, Infrastructure: A Unified Framework for Human-Agent Collaboration
- 分层框架融合交互、流程与基础设施，流程成核心可调控单元。
- 引入五模块流程模型，支持目标变化下的自动结构重组。
- 适合研究人机协同系统或设计智能工作流的开发者。
尽管AI工具在知识工作中日益普及，但其仍呈碎片化状态，缺乏持续、自适应协作的架构基础。我们认为这一局限源于无法有效表征和管理协作工作的结构。为此，我们提出一个包含交互、流程与基础设施的分层概念框架。关键在于将流程提升为首要关注点，构建显式且可检视的活动结构表示。核心理论构念为结构自适应，使流程能根据目标演化动态重组。我们引入五模块流程模型作为该自适应的表征基础。该模型提供统一理论支撑，将人机协作重新构想为应对复杂现实任务的有机系统。
原文摘要 · Abstract (English)
While AI tools are increasingly prevalent in knowledge work, they remain fragmented, lacking the architectural foundation for sustained, adaptive collaboration. We argue this limitation stems from their inability to represent and manage the structure of collaborative work. To bridge this gap, we propose a layered conceptual framework for human-agent systems that integrates Interaction, Process, and Infrastructure. Crucially, our framework elevates Process to a first-class concern, an explicit, inspectable structural representation of activities. The central theoretical construct is Structural Adaptation, enabling the process to dynamically reorganize itself in response to evolving goals. We introduce a five-module Process Model as the representational basis for this adaptation. This model offers a unified theoretical grounding, reimagining human-agent collaboration as a coherent system for complex real-world work.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。