Searching for Synergy in Shared Workspace Human-AI Collaboration

在共享工作空间中探索人机协作的协同效应

Abstract: Automated AI agents are increasingly capable, yet many scientific and professional tasks require human judgment and contextual expertise. We study shared-workspace human-AI teams, where AI agents and human collaborators must coordinate responsibilities before submitting a final answer. 摘要： 自动化 AI 智能体的能力日益增强，但许多科学和专业任务仍需要人类的判断力和情境专业知识。我们研究了共享工作空间中的人机协作团队，在这种模式下，AI 智能体与人类协作伙伴必须在提交最终答案之前协调各自的职责。

Using the Collaborative Gym environment with DiscoveryBench tasks, we examine when adding simulated human collaborators improves performance and when process loss turns additional collaborators into coordination overhead. Across 1,482 sessions, adding relevant collaborators can lower performance when teams lack structure to coordinate their contributions. 通过使用带有 DiscoveryBench 任务的 Collaborative Gym 环境，我们研究了在何种情况下增加模拟人类协作伙伴能提升性能，以及在何种情况下流程损耗会导致额外的协作伙伴反而成为协调负担。在 1,482 场实验中，我们发现当团队缺乏协调贡献的结构时，增加相关的协作伙伴反而可能降低整体表现。

We then evaluate scaffolding that combines shared group memory with simulated human-in-the-loop (HITL) gates, where selected actions require approval from a designated simulated participant. This scaffolding yields higher mean performance, most clearly in three-person teams, with clearer responsibility signals and stronger routing of expertise to team actions. 随后，我们评估了一种结合了共享群组记忆与模拟“人在回路”（HITL）门控机制的支架系统，其中选定的操作需要获得指定模拟参与者的批准。这种支架系统带来了更高的平均性能，在三人团队中表现尤为明显，并实现了更清晰的职责信号传递，以及更有效地将专业知识引导至团队行动中。

Overall, how human-AI teams coordinate and integrate expertise matters as much as the capability available to them. 总而言之，人机团队如何协调和整合专业知识，其重要性丝毫不亚于团队所拥有的能力本身。