Designing the dash in Human–AI Collaboration: A Redesign Probe for Working with Tensions in Care Documentation Practices
TL;DR
An empirical account of how professional judgment is negotiated through AI-generated representations is contributed and a redesign probe is introduced as a way to engage with AI integration in practice.
Abstract
Designing Human–AI collaboration in healthcare requires understanding how generative AI (GenAI) reshapes how care is represented and made accountable through documentation. Instead of treating GenAI as a tool for automation, this paper examines how its introduction reconfigures care documentation practices by reshaping judgment and responsibility in nursing work. We report on a speculative co-design workshop with eight hospital nurses, where GenAI was used as a redesign probe to explore how documentation tasks might be distributed across four roles: human, system, shared, and invisible agent. Through AI-mediated scenario reconfigurations, nurses articulated which forms of judgment and relational work could be supported by systems and which remained inseparable from human practice. AI-generated scenarios were often operationally plausible yet professionally misaligned, surfacing tensions around responsibility, visibility, and trust. The paper contributes an empirical account of how professional judgment is negotiated through AI-generated representations and introduces a redesign probe as a way to engage with AI integration in practice.
