Paper Number
1979
Paper Type
Short Paper
Abstract
Generative AI challenges the boundaries of what machines can do and will lead to new configurations of human work. Despite its benefits, current AI models can produce errors. Therefore, when users delegate parts of their work to generative AI, they must continue taking responsibility for the overall outcome. We present a lab-in-the-field experiment with three different scenarios where users either delegate work to an AI and implement changes afterward (“AI-human pattern”), where users create a draft and delegate it to an AI for improvement (“human-AI pattern”), or where humans work alone (“control”). Particularly, our findings point toward an interesting antagonistic effect: Working with AI increases outcome quality, however, individuals also perceive less responsibility for their work. The chosen interaction pattern therefore is pivotal to the success of the collaboration between humans and AI. The findings of this research-in-progress constitute the starting point for more effective design of work configurations and the future of work.
Recommended Citation
Henkenjohann, Richard and Trenz, Manuel, "Challenges in Collaboration with Generative AI: Interaction Patterns, Outcome Quality and Perceived Responsibility" (2024). ECIS 2024 Proceedings. 3.
https://aisel.aisnet.org/ecis2024/track05_fow/track05_fow/3
Challenges in Collaboration with Generative AI: Interaction Patterns, Outcome Quality and Perceived Responsibility
Generative AI challenges the boundaries of what machines can do and will lead to new configurations of human work. Despite its benefits, current AI models can produce errors. Therefore, when users delegate parts of their work to generative AI, they must continue taking responsibility for the overall outcome. We present a lab-in-the-field experiment with three different scenarios where users either delegate work to an AI and implement changes afterward (“AI-human pattern”), where users create a draft and delegate it to an AI for improvement (“human-AI pattern”), or where humans work alone (“control”). Particularly, our findings point toward an interesting antagonistic effect: Working with AI increases outcome quality, however, individuals also perceive less responsibility for their work. The chosen interaction pattern therefore is pivotal to the success of the collaboration between humans and AI. The findings of this research-in-progress constitute the starting point for more effective design of work configurations and the future of work.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.