While generative artificial intelligence (GAI) has shown promise in various domains, it faces challenges such as knowledge deficits and hallucination issues. To tackle these problems, we propose a multi-AI collaboration system. Our research employs a mixed-methods approach, combining Design Science Research (DSR) with quantitative and qualitative methods. We investigate the design principles for efficient multi-AI collaboration systems and their impact on team trust compared to single AI collaboration. The results indicate that multi-AI collaboration systems significantly enhance users' trust in AI, particularly in the perception of capability. Furthermore, a number of key design principles for multi-AI collaboration systems were identified. These findings contribute to the development of more effective and trustworthy AI collaboration systems, paving the way for improved human-AI teamwork.
목차
Abstract Introduction Methods Findings and Future Directions Acknowledgments References
저자
Ang Zeng [ School of Information, Renmin University of China ]
Xusen Cheng [ School of Information, Renmin University of China ]
Corresponding Author
Kanliang Wang [ School of business, Renmin University of China ]