Human-AI Interaction Patterns in Creative Domains and Their Time-Based Visualization

Titelangaben

Leger, Rebecca ; Buschek, Daniel:
Human-AI Interaction Patterns in Creative Domains and Their Time-Based Visualization.
In: Proceedings of the Mensch Und Computer 2025. - New York, NY, USA : Association for Computing Machinery , 2025 . - S. 422-433
ISBN 979-8-4007-1582-2
DOI: https://doi.org/10.1145/3743049.3743081

Volltext

Link zum Volltext (externe URL):

Angaben zu Projekten

Abstract

The ongoing expansion of AI systems in creative domains brings new dynamics of human-AI interaction that have been difficult to describe and categorize, highlighting the need for new approaches to discussing human-AI interaction. This paper introduces an easy-to-use time-based visualization method to better understand and communicate human-AI interaction dynamics. We applied this method to our dataset of 337 AI systems in creative contexts, in particular music and writing. By clustering the visualizations, we identified seven recurring patterns of human-AI interaction, each representing distinct action and communication dynamics. This contribution offers a shared vocabulary for designers, engineers, and researchers, aiding in the description and categorization of human-AI interaction systems. In a broad view, we contribute deeper insights into the components and types of human-AI interaction, inspiring future design of human-centered AI systems in creative domains with the objective of catering to the diverse requirements of user groups and their respective workflows and demands.