Examining Human and AI Interactions in AI-Based Application Systems—Realizing Local and Remote Experiments in Industry 4.0 Settings

Abstract More and more teams are working with artificial intelligence (AI) in a business context. However, the findings on the influence of continuously learning AI-based systems on humans are rudimentary. In particular, the quantified effect of a learning, unlearning, forgetting and amnesic AI on human workers has not yet been researched. This article therefore constructs an AI Interaction Instrument (A3I) that enables interaction modeling of learning and forgetting human workers with learning and forgetting AIs. The A3I is illustrated in initial experiments with a focus on AI trust, AI frustration and variations in AI reliability in an Industry 4.0 context. In a design-science-oriented manner, the A3I is developed as a research tool that enables the quantification of the influence of forgetful AIs on forgetful humans in experiments using empirical data analysis tools. The objective of intended studies refers to the exploratory examination of evolving AI on human forgetting and vice versa. The results demonstrate the technical functionality and requirement fulfillment of the A3I tool as well as its scalable experiment routines that allow experiments to be conducted with any number of probands. Although the data collected has not yet been analyzed, this study lays the foundation of AI as a tool for managing the learning and forgetting of human workers with regard to the needs of companies. The implication for practitioners and society is, if corresponding effects are statistically proven, that AI can be a tool to intentionally guide human forgetting by design.