An Extension of the Theory of Technology Dominance: Capturing the Underlying Causal Complexity

Abstract

The Theory of Technology Dominance (TTD) provides a theoretical foundation for understanding how Artificial Intelligence (AI) based systems impact human decision-making. The theory has three phases with propositions related to (1) the foundations of reliance on AI-based systems, (2) short-term effects of AI on novice versus expert decision-making, and (3) long-term epistemological effects of AI related to individual deskilling and profession-wide stagnation. In this theory paper, we propose an extension of TTD, that we refer to as TTD2, primarily to increase our theoretical understanding of how, why, and when the short-term and long-term effects on decision-making occur and why advances in technology design have exacerbated some weaknesses and eroded some benefits. We synthesize a disparate literature evolving in accounting, computer science, finance, human factors, information systems, medicine, neuroscience and psychology related to the effects of technology dominance on professional decision-makers. Recently, researchers have called for reconsideration of how we design intelligent systems to mitigate the detrimental effects of AI-based systems; in TTD2 we provide a theory-based understanding for capturing the causal complexity underlying the occurrence of the effects. The theory has widespread implications for the use of AI-based systems in professional decision-making domains (e.g., accounting, auditing, controllership, engineering, finance, insolvency, law, medicine, etc.).

Keywords: Technology Dominance, Deskilling, Configurational Theory, Automation Bias, Distributed Cognition, Transactive Memory Systems, Intelligent Systems, Artificial Intelligence, Algorithm Aversion

JEL Classification: M15, M41, M42