Decision-Making Dynamics in Generative AI Environments

Abstract

This essay examines the shifting nature of decision-making in generative AI environments through a teaching-focused perspective, grounded in real classroom experiences at IIT Delhi. Incorporating interviews, classroom discussions, and foundational theories from consumer behavior and cognitive psychology, the authors explore how individuals interact with AI-generated content under constraints such as time, energy, and situational context. The analysis reveals how users alternate between satisficing" and "optimizing" strategies based on the stakes involved, cognitive effort required, and social factors. Drawing on Herbert Simon’s concept of bounded rationality, the study illustrates how AI systems can both expand decision complexity and improve efficiency. Student experiences with tools like ChatGPT bring to light common cognitive patterns—anchoring bias, trade-off reasoning, social influences, and mixed use of compensatory and non-compensatory decision strategies. These insights suggest that while AI reshapes the decision-making process, it cannot supplant human judgment. Recognizing these cognitive tendencies can help both users and designers foster more deliberate, context-aware interactions with AI tools.