Knowledge of Artificial Intelligence Predicts Lower AI Receptivity

Abstract

As the field of artificial intelligence (AI) continues to progress and transform society, understanding the factors that influence receptivity towards AI has become increasingly important. Yet, little is known about how systematic differences across consumers predict AI receptivity. The current research investigates whether and how knowledge about AI influences consumers’ receptivity towards AI. To this end, we develop and validate an AI Literacy Test (AILT), a novel instrument designed to measure individuals’ objective knowledge about AI and algorithms. In contrast to three surveys documenting that most people expect greater AI knowledge to predict greater AI receptivity, the authors find the reverse; people with greater AI knowledge have a lower preference for using AI-based products and services. This reduction is not indiscriminate and is particularly pervasive for tasks that require more subjectivity to be performed well. However, greater knowledge of AI does not lead to increased AI utilization propensity among even highly objective tasks. These findings suggest that there may be unintended consequences of policymakers’ efforts to educate the public about AI, and that companies marketing AI product and services may need to re-evaluate which target segments may be more likely to adopt their technologies.