Testing the Effectiveness of Using Psychology-Intuitive Information to Promote Responsible AI: Information About the Training Set Shapes Perceptions of AI Algorithms in an Experimental Setting

Abstract

Most of the tools currently available to promote responsible AI take a top-down approach, overlooking the role of intuitive perceptions of end users. However, emerging evidence shows that end users have preferences for specific algorithms, highlighting an alternative way to promote responsible AI. Nevertheless, the large number of elements impacting the performance and reliability of AI algorithms makes aligning end-users’ intuitions with the actual characteristics of these algorithms particularly challenging. In this paper, we propose building on decades of psychological research to identify which information that is important from a technical perspective can also be used intuitively by individuals. Through a set of five studies involving over 1,000 participants, we demonstrate that providing information about the quantity and quality of an algorithm’s training set can align end-users’ perceptions with the actual reliability of AI algorithms, thereby offering a new tool to promote responsible AI.