AI Humanizers, Pragmatists, Skeptics: A Cluster Analysis of Normative Attitudes for AI’s Capabilities and Roles

Abstract

This study explores people’s attitudes about artificial intelligence’s (AI) capabilities through a normative lens. Drawing from literature on phenomenological, ontological, and heuristic judgments about machines, we aim to better understand what capabilities people think AI should be able to have and how that relates to enthusiasm about having AI take certain roles in their lives. Survey participants (N = 601) were grouped via two-step clustering based on their attitudes, revealing three clusters: AI humanizers, who expect empathy and social interaction from AI; AI pragmatists, who expect cognitive capabilities; and AI skeptics, who prefer AI with limited human-like traits and emphasize task-oriented roles. Further analyses showed that AI cluster membership, along with AI threat perceptions, had a significant influence on enthusiasm for AI across different roles.

Ähnliche Arbeiten