Abstract

The present study sheds light on a new psychological construct called “AI pessimism aversion” (AIPA) describing an overly optimistic view of the benefits of AI by neglecting its potential dangers. In an international sample of N = 543 participants, we observed that the construct of AIPA strongly overlaps with single-item measures for positive and negative AI attitudes. Furthermore, the structural equation model suggests that the positive association between the internal locus of control (a personality measure describing persons who see themselves as steering their own lives) and positive AI attitudes was in part mediated by AIPA. We also assessed risk aversion in the present work, because high aversion of risky situations might go along with lower AIPA (the coming AI wave is characterised by high uncertainty). Surprisingly, this was not the case, and these two constructs seem to be (rather) independent. The present work should be a starting point to dive deeper into the study of AI attitudes, a highly relevant research theme for societies standing at the eve of the coming AI wave.