Are AI-Synthesized Faces More Trustworthy?

The video is part of the Journal Club series, exploring a paper on AI-synthesized faces and their perceived trustworthiness compared to real faces. The paper conducts three experiments using Mechanical Turk participants to assess the ability to distinguish between real and synthetic images and to rate trustworthiness. Results show synthetic faces are rated more trustworthy, with gender and racial biases evident. The video discusses the implications of these findings, the demographics of participants, and raises questions about cultural biases. The paper's data is available on OSF and GitHub for further exploration.

Discuss the implications of the findings related to gender and race in trustworthiness ratings.

What demographic factors were considered in the study, and how might they affect the results?

What further research questions arise from the findings of this paper?