Saumure Roger, De Freitas Julian, Puntoni Stefano
Department of Marketing, The Wharton School, University of Pennsylvania, Philadelphia, PA, USA.
Department of Marketing, Harvard Business School, Harvard University, Boston, MA, USA.
Sci Rep. 2025 Jan 8;15(1):1326. doi: 10.1038/s41598-024-83384-6.
A preregistered audit of 600 images by generative AI across 150 different prompts explores the link between humor and discrimination in consumer-facing AI solutions. When ChatGPT updates images to make them "funnier", the prevalence of stereotyped groups changes. While stereotyped groups for politically sensitive traits (i.e., race and gender) are less likely to be represented after making an image funnier, stereotyped groups for less politically sensitive traits (i.e., older, visually impaired, and people with high body weight groups) are more likely to be represented.
一项对生成式人工智能生成的600张图像进行的预先注册审核，涵盖150个不同的提示词，探讨了面向消费者的人工智能解决方案中幽默与歧视之间的联系。当ChatGPT更新图像使其“更有趣”时，刻板印象群体的比例会发生变化。在使图像变得更有趣之后，政治敏感特征（即种族和性别）的刻板印象群体出现的可能性降低，而政治敏感程度较低的特征（即老年人、视力障碍者和高体重人群）的刻板印象群体出现的可能性增加。
