"Researchers have analysed hundreds of thousands of images from places like IMDb and Google Image Search as well as the kinds of online texts used to train large language models. They found a persistent age gap between men and women when it comes to how they are shown and talked about online."
"This age gap, where women are stereotyped as being generally younger than men, could be impacting both the gender pay gap and the way ChatGPT ranks resumes."
Hundreds of thousands of images from IMDb, Google Image Search and textual corpora used to train large language models were analysed. A persistent age gap appears across visual and textual online materials, with women consistently portrayed and described as younger than men. Visual cues and language both reinforce a stereotype of female youthfulness. These representational patterns can influence material outcomes by contributing to the gender pay gap and by biasing algorithmic systems that evaluate candidates, such as resume-ranking processes used by models like ChatGPT. The age gap is evident across multiple data sources and modalities.
Read at Nature
Unable to calculate read time
Collection
[
|
...
]