AI's Beauty Standards: A Look at Virtual Facial Preferences

AI has been trying to figure out what makes a face attractive. Researchers used five different AI models to create images of what they think are "ideal" and "attractive" faces. They then measured these virtual faces to see if they match up with old-school beauty standards. Most of the faces generated were female and Caucasian. The AI's idea of an attractive face had a smaller middle part and a bigger lower part compared to classical beauty standards. The sides of the face were also narrower in the AI-generated images.

The study found that AI's beauty standards are quite different from the old rules. This could be because the AI models are biased or because beauty standards are changing. It's also possible that the AI is just picking up on trends from the data it was trained on. The results show that AI might not always get it right when it comes to beauty. It's important to think about where these models get their information and if they're really capturing what people find attractive. After all, beauty is in the eye of the beholder, and AI might not have eyes at all.

Actions