A study conducted in Denmark has found that artificial intelligence (AI) algorithms can accurately predict a person’s political ideology based on their facial characteristics. The study discovered that right-wing politicians tend to have happy facial expressions in photos, while those with neutral expressions are more likely to identify as left-wing. By analyzing a single photo of a person, AI can predict political ideology with a 61% accuracy rate, using deep learning, a method that teaches computers to process information like humans do. This groundbreaking research was detailed in a paper published in Scientific Reports.
The researchers aimed to determine which information contributes to the success of these predictive techniques. Humans have the ability to make immediate judgments about personality, intelligence, and even political ideology based on a person’s face. Study author Stig Hebbelstrup Rye Rasmussen and his colleagues at Arhus University explored whether computational neural networks, which mimic the structure and function of the human brain, can make similar predictions based solely on a photo.
To train the neural network, the scientists used thousands of photos of politicians from Denmark’s 2017 municipal elections, specifically choosing candidates who were explicitly left or right-wing, of European ethnic origin, and without beards. The photos depicted only the candidates’ facial features, without any backgrounds that could impact the predictions. The final dataset included 4,647 photos of political candidates, with 1,442 female politicians.
Microsoft’s facial expression recognition technology was used to measure the emotions seen in the photos, along with other algorithms that evaluated attractiveness and masculinity. The researchers also tested the accuracy of the algorithm using a few photos of Danish parliamentarians.
Ultimately, the study found that the AI trained on this data could predict ideology with a 61% accuracy rate, surpassing chance predictions. The researchers emphasized the threats to privacy posed by deep learning approaches, noting that using publicly available data, they were able to predict a person’s ideology 60% of the time. The research also revealed that female politicians who were considered more attractive were more likely to be conservative, while attractiveness and masculinity did not correlate with political ideology for male politicians. Additionally, right-wing politicians were more likely to have happier facial expressions, while neutral expressions indicated left-wing party affiliations. The study also suggested that contempt displayed on a person’s face was more common among left-leaning women.
Overall, this study provides insight into the use of AI algorithms to predict political ideology based on facial characteristics. It highlights the connection between predicted ideology and identifiable features of the face, such as attractiveness. The privacy implications of this research are significant, as it demonstrates the potential of deep learning algorithms to reveal personal information.
Denial of responsibility! VigourTimes is an automatic aggregator of Global media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, and all materials to their authors. For any complaint, please reach us at – [email protected]. We will take necessary action within 24 hours.
