AI able to establish a person’s political affiliation based on photography with liberal discoveries facing the camera, while conservatives look disgusted
- Stanford experts have built an AI capable of guessing political affiliation through a photograph
- He was trained with over a million images on dating sites and Facebook
- AI focused on the orientation of the head and facial expressions when guessing
- It has been found that most liberals look at the camera, while conservatives seem disgusted
Stanford research that made headlines in 2017 for designing an AI that uses “facial cues” to determine a person’s sexual preference has returned to what could be another controversial system.
Dr. Michal Kosinski claims that he has a facial recognition algorithm capable of identifying whether a person is liberal or conservative based on a single photograph – and with an accuracy of over 70%.
The technology, which is based on AI 2017, was trained with more than a million images from dating sites and Facebook and was programmed to focus on expression and posture.
Although Kosinski and his team were unable to identify the exact characteristics of the algorithm associated with a political preference, they did find in the images some trends such as head orientation and emotional expression.
Some examples include people who looked directly at the camera being labeled liberals, and those who looked disgusted were considered more conservative.
Scroll down to view
The technology was trained with more than a million images from dating sites and Facebook and was programmed to focus on expressions and posture. The machine learning system harvests and resizes the face to reduce the capture of non-facial features
The study, published in Nature, states that when people are asked to distinguish between two faces – one conservative and one liberal – they are correct about 55% of the time.
“Because people miss or misinterpret some of the cues, their low accuracy is not necessarily the limit of what algorithms could achieve,” the study said.
“Algorithms excel at recognizing patterns in huge data sets that no human could ever process and surpass us more and more in visual tasks, from skin cancer diagnosis to facial recognition to face-to-face judgments of intimate attributes, such as sexual orientation (76% vs. 56%) 7, personality (64% vs. 57%; derived from Pearson’s RS) and – as shown here – political orientation. ‘
The researchers used a sample of 1,085,795 participants from the US, Canada and the UK, along with their political orientation, age and gender.
Stanford research that made headlines in 2017 for designing an AI that uses “facial cues” to determine a person’s sexual preference (pictured) has returned with what could be another controversial system.
The study notes that its ethnic diversity included more than 347,000 non-white participants.
The machine learning system harvests and resizes the face to reduce the capture of non-facial features.
When it came to identifying images in the US, AI was 72% accurate.
Similar accuracy was observed in the Canadian sample, 71%, and in the United Kingdom, with 70%.
The researchers used a sample of 1,085,795 participants from the US, Canada and the UK, along with their political orientation, age and gender. When it came to identifying images in the US, AI was 72% accurate. Similar accuracy was observed in the Canadian sample, 71%, and in the United Kingdom with 70%
The highest predictive power was provided by head orientation (58%), followed by emotional expression (57%).
Liberals tended to confront the camera more directly, were more likely to express surprise, and were less likely to express disgust – those with a look of disgust were labeled conservatives.
“In other words, a single facial image reveals more about a person’s political orientation than their answers to a rather long personality questionnaire, including many seemingly political-oriented elements (e.g.,” I treat all people in the same way. equal ”or“ I think that a lot of tax money also supports artists ”), the study shows.