AI can determine a person’s political affiliation based on their photo with 70% accuracy

Stanford’s research that made headlines in 2017 for designing an AI using “ facial cues ” to determine a person’s sexual preferences is back with what could be another controversial system.

Dr Michal Kosinski claims to have a facial recognition algorithm that can identify whether a person is liberal or conservative based on a single photo – and with over 70% accuracy.

The technology, which is powered by AI from 2017, was formed with over a million images from dating sites and Facebook and programmed to focus on expressions and posture.

Although Kosinski and his team were unable to pin down the exact characteristics of the algorithm associated with political preference, they did however find patterns such as head orientation and emotional expression in the images.

Some examples include people who looked directly at the camera were characterized as liberal and those who displayed disgust were viewed as more conservative.

Scroll down for vidoe

The technology was formed with over a million images from dating sites and Facebook and programmed to focus on expressions and posture. Machine learning system crops and resizes the face, to reduce the capture of non-facial features

The study, published in Nature, indicates that when humans are asked to distinguish between two faces – a conservative and a liberal – they are right about 55% of the time.

“ As humans can miss or misinterpret some of the cues, their low accuracy doesn’t necessarily represent the limit of what algorithms could achieve, ” the study reads.

“ The algorithms excel at recognizing patterns in huge datasets that no human could ever process, and increasingly outperform us in visual tasks ranging from skin cancer diagnosis to facial recognition in passing through judgments on the face of intimate attributes, such as sexual orientation (76% vs. 56%) 7, personality (64% vs. 57%; derived from Pearson’s r) and – as shown here – orientation Politics “.

The researchers used a sample of 1,085,795 participants from the United States, Canada and the United Kingdom, along with their self-reported political orientation, age and gender.

Headline-grabbing Stanford research in 2017 for designing an AI that uses ‘facial cues’ to determine a person’s sexual preferences (pictured) is back with what could be another controversial system

The study notes that the ethnic diversity of the same included more than 347,000 non-white participants.

The machine learning system crops and resizes the face, to reduce the capture of non-facial features.

When it comes to identifying American images, the AI ​​was 72% accurate.

Similar precision was observed in the sample from Canada, 71 percent, and the UK with 70 percent.

The researchers used a sample of 1,085,795 participants from the United States, Canada and the United Kingdom, along with their self-reported political orientation, age and gender. When it comes to identifying American images, the AI ​​was 72% accurate. Similar precision was observed in the sample from Canada, 71%, and the UK with 70%

The strongest predictive power was head orientation (58%), followed by emotional expression (57%).

Liberals tended to face the camera more directly, were more likely to express surprise, and less likely to express disgust – those who looked disgusted were referred to as conservatives.

“ In other words, a single facial image reveals more about a person’s political orientation than their responses to a fairly lengthy personality questionnaire, which included many elements ostensibly related to political orientation (e.g. “ I treat all people the same ” or “ I also believe that a large part of the tax money goes to support artists ”), the study reads.