Would artificial intelligence be able to recognize our political opinions by reading our faces?

An AI could determine our sexual orientation or political opinion with a photo. Photo credit: Shutterstock / Trismegist san

Artificial intelligence might enable us to see the warning signs of Alzheimer’s disease on our faces, but it might well enable something completely different … If we are to believe Michal Kosinski, a Stanford-based artificial intelligence researcher, it might as well Determination used our sexual orientation or our political opinions …

Exposed by artificial intelligence! However, we thought that “at the top of the customer” judgments were no longer relevant … Should we choose to protect our privacy? Is artificial intelligence going too far? So many questions for now with no real answers!

These researchers could therefore demonstrate that a face examined by a device equipped with artificial intelligence provides information about our personal orientations. Michal Kosinski’s report, published in the very reputable journal Nature, tends to show that an algorithm can guess 72% if a Caucasian person is homosexual! And that simply by analyzing the facial features. But be careful, Michal Kosinski publishes this report in order not to use this algorithm, but to show possible deviations.

Other scientists, however, point out the dangerousness of this study … Because it comes back on the scene, a science that was thought to have been finally buried: phrenology. This 19th century theory claimed we could tell the character of an individual from the bumps in their skull!

Researchers claim that the only way to judge the dangerousness of criminals is by analyzing their facies … Photo credit: Shutterstock / Artem Oleshko

But Kosinski wouldn’t be the only phrenologist!

As early as 2019, researchers at the University of Harrisburg claimed that they could only assess the dangerousness of child criminals by analyzing their facies! In China, face recognition software is said to be used to identify Uyghurs. American students would be monitored during exams thanks to artificial intelligence that would read scams on their faces.

On the way to an invasion of our privacy?

As with all new technologies, the limits of artificial intelligence are gradually emerging … Should we continue to go out masked, not to escape the virus but to protect our privacy and free will?