Artificial Intelligence: Why Does Twitter Seem To Prefer Whites To Blacks?

Is Twitter’s algorithm racist? The social network would tend to favor images of white people over black people in photos. In reality, the explanation would lie in the way in which artificial intelligence was trained.

There is a mystery on Twitter. Questions related to the way in which the system selects images for display to users.

Typically, when you post an image on Twitter, it is automatically cropped and it’s a slightly different thumbnail than the original that appears in the News Feed. To see the whole image, you have to open the message. However, users have found that if you post an image containing two faces, one white and one black, Twitter systematically crops the image so as to present the white face, the black remaining invisible. The affair began with a portrait of US Senator Mitch McConnell (White) and that of former US President Barack Obama (Black). It’s the same every time, whether white is placed at the top and black at the bottom or vice versa. That’s not all: between a man and a woman, the man is also privileged.

Algorithmic biases

Twitter acknowledged the problem and apologized. But for now, there is no precise explanation. Research is underway. This case perfectly illustrates what is called the question of “the ethics of algorithms”. To fully understand, you need to know how cropping images on Twitter works. This is automatic. It’s run by what’s called an artificial intelligence neural network, a program that has trained from millions of images.

The most likely explanation is therefore that this one has trained with many more white faces than black faces and therefore has difficulty recognizing a black face in a photo.

An algorithm sees only pixels

Remember that an algorithm does not “see” a person in an image, but it only detects pixels without understanding what it is. Strictly speaking, an AI is therefore not racist. However, it can reproduce racist biases, conscious or unconscious, coming from those who programmed it. It can also suffer from technical faults which resemble, on arrival, racism. There have been issues like this before with Facebook, as well as with Zoom video conferencing software which tended to erase the faces of black people in its virtual background integration system.

This type of problem is not insoluble but it must be taken into account from the creation of algorithms and during training of AIs, so as not to virtually reproduce discrimination.

Các tin khác