By akademiotoelektronik, 05/07/2022

Artificial intelligences would be racist

Bynick Olaizola January 28, 2022

In recent years, artificial intelligence has experienced remarkable development.It is integrated into all the IT tools that we use in our everyday lives, to a point where we are more and more dependent.Besides, because of the place she won within the community, she was able to group a huge amount of information on her environment.It needs it to improve its performance and its operation.

The intelligent system has the ability to distinguish discrimination on race, sex and age.With its current level, AI seems to have categorized its apprehension of the human race.This is one of its weak points of today's algorithms.It is true that AI offers many possibilities for humanity, but it also makes it lower and vulnerable.

Studies by researchers Inioluwa Deborah Raji from the University of Toronto and Joy Buolamwini of MIT show that the results are better on clearer skin faces.

Discrimination related to race, sex and age

This is one of the most common problems with artificial thought is faced every day.The results of the study showed that the intelligent system discriminates women and people of darker color.Naturally, this discriminatory AI approach is potentially dangerous.It thus constitutes a real problem to be resolved in order to prevent even more ills in current societies.

The AI recognition system would also have a discriminatory approach at age level.This discrimination is called agrism.It is defined as segregation behavior against the elderly.This identification system can corrupt information from a large number of people at the advanced age, which can be formidable for the future of modernized society.

No freedom for AI yet!

This "discriminatory attitude of AI" would be due to the massive data collection with which the electronic brain trains.Some data provided to them would not be in with what is "politically correct".

In any case, this flaw tends to prove that one should not yet entrust a total freedom to the intelligent machine, because it also makes mistakes.For the time being, artificial intelligence seems biased by prejudices and stereotypes deeply anchored in the collective unconscious.

Source: Techexplore