Two scientists of the university of cambridge have a treatise entitled "the whiteness of ai" published, which shows how soft stereotypes prague the image of systems with artificial intelligence. The article works both with real and with fictitious intelligent machines.
The two scientists divide the machines into the categories humanoid robots, chatbots, virtual assistants and stock images. They offer three interpretations for the fever stereotyping of the ki systems.
Racial mapping for machines
Machines can receive a raced mapping over the anthropomorphism, so transfer human attributes to the artificial systems. In addition to the obvious visual characteristics that are more likely to correspond to women’s models, the human voice or the type of interaction also plays a role.
Also very scary robots such as nao, pepper and pr2 consist of largely waist materials. Even more profound is the assignment in the sophia presented by hanson robotics, which received the staupburgschaft of saudi arabia in 2017. Despite her asian origins, the robot woman built in hong kong is clearly caucasian wise.
Sophia is probably clear with or without torso as "woman" classify.
Basic and follow
The treatise examines three possible reasons for the dominant female picture. On the one hand, the race mapping often reflects milieus in which they arise. As a second reason, the authors drove up that machines that are intelligent, professionally and smooth, from the point of view of many women people with these attributes of their own breed are attributable.
As a result, the authors warn that the seeds of stereotypes could remove colored people from the sowing utopia. Overall, the presentation of prejudices and the bias in machine learning systems urge. As a result, decisions can fall again, the ethnic groups disadvantage.
Significantly, the sentence brings the sentence: "if women imagine people who are pronounced by supposed beings, these beings will not meet the breeds that they have previously classified as inferior. A woman’s audience can not imagine that it is surpassed by black machines."
The result of a picture search on google "artificial intelligence robot" occupies the seeds of stereotypes.
Self-knowledge is the first step in recovery
As an important first step to break up the structures, the article sees the realization and the keynagation that they exist. The seeds stereotypes were not allowed to remain invisible. It is unlikely that the majority of sewing viewers in human machines recognize a racial mapping, as they only see what ‘human’ mean. For people without woman skin color, the senior stereotypes are never invisible.