Ana Paula da Rosa 242 Another controversial program involving images is Lensa, an app launched by Prisma Labs in 2018. The app re- quests the submission of ten images, preferably of the face, and recommends not sending pictures in which there are groups of people. From this set, artificial intelligence will create im- ages of the user. Just as imperfections can be erased, filters can transform the user’s image into an Avatar or artwork. This year, the app introduced the Magic Avatars option, allowing users to transform their photos into paintings or works of art. Numer- ous individuals adorned their profiles with images generated by the app. The play, the amusement of the machine that envisions us, carries numerous questions: from exclusion to access, the amplification of social stigmas by beauty ideals that emphasize historically constructed markers (of race and gender), the en- couragement of databases in a voluntary servitude even if the program claims data deletion within a few hours. Acknowledged as the most downloaded app in December 2022 in the Photo and Video category, it is relevant to consider that it is not only about access, although this revolution underscored by Véron (2013) entails many continuities and gaps. We step into a game: I do not envision you, if you do not pay. If you do not succumb, I will remove you from the flow. If your features do not conform to the cultural standard, I cannot attribute a 1920s painting-like face to you. The generated avatar can even acquire diverse connotations, including sexualization. Additionally, there are strains regarding the appropriation of digital artist styles and the users’ images granted. The debate on a racialized imaginary in artificial intelligence is extensive, as exemplified by works such as Cave and Dihal (2020), which raise issues with the images generated by AI and the mobilized imaginaries. This reflection includes avatars created for our interaction, such as Magalu’s Lu, and the patterns of Instagram fil- ters chosen as models of an ideal. Consequently, the racial social structure that defines attributes intensifies the erasure of differences in the digital environment as much as, or even more, in the physical environment of streets and cities. These cases raise multiple issues that this text cannot fully address. However, they may indicate a key point about the stigmas and stereotypes of the machine. When we consider the
RkJQdWJsaXNoZXIy MjEzNzYz