Is the collective unconscious real? Carl Jung’s theory of a shared history of thoughts and behaviors manifests in the form of crowdsourced data and its interpretation by artificial intelligence (AI).
Working in performance and new media, San Francisco-based artist Avital Meshi strives to understand widespread cultural attitudes about race, gender, beauty, and sexuality. Assembled from hundreds of AI-generated self-portraits, Avital’s interactive wall mural This person is not me invites viewers to consider personal and pervasive stereotypes surrounding certain words: rich, poor, stupid, unemployed, transgender. “[These] words […] may elicit stereotypical imagery, which has the potential to amplify norms as well as biases,” she writes in her artist statement. “For instance, we can ask how words such as ‘beautiful’ or ‘ugly’ are being visualized. And what does it mean about the way our society creates standards for these words?”
Using online AI platform Astria to generate the images for This person is not me, Avital fed the algorithm images of herself, training the program to recognize the specific planes and curves of her face. Once trained, Avital gave the algorithm a prompt—”Black,” “sexy,” “smart,” “Asian”—” and the algorithm spit back an image of the artist based on that prompt. “The pairing of text and image to create so many images of myself manifests a ‘multiverse aesthetic’ through which I can see myself within the framework of infinite identities,” Avital says. “When I look at these images I can recognize myself, but at the same time I realize that they all depict an ‘other’ whose elements are foreign to me.” Participants often try their hand at guessing which prompts led to which portraits before accessing the mural’s augmented reality layer, which reveals the underlying prompt.
“This kind of identity tourism does not come without critique,” she continues, noting that AI-generated images draw from a well of “embedded values and structural thinking in our society.” In other words, AI-powered imagery reflects our own biases, the algorithm’s rough approximation of our collective ideas about identity and social structure. “The idea of seeing myself as one of these identities does not imply that I can understand the experience and hardship of living under such stereotypical conception of one identity or another,” she says, stressing the importance of individual experience divorced from the collective unconscious.
“When I look at these images I can recognize myself, but at the same time I realize that they all depict an ‘other’ whose elements are foreign to me.” — Avital Meshi
Avital Meshi: Website | Instagram
All photos published with permission of the artist(s).