Biz & IT
The Magic and Minstrelsy of Generative AI
/var/www/vhosts/lawyersinamerica.com/httpdocs/app/views/singleBlog/singleBlogView.php on line 59">
culture
Apr 2023
Fantasy is irresistible. It is the fabric of dreams and the substance of imagination. Fantasy drives innovation. It lures the future, eternally on the horizon of our lives, within reach. On the social internet, fantasy is fact. It is where we work and idle. Fantasy is how we communicate, the surreality of memes, images, GIFs, and video our mother tongue, the language of this warped and warp-speed generation. What I love and loathe about the social internet is its fondness for play. To exist across its sprawling ecosystem of platforms and apps is to understand it as a playground for many things, but especially identity. On the social internet, legibility is a lark.
Until now, I had never felt a true sense of alarm. TikTok has an obvious lust for appropriation, and before its ascendance there were whiffs of digital immoralism on just about every platform, each one unique to the sort of tangled socializing it supplied. But social apps have limits. At stake now, as generative AI becomes what Bill Gates christened "the most important advance in technology since the graphical user interface," is a war over images. The visual vernacular we use, and have gotten accustomed to, is on the verge of mass contamination.
I came of age in the era of AOL chat rooms. For a time, beyond the dominion of my youth in California, Instant Messenger was a telescope into the wider world. Anonymity was a given, and we relished the masks we wore, happily scavenging the nascent paradise of the dotcom meadowlands, bonding with friends and strangers, unaware of what was to come, of how those masks would be turned against us with such ferocity. This was my base understanding for how I should, and would, engage the internet. It was rooted in misdirection, cushioned in a kind of innocent magic. I loved everything about it. Today that's par for the course, though the aftershocks are more far-reaching and harmful.
Last month, out of the blue, I heard from an ex. They were upset that I hadn't notified them of my trip back home. "You in LA?" the text read. When I expressed that I wasn't and asked why they assumed I was, from the chat bubble appeared a shirtless image of me. It was a screenshot from a hookup app, where I had posted that same image years before. Except it wasn't my profile (if only I was 6' 3").
Catfishing is now a common obstacle of digital exchange, a concept popularized ad infinitum in television (Catfish; Inventing Anna) and music. In the era ahead, however, ownership over one's identity will adopt a profoundly more noxious guise. As our interactions become even more immersed in virtual worlds, blurring the lines between fact and fantasy, the vernacular of them will take a costumed appearance. In everything from videos and voice notes, these manipulations will draw on the exaggerations of the human form. The sentinels of AI have arrived. Legibility is the next great battleground.
Writing in her 2002 seminal essay "Future Texts," Alondra Nelson prophesied as much. "In these politics of the future," she wrote, "supposedly novel paradigms for understanding technology smack of old racial ideologies." Twenty years on, the myth of utopia remains. What the vanguard tools of AI attempt to capture and replicate, with programs like ChatGPT and Midjourney, are the very things that make us stubbornly human: how we communicate and what we look like, our modes of performance, our need for sustained connection. That race and gender would somehow take on a less perilous meaning in the future, Nelson suggested, was a lie.
The time ahead will be one of wonder. It will also be one of wreckage.
The great providence of technology is also its most fear-inducing aspect: the immensity of what it can achieve. That is where we stand now, at a crossroads, and many of us--dare I say most of us, and especially those of us who dwell in the margins of society--will fall victim to more intelligently designed schemes, from email fraud and identity theft to online harassment. Facial recognition technology has proven to hold racial bias. Others, from writers to fashion models, will be tossed aside in a great labor overhaul the AI renaissance will herald. To borrow a term from early techno-prophets, the digital divide ever widens.
That's not to say the AI arms race is without genuine benefit. I find assurance in its more practical use. "It will change the way people work, learn, travel, get health care, and communicate with each other. Entire industries will reorient around it," Gates wrote in a blog post. But that reorientation comes at a cost.
The wonder of AI is all around us. I can't deny it, and I don't want to. I find it astonishing how it can so quickly, and precisely, evoke the pastel enchantment and peculiarities of a Wes Anderson film with just a few simple prompts. Or how the vision of Pope Francis in a white puffer jacket--already one the year's best memes--dares us to question the line between reality and fantasy. (A Hypebeast Pope? Yes, please!) It's digital satirism at its most refined. It's the unreal made real. In part, it's what the internet is for: to give fantasy material substance. To give it legibility. Even I have to admit there is beauty in the deconstruction of the given.
But chaos is also inherent. Generative AI pulls from--learns from--the ugliness of human error. There is a concurrent danger in that sort of fantastical authorizing, the way images can swiftly curdle into a more depraved form. I've written at length how TikTok emboldens digital Blackface, and I worry that the currency of cultural theft will only become more ubiquitous in this next era of AI.
Already we are experiencing the stink of its machine learning around identity, in how it treats Blackness, of how it happily twists fantasy into gross minstrelsy. In one video--captioned "Will Smith eating spaghetti!"--the actor is evoked in ape form. In another, Morgan Freeman is turned into a political puppet, made into a clown mouthpiece by a win-at-all-costs conservative movement. The images are unfading. They are visceral in their presentation, their distortion a dagger.
That shouldn't come as a total surprise. The minstrel figure has haunted the registers of American culture since the 1840s, when white men realized there was money to be made in the exploitation of Black storytelling. They deformed it into a kind of entertainment. They called it minstrelsy. "Playing [n****r] is first-rate theater," Margo Jefferson wrote in 1973 of the old practice, which was acted out by white performers. Black life became a metaphor for everything wrong with America, and with it the image of Blackness took on a mutation in pop culture: It was something to be worn, not respected. Over the years, that image transformed, slinking from live theater and movie screens into 30-second TikTok videos, its appearance ever oscillating between analog and digital realms.
This coming age of new minstrelsy will assume an even more cunning chameleon form, adaptive and immediate in its guile, from humanistic deepfakes and spot-on voice manipulations to all manner of digital deceit. Only, everyone will have a target on them. In the future ahead, our images will no longer be ours alone. Authenticity will be punctuated with a question mark. In the future ahead, the masks we wear will be plentiful and perverse on the social internet. Legibility, and the right of ownership, will be harder to control. In some corners, the innocence of creation will remain. The awe of what AI will achieve is undeniable. But so is the fear.
