Could this James Baldwin typewriter be an okay use of AI?

A new interactive exhibition about the writing of James Baldwin features an AI programmed typewriter that directs users to writing drawn from Baldwin’s books, essays, and letters. If you, like me, are aghast that “AI” and “James Baldwin” are in the same sentence, take a closer look. The issue here is not with this project, which is building something artistic and beautiful, but with a broader, pro-AI tech culture that has discredited AI and its public image.

The experiential exhibition is titled “For Those Who Come After” and is part of “KIN: A Festival of Memory and Imagination” in New York, as covered in Fast Company. The installation and its software is designed by Kinfolk Tech, “a nonprofit using immersive technology, public art, and community-led design to help displaced and excluded communities tell their own stories—and shape their own legacies.” They’ve created augmented reality monuments of Black, Brown, Indigenous, and queer figures, run fellowships on community memory, and created archival projects around story telling. The Baldwin installation is centered on a physical station: “a desk, a typewriter, and three card stacks labeled Wisdom, Courage, Love.” The visitor sits in front of the typewriter and asks the question.

Using what seems to be a bespoke AI model, the typewriter points you to three cards from the stacks that are relevant to your question, and then types out “a poetic reflection, generated from the spirit of Baldwin’s archive.” My knee-jerk is to recoil from anything that brings AI close to Baldwin’s work. But the closer I looked, the more it seems clear that the emphasis is firmly to bring people closer to Baldwin’s writing—elsewhere in the room, you’ll find a radio, TV, and record player that all reproduce Baldwin’s words. Kinfolk’s founder Idris Brewster told Fast Company that the project is designed as “a ‘poetic intervention’ to slow people down long enough to reflect on the partnership between humans and machines.” Kinfolk is clearly doing interesting, careful, and artistic work, so I have to question my AI anger, which feels more appropriately directed at the tech bosses, boosters, and acclerationists who have so thoroughly poisoned the well about this technology.

This also happens to be, as Vox put it, “the week the AI freakout went mainstream.” It’s been a deeply infuriating experience to hear higher-ups at these tech companies tell us they believe themselves to be engineering our doom. It’s the sort of thing tech has gotten very good at: inventing a problem and then throwing the consequences to the public to tidy up. And what are we to make of it?

Are they lying to us to boost their stocks and their egos? Or are they really going to work every day at the Armageddon factory and expecting sympathy? Fittingly, this is a very American intersection of wealth and power, the exact kind of problem about which James Baldwin was always so eloquent and incisive.

My armchair psychology is that this is a phenomenon similar to how certain men in Hollywood are bizarre about method acting, which has “become wrapped up in a brand of identity politics that tries to make the art form resemble more traditional forms of male labor,” according to Angelica Jade Bastién in The Atlantic. I sense an similarly insecure defensiveness behind these men crying “we’re too smart!” and “we’re building a bomb!” when in practice, they’re just on the computer. Quieter, more humanistic uses of machine learning or LLM technology must now contend with a public whose idea of AI has been shaped by the loudest, maximalist voices.