In a recent piece for The New Yorker, the science fiction writer Ted Chiang likened ChatGPT to a “blurry JPEG” of the internet, a compressed average of our digitized collective memory. Chiang’s metaphor is disparaging, pointing to the ways that large language models deal in vibes, averages, and approximations rather than writerly precision. ChatGPT is blurry, for Chiang, not just because it sometimes lies, but because it cannot experience the world and extract something new from that experience, the way a human artist or a human science fiction writer might.
But looking at the world with blurred vision can also be a way to take in a scene in its entirety without getting caught up in the particulars, to absorb more information at once than we’re usually accustomed to. The daubs and smudges of an impressionist painting are compression artifacts by another name. To actually see a blurry JPEG of the whole internet sounds like an ecstatic vision. Why haven’t we seen a blurry JPEG of the whole internet yet?
For now, the best substitute we have might be the works of New York City-based digital artist Ezra Miller. In his generative artworks, which resemble alien screensavers glimpsed through stained glass, Miller makes the recursion and recombination that are for Chiang symptomatic of the worst ways the algorithmic internet has blunted human creativity grounds for the sublime.
Across his multivariate practice—which encompasses real-time generative art, interactive websites, live performances, and immersive experiences—Miller is above all interested in how new technologies can enable genuinely new kinds of aesthetic experience to emerge. He’s messed around with WebGL on Tumblr, minted out a collection of generative NFT works, engineered an infinite “multi-modal painting simulation,” and, most recently, experimented with AI to create haunted apartment hunting videos and transform curbside puddles into portals in which the chiffon fabric of reality seems to fray and flicker.
In this informal conversation recorded at Dunkunstahalle project space in downtown New York, he’s joined by friend, writer, critic, and occasional meme, Dean Kissick. Kissick is the New York editor of Spike Art Magazine, where he wrote the monthly column The Downward Spiral between 2017 and 2022. He also hosts Seaport Talks, a monthly series of conversations with critics at T.J. Byrnes of Manhattan. The pair discuss the lootbox mechanics of generative art, the gap being “digital art” and a “digital art experience,” and the overlooked potential of AI to create a Northern-Rennaisance-style Walgreens.
WATCH AND MINT THE FULL CONVERSATION HERE.
This article can be collected here.