Palette

THE WORLD’S LARGEST GIF

🚫 Marius Watz Interviews Jason Salavon on Generative Art vs. Data-Driven Art

👨‍💻 42-year-old programmer | 👨‍👧‍👦 Father of two | 📚 Teacher | 💡 Passionate about technology and financial markets
 |
I love simplicity. ❤️

1️⃣ Jason Salavon is an artist of contradictions with the undertones of a dark trickster. His work reveals a keen mind for computational systems, but is equally rooted in art history and evolving concepts of image production.

2️⃣  a direct line from painting and photography to generative art, and finally, generative AI image generators. His back catalog ranges from sublime abstract visuals to conceptual gambits that prove more poignant upon closer inspection.

4️⃣ ... while also recognizing that his practice was referring more to contemporary art discourse than the somewhat insular Ars Electronica/MIT Media Lab futurism that I spoke natively.

5️⃣ Initially, I was attracted to the formalist delights of Shoes, Domestic Production, 1960-1998 (2001), and later American Varietal (US Population, by County, 1790-2000) (2009) — visually stunning pieces that are technically visualizations of the most mundane of data sources.

6️⃣  I loved the colors and abstract visuals, but also the implication of an in-joke. Visualization can be accurate, while also being embellished to the point of transforming the pedestrian into rococo.

7️⃣ His “amalgamations” — 76 Blowjobs, The Class of 1988 & The Class of 1967 (1998) and his early viral hit, Every Playboy Centerfold, The Decades (normalized) (2002) — take this data-based approach further, blending it with concerns around photography and memory.

8️⃣ These series are conceptually and formally tight — a chef’s kiss blend of pop culture with a note of postmodernism.

9️⃣ But Salavon also has works that are almost nihilistic in their anti-formalism. Spigot (Babbling Self-Portrait) (2010) is an interactive software visualizing Salavon’s Google search history in a blunt form — basic colored squares and the generative faux pas of seemingly random RGB colors (they’re not).

🔟 100,000 Abstract Paintings (2001) and Golem (2002) are generative systems, but compared to the current mode of long-form generative art, they render the individual outcome practically meaningless. Here, long-form, becomes a way of emptying the image of value.

1️⃣ 1️⃣ TODEM (Tapestry of Decadent Meritocracy) (2023) combines all these threads — data as the framing of a narrative, conceptual formalism and photography as carriers of meaning, but parsed through a mostly automated logic that reminds the viewer that what they’re looking at is a kind of meta-text.

1️⃣ 2️⃣  The fractalized image grid is at once a compositional device and a parameter of the algorithm, but it also provides a pragmatic marketing device: value per pixel.

1️⃣ 3️⃣ TODEM also contains hints of Salavon’s dark humor. The images might be generated from prompts taken from Wikipedia, but they are not neutral renditions of history. There is a late-capitalist Americana being performed here.

1️⃣ 4️⃣ I can’t resist making connections to J.G. Ballard’s book, The Atrocity Exhibition (1970) and, more darkly, Why I Want to Fuck Ronald Reagan (1968). The generated images seem to reference the last few decades of (gritty) American movie aesthetics blended with 1950s nuclear optimism.

1️⃣ 5️⃣ In sum,TODEM is an ambitious project,with generative AI applied as a methodology to create a whole visual world, multilayered fractal rather than a simple image.It is also a perfect extension of Salavon’s work to art on the blockchain, embedding a value matrix and minting logic directly into the work.

The text of the interview script is available at the following link:

The text of the interview script is available at the following link:

https://www.rightclicksave.com/article/the-worlds-largest-gif

3️⃣ I first got to know Salavon’s work in the early 2000s, at a time when finding other digital artists was not a trivial task. I saw that he clearly shared some of the same interests as the proto-generative scene I was embedded in at the time (soon to be centered around Processing), ...