Ask what “digital art” means and you get a list of tools. Lev Manovich, who has spent three decades building the vocabulary this field argues in, thinks that’s the wrong question entirely. In “Seven Worlds of Digital Art: A Visitor’s Guide,” published in Right Click Save on September 8, 2026, he proposes replacing the technological definition with a sociological one, borrowed from cultural sociology: where does the work circulate, who values it, and what are the criteria for judging it?
Run that test and the category splits into seven largely non-overlapping worlds.
The seven
1. Historical Digital Art. The 1960s-onward canon — Vera Molnár, Manfred Mohr, JODI — now absorbed into art history. Values are historical influence, documentation, preservation, conservation.
2. Digital Practice in Contemporary Art. Work addressed to art audiences and judged by contemporary-art standards: Hito Steyerl, Trevor Paglen. Validation runs through museums and biennials, and Manovich is blunt about the entry requirement — what’s rewarded is “a position that art criticism can take up, critical or formal; technical novelty on its own counts for little here.”
3. The Specialist World. Ars Electronica, ISEA, the creative-coding communities. Values are technical experimentation, research, and collaboration across art, engineering, and science. Recognition comes from peer juries rather than the market.
4. Code Art. The artist writes a program; the program makes the work. On platforms like Art Blocks, “each output is generated at the moment a collector buys it” — the artist “sets the rules and the range of variation; nobody, the artist included, sees the individual work before it exists.”
5. Platform Digital Art. By volume, the largest world by far. ArtStation, Instagram, marketplaces — circulation “without the curators, galleries, and museums of the art world in between.” Values are virtuosity and a recognizable personal style.
6. Immersive Spectacle. teamLab, d’strict, Refik Anadol Studio. “Values are scale, sensory impact, and visitor numbers.”
7. Commissioned Digital Media. Interface design, film visuals, motion graphics — work that “begins with a brief specifying what the designers must accomplish,” judged by clients through revision cycles, where “recognition rarely reaches the public.”
Why the sorting matters
The framework’s real payload is the claim that these worlds are close to sealed. Each has its own audiences, gatekeepers, reward structures, and standards of quality — so a body of work can be canonical in one and illegible in another, and neither judgment is wrong. A generative artist with a sold-out Art Blocks drop has demonstrated nothing that world 2 recognizes. A festival-circuit installation with three Ars Electronica honorable mentions has demonstrated nothing that world 5 recognizes. The frustration each camp feels about being ignored by the others is, on this account, a category error rather than an injustice.
It also explains a recurring pattern in how this work gets covered. Critics trained in world 2 review world 6 spectacles and find them empty — correctly, by their criteria, and irrelevantly, by the criteria that produced them. Technologists dismiss world 2 as decades behind the state of the art, which is true and beside the point, since being current was never what that world measured.
Where it’s uncomfortable
The taxonomy is descriptive, and Manovich mostly resists ranking the worlds — but descriptive frameworks have a way of hardening into hierarchies once they’re in circulation, and the ordering here runs suggestively from museum to marketplace. Worlds 5 and 7, which contain the overwhelming majority of people who make images with computers for a living, are also the two where he notes recognition is thinnest.
There’s also a question the framework raises and doesn’t settle: whether the boundaries are as stable as the model needs them to be. Anadol sits in world 6 with a foot in world 2’s institutions. Art Blocks artists have been acquired by museums. Sealed worlds are a useful simplification, but the interesting careers tend to be the ones leaking across them.
For anyone who has tried to explain to a gallery why an algorithm is the artwork, or to a developer why a beautifully-shaded render isn’t getting into a biennial, it’s a genuinely clarifying piece of writing.