One output is not the work

Ringers was released on Art Blocks Curated in January 2021: one algorithm run a thousand times, a continuous line wrapped around a grid of pegs, with the peg count, the arrangement and the wrap direction varying and the rules never. Every iteration begins from the same premise and resolves differently.

Any single output reads as a finished composition, and the market prices it as one. The auction record for the collection, and the sale this site leans on elsewhere as the cleanest reading of conviction in the whole category, is Ringers #879, the Goose: $6.2m at Sotheby’s in June 2023. But the artwork is the system, and a single image says nothing about the size of the space it came from.

Matt DesLauriers has named the tension: generative art keeps gesturing at something close to infinity, while everything about collecting, and most things about exhibiting, pull toward the finite output you can keep. Museums inherit the tension whole. Acquire one Ringer and you own a composition; the thing the wall label describes is a process.

Asked by institutions how Ringers should be shown, Cherniak has given the same answer for years: as prints, hung in a carefully chosen grid. One print is a composition. Several, selected and hung together, begin to behave like evidence: what holds from sheet to sheet is the rule, and what moves is the range.

The design literature had a name for this before the medium existed. Jared Tarbell, interviewed by Jason Bailey, pointed generative artists at Edward Tufte’s small multiples: repeat a form with controlled variation and the viewer stops seeing pictures and starts seeing a system. Armin Hofmann’s Graphic Design Manual taught systematic composition before computers could explore it at scale. Lygia Pape’s Livro do Tempo hung a family of related forms, and a colour logic, as one work. The grid is not a hanging convenience. It is the format in which a rule set becomes visible.

This building already hangs that idea a century deep. The Bechers photographed water towers the same way for decades and let difference carry everything; the lineage page files every long-form generative grid under typology, and Ringers is the cleanest proof it has.

The print Ringers #962: The LACMA Iterations, made as that output entered the collection of the Los Angeles County Museum of Art, moved from showing outputs to showing the controls. #962 sits at the centre of a three by three grid. The eight outputs around it were made by changing exactly one variable, the percentage of available pegs the line may use, while everything else holds still. The edition of 726 sold out, with proceeds supporting the museum.

Two decisions on the object mark it as a display argument rather than merchandise. Every print carries a visual explanation of the process, and every print carries pseudocode. The code is not a technical footnote on a certificate; it is printed as part of the work’s substance, which is where it belongs in a medium whose material it is.

A sweep through one dimension is a deliberately small sample of the system, and the constraint is what makes it legible. The viewer moves from seeing an image to seeing a setting on a machine, and then to wondering what the other settings do. That wondering is the exhibition working.

The sweep has a direct precedent, and it was built as a teaching tool. Around 1999, for his Design by Numbers classes, John Maeda built ParameterVision: take a student’s coded graphic, map its two variables to the horizontal and vertical of the mouse, and display the outcomes as a grid of simultaneous states rather than one state at a time. The surviving documentation undersells it: the working grids were larger, and the cells could animate. Tom White, who saw it in use, described it as “looking through the compound eye of a fly”. Casey Reas read it as a computational extension of sight, a two-dimensional parameter space made to fit a single image plane.

Maeda traced the idea past photography’s founding argument about time. Eadweard Muybridge cut movement into separate frames; Étienne-Jules Marey’s chronophotography let successive moments coexist on one plate. ParameterVision carried that from time into computation: not many moments of one movement, but many possible states of one program.

White later ran the approach on Ringers itself, arranging outputs along the system’s principal variables and building a local parameter map around Ringers #879. No machine-learning interpolation, no images the original code could not produce: a sweep through the shipped algorithm’s own space. The record-setting output stops reading as a miracle when its neighbours hang beside it. It has a neighbourhood, and you can see it.

Tyler DeWitt, the artist 0xDEAFBEEF, pointed Cherniak to Andy Lomas’s writing on exploring complex generative systems, which supplies the general principle: past a certain complexity, the instruments an artist builds to navigate a possibility space stop being tooling and become part of the practice. Hold on to that principle. The third answer is made of it.

By Cherniak’s account, institutional reality was moving somewhere his print grids were not. Museums had approval and funding for large LED walls and were looking for motion work to put on them. The infrastructure arrived before the art, which is backwards, and also an opening: the question stopped being how Ringers should be exhibited and became whether it could inhabit those rooms without turning into decoration.

The conversation that turned the opening into a work started with Tal Lanir, building a new-media programme at the Tel Aviv Museum of Art, who wanted the ideas in the LACMA print carried into motion, and who pushed for sound. Cherniak’s refusal held throughout: no animated Ringer, no decorative loop. If the work moved, the motion had to reveal something a still could not.

Ringers in Motion is the result: less a video than a control room for a mature generative system. Beneath it sits a staging harness the artist spent his tool-building years on, specialised for Ringers: the algorithm decomposed, sequenced, transitioned, held on particular states, and kept reliable across an institutional day. As described at its premiere, the work contains almost fifty distinct scenes and runs for more than an hour before any of those animations repeats, all of it generated live rather than played back. It also carries explicit safeguards against swastika-like configurations, which is a fact about exhibiting line-wrapping geometry in 2026 as much as about this algorithm.

The sound is built from the notes of Hatikvah, slowed and fragmented until, in the artist’s description, it sits closer to bells than to a melody, familiar tones returning without quite resolving. Cherniak reads the line’s wrapping through his own references: the eruv, tzitzit, tefillin, forms in which a drawn boundary makes a shared space rather than an exclusion. He says those readings were recognised in the work late, not designed into it early. They are reported here in his frame on purpose; translated into curatorial neutrality they would stop meaning anything.

The LACMA print is small multiples in space. Ringers in Motion is small multiples in time: the work can cross to a distant region of the parameter space, then linger in the neighbourhood of an output you recognise. A print grid is a map of the system. The screen is the same map with a cursor moving across it.

When curators now ask how Ringers should be displayed, there is no longer one answer. Single outputs stand as blockchain-native works, their transactions part of the object’s history. Curated groupings expose relationships across the thousand. Early variations trace the experiments that led to the release. And the software work holds all of it at once: individual outputs, their place in the larger field, the parameters, the neighbourhoods. The work does not contain every possible Ringer, and nothing could; it holds the rules, the relationships and the histories instead. The display question stopped having a correct answer and became a generative space itself.

This gallery holds Ringers in its canon and hangs it at the front of the generative room, and the reasoning above is also this building’s reasoning. The rooms here are grids because a curated grid is the argument: repetition states the rules, variation shows the range, and that is how you exhibit a system when all you have is a wall.