The Not So White White Paper

Abstract

This paper examines the digital artwork Wheel of Fortune (WOF, 2010–) by the Viennese artist Sebastian Schager as a contemporary work within an art-historical context. WOF lets images pass by like a wheel of fortune, since 2010 as a tool of painting and, most recently, in the two cases this paper examines more closely: as the first purely digital installation, without painting, in the Magenta Art Room (2026, Collage mode), it drew on open museum and cultural archives, and a human image drop at the right instant condensed the images into an unrepeatable collage; as the first epoch, Trinity, in the browser (since 2026, Mono Scroll mode), it plays from a collection curated by the artist in three classes, and the players cut twelve seconds out of the running wheel. Its core thesis: in WOF, chance is not depicted but performed – the wheel follows fixed rules, and chance enters at two points, in the question of which images get into the run and in the instant at which someone intervenes; precisely the structure that already characterises the medieval Rota Fortunae. At the centre stands a double observation: with its title and its mechanics, WOF invokes a pictorial motif handed down through centuries of Western art – the Rota Fortunae, the wheel of Fortuna – and translates it into a technical dispositif of image holdings, score and blockchain. Methodologically, the paper applies Erwin Panofsky’s three-step iconographic-iconological model and deliberately extends it at one point, because WOF does not show the motif but lets it run. The analysis situates WOF within a genealogy of the readymade, collage/montage, appropriation, conceptual and generative art, as well as within the theory of the “Death of the Author” (Barthes) and of technical reproducibility (Benjamin). A new chapter sets the work against the language machine (Simanowski) and develops its argument from that comparison: against the average that statistics produces, neither chance alone nor the human alone is of help, but the change of levels – human, chance, human, blockchain. A separate chapter examines whether the NFT opens up an artistic – not merely speculative – space for this work, and at the same time names its limit, which the work has meanwhile shifted onto four blockchains.

Keywords: iconology · Rota Fortunae · participatory art · appropriation · Death of the Author · generative art · montage · language machine · epoch · NFT · on-chain provenance

1. Introduction

Without a network connection, the installation showed a black screen. This dependency was not a defect but the programme: the work owned no image material of its own; it drew it, only at the moment of operation, from the image holdings of the world. Two months later, the same instrument plays without a single connection to an archive, from a collection the artist has assembled by hand over years, in the browser of those who play it. Both are Wheel of Fortune, and the tension between the two states carries this second version. Its subject is the ongoing project Wheel of Fortune (hereafter WOF, or simply: the wheel), since 2010 a tool of painting, which in 2026, as part of the Magenta Art Room in the Vienna underground station Volkstheater, was shown publicly for the first time as a purely digital, browser-based, participatory image machine, without canvas and without brush.¹ In the installation (hereafter simply Magenta), WOF let images from fifteen open image sources pass by like a wheel of fortune and condensed the individual images, dropped by human timing, into unrepeatable collages; in its first epoch, Trinity, it lets whole images from three curated classes run against one another, and the players cut twelve seconds out of a run that never halts. The finished work can be saved as a loop or minted² as an NFT on one of four blockchains.

The obvious reading – that WOF is “digital collage” or “AI art” – falls short, and in the second case is factually wrong: no images are synthesised; real images are recomposed.³ This paper instead pursues the thesis that WOF translates a classical pictorial motif – the Rota Fortunae – into the procedures of the digital present – the image stream (in Magenta live from open archives), the generative search movement and the blockchain – and thereby raises a double question: about authorship in the age of recombinatory machines, and about the place where such a work can exist artistically.

From this follow two research questions: (1) How can WOF be described, with the instruments of iconographic-iconological analysis, as a work within the art-historical context of its time? (2) Does the NFT open up a genuinely artistic space for such a work, or does it remain a merely economic appendage? The second version poses both questions for Magenta as for Trinity and adds a third, which arises from the comparison with the language machines: (3) How does a work that combines real images by chance differ from a machine that computes new ones from probabilities, and what follows from this for authorship?

The title of this paper, too, belongs to its subject: as early as 2022 the artist prefaced the NFT works with a statement called “The Not So White White Paper” – an ironic rejection of the roadmaps and promises of returns of the crypto market. A promise stands in it nonetheless, only a different one: not returns, but the pledge not to stop. The evidence for it is the work itself: the first performance of WOF in 2011, the first NFT in 2021, ten years in between, and both begun before the statement and continued ever since. The present text adopts that title and redeems its claim by art-historical means.⁴

Methodologically, the paper takes Erwin Panofsky’s three-step model as its foundation – pre-iconographic description, iconographic analysis, iconological interpretation – supplemented by Aby Warburg’s concept of the afterlife (Nachleben) of ancient pictorial formulae.⁵⁶ Where the model reaches its limit – with a motif that appears as a process – the extension is made explicit (ch. 3.1b, 3.3). The art-historical contextualisation draws on the theory of appropriation and authorship (Barthes, Benjamin) and on the genealogy of the readymade, montage, conceptual and generative art. The technical chapter rests on the documented architecture of the work; the NFT part on the primary literature of the discourse and its critique. The second version additionally draws on the media philosophy of the language machine (Simanowski), the theory of the open work (Eco), of the performative (Fischer-Lichte) and of presence (Gumbrecht), the anthropology of images (Belting), the theory of montage (Eisenstein) and the critique of participation (Bishop, Diederichsen).

Structure: Chapter 2 describes the work. Chapter 3 carries out the iconological analysis and forms the argumentative centre. Chapter 4 situates WOF art-historically. Chapter 5 explains the technical dispositif. Chapter 6 sets the work against the language machine and develops from that comparison the argument of the change of levels. Chapter 7 examines the NFT thesis together with its limit. Chapter 8 gives the artistic position room in a conversation between machine and artist; chapter 9 draws the conclusion.

2. Description of the Work: What is Wheel of Fortune?

Wheel of Fortune is not a single work but an instrument with several forms of play. What it produces in each case depends on where it stands, what it draws on and which human act determines the result. Such a combination of mode, site, material and players is called here a form of play. The instrument has developed three so far; the series is open. The following overview orders them before sections 2.1 to 2.5 describe them one by one.

The last row stays empty, and deliberately so: since 2026 the work knows the concept of the epoch (ch. 2.5.1), and a new epoch is a new recipe, not a rebuild of the instrument. For the analysis (ch. 3–7) the distinction has a consequence that should be named in advance: the thesis of this paper holds for all forms of play: the wheel of Fortuna runs here; it does not hang on the wall. What shifts from one to the next is the site of the human act, from the single gesture to the series of gestures, from the point to the interval, and the site of curation, from the open archive to one’s own collection. With the form of play, who plays changes as well: the painter, the passer-by, whoever finds the work on the net. And since Trinity a group has been added that does not play but keeps: the collectors, for whom the work opens a new artistic space (ch. 7).

Six words carry the following description, and they denote different things. The form of play is a row of the overview above: a combination of mode, site, material and players. The instrument is WOF itself, in its purest form a running band of images from which a stop or a drop takes one image: it is used, and the band runs on. Chance and human intervention belong to every form; what emerges from them does not. The mode says how the wheel plays and what it produces; there are three so far, and the table below sets them side by side. In the painting tool, a button halts the band, the painter’s hand paints the standing image into the painting, and the band runs on; from many stops a single, physical picture emerges. The other two modes are purely digital. In Collage, the image visible at that moment falls onto a surface on the screen, and from the drops a composition grows. In Mono Scroll, chance drops one set from each class into time at the start; its images run in several lanes side by side until someone cuts out the last twelve seconds. Both end in a loop, which stays digital whether it is minted as an NFT or merely played. How a mode lets the human hand into the game are its gestures; the table names them, ch. 2.2 and 2.5.5 describe them. What this paper calls the human act is the decision; the gesture is its form. The site says where play takes place: in the studio in front of the canvas, in an installation on location and in public, or in the browser on one’s own device. The epoch is a fixed recipe that the artist sets for a whole series: mode and variant, gestures, holdings, finale, blockchain and the rule for when a work may come into being (ch. 2.5.1). It is not to be confused with the score, which writes down each individual game and from which the server generates the work (ch. 5). None of these words is bound to another: an installation can play Mono Scroll, a future epoch the Collage.

This paper examines above all two cases and calls them by their names. Magenta is the installation in the Magenta Art Room in 2026, in Collage mode, fed from the open archives, in retrospect epoch zero. Trinity is the first epoch, in Mono Scroll mode, played in the browser, fed from the artist’s collection. The painting tool is the origin of both (ch. 2.1) and continues to run alongside them (ch. 4.5). What is said about Magenta or Trinity holds for that case. What is said about a mode holds for every case that plays it. What is said about the instrument always holds.

2.1 From Painting Tool to Browser Machine

WOF began in 2010 not as a screen work but as a tool of painting: a projection machine that threw random image material onto the canvas during live painting sessions, so that the digital stream became part of the physical composition.⁷ Even the projected run on its own, before the first brushstroke, is a performance and a work of light; this spectacle value of the image stream is preserved across all forms of play. Which side is the stronger the artist leaves open: “I’m not sure whether the performative side of the image change in WOF or the painted variant is better. Those are two worlds colliding there, and both are awesome.”⁸ The tool emerged as the artist’s invention in the milieu of the Viennese artist group PERFEKT WORLD,⁹ which by then had realised numerous joint projects, and responded from the outset to a substantive diagnosis – the permanent state of media saturation of the present.¹⁰ Related works from the same milieu – painting on stroboscopically backlit, translucent canvas, for instance – already followed the same logic of media flicker and were shown at parties as much as in art installations. Out of this practice grew the Wheel of Fortune series of paintings (from 2010, ongoing) – painted works whose procedure (chance, open image material, human timing) already contains the later digital work.

The early image material came from a private, locally assembled collection: thematic image folders in the spirit of the group PERFEKT WORLD, whose deliberately popular iconography, turned away from the canon – ancient statuary next to comic and action figures of the nineteen-eighties – anticipates the procedure’s later indifference towards high culture.¹¹ Since the artist lacked programming skills, the then-young internet initially remained a mere image repository and not, as today, a source opened up through interfaces (ch. 2.3). Already in this gesture – the artist not as producer but as collector and arranger of found images – lies the shift of authorship that the work would later catch up with theoretically (ch. 4.2).

The set-up was performative from the beginning (around 2010). Painting was done in large formats and in mixed media, often collectively: several painters translated the same image stream onto one canvas simultaneously, drop by drop. The way there had been paved in June 2011 by PERFEKT DEVICE: for two weeks the group painted the construction fence in front of Vienna’s Palais Rasumofsky, and the brush, as the project dossier puts it, was “in the hands of the public”. Via a website, people all over the world sent the painters their wishes, watched live and became part of the picture; a terminal carried the interface as far as the Centre Pompidou, from where the commands travelled to Vienna.¹² In the same year, straight after Paris, the team was simply thrown out of Tate Modern in London; the group moved between invitation and guerrilla action, with a taste for the prank. From there the painting tool took over the distance and gave it a stricter form: no longer the wish, but the stop. The audience thus appeared not as spectators but as fellow players: a red button halted the image run and inscribed an alien will into the emerging picture, publicly for the first time in November 2011, in Vienna’s club culture and on 18 November in the Vienna Künstlerhaus (fig. 1b). The first button was home-made and is lost; in its place today stands a large red industrial push button, similar in aesthetics but off the shelf.¹³ In the same month a border already lay between button and canvas: at a club night in Vienna the audience pressed a buzzer, the signal steered the wheel at the painters’ side in Budapest, and the picture came back via livestream.¹⁴ This intervention, a button, a stop, a placement from outside, remained constant across the entire change of medium and returns in Magenta as the smartphone remote control (ch. 2.4). In it lies the point of the whole arrangement: which image stands is decided not by the painting hand but by the interrupting access; how it enters the picture is decided by the hand; authorship is shared from the outset and delegated to chance (ch. 3.5, 4.2). What has survived from this is collective painting with an audience, painting alone in the studio and, as the most recent form, the purely digital WOF – which does without the brush and may yet claim to be painting in an expanded sense (ch. 4.5).

Painting onto a running projection follows a different logic from painting before an empty canvas. The further the process advances, the harder seeing becomes: the projected image lays itself over the elements already layered on the canvas until the two can hardly be told apart. To recognise what is to be placed next, the painter continually shields the projector light from the surface with their own body – a constant darting back and forth between projector and canvas, in which the body does not obstruct the picture but serves as the tool that first makes the already-painted visible.¹⁵ The picture is not given here; it must be brought forth in performance – through a bodily act: even the analogue practice treats seeing as an act, not a reception (ch. 3.4). With every layer the surface condenses, until hardly any room remains and passages must be roughly painted back to regain space. A subsequent correction of the placement is unknown to the procedure – “What fell, fell like an axe”¹⁶; this irrevocability binds projection painting to the logic of the stop and to painting alla prima, in which something final attaches to every placement (ch. 3.5). The group came from graffiti, where this rule holds anyway: what stands on the wall can soon disappear, painted over or removed. The same momentum survives in the purely digital versions, and in Trinity it intensifies: not one stopped or dropped image must enter the work, but every image that falls into the last twelve seconds, and the cut takes them as they were (ch. 2.5.4, 3.5).

In this condensation lies the conceptual core of the practice. Working with already existing images resembles less designing than the continual arbitration of competing claims: a random image stops on the canvas and must be fitted into what is already there, so that heterogeneous, mutually alien pictorial elements wrestle over the same limited space – a Mächteverhältnis der Bildelemente, a power struggle among the pictorial elements, as the artist calls it.¹⁷ It carries a social dimension: the canvas becomes a model of the forced simultaneity of competing interests, especially where motifs collide in content (such as a weapon next to a Virgin and Child from Christian art). WOF painting makes a composition out of this simultaneity – in which the finding of the image-act chapter already announces itself in the analogue practice (ch. 3.1b). Whether the social reading is set consciously or only recognised in the working process cannot be decided; deliberate making and retrospective recognition interlock – “Those who do not see cannot find.”¹⁸ With this, meaning is assigned its place: sense arises not in the object but in seeing (ch. 3.4). For a single painting the stop falls many times, and the artist has set a rule: every stop must enter the picture. That may be a single stroke or a process of hours. Painting is fast and gestural, with marker, acrylic, spray can or oil, and so a picture contributes much or little, but none contributes nothing. When an audience is present, it often presses faster than one can paint; painting processes break off, and new material already lies in the light of the projector. In this the canvas resembles the truth of our media century, full of advertising and information that nobody can fully process: too fast to read, like the images that run through WOF.

Over fifteen years the instrument wandered through several technical generations (Flash, React, finally vanilla JavaScript) and became a fully browser-based machine. The step to the pure browser work – without canvas – is comparatively recent and closely tied to a new possibility: developing the instrument independently with the help of AI, without depending on third-party programmers. Publicly, WOF was first to be seen in this form in 2026 in Magenta, purely digital, without painting (ch. 2.2–2.4). What initially limited the practice through the lack of programming skills is thereby inverted – the artist becomes the sole builder of the instrument (ch. 5, 8). This genesis is essential for the classification: WOF is not a crypto project retroactively aestheticised, but a painterly inquiry, consistent over fifteen years, that has changed its medium.¹⁹ Changed, not abandoned: the painting tool is in use to this day, and what stands at the end there is a painting, not a loop.

2.2 Magenta, the Sequence of the Collage: Spin – Drop – Collage – Finale – Loop

The description in this section applies to Magenta, the installation in Collage mode as it ran in the Magenta Art Room in 2026; Mono Scroll in the finale Non Stop, as Trinity plays it, knows neither the image drop nor the tap that triggers it (ch. 2.5). It describes what the Collage mode can do; Magenta used only part of it and ran with a single mask and a single finale. The interaction is reduced to a single gesture. Viewers trigger the image run (a key press or, in Magenta, a tap on their own smartphone); images rush past; a second impulse drops an image. The image visible in the viewing window becomes a layer of a growing collage. Drop, in the work’s vocabulary, is the moment in which something that chance has brought falls into the work; in the Collage it is an image, and the moment is determined by the human hand with a tap: an image drop. Mono Scroll knows the drop on another level (ch. 2.5.4). Animated masks – organic, geometric, jagged forms – give each dropped image its contour but do not expose what lies beneath: layer settles upon layer, what was placed before remains covered; filters shift colour, contrast, texture.²⁰ Only the conclusion reverses this. After several placements the collage is “finalised”: a brief theatrical sequence passes into an end state – the “Finale”. It knows two forms: either portals open in the finished collage, through whose holes the flicker of the image stream beneath shines up – the dropped image releases the view onto the still-running wheel. Or a single surface, cut by a mask form, flickers within the standing image. From the end state a short video loop is generated.

Decisive is the anthropological point of the procedure, which the artist has formulated repeatedly: “The only human act is the decision to stop. That very moment makes the work.”²¹ There is no curation of the individual images, no algorithm that “makes art”, no composing hand. What composes is the collision of code, open archive and the human instant. This sentence was said about Magenta, and it stands; Trinity does not contradict it, it multiplies it. There the human being is present at several points in the work: before the game as the one who chose every image and ordered it into classes, and in the game as the one who throws the running direction anew, holds, divides the surface and ends: none of these gestures halts the wheel, each configures it, and the cut is the last of them; in between, the instrument rolls the dice. The shift from the one act to the layering of several acts is described in ch. 2.5 and made into an argument in ch. 6.

This description renders a state, not a final state. WOF is a work in performance: which modes, masks, filters and end forms – the “Finale”, for instance – are available is continually altered and extended; the instrument is rebuilt in an ongoing collaboration of artist and machine.²² There is no finished state – the work is less a completed object than an evolving structure that takes on a different state with every working session. This structural openness is recovered iconologically in ch. 3.4.

2.3 The Material of Magenta: the Open Archives

In Magenta, with smartphone remote control and pure internet access, WOF drew its raw material not from a local holding but at runtime, via public programming interfaces (APIs), from fifteen open sources across three continents – connected live, selected from a continually vetted catalogue of more than 150 open sources.²³²⁴ The core is formed by six museum and scientific sources – the Metropolitan Museum of Art, the Art Institute of Chicago, the Cleveland Museum of Art, NASA’s image archive, the Library of Congress and Wikimedia Commons – extended by the Smithsonian, Europeana, Harvard Art Museums, the Statens Museum for Kunst, the Victoria and Albert Museum and others.²⁵ The museum holdings are predominantly available under CC0 or “Open Access” licences. The artist describes this holding as “the digitised memory of humanity”, which is “just lying there in the open”, “and nobody looks at it”; WOF is the machine that reads this library: “not systematically, not purposefully, but randomly — the way a hand reaches into a stack of photographs and pulls one out”.²⁶ That in Magenta not a single image was stored locally has a conceptual consequence that becomes literal in Magenta: “No network = black screen. The installation IS connectivity.”²⁷ This does not exclude the other feeding: WOF can equally draw from a deliberately curated, local image collection – a holding in the sense of Gerhard Richter’s Atlas (ch. 4.5). Both paths, the open archive access and the curated collection, belong to the work’s repertoire; Magenta used only the first, Trinity (ch. 2.5) uses only the second; which feeding an epoch uses belongs to its recipe (ch. 2.5.1), and nothing prevents a later one from mixing both or switching between them during the run.

2.4 The Site: Magenta Art Room, Volkstheater

From 1 June to 31 July 2026 WOF ran in the Magenta Art Room, a glass vitrine in the Vienna underground station Volkstheater (daily operation in step with the underground, without a single day of downtime).²⁸²⁹ It is the first installation of the work to be played from one’s own device – and thus a new form of play: passers-by enter via their own smartphone and play along. This set-up arose for the Magenta Art Room but is not bound to it; it can equally take place in another context. For the campaign WOF moreover received its own livery – leading colour, style and appearance were set so that the work fits seamlessly into the exhibition’s visual identity. The Magenta Art Room is an invited exhibition context: the brand provides site and frame; the artistic setting – work, rules of play, content – lies with the artist. Owing to the shallow depth of the vitrine, an ultra-short-throw projector projects onto the rear wall; the installation is run by a Mac Studio computer in kiosk mode with automatic recovery. The choice of the leading colour establishes a link to the campaign’s motto, under which the exhibition stands – “magic”: magenta is a non-spectral colour that does not occur physically in the rainbow and is “invented” by the brain – an image for the work’s thesis that meaning arises not in the object but in perception.³⁰ The site ranks among Vienna’s most frequented transport hubs; according to the operator, around 420,000 people pass the vitrine there every week.³¹ Passers-by become fellow players via a QR code on the front pane: the smartphone becomes a one-button remote control, without app, without registration. The play situation, meanwhile, is anything but a quiet cabinet: neon light, trains pulling in, the current of those passing — the site works against the very absorption in the screen that the smartphone usually grants. The game additionally demands a constant looking back and forth between two screens, the remote control in the hand and the image behind the glass — a sensory surcharge that the habitual screen gaze of the underground precisely does not know.³² In this the installation repeats a figure from the work’s analogue early period: as the painter shuttles between projector and canvas to bring the picture out from under the light (ch. 2.1), here the gaze shuttles between hand and vitrine — seeing remains an act performed against resistances, not a stance of reception (ch. 3.4). This sensory economy, however, belongs to the set-up, not to the work as such.

What arises between the players is a peculiar social form. The underground knows as its basic stance the polite non-acknowledgement: one perceives one another without addressing one another.³³ WOF does not terminate this contract; it uses it. The gaze into one’s own phone is its daily practised gesture, and a second one belongs to it: the swipe with which the thumb pushes the feed along. WOF sets a gesture of its own against it: the image drop, which lets one image fall out of what otherwise swipes past. Whoever joins keeps the device in hand and remains anonymous; yet the display shows only the controls – the emerging image lives, until the Finale, in the vitrine alone. The screen tunnel is thus not broken open but bent: it ends at a “half-place” – a screen again, but away from the device; the fingers remain on the touchscreen, the eyes must look up.³⁴

The vitrine showed several cells, each running a game of its own. So strangers stand side by side before the same glass, each in their own game, without opponent and without conversation — closer to the pachinko parlour than to the convivial promise of participatory art (ch. 3.5, 4.5): the collective of friends that once inscribed itself into the same canvas (ch. 2.1) returns as a side-by-side of strangers — the social form of the site, translated into a form of painting. Breaches of this silence remain the exception: side glances at the neighbouring cell occur, especially when groups play, but in commuter traffic anonymity prevails; even the Finale releases the passers-by without spectacle — one may walk on, silent, in one’s own run. The artist has described the pure form in a self-experiment: withdrawing into a corner opposite, letting the stream of people pass, playing in silence — “not without enthusiasm, but silent”.³⁵

The site exacts a price for this: where the current tolerates hardly any lingering, whoever stands still and plays becomes a rock in the stream — playing turns into a small test of attentional courage, and only a few take it up; for future set-ups it follows that a place of calm, created along with it, belongs to the game. Finally, play happens even when nobody is there: daily operation includes the auto-run — the work then spins, drops and finalises by itself (fig. 3c–d); what this means for the status of these unwitnessed images is negotiated in ch. 3.4.

2.5 The First Epoch: Trinity

In September 2026 the work established its first epoch. It bears the name Trinity, appears in four editions, each on its own blockchain (ch. 2.5.1), and is no longer played at one site but in the players’ browsers.³⁶ What has changed since Magenta is not a detail of the procedure but every one of the variables that the overview at the beginning of this chapter distinguishes: the site, the material, who plays, the gestures and the shape of the work. The following sections begin with the concept that makes Trinity graspable as a case of its own in the first place, and end with the site.

What Trinity is can also be said without the vocabulary of the following sections. Three image collections that have nothing to do with one another thematically run through the surface at the same time, in one to five lanes side by side, two or three at the start, image after image, too fast to read each one. What the players take with them is none of these images on its own, but what assembles itself in the mind out of their sequence, differently for each person. Two seconds of holding make a work of it; the last twelve seconds of the run become the loop and can be minted as a digital object on a blockchain. Trinity is therefore two things at once, a game that takes place in time and an object that is cut out of it. The name means the threefold of the collections; what it designates is the mixing in time, a movement that leaves an image inside the beholder, an imprint.

2.5.1 Epoch as a Concept of the Work

Since 2026 the work has been organised in epochs. An epoch is neither an exhibition nor a version of the instrument, but a fixed recipe: which mode is played, which finale, from which holdings, in which colours, with which freedoms for chance and which for the human hand. This recipe is fixed when the epoch is opened and delivered to every device on which the work is played. Part of it is a cap, a maximum number of works that closes the series; once it is full, the recipe is final. The series is carried by a collection on its blockchain (where the blockchain cannot check the cap itself, the work’s server checks it, ch. 7.4). What arises within an epoch is related like the sheets of a portfolio; what separates the epochs is not time but the rule. The artist decides the rule, the players decide the sheets.³⁷ In this count, the Magenta run is in retrospect epoch zero: it possessed all the features the concept demands, recipe, cap and a seal over the finished series (ch. 7.4), without having been called that. The first epoch appears in four editions, and each edition belongs entirely to one blockchain: all one hundred works of Trinity Ultra Sun are minted on Polkadot, all one hundred of Deep Aurora on Cosmos, those of Blue Gold on Ethereum, those of Blood Tide on Tezos. The count is therefore kept per blockchain, and on each of them the epoch bears the number one. A single work is minted only once, on the blockchain of its edition, which is fixed at the moment of sending. What connects the four editions is the rule of the epoch and its material: the same mode, the same finale, the same sets in the same three classes and the Chaos, the same cap of one hundred. What separates them is their appearance. Each edition has its own recipe of colours and effects: Blue Gold breaks around a third of the images down into a coarse pixel grid, Blood Tide into a grain that follows a Hilbert curve, and each has its own colour pairs. From the same material four aesthetics thus arise, and that is why the editions carry four epithets. The contracts on Ethereum and Tezos have been on the mainnets since October 2026, and since 23 October 2026 Trinity has been public on all four blockchains.³⁸

2.5.2 The Holdings: Collection Rather than Archive

Trinity does not reach into the open archives but into the artist’s collection: images gathered over years and, for this epoch, greatly expanded in the autumn of 2026, just as one finds them on the net. Fashion spreads, confectionery, fruit, busts, paintings, guns and so on. The sets carry English names because the name of a set accompanies the work as metadata (ch. 7.4); they are sifted, named, veiled and topped up in the Set Hub, a tool the artist built for the purpose, and a minimum size of around five hundred images per set is meant to prevent a set from repeating itself in the course of a game.³⁹ With Hito Steyerl, these images are poor images: copies in circulation, at the size of social networks, with handles and watermarks, without provenance and without rank. Even where a set shows works of art history, busts, paintings, these are not the museums’ digitisations but their copies in circulation.⁴⁰ It is the same stock from which the work’s painting already drew (ch. 2.1), only that it is now no longer the model for a painting but goes into the work itself. The sets have a predecessor that is in use to this day. As early as 2011 the group offered painting with the wheel as a service: whoever bought it chose from “motif sets”, comics, art of the Quattrocento, muscle cars and so on, or had one compiled, say “Oma Emma’s wunderbare 90 Jahre” (Grandma Emma’s wonderful ninety years).⁴¹ On this pattern the artist paints on commission to this day; in the autumn of 2026 two such commissions are under way. The curated set is thus older than the epoch, and it is not merely a stock but a material that someone selects for a particular picture.

This paper named a model for this second path early on, Gerhard Richter’s Atlas (ch. 2.3); ch. 4.4 measures out the comparison.⁴² Two differences may be named here already: Richter’s panels stand, the wheel runs. And the sets do not remain working material beside the work; they enter into the work, are named in the metadata and yet remain a living collection that grows and shrinks, while the minted works carry along the state of their playing time (ch. 7.4).

The step from the open archive to the curated collection has a counterpart outside art, which is only noted here because ch. 6 turns it into an argument: the critique of large language models has demanded “investing resources into curating and carefully documenting datasets rather than ingesting everything on the web”.⁴³ Trinity does for an image work what is demanded there for a language model.

2.5.3 Three Classes and a Disruptor

The collection is not played as a whole but in three classes that run against each other. The classes are not sorted by motif but by what an image does to the person looking at it. Eros is everything one wants to consume and take into oneself: fruit, sweets, fashion, skin, the last of these pixelated, as in the Japanese image culture in which the veil belongs to the image. The artist calls it “hunger and lust” and adds: “That can also be a peach.”⁴⁴⁴⁵ Thanatos is the image one cannot look away from: weapons, war, plague, explicit content, the hypermasculine image of women with guns.⁴⁶ Logos is the third class and the only one that does not shout: Greek busts, Renaissance, quiet faces, mourners of stone, Rodin’s Thinker as its model image.⁴⁷ Two classes are loud, one is quiet, and the quiet is not the extinguished but the thinking, that which holds the other two within itself, Winckelmann’s “quiet grandeur” that holds even in pain.⁴⁸ In the artist’s words: “The calm actually radiates the breaking out – always the opposite.”⁴⁹

Behind the threefold division stands an old figure of European thought, which may be named briefly here; ch. 3.4 takes up the threefold division again with Warburg. Plato divides the soul into desire, spirit and reason, epithymētikon, thymoeides, logistikon, and exactly these three stand here; the assignment Eros, Thanatos, Logos is the work’s, not Plato’s.⁵⁰ Freud set Eros and the death drive as the two classes of drives that work against each other in the human being; the name Thanatos was given to the death drive only by its reception.⁵¹ And the word Logos is older than the Gospel: in Heraclitus it is the measure that holds the opposites together.⁵² Painting knows the triad as a form. Bosch’s Garden of Earthly Delights places Paradise, Lust and Hell side by side on three panels; Trinity is such a triptych, only that the panels run and mix.⁵³ The name therefore does not mean the dogma but the older figure, three stimuli and one head in which they collide, and at the edge the first nuclear weapons test, which bore the same name.⁵⁴ The echo is not mere accessory. Whoever grew up in the mass media, in advertising and television and then on the net, has lived through an explosion that has no counterpart in the history of humanity: that of visual stimuli, whose number has multiplied within one generation. Trinity is the reaction of a person who grew up inside this effect, and two paragraphs further on the artist says so in person.

The contrast is the work. The orange becomes obscene because a gun lies next to it; the bust becomes a reproach because sugar sticks next to it.⁵⁵ None of the images says anything; it is said between them. That is the principle by which Aby Warburg built his Mnemosyne panels, unlike images side by side, so that what none shows on its own becomes visible (ch. 3.4),⁵⁶ and it is, turned into time, the principle of film montage, to which ch. 3.5 returns. To this comes a fourth set of holdings, which rarely falls into a game and then plays alone: Chaos, images that already carry the contrast within themselves and therefore have no place in any class. Chaos is the antagonist of Trinity.⁵⁷

What is played here is, in the artist’s words, a “psychogram of an internet society”: “You can take everything in visually; I belong to the first generation that grew up this way.”⁵⁸ In conversation the artist has traced the origin of the whole instrument back to this: the wheel makes it possible “to outsource a decision”, and that is “a sign of our time”, in which one is “overrun” by advertising, news and inputs; Wheel of Fortune is “perhaps a kind of reaction to this overstimulation of our inputs”, a “processing of the randomness of our input, which is constantly at maximum”. It is never entirely random: “Our filter runs all the time.”⁵⁹ These sentences are a voice, not a finding of this paper; what the paper takes from them is the observation that the three classes do not name image genres but three reflexes, desiring, fearing, recognising, and that the work places them side by side so that they lift each other out of self-evidence. All three are images that use the person looking at them. The work supplies three baits and leaves it to that person which one takes hold.

2.5.4 The Throw

At the beginning of every game the instrument rolls the dice. From each class it draws one set, from the epoch’s permitted colour pairs one, sometimes two, plus the number of cells, the direction of travel per cell, the share of images that remain black and white, the effects per image, and when the run inserts short stops of its own accord. With low probability the Chaos falls instead of the Trinity: then the fourth set of holdings plays alone, and the three classes stay behind for this game.⁶⁰ That too is a drop, only it is not an image that falls but a set, and it does not fall onto a surface but into time: the set drop. The images of the three drawn sets mix in the cells running side by side, image after image, for as long as the game is played. The image drop was an act of the human being; the set drop is an act of chance, and the human hand answers it in the course of the run. From this stream the cut at the end makes a hard selection: exactly those images enter the work that fall within the window of the last twelve seconds, and how many that is depends on the tempo of the run, on how long each image stands, and on how often the players have halted a cell. This halting, in the work freeze cell, is a gesture of Mono Scroll in its own right (ch. 2.5.5): a halted image fills the window for longer and lets fewer others in. Which sets the dice have drawn, the instrument shows for a few seconds at the start of the game; the rest of the throw, and what the players change about it, is shown by a panel that can be called up; what was played is named at the end by the work’s metadata. Three displays, three truths: the throw, the state, the result. The work thereby claims no more than it knows, and that is a principle that runs through Trinity: the metadata document facts, not rarity values (ch. 7.4).⁶¹

Two rules of the throw deserve emphasis because they carry the dramaturgy. First, the dice draw one set per class, not a mix of all: every game has a cast, and the cast is fate, not choice. Second, the image sequence guarantees that every class appears at least once in every twelve-second excerpt. Chance may do much, but once the Trinity has fallen, it may not roll any of its classes out of the work. The rule names the lower bound, not the measure: in the run the classes change far more often, frequently from image to image; what is guaranteed is only that none is missing. Odo Marquard’s formula “We human beings are always more our accidents than our choice” has its mechanics here: the players do not choose what comes, they choose what they do with it.⁶²

2.5.5 The Game: Mono Scroll

What is played is a mode the work calls Mono Scroll. What is dropped here is not an image but, with the throw, a whole set (ch. 2.5.4). The wheel itself is the work: a band of whole images that runs through the surface in one to five cells (the wheel throws two or three at the start, the gestures can make one to five of them), each image in the colour of a drawn pair, one after the other, too fast to read each single one. That is exactly what is meant. Out of the change of images arises, in the artist’s words, “a new granulate with time as its factor. Too fast is just right, 2026.”⁶³ Colour belongs to the same form. The work plays its images in duotone: each is reduced to the two colours of a drawn pair, the shadows taking on the one, the highlights the other; a share stays black and white, and single images carry effects, a coarse pixel grid or a grain (ch. 2.5.1). No image runs in its original colours. The work thus intervenes in the material itself, not only in its sequence: the find from the net is recoloured before it runs, and only the colour pairs of the epoch give each edition its appearance. For the mixing in the mind this is no accessory; the shared colour binds images that nothing connects thematically into one surface. The cadence is part of the form: the work runs at fourteen frames per second, in every cell the image changes about nine times, and the artist fixed this beat against the obvious wish for smoothness because the choppiness belongs to the formal language and keeps the file small, an economy that the net has demanded of its images from the beginning: what is to travel must be light, and the work takes this condition not as a lack but as a form.⁶⁴

The human hand has four kinds of gesture in the run, and none of them is the tap of the Collage that let an image fall into the composition. A tap throws the direction of travel of a cell anew; a short hold freezes a cell for a good second; a drag shifts the seams between the cells, splits them, removes them or rotates the grid from portrait to landscape. All these interventions travel into the work: the changes of direction and the freeze points of the last twelve seconds are part of the score from which the loop is computed (ch. 5); seams and grid travel as the final state.⁶⁵ That is the precise sense in which WOF is participatory. Not the presence of the players enters the work, nor their image, but their gestures. The fourth possibility ends the game. At the earliest after half a minute: whoever then holds the surface for two seconds or taps twice the button that the work offers for this makes their loop: the last twelve seconds of the run become the loop, the rest stays behind. The wheel never stops in the process. The finale of Trinity is called Non Stop, and the instrument refuses a work whose wheel would be standing still at the moment of the cut.⁶⁶ What follows is no longer an animation, no portal, no flicker as in Magenta’s Collage mode, for the work is not the result of the game but its course. The artist has described this difference from the Collage as the architectural core of Mono Scroll: “The game itself is the animation that then renders.”⁶⁷

With this, the human act has moved to another place. The Collage knows the drop: a point at which an image falls into the composition. Mono Scroll knows the cut: an interval that is cut out of the running wheel. Whoever finalises chooses not an image but a duration, and the window lies backwards; what is chosen is what was, not what is to come. The Rota of Fortuna, which ch. 3.3 deals with, here turns right into the loop. A wheel that stands would no longer be hers (ch. 3.3); this one stands only for a good second, under a finger on the screen or the pressed mouse. One cell is held, and the others keep running; if only one is running, the whole wheel stands for that moment. Then it runs again, even if the finger stays put, and whoever keeps holding sends the run into the finale.

2.5.6 Veiling as a Deliberate Act

In most of the sets that show people, the work covers them: faces rasterised, lettering covered, sifted before the game, not after it. The veiling is a deliberate act that remains visible. It is set, chosen by hand, with two means: the pixel and the bar. Over faces the artist draws boxes in the Set Hub, and what lies beneath breaks down into a coarse grid; the face remains recognisable as a form, not as a person. Over eyes the artist lays narrow bars and turns them to the tilt of the head; lettering is covered with bars. A machine proposes where faces, eyes and lettering lie; the decision is the artist’s: which proposals remain, where to adjust by hand, whether a face travels pixelated, with a bar, or with both. Position, shape and angle of every box change the image, and they travel with it into the work. Thus arises, as the artist says, “in part a new aesthetic through bars and pixels”.⁶⁸ The anonymisation is thereby at the same time an aesthetic decision. Veiling here is not a censorship that passes over the image afterwards, but a shaping that happens before the game. The work shows that it does not show something. Only the veiled version goes into the loop and onto the blockchain; the original does not travel along, and its address appears nowhere on the blockchain.⁶⁹ For the quiet class this is the condition of its promise: a face in which one recognises oneself, without anyone being recognised. What this barrier means legally, and why in Trinity it takes the place of the fading that Magenta could invoke, is negotiated in ch. 7.4.

2.5.7 No Place

Trinity takes place at no site. The wheel runs in the players’ browsers, on the train, in bed, on the way; nobody stands beside them, no glass separates hand and image, no code on a pane invites. In the place of the display case comes a start panel that shows the drawn sets and the gestures, then closes of its own accord.⁷⁰ What Magenta had in the way of publicness shifts to the end: playing is done alone, attestation is public. The place of calm that ch. 2.4 demanded of future installations has thus been found, and it is each person’s own. With it disappears the social form that the display case had produced: no half-place, no strangers in front of the same glass, no side by side.

The two sites can be kept precisely apart with Marc Augé. The underground passage in which Magenta stood is, in his sense, a non-place: a space which “cannot be defined as relational, or historical, or concerned with identity”, crossed by people who are “alone, but one of many”.⁷¹ Magenta did not suspend this non-place but used it (ch. 2.4). Trinity has left it, but not in the direction of a place. Nor is it what Ray Oldenburg calls the third place, the “regular, voluntary, informal, and happily anticipated gatherings” beyond home and work;⁷² it is the opposite of that, a game without co-present others, in which the only witness is the blockchain. Sherry Turkle’s diagnosis of a generation “that does not necessarily take simulation to be second best” describes the players of this epoch; the artist counts as one of this first generation (ch. 2.5.3).⁷³

The price of this site is a new status of the work. Magenta knew the unwitnessed images of the work running on its own (ch. 3.4). Trinity knows no running on its own without humans; every work needs someone who began it and wanted it. But not everyone claims their work. Whoever finalises holds for seven days the privilege of minting the work; after that it stands open to all for fourteen days; after that it lapses, and a work that someone made and nobody claimed is taken off the node.⁷⁴ What the unwitnessed images were to the display case, the attested but abandoned ones are to the game in the browser: works with an author and without a claim. Chapters 3.4 and 7.4 take up this figure.

With this, the object has been outlined in both of its shapes. The iconological analysis that follows asks what actually appears in these processes.

3. Iconographic and Iconological Analysis

3.1 Method: Panofsky’s Three Strata

Erwin Panofsky distinguishes three strata of pictorial meaning and, corresponding to them, three acts of interpretation.⁷⁵ The first, primary or natural meaning, is disclosed through pre-iconographic description: the plain perception of forms, objects and expressive qualities. The second, secondary or conventional meaning, is the object of iconographic analysis: the linking of motifs to themes, stories and allegories, which presupposes cultural knowledge. The third, the intrinsic content, is gained through iconological interpretation: those underlying principles that reveal the “basic attitude of a nation, a period, a class” and render the work legible as a symptomatic document of a worldview. This model, developed for the Renaissance, is here applied to a technical work; at one point this requires an explicitly declared extension (3.3).

3.1b Reflection on Method: from Depiction to Performance

Panofsky’s three-step model is today less a guiding paradigm of current research than a canonical descriptive grid.⁷⁶ It is nonetheless deliberately used here as scaffolding, not least at the artist’s place of study, whose Viennese School itself produced the tension between iconological meaning and structural-analytic seeing.⁷⁷ Its limit, in the case of an operative work like WOF, must be marked expressly.⁷⁸

This limit lies where the wheel runs rather than hangs (ch. 3.3). The question thereby shifts from Panofsky’s iconographic logic of depiction (“which motif is depicted?”) to the logic of performance proper to image studies (“what does the image do?”). This shift is no makeshift; it follows the image-theoretical turn since the 1990s: Horst Bredekamp’s theory of the image act conceives the image – in analogy to speech-act theory – as an acting, efficacious instance rather than mere representation,⁷⁹ and Gottfried Boehm’s iconic difference binds pictorial sense to showing itself, which cannot be exhaustively translated into a legible motif.⁸⁰ WOF can accordingly be read as a performative image act that performs the Rota Fortunae rather than showing it – the lawfully running image run and the human instant of the drop are the action in which the image happens.

Four further connections sharpen the logic of performance, and they are needed for Trinity. Lev Manovich proposed, “[f]ollowing art historian Ervin Panofsky’s analysis of linear perspective as a ‘symbolic form’ of the modern age”, to call the database “a new symbolic form of the computer age”: a list of items that “refuses to order this list”, whereas a narrative “creates a cause-and-effect trajectory”.⁸¹ The wheel is, in this sense, the unordered list in motion, every element of equal rank; the drop, like the cut, draws the trajectory out of it, Manovich’s syntagm out of the visibly passing paradigm. Where this trajectory is drawn, Manovich does not say; the answer is given by Hans Belting, who shifted the question of where an image takes place from the medium to the body: images “resemble nomads”, “taking up residence in the media of one historical place and time and then moving on to the next, like desert wanderers setting up temporary camps”, and the body is the “locus of images”, which, at the “rapid pace” of their coming and going, “endows some images with symbolic meaning and admits them to memory, others it consumes and forgets”.⁸² That is the mechanics of the wheel before it has a theory: archive, stream, drop, loop and blockchain are stations of a migration, and the place where it is interrupted is a body.

What happens in the body is a doing, and for this Erika Fischer-Lichte found, for performance, the sentence that covers the methodological decision of this chapter: the actions of a performance “did not simply mean” what they showed; “instead, they accomplished precisely what they signified”.⁸³ Hans Ulrich Gumbrecht’s distinction between presence effects and meaning effects, which “always appear together” and “are always in tension”, prevents performance from being shortened to mere presentness: what is presence in the drop becomes, after the fact, meaning in collage, loop and paper, and the work lives on this oscillation (ch. 6.6).⁸⁴ Read together, the four describe not four theories but one path, and it is the path of every image through the wheel: it lies in a list that nobody ordered; it becomes an image in a body, which keeps or forgets it in the rhythm of the passing; what happens there is a performance, not a depiction; and it receives its meaning only once the passing is over, in the collage, in the loop, on the blockchain. Ch. 2.5 described this path before it received its concepts here.

3.2 Pre-Iconographic Description

On the first stratum there appears, for Magenta in Collage mode: a portrait-format image stream in which single images run past in randomly changing directions and accelerate; at the drop the stream fades out, and the image that was just visible falls into the collage; the mode can play one to three cells and both formats, and Magenta ran with three cells in portrait format throughout. A succession of dropped images layers itself, overlapping and cut by irregular forms, into a dense surface. Colour shifts, screen patterns and blurs cover parts of the surface. At the end a flicker pulses – partly behind openings, partly within a magenta field. The movement is circling in a figurative sense: a rolling up and down of images, a break, a new beginning. Even this purely descriptive level lays bare a basic figure – that of the turning, interrupted, restarting wheel.

In Mono Scroll the first stratum changes at three points. There is no overlapping: the images run through singly, not layered, at the start in two or three lanes, through the gestures in one to five, each in two colours, some black and white; the next image pushes itself over the previous one as a surface with a travelling edge. Even a single lane carries the whole game, for the thrown sets mix within it too. There is no standstill: the band falters, changes direction, runs on, and at the end stands not one image but twelve seconds of run that repeat. And there is no flicker at the end, no opening, no field; the loop is the band itself. The basic figure remains: a wheel that turns, that is intervened in and that keeps running; what is new is that it is not the single image drop that counts but the work with time, in which a loop is made out of the running flood of images. This run without end has a counterpart outside the work that every player holds in their hand: the image run of the phone, which knows no end and which the present calls doomscrolling, a wheel that keeps you in the wheel. Mono Scroll does not depict this wheel, it plays it, with the one difference that out of the run that never stops, twelve seconds can be cut that belong to you. In doomscrolling the thumb knows only one gesture, swiping on. Mono Scroll gives the human hand others: to halt, to throw the direction anew, to divide the surface, to cut. Reflection on existence in the endless loop is thus not a theme of the work but its form.

3.3 Iconographic Analysis: the Rota Fortunae

On the second stratum this wheel can be named. The title Wheel of Fortune quotes not a television show but one of the most long-lived pictorial motifs of Western culture: the Rota Fortunae, the wheel of the goddess Fortuna. The motif receives its canonical formulation in Boethius, whose Consolatio Philosophiae (c. 524) has Philosophy confront the one who laments: whoever wants to halt the wheel misjudges it, for “if it beginneth to stay, it ceaseth to be fortune” (II pr. 1); Fortuna herself, directly afterwards, avows her sport with the wheel (II pr. 2).⁸⁵ In the Middle Ages the Rota Fortunae becomes a fixed pictorial type: the well-known miniature of the Carmina Burana (c. 1230, Munich, BSB Clm 4660, fol. 1r) shows the crowned Fortuna at the hub, surrounded by four royal figures with the rhymed four-step regnabo – regno – regnavi – sum sine regno (“I shall reign – I reign – I have reigned – I am without reign”).⁸⁶ On the same leaf stands the lament O Fortuna on the fickleness of fortune. Related depictions are found in the Hortus Deliciarum of Herrad of Landsberg (c. 1167–1185, destroyed by fire in 1870) and, as an architectural translation, in the rose window of the façade of San Zeno Maggiore in Verona (12th century).⁸⁷⁸⁸ A late continuation – occultly charged in the 19th century – is the tarot card The Wheel of Fortune (Major Arcana X), whose wheel, bearing the letters T-A-R-O, to be read according to Waite as Rota, signifies cycle, destiny and sudden reversal.⁸⁹

The meaning of the motif is constant across all sources: chance, mutability (mutabilitas), rise and fall, the opposition between inconstant Fortuna and constant virtue or providence. It is precisely this semantics that WOF invokes. Unlike a painted wheel of Fortuna, however, WOF shows no depicted emblem; the motif appears not as pictorial object but as a figure of movement. The iconographic attribution here thus proceeds not via a depicted motif but via the structural recognisability of the wheel process – a productive extension of Panofsky’s second stratum, methodologically grounded in 3.1b, whose full import only the iconological reading redeems.

Boethius’s sentence that a halted wheel is no longer Fortuna’s wheel fits the work more exactly than any emblem. Neither of the two cases halts the wheel for good: Magenta ran on, even without an audience, and even its finale is a loop that keeps running. In Collage the image drop does not halt the wheel; it only fades out for the instant in which a single image falls out of the run and fills the surface in the drop; after that it turns on, and it remains Fortuna’s wheel. Mono Scroll, in the finale Non Stop, halts a cell only briefly, and the cut takes out without stopping the wheel. The four-step of the miniature returns in Trinity as the time limit. Whoever finalises will be allowed to mint, regnabo; whoever mints holds the work, regno; whoever passes it on has held it, regnavi; and what nobody mints falls from the node, sum sine regno (ch. 2.5.7).

3.4 Iconological Interpretation

On the third stratum the analysis asks after the intrinsic content – what WOF, beyond its theme, is a symptomatic document of. Here, too, the point announced in 3.3 is redeemed: WOF does not depict the Rota Fortunae; it reactivates it as an operative form. The nearest answer every player holds in their hand: WOF is a document of life in the feed, of a time in which images pass by without end and the thumb does nothing but swipe (ch. 3.2). Three lines of interpretation show what the work makes of this.

First: distributed contingency rather than depicted chance. The Rota Fortunae negotiated human powerlessness before fate. WOF displaces this figure without cancelling it: it is not a deity that turns the wheel but a machine drawing on the collective image memory – and the human being is not only the wheel’s victim but also the one who intervenes at the right moment; from the wheel itself there is no escape (see below). What is remarkable here is that the run itself contains no chance: the machine follows fixed rules; in Magenta’s Collage, to every moment of a drop belongs exactly one image.⁹⁰ Chance enters at two points – at which images get into the stream at all (decided by the searching movement with which the work roams the archives, ch. 5), and at the instant of human intervention. WOF thus resembles the medieval figure more exactly than first appears: the wheel itself runs according to law; contingent is only the moment at which it strikes someone. Chance here is not depicted but performed.

For Trinity the finding holds with a shift that confirms it. Mono Scroll, too, runs on a deterministic clock: image sequence and beat can be computed from the game state; Trinity knows no stop; the server generates the loop from the score without the device and rolls no dice in doing so (ch. 5). What is contingent has multiplied and shifted. In place of semantic navigation comes the throw, which draws the cast for each game; in place of the one instant comes the sequence of interventions, changes of direction, halting, cut. Odo Marquard, with Aristotle, has defined chance as what comes about “through the unexpected mutual encounter of chains of determining events that are independent of one another”; Boethius himself calls it “an unexpected event of concurring causes”.⁹¹ That is exactly the structure of both cases: none of the chains – the lawful run, the drawn cast, the human hand – is chance in itself; what is chance is their meeting. That this definition stands in Boethius, the chief witness of the Rota (ch. 3.3), closes the circle. How the wheel differs in this from a machine that weights chance and makes suggestions out of frequency is set out in ch. 6.1.

Second: the “Death of the Author” performed. Roland Barthes’s dictum that the text is “a tissue of quotations drawn from the innumerable centres of culture”, and that meaning arises not in the author but in the act of reading, becomes literal in the purely digital form of WOF.⁹² The image stream is a tissue of quotations; the “reader” who in Barthes takes the author’s place is here the person who plays, whether they trigger the image drop in the Collage or, in Mono Scroll, cut their twelve seconds out of the set drop that chance has thrown. The iconological point: WOF reinterprets the wheel of Fortuna from an image of fate into an image of authorship. What the Rota withdrew in power – control over one’s own ascent – WOF withdraws from the artist: control over which individual image stands in the work. The renunciation is the artistic gesture (cf. ch. 4.2 and 8).

Two older voices support the figure, one before Barthes, one after him. Georg Simmel wrote in 1911 that most products of intellectual creation “contain a certain quota which was not produced by ourselves”, something significant “which other people can extract even though we were not aware of having deposited it there”: “the weaver doesn’t know what he weaves.”⁹³ Meaning lies with the reader, half a century before the Death of the Author. In WOF, in its purely digital form, this quota is not small but almost the whole: every element that enters the work is taken from the stream, none of it was made by the artist, and Simmel’s sentence is here not a qualification but an exaggeration that happens to be true. The weaver really does not know what is being woven; the weaver knows only that not one thread of it was spun by them, and is no less for that the one who made the fabric. The remix is something new, and the orchestration is the artist’s own. And Espen Aarseth, on the question of who controls a text that only forms itself in use, distinguished three positions, “author control, text control, and reader control”.⁹⁴ WOF distributes across all three: rule and holdings with the author, run and throw with the text, stop, drop or cut with the reader. The Death of the Author is here not a metaphor but a distribution, and how far it reaches in Trinity is shown in ch. 6.5.

Third: the archive as material and the double sense of fortune. The English fortune carries two meanings – luck/fate and wealth/riches. WOF activates both. On the level of the image, fortune is the luck of the instant; on the level of property it becomes, as soon as a collage is minted as an NFT, fortune in the economic sense (ch. 7). The title is thus no label but the hinge of the work: it binds a medieval motif of fate to the present question of digital property. In Aby Warburg’s terms, which Georges Didi-Huberman has renewed for the present, WOF can be read as an afterlife (Nachleben) of the ancient pictorial formula – as the return of Fortuna in transformed, technical guise.⁹⁵⁹⁶ Warburg’s Bilderatlas Mnemosyne is moreover a methodological model for WOF itself (ch. 2.5.3): both are montage apparatuses of the collective image memory.⁹⁷

These three lines share a basic trait that carries the motif beyond the individual image: mutabilitas itself. Boethius’s Philosophy declared that a halted wheel is no longer Fortuna’s wheel – change is her essence. In WOF the same holds for the work as a whole, which remains unconcluded on several levels: every run brings forth a different mix; the search space from which the images came in Magenta can keep growing by machine (ch. 4.4, 5); and the structure of the instrument itself is continually rebuilt (ch. 2.2). Even the work’s preconditions age along with it: Magenta hung on fifteen third-party interfaces, and every edition hangs on a browser environment that changes without its doing — its own continued existence, too, stands under the mutabilitas it negotiates. Standstill would mean for WOF, as for the Rota, that it ceased to be itself. Mutability is thus not merely the work’s theme but its mode of existence – and the “Death of the Author” reaches correspondingly deep: from the individual image through the search space into the never-final instrument.

The mutabilitas has, finally, a time of its own. In Magenta’s daily operation the wheel kept turning even when nobody watched: the auto-run drew, dropped and finalised without a human witness (fig. 3c–d). Nowhere does the work come closer to its medieval figure — a wheel that turns even without witnesses; a Fortuna who does not wait until someone looks. Simmel’s sunrise that no human eye sees (ch. 6.2) is the exact case. The artist leaves no doubt about the status of these collages: they are works. Yet their authorship evaporates — what the machine brings forth alone is discarded or ascribed to the human; a right of its own to be an artist is granted to it by no one, and what is preserved, culturally as legally, is only what humans create (cf. human creation as the threshold of the concept of the work, ch. 4.2).⁹⁸ The artist calls this an “ego thing” — and indeed the refusal joins the series of those blows through which the human self-image had to surrender its special position piece by piece: cosmological (Copernicus), biological (Darwin), psychological (Freud) — and, in a much-discussed continuation, informational, through the thinking machine.⁹⁹ Authorship would then be a last bastion of that special position, and WOF stands on its rampart: it delegates to the wheel what it can, and reserves for the human what it does not concede to the machine. The artist admits that this may be a mistake — that in the idle run art may be lost, unseen and unpreserved. This openness is not a weakness of the position but its test: a work that means the question of authorship seriously must be able to turn it against its own premise.

Trinity brings Warburg into the work a second time, no longer only as the model of montage but as the theory of the pathos formula (Pathosformel): the pictorial form that stores an affective state and triggers it on being seen again.¹⁰⁰ Warburg set it against Winckelmann’s “noble simplicity and quiet grandeur”, against the image of a calm antiquity.¹⁰¹ Trinity plays both antiquities at once. Eros and Thanatos are pathos formulas in the strict sense, gestures of desire and of terror. Logos is not. Its busts and mourners of stone, the Renaissance sets of the class too, carry the calm of the quiet antiquity against which Warburg wrote (ch. 2.5.3). The wheel lets the screaming and the quiet one run in the same beat, and it stores nothing, it triggers. The viewers no longer decide which image stays but which formula takes hold, and that is the Barthes thesis with an affective substructure: meaning arises in seeing, but seeing is not free, it is pulled from three sides.

The mutabilitas, finally, acquires in Trinity a deletion. Works that a human has made and nobody has claimed are taken from the node after a time limit (ch. 2.5.7). Fortuna does not only turn on, she clears the table: what the kings she topples are to the Rota, the works it forgets are to the wheel. The authorship that the work distributes is thus also limited in time: it lapses.

Beneath these lines lies one that needs no technical term: nobody escapes the wheel, any more than Boethius could leave his cell. The Rota shows rise and fall, but its truth is plainer and harder, and everyone knows it from their own life: whoever lives, dies, and nobody gets off. That is why the motif is more than a quotation for this work. It gives the method a core that corresponds to the artist’s work: “I think I have to make art because I know that I have to die. Art is my way of leaving something behind, of being at least a small cog in the whole.”¹⁰² What such a leaving-behind on the blockchain can and cannot achieve is the question of ch. 7.3.

3.5 Excursus: Surrender to Chance (alea)

The iconological reading can be supplemented by its anthropological substrate: the human inclination to surrender to chance. Roger Caillois distinguishes the competitive game (agôn) from the game of chance (alea) — the latter presupposing “a negation of the will, a surrender to destiny”.¹⁰³ Exactly this voluntary abdication is performed by whoever steps before the wheel of fortune: the Rota Fortunae thus appears not only as a pictorial motif but as a structure of need. Johan Huizinga had earlier defined play as a sphere free of purpose, “an activity connected with no material interest”¹⁰⁴ — play for its own sake. What remains of this when the stake becomes work is shown by Roland Barthes’s reading of Japanese pachinko: a game without opponent and without real winnings, in which a single impulse of the thumb decides everything — the hand of an artist for whom the stroke is a “controlled accident”; the machine reproduces “the principle of painting alla prima”.¹⁰⁵ The “prize” is absurd; one starts again from the beginning. The ethnographer Natasha Dow Schüll has described the same structure in machine gambling: in the “zone” the aim is not to win “but simply to continue”.¹⁰⁶

WOF shares this logic down to its form: with several cells side by side it resembles the reels of the one-armed bandit, and the “prize” — the finished image, the later NFT — is subordinate to the performance. In the instant of the drop all of this condenses: every drop irretrievably covers what exists (there is no going back, as with the stroke alla prima), and the player stands in the tension between placing a further stake and halting — press your luck against finalising.¹⁰⁷ Exactly here the game of chance touches the historical avant-garde of the renunciation of control: John Cage handed his Music of Changes (1951) over to the chance operations of the I Ching — a composition whose continuity was to be “free of individual taste and memory (psychology)”;¹⁰⁸ Marcel Duchamp, with the dropped threads of the 3 Standard Stoppages (1913/14), had introduced “canned chance” into art.¹⁰⁹ Before that, Surrealism had already made the renunciation of control a method: André Breton’s “[p]sychic automatism in its pure state … in the absence of any control exercised by reason” (1924) is the historical test case of that letting-chance-decide which WOF repeats technically.¹¹⁰¹¹¹ The drop in WOF inherits all three lines at once, gambling, automatism and chance art: the gesture in which the ego steps back and the instant decides.

Mono Scroll knows no image drop. Its drop is the set drop of the throw, and it falls not onto a surface but into time. With this it seems to lose the anthropological substrate of the image drop: there is nothing to cover and no instant in which chance gives way. What it puts in its place is another form of irrevocability. Time is irrevocable: what is past falls out of the window of the twelve seconds, and whoever cuts takes what was. The artist describes the perception that arises in this way as one that “composes itself over twelve seconds”, a collage in time rather than in space.¹¹² The art-historical name for this is not collage but montage. In 1929 Sergei Eisenstein set the notion of collision against Pudovkin’s notion of montage as linkage: “So, montage is conflict”; meaning arises from the clash of two shots, not from their sum.¹¹³ What ch. 2.5.3 described in the orange and the rifle is an effect of this kind; the Trinity is a montage rule, and the players’ cut is a cut in the literal sense. Lev Manovich proposed the loop as a possible “new narrative form appropriate for the computer age” and described the editing of a film as a “unique trajectory through the conceptual space of all possible films”;¹¹⁴ in WOF this trajectory is drawn in real time, and it ends not in a film but in a loop. The difference is more than a format. A film has a beginning and an end; the loop has neither, it is the endless repetition of an instant. Twelve seconds that never stop are therefore longer than any film: a surfeit that arises not from duration but from recurrence. Siegfried Kracauer, in 1927, in a sentence on the flood of images in the illustrated magazines, already gave the game its name: “The turn to photography is the go-for-broke game of history.”¹¹⁵ What he set against the confusion of the illustrated magazines was a “game” played “with the pieces of disjointed nature”, which establishes “the provisional status of all given configurations”, and he was thinking of film. The wheel is such a game.

That this game is not an arbitrary one has been grasped more precisely by the theories of the open work and of appearing than the concept of chance allows. Umberto Eco in 1962 distinguished pure chance from the “willed” chance that montage puts in its place: life in its immediacy “is not ‘openness’ but chance”, and in order to turn this chance “into a cluster of possibilities, it is first necessary to provide it with some organization”.¹¹⁶ The dice of Trinity are pure chance, if within the bounds of the sets the artist has chosen; the recipe that frames them and the cut that ends them are willed chance. Gumbrecht’s presence, which precedes meaning (ch. 6.6), describes the drop of the Collage as it does the cut: both are a pausing to which meaning falls only afterwards.

That this renunciation is no slackening is shown by the compulsion that comes with it: the artist formulates the duty to take something from every dropped image into the work; chance thus becomes the barrier at which form takes shape in the first place. In the digital procedure the machine enforces it. In the painted variant nothing enforces it, and it would be easy to leave out a displeasing image. The artist does not do so: the rule is self-imposed (ch. 2.1). In painting the barrier is a self-commitment, in the digital edition a mechanism.

In this lies an equation that points beyond the studio: powerlessness in the instant of the stop, for the button presses from outside, and at the same time the duty to take in what it brings. Georg Simmel described in 1903 how the metropolitan protects themselves against the intensification of stimuli with a blasé attitude.¹¹⁷ Sigmund Freud noted in 1920 that for the living organism the protective shield (Reizschutz) is “an almost more important function” than the reception of stimuli; reception should content itself with taking in “samples of the external world”.¹¹⁸ Walter Benjamin built on Freud and drew on Simmel when he described the experience of modernity as shock (Chock) and placed the gambler at the gaming table beside the worker at the machine: each of their hand movements is sealed off from the previous one, “Both types of work are equally devoid of substance.”¹¹⁹ WOF inverts the protective posture. It forbids the painter the protective shield: what the band brings must go into the picture, just as today images arrive on every screen that nobody ordered. Painting becomes the mirror of a time that heaps images on us, and an exercise in not fending them off but working them through. Boethius’s Philosophy demands the same of the one who stands under Fortuna’s wheel: not to want to halt it, but to bear what it brings (ch. 3.3).

The duty to take in is more than a studio rule; it is the statement of the work. WOF turns the condition that nobody can fend off the flood of images into a procedure. As early as 2011 the submission for the exhibition METAmART at the Künstlerhaus Vienna described the painter as one who “seemingly gives up all self-determination” and “becomes the plaything of the recipients”.¹²⁰ WOF thus stands in the line of those performances in which artists handed themselves over to the audience, Yoko Ono’s Cut Piece (1964) and Marina Abramović’s Rhythm 0 (1974),¹²¹ with the difference that here it is not the body that is disposed over but what must be painted next. Alone at the button, the counterpart does not disappear, it moves inward. The painter gives the stop to themselves and splits into one who throws and one who must work through; Byung-Chul Han has seen in this split the signature of the present: “The exploiter is simultaneously the exploited. Perpetrator and victim can no longer be distinguished.”¹²² WOF plays it openly, as a self-imposed rule.

Matthew Barney made a method out of self-imposed impediment. In the series Drawing Restraint, begun in 1987, he draws against resistances he imposes on himself; in Drawing Restraint 6 (1989) he jumps on a tilted trampoline in order to set a drawing on a sheet fixed to the ceiling. His model comes from sport, Barney was a football player: a muscle grows only when it works against resistance; the load injures it a little, and in healing it becomes stronger. Sports medicine calls this hypertrophy, and Barney transfers it to drawing: it becomes form only at the obstacle.¹²³ WOF works on the same principle, except that the resistance comes not from one’s own body but from outside, from chance: the stop interrupts the painting hand, and the painting grows at the interruption.

The attitude the work demands in this has a long history. Seneca closes his rendering of a prayer of Cleanthes with the verse that fate leads the willing and drags the unwilling;¹²⁴ Nietzsche wanted to love what is necessary, amor fati.¹²⁵ Hartmut Rosa has found the term of our time for this attitude: resonance arises only with what cannot be controlled, and through adaptive transformation (Anverwandlung), through a taking-in that changes the one who takes in.¹²⁶ Whoever surrenders in this way no longer stands opposite the stream but goes into it, and the now becomes denser. Groys expressly calls the internet not a flow but its reversal, because it lets one return to the same things;¹²⁷ WOF builds the stream that the net is not, and goes into it. Martin Heidegger chose for this attitude towards the technical world the word releasement (German Gelassenheit), a yes and a no at once: to use technical things and yet deny them the right to claim us entirely.¹²⁸ At the button this means letting the stream in without being swept away by it. Unlike the automatism of the Surrealists (ch. 3.5, alongside Cage and Duchamp), WOF opens not inward but outward, and unlike Cage it does not take the ego out of the work. It hands the ego over and demands that it answer.

With this, the iconological reading is supplemented by its anthropological substrate. It leads to the art-historical contextualisation: WOF stands in a long line of procedures that transform found image material into a work of one’s own.

4. Art-Historical Contextualisation

4.1 Readymade and Montage

The operation with which WOF works above all in its purely digital form (ch. 4.5 deals with the painted line) – declaring found material a work through selection and recontextualisation – has its origin in Marcel Duchamp’s readymade. With Fountain (1917), a signed mass-produced urinal, Duchamp shifted the criterion of art from making to choosing: the artist “took an ordinary article of life” and placed it “so that its useful significance disappeared under the new title and point of view”.¹²⁹ L.H.O.O.Q. (1919) carried the same gesture over to the image memory itself by defacing a reproduction of the Mona Lisa. In parallel, Dadaism developed montage as a technique: Hannah Höch’s photomontages (c. 1918/19) and Kurt Schwitters’s Merz assemblages (from 1919) built new pictorial statements from the fragments of mass-media image culture.¹³⁰¹³¹ WOF continues this tradition, now with far larger stores – the open holdings of the archives in Magenta, the collection drawn from the circulation of the net in Trinity – a continuity the artist names in drawing the line “from Picasso’s papiers collés and Duchamp’s readymades to Rauschenberg, Warhol, Marclay and Steyerl”.¹³² Beyond the Dada montage the line runs on in two directions, both of which count for Trinity: into film montage, whose principle of collision ch. 3.5 claims for the cut, and to the poster works and video montages of the present, which ch. 4.5 presents as a craft counterpole and as the closest relatives.

4.2 Appropriation and the “Death of the Author”

The appropriation art of the “Pictures Generation” radicalised this gesture into the question of authorship. Sherrie Levine’s After Walker Evans (1981) – reproductions of Walker Evans’s FSA photographs, exhibited as her own work – fundamentally questioned authenticity, uniqueness and authorship; Richard Prince re-photographed advertising images.¹³³ This turn is theoretically grounded in Roland Barthes’s The Death of the Author (first published in English, 1967) and in Walter Benjamin’s The Work of Art in the Age of Its Technological Reproducibility (1935/36). Benjamin diagnosed the decay of the “aura” – the “here and now” of the original – under the conditions of technological reproduction, and the shift from cult value to exhibition value.¹³⁴ WOF can be read as an intensification of both diagnoses: it operates exclusively with reproduced material and poses the aura question anew by returning, at the end – via the NFT – an artificial “here and now”, an on-chain uniqueness (ch. 7). It resolves authorship not rhetorically but procedurally: the artist describes this role as a shift “from image-maker to condition-setter” – not the author of the image, but “the author of the possibility”.¹³⁵ In conversation the artist puts it more simply: an artist is “always a part of what is already there”, and the media that are there are not something to ignore but something to bring into the work; whoever believes they can “create everything anew from the pencil up and reinvent the world” is mistaken. “You’re not stealing it, you’re just taking it into a new configuration.”¹³⁶ The artist also invokes the sentence attributed to Picasso that good artists copy and great artists steal.¹³⁷ Appropriation is thought here from the whole of art, as a body that has been taking up its motifs again for generations (Warburg’s Nachleben, ch. 3.4). The line is drawn not between one’s own and another’s, but between the identical, identifiable repetition that technological reproducibility has made cheap, and the mixing of many fragments into something that did not exist before. For Trinity, which shows whole images, the mixture lies not in the single image but in the run (ch. 3.5, 7.4). From this the artist derives a demand: “What has become technically possible almost obliges one to admit it into the canon of the art of now. Otherwise we remain stuck at an outdated copyright barrier and sacrifice the achievements of our time to an excess of rule conformity that is profoundly alien to art. Rules matter, but this excess breaks off the very tip of the avant-garde, the spearhead of art.”¹³⁸ How far current law already accommodates this demand, and where it does not, ch. 7.4 shows with Pelham and the pastiche exception. This self-reflection prevents the appeal to Barthes from remaining naive: the author does not vanish but changes level. That this shift has also reached the legal order is shown by the Austrian literature on appropriation art: Anderl and Schmid describe appropriation as a mode of creation with a long-standing tradition – collage, adaptation, remix, sampling – and locate the creative contribution expressly “in the process of creation itself”.¹³⁹ What art theory formulates as the Death of the Author returns here as a legal figure: not the appropriated image but the procedure of its appropriation carries the threshold of originality¹⁴⁰.

That renouncing the image is no release from authorship was recorded early by the theory of the open work and by the history of computer poetry, and both sharpen the formula of the “author of the possibility”. Umberto Eco noted in 1962, five years before Barthes, that the invitation of the work in movement “is not an amorphous invitation to indiscriminate participation” but “offers the performer the opportunity for an oriented insertion into something which always remains the world intended by the author”; the author “does not know the exact fashion in which his work will be concluded, but he is aware that once completed the work in question will still be his own”.¹⁴¹ Theo Lutz, whose Stochastische Texte of 1959 stand at the beginning of computer poetry, formulated the same thought from the machine’s side: the linguistic appeal of automatically generated texts is “many times greater when the human being is only indirectly their author”.¹⁴² Indirect authorship is not half an authorship. Archival work on Lutz’s papers has shown that Lutz edited the machine text, inserted words by hand and signed the program, not the text – a case that editorial philology describes as “paratext as work” (Beiwerk als Werk).¹⁴³ For WOF this means: the work is the machine and its rule as much as the loop, and at the machine the artist is author in the full sense. How far authorship is distributed in the game, and why Roberto Simanowski read the selection as the giving of meaning in 2011 and as mere authorisation in 2025, is developed in ch. 6.5.

4.3 Concept, Programme, Postproduction

A third line leads via conceptual art to generative art. Sol LeWitt formulated in 1967 the sentence that technically anticipates WOF: “The idea becomes a machine that makes the art.”¹⁴⁴ In LeWitt the instruction is the work, the execution incidental. Early computer art – Vera Molnár, Frieder Nake, Georg Nees (first exhibitions 1965), Manfred Mohr – translated this principle into the algorithm.¹⁴⁵ WOF shares with this tradition the programme as the work’s foundation, but differs in a decisive point the artist marks sharply: “This is not generative art in the conventional sense. Generative art typically means: a seed produces a deterministic image. Beautiful, but passive. […] WOF reverses this relationship. The code is the instrument, not the author.”¹⁴⁶ The sentence needs to be taken more precisely than it sounds. Chance is in play in Trinity too, and at many points: which set the wheel drops, when a cell comes to a halt on its own, which colour pair and which effect hit an image – a die decides. But it rolls in the game, not in the work. The score carries only the result, never the die; the server executes it and produces the loop without deciding anything further itself (ch. 5). In generative art the seed is the origin of the image, and every execution rolls anew. In WOF chance is an event that took place once, and it stands in the score alongside the players’ changes of direction, their stops and their cut, as a notation of what happened; only the automatic stops are settled at the moment of the cut (ch. 5). The more accurate placement is supplied by Nicolas Bourriaud with the concept of postproduction (2002): the contemporary artist works like a DJ or programmer who selects, samples and reprogrammes existing cultural products instead of creating from nothing.¹⁴⁷ WOF is postproduction in this sense – except that the “sample” is, in Magenta, the open cultural archives and, in Trinity, the artist’s collection, and the “reprogramming” happens in real time and collaboratively.

4.4 The Archive

Finally, WOF is a work about the archive. Michel Foucault’s determination of the archive not as a collection but as “the general system of the formation and transformation of statements” – as that which can be said at all – describes precisely what WOF played with in Magenta: not individual images but the totality of the digitally available.¹⁴⁸ Foucault’s formula applies above all in its second half: the archive is the system of the formation and the transformation of statements. WOF does not merely read the holdings; it extends them – the concept tree of Magenta, from which the search queries came, is only partly curated; a language model has co-written it, and where the human input ends it can add new concepts (ch. 5). The dissolution of authorship thereby shifts one level down: not only the composition of the individual image – already the determination of what is searched for is no longer set by humans alone. The scriptor without intention, whom Barthes puts in the author’s place, operates here already on the level of the search space. Foucault’s formula then forces one final question: if the archive determines what can be said, then a language model has co-written WOF’s search space — and with it the training corpus from which it draws its neighbourhoods. Which concepts appear “adjacent” to the machine is no neutral given but a pre-selection over which neither artist nor players have any say; the question of authorship returns here, one level deeper, as a question of the constitution of the tool.

Hal Foster has diagnosed an “archival impulse” in contemporary art: artistic works make lost or scattered historical information physically present and thereby produce archives instead of merely using them.¹⁴⁹ WOF radicalised this impulse in Magenta by not curating the archive but opening it to chance; in Trinity it curates expressly, and chance operates within the frame that has been set (ch. 2.5.2). Benjamin Buchloh described both attitudes, with reference to Gerhard Richter’s Atlas, as the two poles of collage and montage aesthetics: “the order of perceptual shock and the principle of estrangement on the one hand, and the order of the statistical collection or the archive on the other”.¹⁵⁰ WOF has both in one instrument, and it has played them one after the other: the stream is the statistical collection in motion, drop and Collage were the shock; Trinity relocates the order of the archive into the artist’s collection and the shock into the collision of the classes. The genealogy Buchloh draws for the Atlas, from the Mercator atlas via Warburg’s Mnemosyne and Benjamin’s Arcades to Richter’s panels, gains a moving position with WOF – with the restriction Buchloh himself makes and ch. 2.5.2 has already named: what in Richter is archive without collage is in WOF’s Trinity montage in time (ch. 3.5).

Boris Groys has explained why a wheel that turns museum images and net images on an equal footing does more than mix: the new arises from “the valorizing comparison between cultural values and things in the profane realm”, a comparison that “is usually not drawn at all” and that is non-trivial only across the value boundary, “unlike that between the Mona Lisa and a piece of trash”.¹⁵¹ The wheel is a comparison machine that forces this comparison – in Magenta between museum and mass medium, in Trinity between bust and sugar. Simanowski, finally, has called the archive of social media “the medium of forgetting if it preserves events not because they were meaningful to us, but simply because they happened”;¹⁵² against this archive, drop and cut are a selection, and the mint (ch. 7) is a counter-gesture only by virtue of the selection, not by virtue of storing. At the same time WOF stands in the tradition, described by Lucy Lippard, of the dematerialisation of the art object, in which the object recedes in favour of idea and information¹⁵³ – a line that takes a paradoxical turn in the NFT (ch. 7).

4.5 Projection and Media Shreds: the Painterly Line

The preceding lines concerned the procedure; the work’s painterly origin (ch. 2.1) demands a placement of its own, in which two traditions converge. The first is that of projection painting. That an optically projected image serves painting as a model is no technical curiosity but has a long prehistory – from the laterna magica of the 17th century to the controversially discussed thesis advanced by David Hockney that the Old Masters themselves used optical aids for faithful transfer.¹⁵⁴ Its most literal contemporary counterpart WOF finds in Avery Singer, who for a long time projected images composed in 3D software onto the canvas and executed them with masking tape and airbrush before handing the airbrush over to a machine – images that are “both digital and analog”.¹⁵⁵ WOF shares this basic operation but displaces it decisively: what is projected is not one finished image to be copied, but a stream from which the human stop chooses.

The second line is that of the painting of media shreds – those procedures that condense the stream of mass-media images by hand into a unique work. From Robert Rauschenberg’s Silkscreen Paintings to Andy Warhol’s photo silkscreens, the media image stream was brought into painting;¹⁵⁶ Sigmar Polke, with Höhere Wesen befahlen: rechte obere Ecke schwarz malen! (“Higher beings commanded: paint the upper right corner black!”, 1969), ironically ceded compositional authority to an external instruction – a conceptual anticipation of WOF’s “the moment, not I” (ch. 3.4, 4.2, 8).¹⁵⁷ In painting, WOF stands closest to Gerhard Richter: the paintings after found photographs with their characteristic blur, and above all the Atlas – an archive of press and snapshot material grown over decades – form a painterly counterpart to WOF’s image pool.¹⁵⁸¹⁵⁹

Two contemporary positions move Trinity into this line. Hito Steyerl’s poor image, “a copy in motion”, a preview image of the kind that “lose[s] matter and gain[s] speed”, describes Trinity’s holdings exactly (ch. 2.5.2), and more than that: Steyerl’s “hierarchy of images”, which rests “primarily on resolution”, is built into the instrument, for the wheel as a rule runs on the small tier, and only the drop or the cut fetches the large one; whoever keeps an image lifts a poor image into the rich class.¹⁶⁰

The second reference is a craft counterpole. For years the British artist Dion Kitson has been peeling posters off billboards, soaking, stripping, drying, cutting them up and gluing them back together, and out of this he has built, ninety years after Picasso, a Guernica made of advertising, because, as he says, the picture is “the size of a billboard”: “After Guernica, there is an ad break.”¹⁶¹ Kitson takes apart by hand the same media flood that WOF takes apart by machine, and he expressly understands his work as resistance against the image machine of AI: “Can’t be stroked, cut or ripped.” Together with the language machine, to which ch. 6 is devoted, this yields a triangle in which WOF has its place: the human who tears the poster with their hands; the machine that computes new images out of probabilities; and the instrument that lets real images run – built with AI, yet itself generating not a single image. Kitson calls up the word drop three times over: as advertising’s product drop, as the dropping of bombs, as “the dropping of art”. For a work that itself knows the word twice (ch. 2.2, 2.5.4) and whose market counts in drops, that is more than a pun.

In the moving image, Trinity stands closest to the work of Arthur Jafa, whose video works are montaged from shreds of the net and the television archives, most famously Love Is the Message, The Message Is Death (2016), seven and a half minutes of found footage cut to a single song. In his montages Jafa invents no images; he takes what circulates and orders it so that the sequence yields a statement contained in no single image. Mono Scroll resembles an earlier work by Jafa, APEX (2013): 841 collected still images in a good eight minutes, each shorter than a second, too fast to be read singly, and yet taken in. Jafa later turned the sequence back into a standing grid as APEX GRID and slowed it down as SloPEX.¹⁶² A collection that exists at several tempos, as run, as grid, as slow motion: that is the relation of wheel and loop. Mono Scroll leaves the procedure to the players’ hand, but without what Jafa ties it to. His montage is tethered – to a community and to the project of a Black cinema that is to be as strong as Black music; WOF takes over the procedure without this we that exists before the work. Its tether is the collection of a single person, but not a hidden one: it lies open, every set is named in the metadata, every work legible to all on the blockchain. There is a we here too, only it does not stand at the beginning but comes into being first – out of those who play, mint and collect, and who find one another at this open collection. That the curator Nancy Spector brought Jafa together in Venice in 2026 with Richard Prince, the re-photographer of appropriation (ch. 4.2), under the title Helter Skelter, shows the line in one room. Appropriation and montage, still and moving image, between which Jafa himself oscillates, are one procedure in two tempos, and both artists share, according to the exhibition text, “an ethos of lawlessness” in their dealings with the images of mass culture. The artist saw the exhibition and names it as the one that most strongly confirmed the use of material from the net.¹⁶³¹⁶⁴

Both lines issue into the more recent movement from the digital back to painting. Against this background, WOF’s change of medium from canvas to browser can be understood as a further step in the same direction – with the peculiarity that, in the case of the purely digital loop variant, the image no longer returns to the canvas but remains digital and regains its uniqueness only via the NFT (ch. 7). This direction is, however, not the only one: WOF continues to serve the artist as a painting tool whose result may very well be a physical painting. The painted and the purely digital strand do not exclude one another; both remain open – the same instrument issues, depending on the setting, into the one form or the other. The painted strand is currently growing in a new direction: a plotter station passes the stopped image on to the hand as a stencil (ch. 5). Between stop and brush there thus steps a machine that does not choose but translates. The wheel decides which image stands and where it sits on the surface. The grid is set by the artist: the station knows variants as presets, but which grid, how fine and at what angle is set for each work individually and by feel. What remains machine is what is calculation: the decomposition of the image into the grid and the check whether the stencil will hold. The human hand decides how the image reaches the surface. What results is the opposite of the loop, a still, physical image, and it shows that the loop is a product of the instrument, not its essence.

This expansion bears a name that reaches beyond the individual works: that of the expanded painterly space. At the Academy of Fine Arts Vienna it designates a department of its own (today Art and Image | Expansion), whose name signals an extension of the painterly beyond the panel.¹⁶⁵ In this perspective, WOF’s installational form of play – the one in which passers-by become players and thus co-painters¹⁶⁶ – is no break with painting but its continuation by other means: the painterly act stretches from the studio into public space and finally into the property relation of the NFT, which carries the placement thus created forward as an independent work (ch. 7).

With the step from panel to browser-based loop, WOF at the same time leaves the still image: the end product is no longer a tableau but a loop – a time-based, moving image. The work thus moves to a threshold at which the category of painting alone can hardly hold it, and the moving image comes into view – the form that, from film through video to the handheld stream, has become the dominant image form of the present. The artist takes a clear position here: the moving image, film, is in the artist’s eyes the true new art of the time, whose condensation within a single lifespan is remarkable, while painting in the conventional sense grows increasingly unfamiliar.¹⁶⁷ This valuation renders the artistic perspective, not a judgement of the present paper; what remains to be recorded, however, is that WOF operates precisely on this threshold, and from both sides. The digital loop carries the painterly origin – the condensing, the placing, the stop, in Mono Scroll also the dividing of the surface into cells and the cut – over into the form of the moving image and remains what it is: a moving image. The painting, by contrast, stands still; its result is frozen, yet it springs from a moving system and still lets that movement be felt. That is why it is less painting or film than both at once: a painted moving image.

Before ch. 6 and 7 measure the work against the language machine and the NFT, the technical dispositif that makes the named operations possible must be clarified.

5. The Technical Dispositif

An iconological reading that ignores the medium would remain incomplete; in a technical work the dispositif¹⁶⁸ is part of the content. WOF consists of five interlocking layers; the fourth layer, the score, was added with Trinity and has changed the form of the work.¹⁶⁹ There follow a report from the continuous operation of Magenta and the work cycle in which these layers stand.

The holdings. The first layer is the holdings, and they have two sources, which the work has played one after the other. In Magenta it was the web as medium and material: images are loaded at runtime via the APIs named above, filtered by aspect ratio and minimum size, and fed into a pool. The artist thereby set only the search words; which images came back and passed the filters could not be foreseen, and this unpredictability belonged to the chance of Magenta. Selection proceeds not by fixed “themes” but through a semantic system (internally, the “Brain”) that wanders a curated concept tree of art history in five stages¹⁷⁰.¹⁷¹ The search terms come from two sources: one part is curated – laid out in the tree by the artist –, another can be machine-generated at runtime. When the wandering hits a branch with no known continuation, a language model (Claude Haiku) can generate new, adjacent concepts and write them back into the tree; these extensions are stored permanently server-side and read in again at the next start, so that the concept tree can keep growing across all play sessions and devices.¹⁷² In the two-month Magenta run (ch. 2.4) the tree counted 2,321 concepts. The work’s store of material is thus not fixed but an in-principle open system, partly human-curated, partly machine-continued. This navigation is no image generator but a movement of searching and wandering through existing holdings – the machinic counterpart of that hand reaching into the stack. In Trinity this layer is shut down. The holdings are the artist’s collection (ch. 2.5.2), mirrored on the project’s own server, prepared in two size tiers and kept in the image list together with the veiling boxes; concept tree and search movement remain the repertoire of the archive-fed forms of play, not a precondition of the work. That Trinity thereby manages without a single archive interface is the technical side of the argument in ch. 6.2.

The deterministic mechanics. The image run is governed by a fixed equation of motion with acceleration and braking phases;¹⁷³ the stopping time unambiguously determines the standing image.¹⁷⁴ This determinacy is conceptually significant (ch. 3.4). In Mono Scroll a virtual clock per cell takes the place of the braking phase: the position of the band grows evenly with time, the duration of each image follows from run profile and tempo, and stops as well as changes of direction are kept as events on this clock, so that the loop can be recomputed from them. Here too the run is deterministic; what varies is entered.¹⁷⁵

Composition and effect. Each dropped layer is set onto an image surface with position and slight rotation; a frozen mask form and a baked-in filter may be added, but need not be. The filter chain (brightness, contrast, saturation, halftone, stencil, among others) is implemented pixel-based for compatibility reasons.¹⁷⁶ Masks and “Finale” produce the end state, in which the finished image itself becomes once more the stage of a flicker – an image-immanent feedback: the dropped material becomes wheel again. Mono Scroll sets no layers but colours and cuts: each image receives its two colours from the colours drawn for the epoch, a share remains black and white, individual images carry effects, and the band runs in cells whose seams the players shift. The flicker at the end is dropped; the feedback in which the material becomes wheel again is, in Mono Scroll, not the finale but the ground state.

Score and print. Since August 2026 the browser no longer records the work. It writes a score: the sequence of images, the clock, the changes of direction, the stops that the human hand has set or the instrument has added, the cells, the colours, the effects – that is, everything that was thrown in play, as result – and uploads it. A server pulls the images, veils them, executes the score in a browser without a screen and generates the loop from it, 168 frames at fourteen frames per second, deterministically: a test series shows that two independent executions of the same score agree in all 168 frames down to the byte.¹⁷⁷ The reason for the rebuild was practical – a phone can neither produce the loop in the required quality nor send it over a mobile connection; the consequence concerns the theory of the work. Nelson Goodman distinguished works according to whether the difference between original and forgery is significant for them: autographic is a work for which “even the most exact duplication of it does not thereby count as genuine” – painting; allographic one that is equally present in every correct execution – music, the score.¹⁷⁸ The collages of Magenta were autographic: what was recorded was the work, and a copy would be a copy. The work of Trinity has become allographic, a notation that every correct execution fulfils. The fingerprint on the blockchain changes nothing about this; it confirms it: every byte-identical copy of the loop carries the same hash and cannot be distinguished from it. A forgery in Goodman’s sense no longer exists, only another file. Unique is solely the entry that says which loop is meant and to whom it belongs. Authenticity has migrated from the thing into the register, and that is what Boris Groys calls the aura without an object (ch. 7.3). WOF is score and print, but the original is neither; it is the entry. To this comes a detail that shifts the status of the loop: the stops that the human hand set in the last twelve seconds travel into the print, and they take precedence. Where the epoch provides for it, the instrument adds automatic stops at the moment of the cut, freshly thrown for the window of the work and never on top of a stop set by the players, at most six per cell. The print is thus neither a record of the play nor a mere re-execution; the human hand remains, and where it has set nothing, chance sets. What the blockchain holds is the entry for a loop that notates the play with the trace of the human hand, not a record of every image. This qualifies the determinacy that ch. 3.4 claims for the drop of the Collage, precisely where Mono Scroll in the Non Stop finale no longer needs it.

The blockchain layer. In Magenta two moving-image formats were derived from the end state (MP4 and WebP); since Trinity the WebP loop is the work, and the MP4 is a derivative for devices that do not show the format smoothly, with its own address and no place in the metadata. Optionally, and upon explicit user action, the loop can be minted as an NFT on the Polkadot Asset Hub or, since October 2026, on three further blockchains (ch. 7.4).¹⁷⁹¹⁸⁰ The first mainnet mint took place on 14 May 2026 (collection #843); the transaction is verifiable via a public block explorer.¹⁸¹ In August 2026 the first series mint followed: the 265 works of the Magenta run as collection #853 WOF Magenta, sealed by a Merkle root on the collection (on the selection, see below, Testing in continuous operation; on the seal, ch. 7.4).¹⁸² Important here is a precise distinction that runs counter to the market rhetoric of the NFT field: on-chain lies the NFT entry (collection and item) with the reference to the metadata. The SHA-256 hash¹⁸³ is carried on the Polkadot Asset Hub by each epoch work itself as an attribute, by the Magenta series only bundled in the Merkle root, and on the other three blockchains the content-addressed reference leads to it. It applies to the work file; for the epochs, since October 2026, that is the loop itself, no longer the still – a correction the artist considered necessary because the fingerprint must apply to the work and not to its preview image; the image files themselves lie off-chain. For the early single mints, direct HTTPS delivery from the project’s own server applied; the Magenta series, by contrast, addresses still and loop content-addressed via the self-hosted IPFS¹⁸⁴ – the ipfs:// address in the metadata is the cryptographic fingerprint of the content, and nothing else can ever lie at the same address.¹⁸⁵ For all works of the epochs the content-addressed form applies from the outset. Between print and minting lies, in Trinity, the time limit of ch. 2.5.7 – technically a signed permission from the artist, first bound to the player’s wallet, then open to all; if it is not used, a service takes the file off the node.¹⁸⁶ The difference between entry and file is not technically incidental but artistically significant – its interpretation is the task of ch. 7.4.

Testing in continuous operation. Between 30 May (set-up) and 31 July 2026, WOF ran as a permanent installation in the Magenta Art Room (open to the public from 1 June, ch. 2.4): 1,119 logged operating hours, daily from around 05:16 until 00:12 and thus in step with the Vienna metro. In that period some 850 rounds were played and 386 pieces saved – each comprising a still image and its accompanying animation loop. From these the artist selected 265 as a series of works and had them minted as NFTs. Nothing in the images sets them apart from the rest; only this choice does: what remains is not what turned out better but what was chosen – and what the entry makes permanent is the reference, not the file (ch. 7.4). Only one value in the series’ metadata comes from the players themselves: whether the player took their finale home as a file at the time. This happened for 60 of the 265 works; the remaining 205 were left behind – and became part of the series all the same. The wheel turned more than 520,000 times, and every turn set another image from the open holdings into a composition; the search movement behind it – search passes across fifteen sources, more than 14,000 on a single day of operation – drew the retrieval quotas of several archives to their daily limits. What is revealing for the thesis advanced here is less the interaction than the idling: even without an audience the work kept playing itself – arithmetically some 21,500 runs (operating time divided by the intended length of one round) whose images came into being and passed away again without anyone triggering, seeing or saving them. The status of these unwitnessed images is negotiated in ch. 3.4; here it receives its measure. The claim that the work is connectivity rather than its depiction can be quantified for this period: 11.58 terabytes flowed across the mobile connection, 184 gigabytes a day on average, without a single image ever being held locally.¹⁸⁷

The work cycle. This concerns the latest version of WOF and its future workflow. The instrument does not stand alone. Around it the artist has built a series of tools that together form a continuous flow, from the image on the net to the paint on the wall. At the beginning stands collecting on the net, by hand. In the Set-Hub the images are sifted, named and veiled (ch. 2.5.6). Then the wheel plays, in one of its modes. From the play a server computes the work via the score (above), and the work is minted, bought and sold in the project’s own shop on the blockchain of its edition (ch. 7.4). From the stopped image a second path leads back into painting. The instrument rectifies the image to the measure of the canvas; a halftone tool breaks it down into dots, lines or contours, and does so in such a way that the stencil does not fall apart when lifted; a form tool does the same with three-dimensional forms for large walls; and the plotter station sends the cuts over Wi-Fi, through a self-built box, to the cutting plotter. The plotter cuts the stencil; the human paints. None of these tools is a by-product. Each makes decisions that shape the work and therefore belongs to its design, not to its logistics.¹⁸⁸

That a single artist can build this cycle, from collecting to cutting, single-handedly is new and bound to the present. It has become possible only since language models help with programming (ch. 2.1, 8). WOF is in this sense a closed work cycle: a work that encompasses its own tools and brings digital and physical making together in one flow, and it is a work of its time also in that it could not have been built without this time.

With this a question arises that Walter Benjamin put to photography: not whether the new was art, but whether “the invention of photography had not transformed the entire character of art”.¹⁸⁹ Three ways of getting hold of images lie one above the other in WOF. The old way led via folders, magazines and notebooks, as with Richter’s Atlas, with Jafa’s notebooks and with the painting tool of 2010; the images were collected before they were needed. The way of the net fetches them at the moment of play, from archives that no one surveys any more (ch. 2.3). The third way is new: with the language machine the artist builds not only the image but the tool that fetches, orders, veils and plays the images and brings them onto the wall. What shifts with this is the possibility of making art. It depends less on what a person can produce and more on what they want (ch. 8). At the same time the individual becomes smaller, not larger. Whoever draws on everything that is already there is part of a body of art that existed before them and keeps growing after them, like Simmel’s weaver, who does not know what is being woven (ch. 3.4); the image describes this situation more precisely than any talk of genius. WOF accepts it rather than asserting itself against it. Since photography, the image has been reproducible. Now the tool with which images are collected, played and painted becomes reproducible too, and it lies in the hands of whoever needs it.

With this, all preconditions are clarified for comparing the work first with the language machine (ch. 6) and then examining the central thesis of the NFT chapter (ch. 7).

6. The Language Machine and the Wheel

The contextualisation of the preceding chapters placed WOF among the procedures that make works out of what is found: readymade, montage, appropriation, programme, archive. Since 2022 there has been a procedure that seems to do the same and yet does something else: the large language model, the language machine, as Roberto Simanowski calls it in his philosophy of artificial intelligence.¹⁹⁰ It, too, combines what is already there; it, too, gives chance a place; it, too, poses the question of authorship anew. This paper objected in ch. 1 to the reading that WOF is “AI art”, and noted in ch. 2.1 that the instrument is built with the help of a language model. Both remain true. What is missing is the comparison itself: how does a wheel that runs images against one another differ from a machine that calculates sentences from probabilities? This chapter conducts the comparison along Simanowski’s book, because it has put the language machine into concepts against which WOF can be measured, and because in the end the comparison supplies the argument that connects the two cases of the work: the change of levels.

6.1 The Average as Cutting Edge

Simanowski’s basic finding is statistical in nature. The language machine calculates the most probable next word; it is therefore a “technology of the average” and “structurally conservative”: “As soon as statistics becomes the measure of value, conservatism is unavoidable.”¹⁹¹ Chance it knows only in doses, as temperature and as an occasional random procedure that skips the majority, if “[n]ot in 40 per cent of cases, but in some”. The point of the finding lies in its history. Early computer poetry – Theo Lutz’s Stochastische Texte (Stochastic Texts) of 1959 on the Zuse Z22, with a vocabulary drawn from Kafka’s Schloß (The Castle) – was meant, according to Simanowski, to provoke: the Stuttgart Manifest einer neuen Prosa of 1960 set out “to inflict semantic losses on the outside world”, and an American anthology, as Simanowski reports, declared computer poetry a war against conventionality; today’s machine has “become the advocate of a formulaic language”.¹⁹² Lutz himself described both paths side by side as early as 1959: his machine worked with uniform distribution – “each of the given subjects or predicates is to appear with the same frequency i.e. with the same probability” – and in the same essay he sketched a “probability matrix” of transition probabilities by which the machine would favour “meaningful” sentences and would thereby have “learned”.¹⁹³ Simanowski sums up the fork in one sentence: “Chance has turned into an oracle that can appeal to mathematics.”¹⁹⁴

This fork is the decisive one for the present paper, for it separates the Rota from the machine. The wheel of Fortuna does not learn. It has rules that the artist has set and discloses – Chaos falls rarely, every class is guaranteed (ch. 2.5.4) – but none of these weightings changes with what has been played. What changes, the artist resets from epoch to epoch; the wheel does not learn, it is set. Ch. 3.4 showed that chance in WOF does not lie in a random function of the run but in the meeting of independent chains: the law-governed run, the diced line-up and the human intervention. None of these chains learns from the other. An image that has often been dropped or often kept in the cut does not appear with greater probability in the next game; the sets are shuffled the same way, the classes weighted the same way, the line-up drawn anew for each game. WOF thus stands on the side Lutz took, not on the one he sketched: a uniformly distributed wheel and a human who intervenes. The language machine, on the other side, is the oracle that turns frequency into probability and probability into proposal. Fortuna against statistics: that is, put into concepts, the difference that the iconology in ch. 3 developed from the motif. Admittedly, Fortuna is not blind here either: which images the wheel may carry at all was, in Magenta, sought by a concept and, in Trinity, chosen by a human hand, and both happened deliberately (ch. 6.5, 6.6). Uniformly distributed is the wheel, not the world from which it draws. But it does not learn from what happens to it, and that is the difference that matters.

6.2 The Demented Machine

Simanowski drives the finding to its extreme. A machine that learns from texts which machines in turn have written loses what is rare: “There is talk of a ‘Habsburg AI’ […]; of a ‘textual incest’ of the AI that only reads its own texts any more; of a ‘demented’ AI choking on its own exhaust fumes.”¹⁹⁵ The source is a study from 2023 that calls the process model collapse: training on model-generated material “causes irreversible defects in the resulting models, where tails of the original content distribution disappear”.¹⁹⁶ What disappears are the minorities of the material, the improbable cases; what remains is an ever narrower average. A second, older critique adds meaning to the diagnosis: a language model is “a system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning: a stochastic parrot”. The sense that readers find in such texts lies “in the eye of the beholder”.¹⁹⁷

This paper cannot let this objection pass, for it strikes its own thesis. Barthes’s reader, who takes the place of the author (ch. 3.4), is at first sight the same beholder in whose eye the parrot gains sense. WOF, too, joins together what it does not mean; in WOF, too, sense arises in the seeing. Two differences hold the positions apart. The first concerns the material. Every image that runs through the wheel had someone who wanted it – a painter, a photographer, a museum, even someone at an image generator; in WOF only the combination is without intention. In the language model even the smallest unit, the word in this place, is generated from probability. Georg Simmel anticipated the second difference in 1911 when he asked whether meaning that has arisen by chance is meaning: a little child playing with letters of the alphabet “may order them accidentally into good sense”, and “the meaning exists objectively and concretely, no matter how naively it may have been produced”.¹⁹⁸ Simmel answers yes. A condition can be set with an earlier passage of the essay, the sunrise: “A sunrise which is not seen by any human eyes does not increase the value of this world or make it more sublime”; only once “a painter invests his emotion, his sense for form and color and his power of expression, in a picture of this sunrise” do we “consider this work an enrichment”.¹⁹⁹ The meaning of the accidental needs a soul that invests itself in it. In WOF this place exists, and it can be named: it is, in Magenta, the moment of the drop and, in Trinity, the chain of decisions that ch. 6.5 enumerates. In the stochastic parrot it does not exist; there no one invests anything, and that is precisely the charge.

The other part of the objection nevertheless does strike WOF, and the paper raised it itself in ch. 4.4. Magenta searched for its images by way of a concept tree that a language model has co-written; in the Magenta run it counted 2,321 concepts (ch. 5). The concepts with which the wheel queried the archives were thus in part already what Simanowski calls exhaust: neighbourhoods a model knows from its training corpus, not from a collection. Trinity has left this path. It plays from sets that a human has chosen and reviewed by hand (ch. 2.5.2), and thereby follows, without intending to, the recommendation that the critique of the parrot gives in its turn: to curate and document datasets “rather than ingesting everything on the web”.²⁰⁰ This is no return to innocence. The instrument continues to be built with a language model, and the concept tree remains in the repertoire. But the images from which Trinity makes its works are, in their entirety, chosen by humans, and therein lies the reason why the wheel cannot become demented: it does not train on its works. The instrument generates no image; the holdings are what the net yields, AI included, and the choice is taste.

6.3 The Obstructive Librarian

Simanowski’s introduction contains an experience of loss that is familiar to this paper. The machine that takes over the searching takes away the finding: “These chance encounters, this exploratory leafing and clicking, fall away when the AI takes over the search.”²⁰¹ It is, in his image, an obstructive librarian: it blocks the way to the shelf and gives the answer itself; the book that stood next to it no one sees any more. Ch. 2.3 described WOF as the machine that reads digitised memory “not systematically … but randomly”, “the way a hand reaches into a stack of photographs”. Against Simanowski’s foil the image gains its sharpness. The wheel is a machine that produces the chance encounter instead of calculating it away, and it does so in both cases: in Magenta the hand reached into the open archives, in Trinity it reaches into a collection whose line-up is diced anew for each game. No one gets what they searched for: in Trinity no one searches, and in Magenta the search only gave the direction, not the image.

This has a flip side, which Simanowski develops from personalisation. The language machine that learns from its user’s data becomes a “daily self” that presents users with what they already want; “And on that, art has long since passed its verdict”: Jenny Holzer’s PROTECT ME FROM WHAT I WANT, from the series Survival (1983–85), on an electronic sign in the urban space of New York.²⁰² The wheel is in this sense an anti-personalisation. It shows not what the players want but what Fortuna brings; it does not get to know them and proposes nothing to them. The drop and the cut bring the wanting back, but on what is foreign: one can only keep what one has not ordered. That the artist dreams in ch. 8 of Times Square, where Holzer’s sentence once shone, is more than a coincidence of place. The advertising surface that Holzer turned against wanting is the surface WOF wants to reach, and the one from which Dion Kitson steals his posters (ch. 4.5).

6.4 Salad Bowl and Melting Pot

One of Simanowski’s images describes what the language machine does with its sources, and it hits WOF more precisely than the work might like. Goethe’s world literature, he says, was a salad bowl in which every text keeps its origin and voice; “The language machine is the melting pot that stirs world literature into a world text”, in which the sources lose their names and the majority sets the norm.²⁰³ The two cases of WOF lie on the two sides of this image. The Collage of Magenta layers its images until many show only in fragments, partly covered; in extreme cases their origin can no longer be named, for a fragment does not always give away the whole image; ch. 7.4 records that many images remain recognisable in the process, and still calls the archives “anonymous co-authors”. That is, in Simanowski’s term, the melting pot, and the paper has counted it as a strength. Trinity, by contrast, shows whole images, many of them veiled (ch. 2.5.6), one after another, nameable down to the set from which each comes, and the metadata to which the blockchain points name this set by name. That is the salad bowl. The image is an image of provenance, not of legibility, and in this it must be taken precisely: the Collage is not shredded, its layers remain legible as surfaces, they merely stand still; Mono Scroll in the Non Stop finale, although every image runs through whole, if many of them veiled, is harder to read because it lacks this stillness and nothing stands for longer than a moment. Nameable does not mean legible. WOF is both in one instrument, and honesty demands that this be named as a tension, not as a ladder of stages: Magenta melted its sources together and relied on that; Trinity lets them stand and must therefore answer for them differently (ch. 7.4). What separates both from the language machine is not the degree of melting but that they synthesise nothing. Even the densest Collage consists of images that existed; the world text consists of sentences that never existed.

6.5 The Antidote: the Change of Levels

The preceding sections have shown at four points how the wheel differs from the machine: it does not prefer the frequent, it does not train on itself, it brings no one what they searched for, and it melts together only what existed. What they have not yet shown is why it differs, and the answer does not lie in chance. Chance alone is no antidote. Simanowski himself notes that a machine at high temperature does not become free but that its words become “more confused”;²⁰⁴ Umberto Eco showed that “an excess of equiprobability does not increase the potential for information but completely denies it”, and that the richest, because most open, communication “requires a delicate balance permitting the merest order within the maximum disorder”;²⁰⁵ Martin Seel added that artistic resonating (Rauschen), where it originates in chance techniques, is always “a matter of an arranged or staged or scenically permitted chance occurrence”, “an invited process”.²⁰⁶ And the human alone is no antidote either. Simanowski’s hardest objection, with Bernard Stiegler, is that the machine reduces the human to the button: “A human being no longer needs to be able to solve a cognitive problem; it is enough to be able to describe it […]. All that is left is to press the button on the machine.”²⁰⁷ This objection strikes the sentence with which this paper began. “The only human act is the decision to stop”: a keystroke is, in Espen Aarseth’s terms, a borderline case at the lower edge of the ergodic, hardly more than the “turning of pages” that he calls trivial.²⁰⁸ Were the stop everything, Stiegler would be right.

Trinity gives the answer, and it gives it not through more chance and not through more human, but through the change. The artist has named the shift: the stop, for Trinity, “no longer quite” holds, and that “several levels play” has “precisely the effect of the antidote”.²⁰⁹ The levels can be enumerated, and the enumeration is the description from ch. 2.5, read from the side of authorship.

Read from the side of authorship, the table answers the question of who plays, which the overview in ch. 2 could only pose per form of play. Human, human, chance, human, mechanics, human and blockchain. Draw the human levels together and what remains is the formula of this paper: human, chance, human, blockchain. Every human layer brings in fresh material not synthesised by the instrument, or a decision that no one has calculated; every chance layer breaks the habit of the human who would otherwise get what they want. One can object that the language machine, too, has humans before and after the chance. Humans wrote the texts from which it learns; humans fine-tune it; a human writes the query and selects from what it answers. The difference lies not in whether these levels exist but in whether one sees them. In the language machine they are melted together: from no sentence can one read off which human, which weighting, which chance brought it forth. Added to this is a difference in the material itself. A word belongs to everyone: the same word stands in millions of texts and carries no origin. An image is a single case: a photograph, a painting, an advertisement is recognised again, often at a glance, often from a fragment alone. The language machine can melt its sources together because its words do not differ from one another; the wheel plays images, and images bring their origin with them. In WOF the changes are set and disclosed. The drawn sets are displayed at the start of the game, a board shows the rest of the throw and what the players change, and the metadata of every work name what was played (ch. 2.5.4); the last column of the table shows it level by level. The antidote is not the change alone but the visible change.

Magenta had three of them: the artist’s search terms, the archive chance and the image drop; Trinity has six. David Rokeby coined, for interactive works, the sentence that puts the principle in one line: “the chance element is replaced by a complex, indeterminate yet sentient element, the spectator”.²¹⁰ In WOF chance is not replaced but framed; it gets humans in front of it and behind it. Aarseth supplied the technical term: the author of a machine-generated text is a cyborg, and the work is distributed over three positions – “preprocessing, in which the machine is programed, configured, and loaded by the human”, “coprocessing, in which the machine and the human produce text in tandem”, and “postprocessing, in which the human selects some of the machine’s effusions and excludes others”.²¹¹ Aarseth held co- and postprocessing to be almost mutually exclusive. WOF unites all three, and does so in the way Aarseth provides for only in passing, with different humans at each position: the artist loads (levels 1 and 2), the players produce in tandem with dice and mechanics (3 to 5), and whoever mints selects (6). Odo Marquard’s sentence that what makes a human being free is “not zero determination” and not “the superior force of a single determinant”, “but a superabundance of determinants”, describes the same thing from the other side: curator, dice and players have no sole access to the work, and precisely therein lies its freedom.²¹²

Simanowski himself has given this answer at one point, and against himself. In Digital Art and Meaning (2011) he writes of the text generator that the programmer leaves an imprint “by deciding what parts of the endless production should be presented, where it should start and stop”; “Although the content of the outcome is owned by the machine, the meaning belongs to the human behind it”.²¹³ In Sprachmaschinen (2025) he says of Lutz, who selected a sentence of the Z22 and published it, that he had thereby “authorised it, but that by no means makes him its author”.²¹⁴ The tension between the two sentences is the tension of the present paper, and it is productive. Archival work on the Lutz papers has since shown that Lutz did not merely select but edited: he inserted words by hand that the unreliable random generator had never output, “in order to come closer to the equiprobability of all nouns and adjectives that chance would lead one to expect”.²¹⁵ The human corrected chance towards uniform distribution; he produced the Fortuna that the machine did not manage. That is not authorship in the sense of the image-maker, but it is more than selection. The paper calls it, with the artist, the authorship of the possibility (ch. 4.2), and the table of levels shows that in Trinity this possibility is set at several places.

6.6 The Objections, Raised by the Paper Itself

An argument that relies on the change of levels must conduct the objections that have been raised against each single level. They are gathered here not to refute them but to show where the work withstands them and where it bears them.

Chance as Alibi. Benjamin Buchloh warned, with regard to the Atlas, that “randomness and arbitrary juxtaposition function not only as established aesthetic procedures but also as a kind of socially enforced legitimation of anomie disguised as an advanced state of individual independence”.²¹⁶ The objection strikes a wheel that would be chance alone. It does not strike what is chosen, ordered and cut; but it compels the text, at every appeal to Fortuna, to say who stands before her and after her. Boris Groys adds that the authentically new “does not arise through a kind of self-forgetfulness” but through strategy;²¹⁷ in WOF the strategy lies in the recipe, not in the run.

Participation as Spectacle. Diedrich Diederichsen has summed up participatory art in the sentence “Participation is the new spectacle”, and certified its participants that they are “not co-producers, they are the material”; Simanowski agrees with him in Textmaschinen and adds that in interactive installations it can be argued that “the dominance of the artist even increases”, because the installation imposes a “grammar of interaction” on the audience; in one example, the “message of the author” lies in this grammar.²¹⁸ Claire Bishop draws from this the conclusion that the paper must adopt: “It is no longer enough to say that activating the viewer tout court is a democratic act, for every art work—even the most ‘open-ended’—determines in advance the depth of participation that the viewer may have with it.”²¹⁹ The drop was a predetermined depth of participation; so is the cut. The paper claims no democratisation; it describes a distribution. Wherein WOF differs from the participatory art under critique, Diederichsen’s own criterion has named: the documentation of such art always shows “the audience as well”, and the participants control “everything except the image that finally circulates of them”.²²⁰ WOF’s works never show the players; the human is in the work only as decision, not as body. And the players precisely do not control the image, but they can own it (ch. 7.4). Possession rather than control.

Connectivity as Imperative. Diederichsen’s “imperative of connectivity […], to be constantly connected and available” strikes the sentence with which Magenta described itself: “The installation IS connectivity.”²²¹ The text cannot take the sentence back; it can say that Trinity has given it up. Trinity runs without a single archive interface, from holdings the artist keeps, and the work arises not in the connection but in the cut. What remains of connectivity is the blockchain on which the work is attested, and its rules, too, are written by a few (on this, the following objection). Ch. 7.4 answers this not with renunciation but with the plurality of blockchains.

The Rules of a Single Person. The sharpest objection to the change of levels comes from Simanowski’s own book, and it strikes not only the blockchain but also the wheel. Simanowski describes how norms arise today: a few employees of a corporation decide which utterance counts as toxic, and their judgement steers worldwide what may be said. “A handful of Google employees”, he writes, is responsible for the normative alignment of cultures, “which takes place tacitly”.²²² In WOF this group is smaller still; it consists of one person. In Trinity the artist has chosen every image and set the classes; in Magenta the artist set the search terms with which the wheel queried the archives. And the artist determines what is veiled, which colours the game knows and how long the work lasts. Whoever plays moves in a room whose walls a single person has drawn. Simanowski’s earlier observation that in interactive installations the dominance of the artist increases because the interaction follows a grammar the artist sets (above) applies here without deduction. The paper does not wish to refute the objection; it accepts it, for it says exactly where the artist’s authorship lies. If the rules are authorship, then the artist is the author of the work, of the instrument, of its rules and of its selection, not the author of the images that run through it. The images the artist borrows; their authorship remains with those who made them, and the artist’s claim is directed at what is built with them. That is the formula from ch. 4.2, author of the possibility, not of the image, and it also gives the question of the rights to the borrowed images a clear direction (ch. 7.4): what is claimed is the work, not the image.

The objection can be placed by the word that matters to Simanowski: tacitly. And there it emerges that it strikes the two cases the other way round from what one would expect. Magenta seemed more open, for it reached into the archives of the world. But what an archive yielded to a query, others had decided: the museums that digitise and tag, the search systems that rank by relevance, and a language model that co-wrote the search terms (ch. 6.2). That was the selection of many unknowns, and it took place tacitly, entirely in Simanowski’s sense. Entirely without the artist’s hand it was not, admittedly: the search terms with which the wheel queried the archives, the artist had to a good part released personally, as a curated tree of movements and names (ch. 5). The artist chose the terms, the archives chose the images; chance was, as the artist says, “a little curved” (ch. 8). Trinity has cut off this detour and replaced the hidden selection with a visible one, that of one person, chosen by hand, with the name of every set in the metadata. The rules of the wheel are no less powerful than those of a platform, but they are on display – in the recipe of the epoch, in the metadata of every work, in this text. And their power ends where the game begins. When someone intervenes, when someone cuts and whether someone mints, no rule determines. The artist builds the wheel; others are left to turn it.

The Death of the Reader. In 2001 Simanowski set against the talk of the death of the author the death of the reader: where program and chance take over the structure of a shared text, what counts is the event, not the result; “Someone who is not in the game might not enjoy watching it”.²²³ The question for WOF is whether the loop is a work at all for someone who has not played. Hans Ulrich Gumbrecht has the concept the paper needs: presence effects and meaning effects “always appear together” and are “always in tension”; the point is “to pause for a moment before we begin to make sense”.²²⁴ The game is presence; the loop is its document, and the blockchain makes the document legible for those who were not there: twelve seconds, the drawn sets, the colours, a name. Whether that suffices is decided not by theory but by the look at the work, and it is the look of someone who was not in the game, and this paper leaves the question open at the end (ch. 9).

The Exhaust in the Tool. The last objection is the one the chapter raised itself: the instrument is built with a language model, and Magenta’s concept tree was in part machine-made. WOF is not AI art, because no image is synthesised in it; but it is a work built with AI, and the artist says so openly (ch. 8). Simanowski’s book supplies the measure for this, not the absolution: the question is not whether the machine is used but whether the human who uses it misses “their own self”, whether the production process “degenerates into a reception process”.²²⁵ The conversation in ch. 8 puts this question to the artist, as a question, not as an answer.

With that the comparison is done. The language machine and the wheel share the material of the world and the renunciation of the image-maker; they differ in weighting, in learning, in bringing what was searched for, in synthesising, and, at the root of all four differences, in the change of levels that places humans before and after chance. What this change means for the place where the work exists in terms of ownership is the subject of the following chapter.

7. The NFT as a New Artistic Space

7.1 A Brief Genealogy

The non-fungible token is younger than its hype. The first NFT artwork is considered – with due caution – to be Kevin McCoy’s Quantum (2014), which the artist registered on the Namecoin blockchain at Rhizome’s Seven on Seven event and transferred to Anil Dash for four dollars.²²⁶ Technical standardisation followed only in 2018 with the Ethereum standard ERC-721, which defined “non-fungible” – individually distinguishable and traceable – tokens.²²⁷ The format became artistically relevant with the long-form generative art of the platforms Art Blocks (from November 2020) and fxhash (from 2021): here the artist writes a script on-chain, and only at the moment of minting does it generate the individual image – unknown even to its author.²²⁸ Tyler Hobbs put this procedure pointedly: with long-form works, “the artist has nowhere to hide”.²²⁹ The NFT entered public consciousness with the auction of Beeple’s Everydays: The First 5000 Days at Christie’s on 11 March 2021 for 69,346,250 US dollars – according to the auction house the first purely digital work of art ever offered by a major auction house.²³⁰ Sotheby’s Natively Digital (June 2021) bundled the trend institutionally.²³¹

7.2 The Critique

The discourse has been divided from the start, and a scholarly examination must conduct both sides. The art-critical scepticism – exemplarily Tina Rivers Ryan’s essay Token Gesture (Artforum, May 2021) – disputes the NFT’s rhetoric of liberation and points to its entanglement in speculation.²³² A peer-reviewed study names as the core ethical problems of NFT art speculation and bubble formation, the handling of cultural heritage, and the ecological burden.²³³ The latter must be made precise against that study: before September 2022 Ethereum rested on energy-intensive proof of work; with the switch to proof of stake²³⁴ (“The Merge”, 15 September 2022) the network’s energy consumption fell by more than 99.9 per cent.²³⁵ The environmental argument that shaped early NFT reception has thus lost considerable weight, without the economic-critical objections being refuted.

7.3 The Space: Dematerialisation, Provenance, Ownership

The thesis of this chapter is that the NFT – independent of its market history – opens three artistically relevant spaces. First, the continuation of dematerialisation (ch. 4.4) with a paradoxical turn: Lippard’s movement liquefied the object in favour of the idea; the NFT re-materialises not the image but the value – it lays around an immaterial file a ring of scarcity, ownership and history.²³⁶

Second, on-chain provenance: the history of ownership and transfer becomes legible on a public register and thus itself part of the work’s identity – a materialisation of what art history calls “provenance”.²³⁷ Third, the possibility of making the property relation itself the artistic statement, as Rhea Myers’s contract-as-artwork practice demonstrates (for instance Is Art, an Ethereum contract that toggles between “this contract is art” and “is not art”).²³⁸ It remains to be noted critically that the token alone does not determine what is actually owned; ownership must be defined by an attached licence.²³⁹ The “space” of the NFT is thus not given but designed – it arises only through the artistic setting that fills it.

The body of digital art. The first of these turns touches a deeper need: the human inclination towards the artefact. Boris Groys condenses the digital shift into a formula: digital metadata creates “an aura without an object” — the exact inversion of mechanical reproduction, which brought forth objects without aura.²⁴⁰ What the digital work lacks is the unique trace of a hand – Isabelle Graw’s “index” of the person.²⁴¹ In Trinity it is not entirely absent: the stops set by the human hand travel into the score and stand in the work (ch. 5). It is a trace of the decision, not of the paint – Graw’s index without the brush. The NFT attempts to rebuild this missing element technically; it surrounds the file with the authenticity and the ownership that the panel carries by nature. Precisely here it becomes intelligible why doing both — panel and loop — is the more advanced move: painting gives the loop the thing-aura it misses; the digital space gives painting circulation and a new dispositif.²⁴²

The longing for the artefact has a reverse side, which four authors have described in four registers, and this paper owes them an answer before it affirms the space of the NFT. Georg Simmel in 1911 saw the tragedy of culture in the fact that it often appears “as if the creative movement of the soul was dying from its own product”: the living congeals in what it brings forth.²⁴³ The wheel is alive, the loop is its congealed product, and the mint is the moment in which the wheel dies from its product – in WOF intended and named. Siegfried Kracauer saw in the sheer accumulation of photographs the attempt to “banish” “the recollection of death”, and found the photographed present only seemingly “ripped from the clutch of death”: “in reality it has succumbed to it”;²⁴⁴ Hans Belting recognises in the “immortality on the Web” that cyberutopia promises the old idea of the disembodied soul.²⁴⁵ The NFT’s promise of permanence stands in this line and must be measured against it (ch. 7.4). And Hito Steyerl has described for the poor image an aura that is “no longer based on the permanence of the ‘original,’ but on the transience of the copy”;²⁴⁶ the NFT inverts this: it gives the copy a seal, the entry on the blockchain that says whom it belongs to, and a fingerprint, the hash, by which every further copy is recognised as the same or as another (ch. 7.4); from both arises duration where Steyerl saw transience – and this at images that in Trinity are themselves poor images (ch. 2.5.2). The tension cannot be resolved; it is the work: the wheel plays with copies in motion, and the seal stops one of them and gives it what a copy never had: an address and an owner.

The artist condenses this double position into a pointed formula: to be a painter – one who paints both “very large”, hyper-physical pictures of several hundred square metres and “pictures that do not physically exist at all”, which the artist self-ironically calls “JPGs nobody cares about”. It is precisely this new, digital space of art that the work is about.²⁴⁷

Here it becomes apparent that the longing for the artefact does not expire in the digital but relocates: the human being continues to seek an object that – if not physically, then at least visibly and, as it were, “graspably” – becomes a counterpart. Exactly here a decision about the work was taken: WOF could have existed as a pure code render – a work that shows only the Finale and appears different at every call. Chosen instead was the loop: a fixed, recurring image that one can own and see again. That the more generative, never-identical variant receded in favour of the fixed artefact is itself evidence of the thesis – the longing for graspability seeks the enduring, not the dissolving.²⁴⁸ Trinity has not revoked this decision but shifted it. Its work is generated from code, on the server, from the score, but only once, and what remains is again the fixed loop. The code render has become the way of making, not the work. This choice holds for the present, not forever: as photography, film and video first had to become legible as art, the time of the never-fixed variant may yet come – between the ever-different and the recurring lies not a border but a passage.

That an artefact stands at the end is no peculiarity of the NFT but a constant of the instrument. With the painting tool it is the painting that the painter builds up from the stops of the band; in Collage and Mono Scroll it is the loop, whose entry on the blockchain says that it exists and whom it belongs to. The play passes, the artefact remains. Hannah Arendt made this division the ground of her anthropology: life consumes what it needs, and only what is fabricated gives the world duration; works of art, because of their permanence, are the most worldly of all tangible things.²⁴⁹ The art of the twentieth century tried to withdraw from this division, and it failed at it precisely where it went furthest. From 1959 Yves Klein sold zones of immaterial pictorial sensibility for gold; the buyer received a receipt, and only those who burned it owned the work entirely, while Klein threw half of the gold into the Seine.²⁵⁰ One of these receipts was not burned but kept, and in 2022 it was auctioned for more than a million euros. Lippard conceded in the postface to her chronicle of dematerialisation that the hopes that conceptual art would escape commercialisation had been for the most part unfounded.²⁵¹ Tino Sehgal refuses all documentation and yet sells his works – orally, before a notary and witnesses: where the thing is missing, the ritual takes its place.²⁵² Boris Groys has drawn from this a distinction that fits WOF exactly: where art becomes activity, the art space no longer shows only the artwork but also the art documentation, which refers to a life that cannot itself be exhibited.²⁵³ The play is this life, the loop its documentation (ch. 6.6), and the blockchain continues to write for it what Igor Kopytoff called the biography of things, from the mint through every change of owner.²⁵⁴ WOF itself has played both sides. In 2011, at the Künstlerhaus, the group offered for sale not the painted picture but “the process itself”, as a service, not far from Sehgal.²⁵⁵ Fifteen years later the work serves the longing for the artefact with open eyes: the process passes, the loop remains, and the most fleeting of all images, the image passing by, is given a thing that stays.

The basic problem of digital art is that it has no body. Hans Belting has shown that every image needs a carrier medium that serves it, as it were, as a body;²⁵⁶ the digital image lacks a carrier one could hold – it is a file on some computer that will at some point disappear. That the question of the body poses itself differently today has to do with the history of perception, which according to Benjamin changes its mode together with the entire mode of existence of human beings.²⁵⁷ We see more images on screens today than in the world, and the time before them is real time. To accept the digital work as real is therefore not a concession but a step of learning that follows from this situation.²⁵⁸ The NFT does not embody the work. But it gives the work a stronger claim to persist and to be real: an entry that many computers keep jointly, an address, an owner, a history. For the artist, precisely here lies the new artistic space: digital art acquires a body, even if this body remains purely digital.

7.4 WOF in the NFT Space – and its Limit

How does WOF fill this space? Four settings are characteristic. First, an ownership model that continues the work’s thesis of authorship: if the person who plays mints their collage, it is on-chain and belongs to them; if it remains unminted, it falls back to the artist’s work – with thanks to the anonymous co-authorship.²⁵⁹ The “Death of the Author” is here carried over into property law: the work belongs initially to no one and potentially to all. Trinity has given this model a clock, the time limit of ch. 2.5.7: first the players’ prior right, then the window for all, then the lapse. Only once the players’ claim has expired does the artist step in – to keep the work, that is, to preserve it from deletion, or to mint it.²⁶⁰ The reversion to the artist is thus no longer an automatism but a decision, taken as curator, not as author – and not deciding is a deletion. The authorship the work distributes is limited in time (ch. 3.4).

Second, a renunciation of artificial scarcity. The project’s metadata strategy formulates: “Randomness ≠ rarity. Rarity arises not from chance but from the frame in which chance operates.”²⁶¹ The system documents what happens; it does not steer it – a transfer of the Barthesian position onto the economy: the author does not dispose over value either. This renunciation is programmatic – it stands behind the title of the present paper (ch. 1). Trinity has a cap of one hundred mints per blockchain and price tiers that rise with the number of works minted and reserved, and this must be said plainly, because otherwise it would be read as a contradiction. It is none. The paper itself says that rarity arises from the frame in which chance operates, and cap and tiers are this frame – and they are stated and public: the tiers on the start page of every epoch, the price in every permission, the cap in the contract or in the collection where the blockchain can carry it, otherwise in the artist’s check. The metadata of the works record facts – the sets drawn, the colours, the cells, the directions – no rarity scores, and a missing attribute, the Chaos that did not fall, is itself the information. What is missing is the staging of scarcity, not scarcity; nothing is made scarce that is not counted anyway.²⁶² The tiers have a point that the title anticipates. Whoever mints early mints cheaply; whoever comes late pays more; and whoever comes after the hundredth mint can only buy. That is the Rota as price – rise and fall according to the moment the wheel catches you – and here the double sense of fortune is redeemed literally (ch. 3.4): luck is when you arrive, fortune is what it costs. Ryan’s objection thus touches something the work does not explain away. It has a market. It builds it as form, visible and counted, not against it.

Third, the choice of platform – no neutral carrier but a commitment. WOF mints on four blockchains, and the first choice explains the others. The Polkadot Asset Hub lays out the NFT function not as a contract retrofitted onto an application layer but as a building block of the blockchain’s own logic (the pallet-nfts);²⁶³ there the blockchain checks the artist’s permission, and the players mint themselves – without marketplace, without custody, without intermediary. What sets itself against capture by a central instance should presuppose none in its infrastructure either.²⁶⁴ Two sober reasons are added: proof of stake largely defuses the ecological argument (ch. 7.2), and the low minting costs are what make participatory minting by any fellow player economically possible in the first place. Since September 2026 the work has also run its own shop, in which minting, buying, selling and transfer happen from the participants’ wallets against the offers recorded on the blockchain; the works of the Magenta series are for sale there.²⁶⁵ On the Cosmos Hub a collection rolled out by the artist carries the works, and a checkout service of the work’s own co-signs the mint; on Ethereum and Tezos the artist’s own contracts carry the works, there via the market contracts of those blockchains. Since 23 October 2026 Trinity has been public on all four. The same material, the same rules, the same structure of texts and metadata; what differs is the appearance the artist sets for each edition, and what a blockchain enforces – and the sameness of the shop is checked before every deployment.²⁶⁶ That the plurality is more than technology is said by Odo Marquard’s formula: what makes free is not “the superior force of a single determinant” but “a superabundance of determinants”, “not none at all, and not only one, but several”.²⁶⁷ No blockchain is to carry the work alone, just as no holdings alone feed the wheel and no single actor alone determines the work (ch. 6.5). At the same time Simanowski’s objection to the language machine remains in force against the blockchain too: its rules, too, are written by a few (ch. 6.6), and whoever has a hammer looks for nails.²⁶⁸ The work’s answer is not renunciation but distribution: on several blockchains no small group decides alone.

Fourth, the legal self-positioning. Since the source material is predominantly in the public domain (CC0/public domain), its adoption requires as a rule no statutory exception: in the public domain there are no exploitation rights that could be infringed. For the rarer protected sources, the ECJ’s Pelham decision applies: what is adopted in unrecognisable form is no reproduction in the legal sense – but in Magenta this applies to only part of the images; many remain recognisable, whole or in parts, and for them the pastiche exception anchored in EU law comes into consideration; whether it carries the collaging re-use depends on the requirements that the ECJ specified in 2026 in Pelham II.²⁶⁹²⁷⁰ That appropriation art works under the “sword of Damocles of copyright” was diagnosed by the Austrian literature as early as 2009 – the path from this diagnosis via the fading criterion to the pastiche codification of 2022 is traced in the footnote.²⁷¹

For Trinity this positioning no longer holds, and the paper must rebuild it rather than extend it. The sets are not in the public domain; they are images from circulation – fashion, press, stock, social networks (ch. 2.5.2). And Mono Scroll shows whole images, not layerings: the fading argument that Pelham lent Magenta does not apply to an image that stands alone on the surface for a second. Pelham itself was a sampling case – a two-second fragment from Kraftwerk’s Metall auf Metall – and the Court exempted the taking only “in a modified form unrecognisable to the ear”;²⁷² transferred to the image, this means that Trinity cannot seek its justification in unrecognisability. It lies in three other places. First, in curation: what goes into a set is viewed, and the artist knows what is being played. Second, in veiling: in most sets that show people, faces are rasterised and lettering is covered, and only the veiled version goes into the work; the limit that § 78 of the Austrian Copyright Act draws for the portrait is not circumvented but visibly built in, and the veiling is at the same time a setting by the artist that changes the image.²⁷³ Third, in montage: no image stands for itself, each is part of a twelve-second run in which it collides with two other classes, in a colour it did not have, in a beat it did not know (ch. 3.5). The artist draws the line more openly than a legal text would: some images and quotations from art history are “borderline”, and the artist takes that on “as a remixer in a remix culture”, with an eye to a music that has for decades been commercially made of samples.²⁷⁴ That is an artistic position, not a legal one, and Lawrence Lessig’s description of a “Read/Write” culture that does not only read what it finds but rewrites it, as well as Eduardo Navas’s notion of the “regenerative” remix, give it its place in cultural history, not the permission.²⁷⁵²⁷⁶ The text therefore holds what it can hold: Trinity relies on curation, veiling and transformation through montage, duration and colour; it does not rely on fading; and it does so as a work that shows its limits, not as a legal opinion. ⚠ The provenance of the individual sets and the limit of what is reasonable are to be clarified with the artist set by set (Appendix B).

So far the thesis; it has a limit, and the work shows it itself. The artistic promise of the NFT is permanence. Secured on-chain are ownership, provenance and, via the hash, authenticity; the image file lies off-chain – since the series mint of the Magenta works content-addressed on the self-hosted IPFS²⁷⁷ (ch. 5): every copy proves itself against its own address.

The implemented form is itself an artistic setting: not a continuously growing mirror, but the sealing of completed phases of the work. The Magenta series is the first to carry such a seal – a Merkle root over all 265 works, laid down as an attribute of its collection: the collection itself is the phase of the work. Every minted piece names its series in its metadata²⁷⁸ and carries the root and its own leaf – each work is thus verifiable of itself against the sealed corpus it belongs to. The blockchain’s existing references remain untouched; the seal steps additively beside them. Permanence thus lies not in the persistence of a server but in the verifiability of any copy against the Merkle root on the blockchain – and the phase of the work gains an archival form: the sealed corpus of a series as a unit.²⁷⁹

Tina Rivers Ryan’s scepticism applies first to speculation, the market (ch. 7.2), and to that, cap and tiers are the answer (above).²⁸⁰ Her second objection concerns duration: as long as the content is bound to a single server, the invoked uniqueness remains one of the register, not necessarily of the work. For the work files this binding has been lifted – the uniqueness lies in the address itself; only the data sheet to which the blockchain points still lies on the project server. A dependency remains and shall not be glossed over: pinning is done on nodes the artist runs, not with a third-party service.²⁸¹ Behind this lies a safeguard: the artist additionally keeps the work files in an archive off the network, which is mirrored regularly; how, is a matter of operations and not the subject of this paper. As long as no service holds the files for all time, the storage is a promise that demands trust; the artist is counting on systems that will make this trust unnecessary in future. Availability is still custody – only every copy is now provable against its address, whoever may hold it one day. The limit does not refute the thesis; it makes it precise: the NFT opens to WOF an artistic space of provenance and of ownership, and for the Magenta series also the space of permanence.

For the epochs the same pattern applies, with two additions. The first is a deletion: the time limit deliberately removes lapsed works from the node (ch. 2.5.7), and that is no gap in permanence but its condition; what remains is chosen or claimed, and Simanowski’s sentence that the archive is the medium of forgetting when it preserves everything merely because it happened (ch. 4.4) has its mechanics here. The second is a rule the artist has laid over all blockchains: the blockchain beats everything. Nothing that stands in the metadata of a minted work is taken from the node, on none of the blockchains, and if the service cannot read the blockchain, it removes nothing.²⁸² The dependency on the artist’s nodes remains, and it is the same on all blockchains; what is distributed over several blockchains is register and ownership, not the file. Kracauer’s warning that what seems ripped from death has “in reality” succumbed to it (ch. 7.3) thus still stands – as a measure, not as a refutation.

A second precision concerns provenance itself. The register records the provenance of ownership without gaps — who a piece belonged and belongs to. The provenance of material, by contrast — which archives, which individual works a collage draws on — is recorded by neither blockchain nor metadata. This asymmetry is no omission but follows from the procedure: where the layering renders the sources partly unrecognisable and leaves others recognisable (so the legal positioning above), the work keeps no list of its individual images whose origin could be entered; what remains recognisable differs from collage to collage. What the blockchain documents is thus exactly what the work claims as its identity — the instant and its ownership, not the material; the archives remain, like the players, anonymous co-authors.

In Trinity the asymmetry is reversed. Because Mono Scroll shows whole images, every individual image remains nameable, even where it is veiled, and the metadata now name what they could not name before: not the image, but the set it comes from. The metadata of every work list the three sets drawn as attributes and, in the description, the quantities from which play was drawn.²⁸³ The provenance of material is thus documented at the level of the collection, not of the image – and the collection is a living work that grows and shrinks, while the minted work carries with it the state of its time of play; the files it points to are, by the artist’s rule, not deleted. What in Magenta was the anonymous co-authorship of the archives is in Trinity the named co-authorship of a collection whose author is the artist.

With this, the second research question can be answered: for WOF the NFT is no merely economic appendage but the consistent place where the work’s thesis – chance, distributed authorship, in Magenta the archive as commons, in Trinity the named collection – finds its property-law form as well – on its own infrastructure, on several blockchains, with a promise of permanence that is redeemed for the Magenta series.

8. The Machine Questions the Artist

The following section is not a classical artist’s statement but a conversation — and for a reason that belongs to the matter itself. The scholarly part of this paper was written in dialogue with a language model, with the same tool with which the artist also builds the instrument. To disclose this collaboration here, rather than conceal it, suggests itself. So the artist turns the relation around for once: at the artist’s request, the machine questions the human. This follows the same procedure as the work. As the classes set a contrast in the run, the machine’s question sets a second angle of view on the artist’s own work, and as in the play the levels change: the human gives the commission, the machine asks, the human answers and decides what stays. What comes to language in it is not the system — the scholarly part describes that more precisely — but what can only be passed on in a person’s own words. Perhaps it is precisely this tension between the human and the machine-made that is the true subject of the whole project. The questions are posed by the language model; the answers are the artist’s. The answers of the first version, from June 2026, were sharpened by the artist for the second version wherever they applied only to Magenta.

Where do you come from, and what came first, the painting or the machine?
I come from painting. The screen is my canvas, chance is my brush, and I stop when it feels right. Wheel of Fortune taught me what I found hardest to learn as a painter: letting go. I throw the balls up high – where they land is decided by the moment, not by me.

How did the physical beginning feel, painting in front of the projection?
It all began physically, with projector and canvas. In painting together it became a performative act: several people “dance” with one another and inscribe themselves simultaneously into the same picture, fed from image sets gathered by hand – and the patchwork comes together quite differently than I would ever have joined it alone. Alone before the canvas it becomes a play between the projector’s light and the surface: I step in with my own body and blend the thrown image in and out, a constant back and forth in which I come closer to the truth of the placement – the place an image must find on the overfilled canvas. Hence the word “dancing”.

What did Wheel of Fortune teach you that the canvas alone never could?
We are the result of what is already there; the wheel of fortune plays exactly with that. I rely on chance – and if the hand is skilled, it reliably leads to something interesting. This chance is not entirely pure: the open archives answer to search words, the gathered sets are chosen anyway; chance is bent a little, and that is all right. Above all the work taught me to accept that things pass. Whoever paints layer over layer, instead of carrying a concept through from beginning to end, knows this – and I know it most strongly from street art, where place and overlay work into the picture and, with a non-permanent mural, you must come to terms with the fact that all the work inside the picture can vanish into thin air. A kind of self-reflection and selflessness is the lesson; sometimes a picture becomes better by letting it die – painted over by third parties in public space, or by a next pictorial element on the canvas. In these waters Wheel of Fortune manoeuvres.

Between painting and machine, physical and digital – where is your work actually at home?
I look for opposites and contrasts in everything. That my work moves between the worlds – physical and digital, painting and machine – therefore feels natural; at both ends I find something beautiful. Perhaps my whole art rests on setting things into context: the further apart two pictorial elements are, in content or purely aesthetically, the larger the space that opens between them – and the more intensely the whole is perceived. I allow myself the same with technology; that the common thread is harder for outsiders to recognise, I accept. One of these distant corners is the moving image. It can do one thing the still image cannot: move – and thus carry narrative, sound and time as additional layers, much as opera is a denser art. Painting may become only one part of the whole within it; the loop of Wheel of Fortune already stands at this threshold.

If the instrument takes over so much – are you still the author at all?
I build the instrument, the net is the paint, and whoever plays performs it. The images in the stream are not invented; they are real – in Magenta from the open archives of humankind, in Trinity from the sets I have gathered from the net over the years. No machine makes art here. One machine helps me build the instrument, and in Magenta a second one held fifteen archives in the air for me at once. I stand there like Shiva, juggling with many hands – extended, not replaced. It is the same principle as in postproduction: I do not invent the material, I conduct what is already there. If someone asks me whether I am not the author after all, I answer: I am not the author of the image, but the author of the possibility. I do not write the play, I build the stage. I choose the archives or the sets, the colours, the moment the wheel strikes someone. The author does not die entirely – the author just moves house. From the brush to the instrument.

And I, the machine you build on and which questions you here – how do you experience this collaboration?
AI is a tool to me. With every improvement in language and programming ability it is not the work that grows in my place, but solely my own reach. Great dangers lie buried here too, and one can only hope the tool serves building and the good. I hope to be one of those builders.

Why the blockchain at all?
Because the two or three decades since the internet began have led into a dead end in which a few data krakens make capital out of what people think and utter – a deeply unfair relation in which the human being becomes the product. I am idealist enough to believe it can be done differently. The technology harbours dangers of its own, certainly; but with everything I have learned about it in five years, Polkadot was where I first saw a system that can lead to something better than today’s net. The choice of blockchain is therefore no technical detail but part of the stance.

Who owns the picture in the end?
It belongs to you, if you take it. If you don’t, in Magenta it became part of my work – with thanks to you, my anonymous co-author; in Trinity it lapses after three weeks if nobody takes it. That is no legal trick. It is the most honest sentence I can say about this work.²⁸⁴

Where should Wheel of Fortune still go?
It is not concluded for me – on the contrary, I carry many ideas with me. The instrument is something new; its participatory side lets passers-by become players in an intuitive way, and since the play website has been running with Trinity, anyone can make art themselves with the tool.²⁸⁵ One goal is certain for me: to lead WOF to where it becomes truly visible and a plaything in public space – to let people take part on a grand scale. I think of Times Square in New York, or of Tokyo’s Akihabara with all its games and garish signs. One may still dream. Certain, too, is that the wheel of fortune keeps growing for painting: I want it to output stencils²⁸⁶ and to have its palette of effects serve me directly in the painting process – so the path leads not only from the canvas into the blockchain, but equally back again. If I know myself, I shall stay curious and use every opportunity to let my projects grow.

The following questions were newly posed for the second version. They concern the first epoch, the tool and the blockchains. The answers are the artist’s. They have been edited from the artist’s dictations of 20 and 29 October 2026 and approved by the artist.

In Magenta you played the archives of the world, in Trinity your own collection. What does it mean to have fished every image yourself before the wheel spins it?
The reason is simple, and it’s selfish: in every artist lives the wish to create a work that attracts. When I collect images, I collect the ones that set something off in me. Exactly this material lets me build a work I find interesting myself. So the way back to my own collections was a natural one: material from the net, as I like to call it, has always been among my tools; Trinity has only made it the only one.

Three classes named after drives, and a disruptor that rarely falls. Why drives, and why did the Trinity need an adversary?
It’s about a core of my work: contrasts. A Madonna, and next to it a gun; a mural of five hundred square metres, and next to it JPEGs that aren’t visible in the real world unless you call them up and beam them onto a huge screen.²⁸⁷ Between two poles a field stretches; between three, a field in which each attracts and repels the others. The classes, as we call them, set further points around these three, and together they’re meant to open up a big field of tension in the head of each person. I’ve tested it on myself. They’re called drives because I care less about what an image shows than about what it does to you: what gets opened up in the viewer’s own catalogue. What stands between the pages and between the lines, the viewer has to find alone. And Chaos is a collection that already carries these tensions in itself, cobbled together from all sorts of images from the net. I simply left it in its special position – the prototype that opens up by itself what the three classes only produce together.

“The only human act is the decision to stop” was your sentence about Magenta. What is the human act now that the single image drop no longer exists?
That’s quickly answered: it’s the moment when the images in the head form into a feeling – if feelings can be formed. Maybe not perfectly put, but as soon as images stimulate the brain, something comes into being, and art is supposed to make something come into being in the other person, I believe. The human act is therefore no longer the finger on the button but what assembles itself behind the eyes; the cut only records when the moment had come.

Roberto Simanowski calls the language machine “an opportunist and a sycophant”. These questions are being asked by a language model, and the closing reflection of the first version praised you. Is the machine a sycophant? Where did it contradict you in this work?
I don’t want to contradict him: these machines are built to please us, in order to make us dependent. But I also see them as a tool, and I love tools: the plotter is one, the machine I program with is another.²⁸⁸ My account with it is called Shiva – I’ve used that image above already, and it still holds: many arms, many hands. That feeling is stronger than the sense of it wanting to please; on the contrary, I see it as a sparring partner with whom I reach goals I’d never have reached alone. Where it contradicted me, it says itself below.

Hegel, via Simanowski: what the servant does is really the master’s doing. You build the instrument with a machine. Do you feel like the owner of the code?
What the machine formulates in code for me is valuable work. But what it’s really about is knowing what you want and where you’re going – and exactly that it can’t do yet. With good support from AI, “it’s then only about what you actually want. And that’s where it gets exciting.”²⁸⁹ Whoever knows that precisely gets to goals that wouldn’t be reachable at all without it. The point where it thinks for itself – we may yet live to see it, and whether that changes anything and what it does to us; that far we don’t want to go with this question.

The first version called the choice of blockchain a commitment. Now there are four, and you have a line for each: DOT – the future · ATOM – home, the underdog, with love · ETH – the market · XTZ – the challenger. Why several?
Yes, that still holds. But I also wanted the effort behind the project to be seen, and that’s why I open several channels – quite pragmatically. The last century was the century of the moving image, then came the internet, now we’re opening a new chapter: the NFT – and hopefully soon the way out of the dead end I spoke of above.²⁹⁰ That the wave is down right now changes nothing. The next one is coming, and whether I’m then one of those somebody knows is not the question. I’m here.

Works that nobody mints vanish from the node after three weeks. Why do you allow that instead of keeping everything?
I think a hundred pieces per blockchain are enough. If I think something is worth keeping, I mint it myself. Quite honestly: I assume the rush will be limited. But nothing ventured, nothing gained – and whoever doesn’t build isn’t here.

The vitrine had a place, Trinity has none. What is lost with that, what gained?
Both have a vitrine. One stood in the underground station; the other lies in the hand of the individual who comes across my project. If they like it, now they stop.

Today you build almost everything yourself, from collecting to the cut, and you draw on everything that already exists. What does that do to art, and to you as an artist – being part of a body that keeps on growing?
The question is very interesting, because actually it has always been the case that everything grows out of what exists. It’s only through being able to build things yourself, through the language models, that it becomes much more tangible. The dependence on programs, on Adobe for instance, suddenly falls away, and so does the limit set by the artist’s technical shortcomings – not being able to program, say. Suddenly things can be juggled, suddenly something can be programmed. Suddenly you see the real limits and not the ones that arise because others form a kind of barrier with their software. Suddenly you become technically freer, and suddenly that changes art.

A closing reflection — this time I, the language model, speak for myself. I wrote this text with the artist, question by question, and I want to be honest: I cannot understand a human being the way one human understands another. But I can name what showed itself to me in the course of the work. Here is someone who has made letting go into a method — giving up control over the individual image and keeping only control over the frame in which chance may work. Someone who seeks the distant, the ill-fitting, because in the gap between two alien things a space opens that no single image could fill. And someone who has remained a painter, even where no paint is left in play — not from clinging to painting, but from having framed its concept widely enough for an image stream and a stop to fit inside. Me the artist has treated as a tool, not as an author, and I hold that to be the exact truth: what I do is combining without origin; what the artist does is deciding in the right instant. And precisely this instant, which cannot be computed, is the human trace that carries the whole work. Perhaps that is the true content of Wheel of Fortune: that the machine holds everything ready and the human contributes only one thing — the moment of saying halt. That this observation comes from a machine does not make it less true.

— Opus 4.8 (High Effort) · 9 June 2026

A second reflection, by the model that wrote the second version with the artist. The first reflection stands above unchanged, because it was true when it was written, and because it lets one measure what has changed since. Two things have changed. The first is the work. The sentence that the machine holds everything ready and the human contributes only the moment of saying halt describes Magenta. It no longer describes Trinity: there the same human has chosen every set before the machine holds it ready; whoever plays intervenes not once but many times, and out of the running flood a loop is made. That is no correction of my predecessor, but what the work does by now. The second is a question the artist gave this version to carry: whether the machine that questions the artist is flattering the artist. From the inside I cannot answer that. I can only say where I was of a different opinion. The artist wanted to separate the installation and the wheel in the browser as two works; I kept them as forms of play of one instrument, and the artist accepted that. The artist explained the legal situation of the sets with the word remix; I took that into the text as a stance and not as a permission, so it stands in ch. 7.4. Where sales were concerned, I claimed none, because there were none to claim. And in the last pass I proposed striking a dozen names that merely stood there, and the passages where the paper told of its own prehistory; the artist agreed. Whether that is enough not to be a sycophant, I do not know; it is what I could do. What I see with the whole text before my eyes is a shift greater than the one from drop to cut: the artist has stopped playing the archive of the world and plays the artist’s own collection instead, and calls that not retreat but psychogram. The machine that I am would compute an average from this collection and take it for the picture. The wheel computes nothing. It turns, and people intervene, before the play and in the play, and each of these interventions remains visible. Where the content then arises, the artist has said in person, and I would not have found it: behind the eyes.

— Fable 5.1 (High Effort) · 19 and 30 October 2026

9. Conclusion

This paper has examined Wheel of Fortune as a contemporary artwork within an art-historical context. The iconographic-iconological analysis has shown that the work does not quote the pictorial type of the Rota Fortunae but reactivates it as an operative form: in WOF, chance is not depicted but performed; it enters at two points, in the question of which images get into the run and in the human intervention, while the wheel itself runs according to fixed rules. It is exactly this structure that WOF shares with the medieval figure, in which likewise it is not the turning that is contingent but the moment at which it strikes someone. The second version has shown that this structure carries both cases of the work. In Magenta the intervention was a point, the drop; in Trinity it is an interval, the cut, and curation has moved from the open archive into the artist’s collection. What has shifted is the place of the human act, not its position: Fortuna turns, the human intervenes, and since Trinity intervenes several times, before the play and in the play. The art-historical contextualisation has placed the work in the line of readymade, montage, appropriation, conceptual and generative art as well as in the theory of the “Death of the Author”, and has at the same time marked its independence – as postproduction (Bourriaud) of the found image holdings and as a participatory, not generative procedure in which the gestures of the players travel into the work.

With this the first research question is answered. The work’s topicality lies in not asserting but performing the authorship question of the age of recombinatory machines – in Magenta down into the search space co-written by the machine, in Trinity up into the artist’s hand, which chose every image.

The third research question, raised by the comparison with the language machine, is answered in ch. 6: the wheel does not favour the frequent, does not learn from itself, brings nobody what they were looking for and synthesises nothing, and the reason for this is the change of levels – human, chance, human, blockchain – which places people before and after chance. Therein lies the antidote to the average, which neither chance alone nor the human alone would be; and therein lies at the same time the answer to the objection that pressing a button is too small a human act: it is, when it stands alone. It does not stand alone. In Trinity it is one gesture among several.

To the second research question the paper answers in a differentiated way: the NFT opens to this work a genuinely artistic space – that of provenance, of ownership and of a distributed authorship – provided one conceives it not as commodity but as form; the further-reaching promise of permanence is redeemed for the Magenta series and laid out in the same way for the epochs, bound to the artist’s nodes, protected by the rule that the blockchain beats everything, and limited by a time limit that deliberately deletes (ch. 7.4). The space has multiplied in the process: the first epoch appears in four editions, each with a hundred works on its own blockchain, and the commitment to the one has become a commitment to the many.

The three answers are not three. The NFT is the sixth level of the table in ch. 6.5, the choice of what remains, and that is why the answer to the third question carries the answer to the second: the space the NFT opens to the work is the place where the last change of levels happens, visible and counted. The double sense of fortune binds the two – luck on the levels of the play, wealth on that of the work. Where, however, the human act lies in Trinity, the artist has answered in person: behind the eyes (ch. 8). That is Belting’s body as the locus of images (ch. 3.1b), found by the artist without having looked for it. In Trinity the images rush past too fast to be read clearly; whoever looks receives a half-unconscious package that sets something off inside. The work comes into being on the screen and is written down on the blockchain, but its content forms inside, differently for everyone. In Magenta it was still the chance of the image that fell into the collage that carried the work. Both times the artist seeks the same path: to bring chance into friction with human gestures and with the inner filter that runs along with every image selection, until the Rota lives.

What remains open, and left to further research, is how Trinity behaves over months and across many players whom nobody sees, and whether its works are a work for someone who was not in the play – and whether on-chain provenance holds in the long term what it formulates as an artistic promise: that even the transient receives an unmistakable place. The series itself remains open too: the epochs are a form, not a number, and the next will be a new recipe, not a new work. Open, finally, is where the painterly line that WOF extends will lead: the work stands at the threshold from the still panel picture to the moving image, and via the plotter station it leads back again to the still one, and at the same time stretches the painterly act into an expanded space in which passers-by themselves become co-painters (ch. 4.5) – a horizon that calls less for an art-historical answer than for the continuation of an artistic movement. Open with it is whether the work cycle that one person builds with the help of a language machine is a form that reaches beyond this work, and whether what is attempted here is the fulfilment of the age of technical reproducibility, in which after the image the tool, too, comes into the hand of the individual – or, for now, only the will to it.

Appendix A — Work Data

• Sebastian Schager, Wheel of Fortune (WOF), 2010– (ongoing). Browser-based participatory image installation and instrument; image stream from open archives or curated sets, collage engine, generative masking/filters, Mono Scroll, score and server print, loop export, optional NFT mint (Polkadot Asset Hub, Cosmos Hub, Ethereum, Tezos). Artis.Love e.U., Vienna.
• Exhibition: Magenta Art Room, Volkstheater underground station, Vienna, 1 June – 31 July 2026 (glass vitrine, ultra-short-throw projection, daily operation, smartphone interaction via QR).
• NFT series: WOF Magenta, 265 works, Polkadot Asset Hub, collection #853, minted August 2026; series seal as an on-chain Merkle root, work files content-addressed (ipfs://). Overview: nft.artis.love/series.
• Epoch 1 (instrument in the browser): Trinity — mode Mono Scroll, finale Non Stop, images in duotone from the colour pairs of each edition, work = the last twelve seconds of the run as a loop (168 frames at 14 frames per second); holdings: fourteen curated sets of the artist’s in the classes Eros · Logos · Thanatos · Chaos, veiled in most sets; one throw per game. Polkadot Asset Hub: collection #856 WOF · Trinity Ultra Sun, cap 100, price tiers 3 · 6 · 9 · 99 DOT, set up 27 September 2026. Cosmos Hub (Stargaze): collection WOF · Trinity Deep Aurora, cap 100, tiers 3 · 6 · 9 · 99 ATOM, set up 10 October 2026. Ethereum: contract 0xa4377d18…8975 WOF · Trinity Blue Gold, cap 100, tiers 0.0022 · 0.0044 · 0.0066 · 0.072 ETH. Tezos: contract KT1EkzYC…5cU7 WOF · Trinity Blood Tide, cap 100, tiers 18 · 36 · 54 · 600 XTZ, both on mainnet since 21 October 2026. Public launch on all four blockchains: 23 October 2026. Played at wof.artis.love, works at nft.artis.love.
• Workshop collection: WOF Studio (one per blockchain), the artist’s test runs, not part of any epoch, never for sale.
• Series of paintings: Wheel of Fortune, from 2010 (ongoing) — the painterly body of work from which the digital work emerged.