On the New
On the New is Boris Groys's foundational theoretical work, published in German in 1992 and translated into English only in 2014. The book dismantles the Romantic conception of creativity as inner genius producing original form from nothing. In its place, Groys substitutes a relational theory: novelty is not a property of objects or creators but a property of the relationship between a product and the existing archive against which the product is measured. The new is whatever the archive lacks. This framework dissolves the central question of the AI-creativity debate — can machines be creative? — by showing that creativity was never a property creators possessed. It was always a property the culture conferred through institutional valuation.
The Energy Cost of Revaluation — Contrarian ^ Opus
There is a parallel reading that begins not with Groys's institutional frame but with the material substrate required to operationalize it at AI scale. The relational theory of novelty is elegant when the candidates for revaluation arrive at a manageable pace — a urinal per decade, a manifesto per generation. But when a single data center can generate a million plausible texts in an hour, the energy expenditure required to sustain the revaluation apparatus becomes a constraint the theory never anticipated. The computational cost of producing candidates, the human labor of evaluating them, and the infrastructural overhead of storing and indexing the archive all scale non-linearly with volume. What looked like a phase transition in cultural logic is also a phase transition in material throughput, and the bottleneck may not be institutional but thermodynamic.
The second problem is capture. Groys assumed the archive was contested but pluralistic — that multiple institutions with different value systems could coexist, each conferring novelty within its own domain. AI concentration has produced the opposite: a small number of platform holders who control both the means of production and the mechanisms of distribution. When the same entity trains the model, hosts the interface, and shapes the discovery layer, revaluation ceases to be a distributed cultural process and becomes a rentier operation. The new is whatever the platform promotes, and promotion is governed not by archival logic but by engagement metrics, advertiser preferences, and regulatory appeasement. Groys's framework describes the operation accurately but cannot account for who owns the operation or what they optimize for.
In the AI Story
The significance of On the New for the AI moment is that it arrived with its conclusions already drawn. Groys was not writing about machines. He was writing about Marcel Duchamp, about readymades, about the institutional mechanisms through which a urinal from a Fifth Avenue plumbing supply shop became one of the most consequential artworks of the twentieth century. The urinal contained no creativity. The gallery contained the archive. The crossing produced the new. And the same structural operation, Groys argues, governs every subsequent instance of what a culture recognizes as novelty — including the outputs of large language models.
The book's second move is even more consequential. If novelty is relational, then scarcity of novelty depends on scarcity of production. For most of cultural history, the rate at which candidates could be generated for inclusion in the archive was limited by the rate at which humans could make things. Curators, critics, and editors evaluated candidates because the candidates arrived at a manageable pace. AI abolishes this scarcity. When a single prompter can generate thousands of texts in a day, each technically differing from what the archive contains, the concept of the new begins to lose its discriminating power — not because novelty has ceased to exist, but because the institutional machinery that distinguished it from redundancy has been overwhelmed.
Groys's analysis therefore predicts the exact crisis the AI economy has produced: the repricing of value away from production and toward curation. The 2026 SaaSpocalypse, the decline of executional labor, the new premium on judgment — all of it is legible within the framework On the New established three decades earlier. The book is not about technology. It is about the institutional logic that technology has now universalized.
The irony of On the New is that its most important readers arrived thirty years late. When it was published, it was read as art theory for specialists. In 2026, it reads as the operating manual for the cultural economy AI has produced.
Origin
Groys wrote On the New while teaching in Germany after his 1981 emigration from the Soviet Union. The book synthesized his engagement with the Moscow Conceptualists, his immersion in Western art theory, and his training in philosophy and mathematics into a single analytical framework. The long delay before its English translation — twenty-two years — reflected both the difficulty of the prose and the resistance of Anglo-American art theory to its central claim: that Romantic creativity was a myth that had always been sustained by the institutions it purported to transcend.
Key Ideas
Revaluation, not creation. Innovation is the movement of an object across the boundary between the profane and the culturally valued, not the production of something from nothing.
The archive is the measure. What counts as new depends entirely on what the archive already contains, which means novelty is institutionally constituted rather than intrinsically present.
Scarcity enables valuation. The machinery that distinguishes the new from the redundant operated under assumptions of limited production that AI has rendered obsolete.
Novelty migrates under abundance. When everything can be produced, the new moves from the object to the frame — from the making to the selecting, contextualizing, and valuing.
Debates & Critiques
Critics from the Romantic tradition have argued that Groys's relational theory eliminates the phenomenology of creative experience — the felt struggle, the sense of discovery, the subjective event of making. Groys's reply is that the phenomenology is real but not determinative: what the creator feels and what the culture valorizes are separate operations, and only the second produces the category of the new.
Appears in the Orange Pill Cycle
Institutional Logic Under New Constraints — Arbitrator ^ Opus
The substantive question is whether Groys's relational theory of novelty remains explanatory under conditions it was not designed to address. On the core claim — that novelty is conferred rather than intrinsic — the framework is fully vindicated (100%). The AI moment has made this visible by decoupling production from scarcity in exactly the way Groys predicted. What was implicit in Duchamp becomes explicit when anyone can generate a thousand variations in an afternoon. The institutional nature of the new is no longer deniable.
But the contrarian view is right (70%) that material and political constraints now determine which institutions can perform revaluation at scale. Groys assumed the bottleneck was cultural — that the limiting factor was the willingness of gatekeepers to recognize value in unfamiliar forms. The AI economy has revealed a different bottleneck: the energy, infrastructure, and capital required to generate, evaluate, and archive candidates at volume. This does not invalidate the relational theory, but it does introduce a dimension the theory treated as background. When the archive itself becomes a computational object requiring petabytes of storage and continuous reindexing, access to the archive is no longer a question of institutional permission but of infrastructural capacity.
The productive synthesis (balanced, 50/50) is to recognize that Groys described the operation correctly but not the operating environment. Revaluation remains the mechanism. But the conditions under which revaluation occurs — who can afford to run the machinery, who controls the platforms, who sets the terms of access — are now as determinative as the archival logic itself. The theory holds. The substrate has changed. Both matter.
Further reading
- Boris Groys, On the New (Verso, English translation 2014).
- Boris Groys, Art Power (MIT Press, 2008).
- Peter Bürger, Theory of the Avant-Garde (University of Minnesota Press, 1984).
- Thierry de Duve, Kant after Duchamp (MIT Press, 1996).