The Museum of Everything

In the AI Story

Every previous technology that reduced the cost of cultural production generated a version of this question, and every previous generation answered it by building institutions of curation. The printing press produced the library, the bookshop, the literary review, the university reading list. Photography produced the gallery, the museum of modern art, the curated exhibition. Recorded music produced the record label, the radio programmer, the music critic, the playlist. Each was a mechanism for filtering the flood of produced material into something a human mind could navigate. The criteria varied — scholarly significance, literary merit, commercial appeal — but the function was consistent: reduce the space of available artifacts to the space of attended artifacts, guided by human judgment about what merited attention.

In each case, the curatorial institution emerged in response to a specific condition: the cost of production fell faster than the cost of attention. When anyone can produce, the question of what deserves attention becomes structurally important. The answer was always supplied by human judgment, institutionalized in structures that translated individual discernment into collective focus. AI has produced the most extreme version of this condition in cultural history. The cost of producing a formally competent artifact in nearly any domain has approached zero. The cost of attention has not changed. The human mind can still hold only a finite number of artifacts in focus. The disproportion between production and attention has never been greater.

Kubler's framework provides the theory of curation required. The artifact that merits attention is the one that occupies a structurally significant position in a formal sequence — not the most beautiful, not the most accomplished, but the one whose position either opens new possibilities or extends existing ones in directions the sequence had not previously explored. The prime object merits attention because it opens a sequence. The replica that extends a sequence in a genuinely new direction merits attention because it reveals possibilities the prime object only implied. The replica that occupies a position already filled — however competently, however beautifully — does not advance the sequence and does not, by this criterion, merit the same order of attention.

The criterion is severe. Applied rigorously, it would empty most galleries, most playlists, most bookshops. The vast majority of artifacts in any domain occupy positions that have already been filled. The severity is the point. In a museum of everything, a criterion that admits everything is no criterion at all. The practical application requires a capacity no algorithm currently possesses: the capacity to read a formal sequence well enough to identify where its live edges lie. This is the curator's skill, and it is the skill the AI age has made most valuable — not selecting for beauty, novelty, or reputation, but selecting for structural position. The curator in the museum of everything is not algorithmic. She is a mind shaped by deep entrance into formal sequences, carrying the structure of those sequences as an internalized map, and able to perceive, amid the infinite abundance, the finite number of artifacts that change the landscape.

Origin

The concept is the book's image for the condition AI-scale generation produces, extending Kubler's framework from cases where abundance was bounded to a case where abundance approaches infinity. The image generalizes across domains — visual art, code, music, text — because the structural problem is the same wherever AI produces formally competent artifacts at scale.

Key Ideas

Infinite production, finite attention. The defining condition of the AI age is the radical disproportion between the cost of producing artifacts and the cost of attending to them.

Curation is not algorithmic. The capacity to identify structurally significant artifacts requires reading sequences deeply enough to perceive their live edges — a capacity built through entrance, not through computation.

The criterion is structural, not aesthetic. The artifact that merits attention is the one whose position advances the sequence, not necessarily the one that is most beautiful or most accomplished.

Historical continuity of the problem. Every previous technology that cheapened production generated curatorial institutions; AI requires a new generation of such institutions operating at unprecedented scale.

The curator's skill is the scarce capacity. In a museum of everything, the capacity to walk through and stop in front of what matters becomes the single most valuable cognitive skill a person can possess.

Debates & Critiques

A live question is whether AI itself can be developed to perform the curatorial function — whether machine learning systems trained on human curatorial judgments can approximate the capacity to identify structurally significant artifacts. The book's position is cautious: current recommender systems optimize for engagement, not structural significance, and the capacity to perceive where a sequence's live edges lie has not been demonstrated by any AI system. Whether it can be in principle remains open; as of 2026, the curatorial function is a human responsibility that AI tools can assist but not replace.