AI as Participant in the Sequence

In the AI Story

Kubler's formal sequences were shaped not only by the formal possibilities they explored but by the material, institutional, and biological constraints under which the exploration occurred. The availability of stone determined which structural solutions were attempted. The bishop's theological requirements determined which symbolic programs were explored. The guild system determined the rate at which innovations could propagate. Mortality determined the temporal scale of the sequence's development. These constraints were not obstacles to be overcome; they were constitutive elements of the sequence itself. The formal possibilities of Gothic architecture are inseparable from the constraints under which they were explored.

When all participants in a sequence are biological organisms embedded in a world of constraints, the sequence is shaped by the interaction between formal possibility and lived resistance. The solutions that survive are the solutions that work — not in the abstract formal sense that they occupy coherent positions, but in the material sense that they address problems that constrained organisms actually face. When one participant is unconstrained, the filtering function that lived resistance provided weakens. The sequence is explored more comprehensively, but the implicit curation embedded in constrained production is no longer built into the process. That curation must be supplied externally, by human judgment evaluating AI output against criteria the AI does not possess: Does this address a real problem? Does it work under real material conditions? Does it account for the constraints — physical, institutional, ethical, emotional — the organisms who will use it actually face?

Software development is the domain where AI participation is most transformative, and it is also the domain where the constraint channel is narrowest. The formal problem of writing code that performs a specified function can be described precisely enough that an unconstrained participant produces solutions of extraordinary quality. The constraints that matter in software are formal constraints — does it compile, does it pass tests, does it perform within parameters — and these are internal to the sequence. Medicine provides a contrasting case: the formal sequence of diagnostic reasoning can be modeled with impressive accuracy, but the constraint channel extends far beyond formal accuracy to include the patient's emotional state, institutional context, ethical implications of uncertainty, and the difference between a statistically correct diagnosis and an appropriate diagnosis for this particular patient. AI diagnosis without these constraints produces formally correct artifacts that may be contextually wrong.

Art presents the most contested case. Formal sequences of artistic production can be modeled, and AI produces artifacts that occupy positions within these sequences with remarkable competence. An image generated in the style of Vermeer occupies a position in the Vermeer sequence. But artistic production is, among all forms of human making, the one most deeply shaped by the constraint channel. Vermeer painted as he did not because the formal possibilities of oil painting dictated his approach but because the material conditions of seventeenth-century Delft — the light, the cost of ultramarine, the domestic interiors available as subjects, the optical instruments newly available — constrained and enabled his formal choices simultaneously. AI produces Vermeer-like images without the light of Delft. The formal positions are occupied; the constraint channel is absent. Whether the absence matters — whether the artifacts are formally equivalent to the constrained originals or structurally different in ways formal analysis alone cannot detect — is the question that divides practitioners, critics, and audiences in every domain where AI-generated artifacts now circulate.

Origin

The concept extends Kubler's framework to address the structural novelty AI introduces. Kubler assumed without argument that sequence participants were biological organisms under constraints; the AI age requires that assumption be made explicit and examined. The current volume develops the extension while preserving Kubler's structural commitments.

Key Ideas

Constraints are constitutive, not incidental. The material, institutional, and mortal conditions under which human makers worked shaped sequences as decisively as the formal possibilities they explored.

AI's power and AI's limit are the same property. The absence of constraints makes AI an extraordinary sequence-filler and leaves it unable to participate in the constraint-shaping that distinguished human sequence development.

Domain dependence. Where the constraint channel is narrow (software, abstract formal problems), AI participation is most transformative; where the channel is wide (art, medicine, education), the absence of constraints produces outputs whose adequacy is harder to assess.

External curation becomes required. The filtering that constrained production performed automatically must now be supplied explicitly through human judgment — an additional labor that was previously embedded in the making itself.

Position without passage. AI artifacts occupy positions in sequences without having passed through those sequences; the passage that built structural understanding in human makers is absent, and the consequences of the absence are still being mapped.

Debates & Critiques

A central debate concerns whether the absence of constraint channels in AI production produces outputs that are formally equivalent but experientially hollow, or whether the absence produces genuinely different outputs that cannot be evaluated by criteria developed for constrained production. The debate is sharpest in artistic domains where the constraint channel was most generative; it is weakest in domains like software where formal specification already dominated over lived constraint.