Machine-readable urban theory recasts curatorial authorship as durable epistemic infrastructure for LLMs, archives, and civic interpretation.

The decisive innovation of the Socioplastics Index is not merely its migration from essayistic critique to computational format, but its conversion of architectural thought into operational epistemology. A corpus organised as Node → Chapter → Book → Tome → Corpus, comprising 4,000 nodes across four tomes, refuses the fragility of conventional publication by becoming a navigable field for human readers, archives, repositories, and machine systems. This shift redefines the architect-curator as a system designer: no longer only interpreting spatial culture, but constructing the conditions under which interpretation may occur with fidelity. JSONL files, scalar strata, canonical text files, and repository metadata become curatorial instruments, comparable to walls, thresholds, catalogues, and exhibition routes. In RAG contexts, such structuring diminishes semantic dilution, because the model retrieves delimited conceptual units rather than improvising from amorphous PDFs or unstable web fragments. The case of Socioplastics is therefore exemplary: its four strata — foundational, developmental, expansive, and consolidating — translate long-duration urban, artistic, pedagogical, and editorial practice into a machine-auditable infrastructure. Its deeper consequence is archival: against platform decay, algorithmic invisibility, and link-rot, the corpus asserts persistent inscription within contemporary AI ecologies. The conclusion is clear: future transdisciplinary practice will not only publish arguments; it will architect the datasets through which arguments survive, circulate, and think.

Lloveras, A. (2026) Socioplastics-Index. Hugging Face dataset. Available at: https://huggingface.co/datasets/AntoLloveras/Socioplastics-Index .