Semantic Infrastructure is a theoretical paper by Lee Sharks that places the Semantic Economy framework in direct relation to the Semantic Web and knowledge-graph traditions. It argues that machine-readable meaning is already material infrastructure and therefore already has a political economy.
The paper begins with the Semantic Web vision associated with Tim Berners-Lee, James Hendler, and Ora Lassila. RDF expresses triples, OWL defines ontological relations, SHACL validates constraints, SPARQL queries graphs, and Schema.org organizes public web markup for machine consumption. Knowledge graphs extend these standards into search, recommendation, enterprise systems, and AI retrieval.
Its intervention is that these systems contain accumulated labor that technical descriptions often render invisible. Ontologies require experts and committees. Open graphs require continuing editors. Webmasters provide structured markup. Data labelers and moderators sustain later AI systems. This work is described as semantic labor, while its durable output becomes semantic capital.
The paper then maps technical and economic categories. Standards form semantic infrastructure. Knowledge graphs store semantic capital. Knowledge panels, APIs, and proprietary graph services can extract semantic rent. Training on open knowledge may liquidate attributable structures into proprietary model weights. Declining maintenance, invisible usage, recursive synthetic training, and volunteer burnout are grouped under semantic exhaustion.
The paper’s strongest claim is that “semantic economy” is not only a metaphor because structured meaning contributes measurable value to search, advertising, product discovery, and AI services while depending on real labor and maintenance. At the same time, the work acknowledges that many extraction dynamics are structural rather than reducible to individual malicious intent.
The article should treat this record as an interdisciplinary synthesis and vocabulary bridge. It does not establish a universally accepted field taxonomy, but it gives the archive’s economic concepts externally recognizable technical referents.