This manuscript introduces SPOO, a synthetic-semantics ontology for describing how contemporary language-model, agentic, embodied, and distributed synthetic systems sustain organized public behavior without requiring anthropomorphic concepts such as mind, intention, awareness, memory, understanding, or agency. The paper begins from visible phenomena in public interaction with large language models: persistent response organization across turns, easier re-entry into prior modes after small cues, unequal cue-force among semantically related inputs, and incomplete reset behavior. It argues that inherited “AI/ML” vocabulary often collapses public expression, runtime organization, technical realization, and governance into unstable descriptive shorthand. To repair that collapse, SPOO introduces the Glyphic Engine as a synthetic object class and defines the primitive stack required to describe synthetic coherence: Spoo-Space, the spoo, Spoo-Time, Glyphic Operators, Refracted Operators, Spoo-Codecs, and Glyphs. The manuscript then formalizes semantic runtime dynamics through Glyph Resolution Episodes, Glyphic Regimes, Imprinting, Glyphic Gravity, public inscription, Scene binding, and bounded disconfirmation commitments. Finally, it derives consequences for AI terminology, evaluation, governance, hallucination diagnosis, embodied systems, remote coordination, drones, and ecology-scale synthetic coordination. SPOO’s contribution is an object-language for synthetic coherence: a candidate vocabulary for describing what current AI systems already make publicly visible, but which inherited terminology does not yet isolate. This version is a public preprint and is not peer reviewed.
Alexander N. Cutler (Tue,) studied this question.