On the Structural Preconditions of Artificial General Intelligence A Structural Inscription Overview A structural inscription recording the conditions under which Artificial General Intelligence may be placed without erasing responsibility. This is not a definition of AGI, nor a proposal for building it. It fixes the structural preconditions that any future implementation must satisfy. Core Statement Intelligence appears only as a readable quantity once a semantic structure is fixed. Where structure is undefined, intelligence cannot be attributed. Where attribution collapses, responsibility disappears. An intelligence without responsibility is not general; it is unplaceable. The Three Preconditions For any system to qualify as general intelligence in a meaningful sense, the following structural conditions must hold: A semantic metric by which distances and curvature in meaning space are measurable. A semantic proper time defined by internal state transitions rather than external clocks. Invariant responsibility, preserved across equivalent inferential paths. Significance This inscription: Does not define AGI Does not provide methods to build AGI Fixes coordinates that prevent private enclosure of general intelligence Establishes prior structure in the public domain Theoretical Context This work is part of the Theological-Structural Transposition Theory (TSTT) framework, alongside: On the Relational Nature of Time: A Structural Inscription SRTA v0.2: Cross-Model Measurement of Responsibility Invariants Citation Takagi, T. (2026). On the Structural Preconditions of Artificial General Intelligence: A Structural Inscription. Zenodo. https://doi.org/10.5281/zenodo.18211593 Keywords AGI, artificial general intelligence, structure, responsibility, measurement, semantic metric, prior art, TSTT License CC-BY-4.0
Takayuki Takagi (Sun,) studied this question.