RFC-038: Domain of Concern Registry — AIGP Dialects as First-Class Artifacts — 2. Motivation
AIGP Specification › RFC-038: Domain of Concern Registry — AIGP Dialects as First-Class Artifacts › 2. Motivation
← 1. Abstract · Section index · 3. Core Definitions →
2. Motivation
2.1 The Assembly Problem
RFC-034 defines what a Domain of Concern is. RFC-035 defines how variables must be declared to be valid measurements. RFC-036 defines how valid measurements combine into concern posture. RFC-037 defines who may observe and render verdicts.
But none of these RFCs produce a distributable artifact. An organization implementing AIGP must:
- Select a Domain of Concern,
- Instantiate a complete Mediation Observation Model with all variables fully declared,
- Bind calculation semantics to the variable set,
- Declare observer requirements and accreditation criteria,
- Set local thresholds, admissibility floors, hard stops, and escalation paths,
- Version the whole thing,
- Distribute it to all governed applications,
- Ensure all subscribers are using the same version.
This is the assembly problem. Without a registry, every organization solves it independently, non-reproducibly, and incompatibly.
2.2 The Reproducibility Requirement
Science solved this problem centuries ago: a research protocol is the declared, complete, distributable description of how an experiment is conducted. Any qualified lab that follows the same protocol should produce comparable results under comparable conditions.
An AIGP dialect serves the same function: any governed system that subscribes to the same dialect version should produce comparable evidence, comparable measurements, and comparable verdicts — because the observation apparatus is declared, not improvised.
2.3 The Interoperability Requirement
Organizations that share a Domain of Concern but use incompatible variable sets, thresholds, or calculation methods cannot compare their governance postures. A financial regulator cannot evaluate two firms’ autonomous-trading governance if one measures 10 variables and the other measures 7 different ones.
A shared dialect version makes cross-organizational comparison meaningful — not by imposing uniform thresholds, but by ensuring the measurement apparatus is structurally compatible.