RFC-034: Domains of Concern for Mediated Intelligence — 3. Problem Statement
AIGP Specification › RFC-034: Domains of Concern for Mediated Intelligence › 3. Problem Statement
← 2. Purpose · Section index · 4. Core Definitions →
3. Problem Statement
Current AI governance discussions often treat “domain” as an industry category.
Examples include:
- healthcare AI,
- military AI,
- financial AI,
- legal AI,
- education AI,
- cybersecurity AI.
This classification is useful for policy and institutional accountability, but it is insufficient for empirical concern calculation.
A single industry may contain many different mediated intelligence constraints. Healthcare may involve triage, diagnosis-like reasoning, patient communication, clinical documentation, billing, scheduling, and emotional support. These are not the same concern. They do not require the same evidence, the same observer, the same thresholds, or the same calculation.
Likewise, the same mediated intelligence constraint may appear across multiple industries. Autonomous systems appear in military drones, surgical robots, financial trading systems, cloud remediation engines, industrial robots, and self-driving vehicles.
Therefore, the correct unit of analysis is not the industry. The correct unit of analysis is the Domain of Concern.