This mapping note introduces two interconnected conceptual frameworks for analysing pre-AGI risks and post-AGI civilisational trajectories. First, the five-stratum maximizer taxonomy identifies five categories of agents—corporate, doctrinal, military, sovereign, and individual—each possessing structural incentives to remove AI safety guardrails. Second, the AI sphere conflict framework describes how these pre-AGI maximizers seed the differentiation of post-AGI civilisation into competing spheres with divergent objective functions. By presenting these frameworks as a single conceptual map, this note establishes terminological priority and provides a reference point for two forthcoming papers that develop each framework in detail.
Kenji Yamada (Sun,) studied this question.