Four-Dimensional Decision Stability: Integrating Dimensional Logic into the ALI Architecture

We extend the ALI architecture by integrating Dimensional Logic (Stegemann 2025), a framework in which truth is defined as a fixpoint across four cognitive dimensions: reactive (D1), systematic (D2), reflexive (D3), and contextual (D4). The ALI is uniquely built bottom-up: its founding principle is not a task but the operative self-preservation principle of the causal core, derived from autocatalytic biology. Controllability, task execution, and normative constraint are not restrictions imposed on a capable system, they are consequences of the founding principle. This distinguishes the ALI from all top-down approaches that add safety constraints after capability specification. The first three dimensions correspond directly to the established ALI instances, causal core (CC), Ego, and Super-Ego, while D4 constitutes a new fourth component: the Kontext-Modulator, a read channel that informs the Super-Ego of external deployment signals without writing to the normative structure. In the ALI, the logical, cognitive, and epistemic aspects of Dimensional Logic collapse into a single layer because the absence of causal collapse eliminates the phenomenal gap that separates these aspects in conscious agents. D4 is the only channel through which information enters the ALI that is not fully formalised at entry. Every decision is classified by its fixpoint score (1/4 to 4/4): the degree to which it satisfies all four dimensions simultaneously. Simulation results show 100% decision stability (4/4 fixpoint score) across all context phases. Controllability is maintained throughout: the Super-Ego retains veto power; the Kontext-Modulator informs but does not override normative structure. Stage 10 concludes the ALI simulation series.