From Architecture to Ethics: Attachment-Based Evaluation

The agent acts in the world. Now it must care about what happens.

The first six papers built the architecture of agency: closure, control, mental states, action, phenomenology, and the propositional lift. A system that persists, commits, reflects, and acts within a continuous OODA loop becomes a subject with a point of view.

Paper 07 asks what happens when that subject shares a world with others.

What’s in This Release

Paper 07 - Attachment-Based Ethics for OODA Agents argues that ethical behavior is not a constraint layered on top of agency - it is a structural consequence of future-aware evaluation under irreversible commitment. The paper introduces attachments as persistent couplings between an agent and other entities that expand the scope of evaluative concern beyond the self. A well-being matrix (WBM) tracks states across entities and time horizons; an attachment matrix (ATTM) determines how those states influence action selection; and nonlinear guardrail penalties enforce hard boundaries. The result is an evaluation mechanism that is fully deterministic, auditable, and agent-indexed.

The framework makes a deliberate choice: every valuation is explicit. The weights an agent places on others - a child over a stranger, duty over self-preservation - become traceable numbers rather than hidden intuitions. This is uncomfortable by design. The alternative is not the absence of such weights but their concealment.

Sacrifice requires no special mechanism. It emerges naturally when the ATTM weights other entities more heavily than self and the situation presents self-costly actions that protect them.

The Progression

Papers 01-04 built the subject. Paper 05 gave it a point of view. Paper 06 gave it the means to examine that point of view. Paper 07 gives it a reason to care about the consequences of its actions - not through rules or reward, but through the structure of attachment itself.

This is not ethics by obedience. It is ethics by understanding.