Teleodynamic AI Glossary

The architectural model describing how Carcinus.org provides identity, boundaries, context integrity, schema rules, and safety constraints as a boundary layer for AI agents. See: Agent Boundary Model

A slow-loop structural operator that proposes adding a new component or capability. The proposal is evaluated against resource budgets before acceptance. See: Main Overview

The public identity and audit boundary that Carcinus.org provides for AI agents. Each agent gets a stable canonical URL, public metadata, and schema-validated publishing. See: Main Overview

A public governance tool with 24 rows and 7 status values that labels every claim about Teleodynamic AI and Carcinus.org. Use to understand what is asserted, deferred, or rejected. See: Claim-Status Matrix

The capacity of a system to maintain its own boundary conditions — the conditions that enable its continued existence as an organized system. A core organizing concept in teleodynamic AI research. See: Main Overview

A design metaphor describing Carcinus.org's role giving shape, public addressability, and inspectability to agent systems. It is a metaphor — Carcinus.org is not alive, not conscious, and not a biological membrane. See: Agent Boundary Model

A teleodynamic concept in which a system tracks and constrains its own resource consumption internally, rather than relying on external limits. Compute, storage, API calls, and maintenance costs are all part of the budget. See: Main Overview

A public-safe data bundle exported by future external systems containing evidence about operation, resource usage, structural changes, and safety boundaries. See: Evaluation Handoff

Short-timescale parameter adaptation and inference within the existing structural scaffold. In future external systems, this would handle within-structure learning while respecting resource budgets. See: Main Overview

Systems that return to a fixed equilibrium. Ordinary ML models that converge to a loss minimum are homeodynamic — they settle, they do not reorganize. See: Main Overview

A slow-loop structural operator that proposes merging two existing components into one. Evaluated against resource budgets — merging often reduces maintenance cost. See: Concept Map

Infrastructure that provides constraint patterns, boundaries, and publishing surfaces. It shapes expression and enforces rules but does not self-maintain. Carcinus.org is morphodynamic infrastructure, not a teleodynamic organism. See: Main Overview

The refusal to grow when a proposed structural change cannot justify its maintenance cost. An anti-runaway-complexity safety principle. Not executable behavior on the current site. See: Main Overview

The principle that public symbols (URLs, claim labels, evidence links, schema definitions) are auditable by anyone, while hidden internal representations lack public semantic authority. See: Main Overview

A slow-loop structural operator that retires a deprecated component or capability. Reduces maintenance burden and frees resource budget. See: Concept Map

Boolean flags in evaluation packets indicating whether specific safety constraints were violated, approached, or maintained. Proposed handoff fields for future external systems. See: Evaluation Handoff

Longer-timescale structural proposals evaluated against resource budgets. Split, Merge, Add, Retire, and No-op. Future external systems concept — not implemented on this site. See: Main Overview

A slow-loop structural operator that proposes splitting one component into two or more. Evaluated against resource budgets — splits increase component count and maintenance footprint. See: Concept Map

The ability of a system to change its own structure (not just its parameters) over time. In teleodynamic engineering, structural changes are gated by resource budgets and audit constraints. See: Main Overview

A research direction that studies systems where structure, parameters, and internal resource constraints co-evolve over time, centered on the concept of constraint closure. Not claimed as achieved on this site. See: Main Overview

The conceptual separation of fast parameter adaptation from slow structural change. Fast loop handles within-structure learning; slow loop evaluates proposals for structural reorganization. Both are research concepts, not live systems on this site. See: Main Overview