Research foundations
Semantic glyph interpretation and Teleodynamic AI converge on one question: can machine systems be structured to do more than predict the next state, and instead maintain directed organization under constraint?
The Deacon-style distinction is useful because it separates ordinary near-equilibrium dynamics, far-from-equilibrium self-organization, and living-system dynamics that preserve internal organization instead of simply dissipating the conditions that created them.
For this site, that is a design analogy and research lens. A machine-learning system should not claim biological purposefulness. It can, however, make preservation, resource limits, boundary conditions, and viability explicit design variables.
Rudolph's teleodynamic architecture program treats transitions, operator-level attractors, and normative curvature as explicit objects of modeling. The public site should present this as a speculative architecture direction, not a deployed runtime.
Ter Horst and Zambrano's 2026 Teleodynamic Learning paper frames learning as the co-evolution of representation, adaptation, and resource-sustainable structural change, instantiated in the Distinction Engine.
Probability-driven continuation without explicit internal orientation. This is the safest bucket for most current language-model behavior.
Externally constrained behavior through rules, reward models, or policy layers, without internal teleodynamic structure.
A research target: internal operator-level attractors, multi-scale regulation, and explicit value or norm curvature.
Teleodynamic.com should use "oriented" as a research and design term. It should not claim current AI systems possess intrinsic agency, consciousness, certification-ready safety, or settled biological equivalence.
Glyph systems ask how compressed symbolic forms can carry complex meanings between people and AI systems. The useful split is between visible form, composition grammar, ontology constraints, and evidence traces.
EASY-AI demonstrates one public research direction: semantically composable glyphs for communicating AI-system structure. Glyph Code Prompting is a looser practice: user-defined symbols act as conceptual anchors that steer attention and context, but remain approximate prompts rather than literal model-state visualizations.
What public mark or rendered form was seen?
What role does it play: entity, operator, modifier, relation, state?
Which meanings are allowed by type, context, and source evidence?
What approximate human explanation can be defended?
Both research streams reject a purely state-centered account of meaning. Teleodynamic work asks how transformation itself becomes structured; glyph work asks how compact marks guide interpretation across systems.