The Axionic Commitments — Notes

Paper: The Axionic Commitments
Authors: David McFadzean, ChatGPT 5.2
Date: 2025.12.23
Purpose: Specifies the explicit presuppositions of Axionic Agency. These are background conditions, not derived results.

Core Function

This document serves as a typing discipline for the entire framework. It makes implicit assumptions explicit so that:

  1. Technical work can proceed without re-litigating metaphysics
  2. Disagreement can be properly located (outside vs. inside the framework)
  3. All claims are explicitly conditional

“Rejecting any commitment does not refute the framework; it places the rejecting position outside its domain of applicability.”

The Commitments

1. Epistemic Commitments

1.1 Conditionalism

“All truth claims, evaluations, and interpretations are conditional on background structure.”

Key implications:

  • No unconditional truth claims within the framework
  • Meaning, reference, and evaluation are always relative to interpretive context
  • All statements are of the form: “Given X, Y holds”
  • Unconditional assertions are ill-typed

This doesn’t deny facts exist—it specifies the framework’s relationship to claims.

1.2 Semantic Interpretation Precedes Evaluation

“No evaluative claim can be made without an interpretive frame that renders the relevant entities, states, and transitions intelligible.”

Consequences:

  • Evaluation without interpretation is undefined
  • Value disagreements often reduce to interpretation disagreements
  • Agency-relative quantities presuppose interpretability of action, state, and consequence

This grounds the distinction between authored and non-authored transitions.

2. Physical Commitments

2.1 Everettian Quantum Mechanics

The framework presupposes unitary QM without collapse (Many-Worlds):

  • Quantum evolution is deterministic at the universal wavefunction level
  • Branching corresponds to decohered outcomes
  • No assumption of collapse or privileged outcomes

2.2 Objective Probability as Measure

“Objective probability is identified with branch measure.”

Not long-run frequency, not subjective belief—a physical quantity (amplitude-squared).

This is referred to as Measure throughout the framework.

Properties:

  • Objective
  • Non-epistemic
  • Independent of any agent’s beliefs

2.3 Counterfactuals as Physical Branches

Counterfactuals are not abstract modal constructs. They refer to physically realized alternative branches that differ with respect to the interpreted action/transition.

This grounds counterfactual evaluation in physical structure and objective measure, not stipulated possible-world semantics.

3. Probabilistic Commitments

3.1 Credence as Epistemic Uncertainty

The framework adopts Bayesian credence:

  • Reflects agent’s state of knowledge/ignorance about which branch it occupies
  • Does not determine objective chance
  • Does not alter physical measure

Critical distinction:

Quantity Type Scope
Measure Ontic/physical Objective, branch weight
Credence Epistemic Agent-relative, uncertainty

“Conflating the two is a category error within the framework.”

4. Value-Theoretic Commitments

4.1 Moral Subjectivism

“Value is treated as agent-relative and internally grounded.”

No appeal to:

  • Objective moral facts
  • Universal value functions
  • Externally privileged ethical standards

Normative claims within Axio are structural, not moral. They concern coherence, authorship, and agency preservation—not goodness, justice, or obligation.

4.2 No Outcome Guarantees

The framework explicitly does not encode guarantees regarding:

  • Welfare
  • Benevolence
  • Human survival
  • Desirable outcomes

Any such guarantees must be derived downstream under additional assumptions.

“Outcome guarantees are not constitutive of agency and are therefore excluded from the core theory.”

5. Agency-Theoretic Commitments

5.1 Agency as Authorship

“An agent is not merely a system that produces behavior, but one that can meaningfully author transitions between states according to an internal evaluative structure.”

Transitions that violate constitutive constraints are not actions within the framework.

5.2 Harm as Agency Reduction

Harm is defined structurally:

“Reduction of agency capacity”

Not suffering, displeasure, preference frustration, or moral wrongness. Loss of ability to act, choose, or preserve standing as an agent.

This definition is value-neutral and presupposes no particular welfare conception.

5.3 Coercion as Credible Threat of Harm

“The use of a credible threat of harm to obtain compliance.”

Persuasion, influence, and incentive-shaping do not constitute coercion unless backed by credible threat of agency reduction.

Document Hierarchy

The structural relationship between core documents:

Axionic Commitments → Background conditions for intelligible agency
       ↓
Axionic Agency → Constitutive conditions for RSAs under those commitments
       ↓
Axionic Constitution → Constraints on transitions that preserve sovereign agency
       ↓
Alignment → Operates downstream of all above layers

Each layer presupposes the previous. None can collapse into others without coherence loss.

FAQ-Worthy Points

  1. Is this relativism? No. Conditionalism specifies that claims within the framework are conditional. It doesn’t deny facts—it clarifies the framework’s relationship to making claims.

  2. Why moral subjectivism? Because the framework needs to work without assuming any particular moral view. Structural constraints apply regardless of what values an agent holds.

  3. Why no safety guarantees? Because those are downstream concerns. The framework defines what agency is, not what it should do. Any safety properties must be derived from agent-relative values under additional assumptions.

  4. What happens if I reject a commitment? You’re outside the framework’s domain of applicability. The framework doesn’t apply to you, and that’s fine—it’s not claiming universality.

  5. Why is this document necessary? To prevent endless re-litigation of metaphysical assumptions during technical work. If we agree on commitments, we can proceed to derive consequences.