The Zx3 Framework

A plain-language guide to multi-valence decision evaluation.

What Zx3 Is

Zx3 is a framework for evaluating decisions, policies, and systems across multiple dimensions simultaneously. Instead of reducing complex choices to a single score, Zx3 preserves the structure of tradeoffs so that genuine conflicts become visible rather than averaged away.

Most evaluation methods assume that criteria are independent — that you can score safety, fairness, and efficiency separately, then add them up. But in practice, improving one often degrades another. Zx3 models these interactions explicitly.

The framework emerged from a simple observation: every consequential decision affects at least three parties. Understanding those effects requires looking at all of them at once, not one at a time.

The Core Structure

Self × Other × System

Every decision is evaluated across three stakeholder perspectives:

  • Self — the decision-maker or direct actor
  • Other — those affected by the decision who didn't make it
  • System — the larger structure (institution, ecosystem, society) that contains both

These perspectives are then crossed with three domains — Governance, Economy, and Culture — creating a structured evaluation space that reveals how a decision ripples across contexts.

The Three Valences

Why [-1], [0], and [+1] Matter

Instead of continuous scores, Zx3 uses three discrete valences for each evaluation cell:

[−1] Harm

Negative valence must stay visible. A decision that scores well on average but causes genuine harm in one dimension should not have that harm averaged away. The [-1] valence is a signal that demands attention, not a number to be offset by positives elsewhere.

[0] Uncertainty

Zero is not "neutral" — it means "we don't know yet" or "the effects are genuinely mixed." Uncertainty is a feature, not a failure. It flags areas that need more information or deliberation before a decision proceeds.

[+1] Concrete Benefit

Positive valence must be concrete and specific. "+1" means an identifiable, verifiable benefit — not aspirational language or projected value. If you can't point to the specific benefit, it's a [0], not a [+1].

How It Connects

The Zx3 framework is one part of a larger research program that spans AI safety, consciousness theory, ethical governance, and measurement formalization. Each piece reinforces the others:

  • The Core Specification defines the invariant structure
  • Assessment tools make the framework practical
  • Research publications provide formal grounding
  • Accessible writing makes the ideas available to everyone

Start Here

Where to Go Next

Depending on what brought you here:

I want to understand the ideas
Start with the free ebook on the home page, then explore the writing as it becomes available.

I want to use something
Tools and templates are coming soon. Check back or subscribe for updates.

I want to read the research
Visit the research page for formal publications and specifications.