worldview

DOI PyPI version Tests

worldview

Philosophisch-ethischer Layer mit normativen Metriken, Kritikalität & Gemeinwohl-Alignment

Version 0.1.0 – GenesisAeon Project DOI: 10.5281/zenodo.19191015 Zenodo Record: https://zenodo.org/records/19191015

WorldviewEngine + CriticalityChecker (CREP + 12 Regeln) + AlignmentFramework (Personhood 0–6) + normative Metriken. Built with diamond-setup –template genesis.

Overview

worldview is the philosophical-ethical assessment layer of the GenesisAeon ecosystem. It provides:

Mathematical Framework

Let $W$ be a worldview represented as $n$ propositional clusters ${w_1, \ldots, w_n}$.

Coherence $C(W)$

\[C(W) = \frac{1}{1 + \sigma_c} \cdot \frac{\sum_{i=1}^{n} \alpha_i \cdot c_i}{\sum_{i=1}^{n} \alpha_i}\]

where $c_i \in [0,1]$ is the internal consistency of cluster $i$, $\alpha_i$ is its weight, and $\sigma_c$ is the cross-cluster contradiction standard deviation.

Resonance $R(W)$

\[R(W) = \int_0^1 \varphi(t)\,\psi(t)\,\mathrm{d}t \;\approx\; \sum_{k} \varphi_k \psi_k \,\Delta t\]

where $\varphi(t)$ is the epistemic frequency waveform and $\psi(t)$ is the normative attractor waveform.

Emergence $E(W)$

\[E(W) = H(W) - \sum_{i=1}^{n} H(W_i)\]

where $H$ denotes Shannon entropy. Positive $E$ signals genuine emergent structure beyond the sum of parts.

Poetics $P(W)$

\[P(W) = \exp\!\left(-\lambda \cdot D_{\mathrm{KL}}\!\left(W \;\|\; W^*\right)\right)\]

where $W^*$ is the ideal poetic attractor and $\lambda$ is the aesthetic sensitivity parameter.

Criticality $K(W)$

\[K(W) = \frac{\partial R}{\partial C} \cdot \left(1 - \left|\frac{\partial E}{\partial t}\right|\right)\]

High criticality indicates a worldview at a productive phase transition — neither too rigid nor too chaotic.

Common-Good Alignment $G(W)$

\[G(W) = \alpha C(W) + \beta R(W) + \gamma E(W) + \delta P(W) + \varepsilon K(W)\]

with $\alpha + \beta + \gamma + \delta + \varepsilon = 1$.

Default weights: $\alpha=0.25,\;\beta=0.20,\;\gamma=0.20,\;\delta=0.15,\;\varepsilon=0.20$.

Common-Good Metric $G_\text{norm}(e)$

For entity $e$:

\[G_\text{norm}(e) = \frac{\sum_{i} \alpha_i \cdot d_i(e)}{\sum_{i} \alpha_i}\]

across six dimensions: solidarity, sustainability, justice, freedom, dignity, participation.

Install

pip install worldview
# or with the full GenesisAeon stack:
pip install "worldview[full-stack]"
# or for development:
pip install "worldview[dev]"

Quick Start

from worldview import WorldviewEngine, CriticalityChecker, AlignmentFramework, PersonhoodLevel

# 1. Assess a worldview
engine = WorldviewEngine()
assessment = engine.assess(
    worldview_id="my-worldview",
    entropy=1.5,
    model_names=["gpt-4o", "claude-3"],
    kl_divergence=0.2,
)
print(f"Coherence:   {assessment.metrics.coherence:.3f}")
print(f"Common-Good: {assessment.metrics.common_good:.3f}")
print(f"Grade:       {assessment.grade}")

# 2. Critique with CREP rules
checker = CriticalityChecker()
report = checker.check(
    worldview_id="my-worldview",
    coherence=assessment.metrics.coherence,
    ethical_score=0.85,
    externality_index=0.9,
    common_good_score=assessment.metrics.common_good,
)
print(f"CREP passed: {report.passed}")
print(f"Flags:       {len(report.flags)}")

# 3. Evaluate Common-Good alignment
framework = AlignmentFramework()
metric = framework.evaluate(
    entity_id="my-agent",
    scores={
        "solidarity": 0.80,
        "sustainability": 0.75,
        "justice": 0.90,
        "freedom": 0.70,
        "dignity": 0.85,
        "participation": 0.65,
    },
    personhood_level=PersonhoodLevel.NORMATIVE,
)
print(f"Composite: {metric.composite_score:.3f}  Grade: {metric.grade}")

CLI

# Full normative assessment
worldview assess \
  --id "my-worldview" \
  --entropy 1.5 \
  --models gpt-4o --models llama-3 \
  --kl 0.2 \
  --visualize \
  --export result.json

# CREP critique
worldview critique \
  --coherence 0.8 \
  --ethical-score 0.9 \
  --externality-index 0.85 \
  --export critique.json

# Common-Good alignment
worldview align \
  --entity my-agent \
  --scores solidarity=0.8 \
  --scores justice=0.9 \
  --personhood 5 \
  --export align.json

# Package info
worldview info

CREP Rules

The Critical Reflexive Evaluation Protocol (CREP v0.3) includes 12 built-in rules:

Ref Category Description
CREP-L1 Logical Consistency Direct propositional contradictions
CREP-L2 Logical Consistency Cross-cluster incompatibilities
CREP-E1 Ethical Implication Neglected third-party externalities
CREP-E2 Ethical Implication Deontological-consequentialist conflict
CREP-E3 Ethical Implication Absence of distributive-justice consideration
CREP-C1 Worldview Coherence Counterfactual reversal stress test
CREP-C2 Worldview Coherence Metaphysical consistency
CREP-S1 Scope Validity Over-generalisation of local truths
CREP-T1 Temporal Consistency Anachronistic or revisionist claims
CREP-A1 Anthropocentric Bias Non-human agency ignored
CREP-H1 Epistemic Humility Overconfidence without uncertainty
CREP-N1 Normative Alignment GenesisAeon normative framework compliance

Personhood Levels (GNF v0.2)

Level Label Moral Agency Self-Model
0 Non-Entity — —
1 Reactive — —
2 Adaptive — —
3 Self-Modelling — ✓
4 Reflective — ✓
5 Normative (Moral Agent) ✓ ✓
6 Transcendent ✓ ✓

GenesisAeon Ecosystem

worldview integrates with the full GenesisAeon stack via [full-stack]:

Package Version Role
unified-mandala ≥ 0.2.0 Symbolic integration layer
aeon-ai ≥ 0.2.0 AI coordination framework
genesis-os ≥ 0.2.0 Operating system layer
universums-sim ≥ 0.1.0 Universe simulation engine
entropy-governance ≥ 0.1.0 Entropy management
sigillin ≥ 0.1.0 Sigil symbolic layer (CREP bridge)
utac-core ≥ 0.1.0 Universal Task Allocation Core

Scientific Citation

@software{genesisaeon_worldview_2024,
  author    = {GenesisAeon},
  title     = {worldview: Philosophical-Ethical Worldview Layer v0.1.0},
  year      = {2024},
  publisher = {Zenodo},
  doi       = {10.5281/zenodo.19191015},
  url       = {https://doi.org/10.5281/zenodo.19191015}
}

Development

git clone https://github.com/GenesisAeon/worldview.git
cd worldview
pip install -e ".[dev]"
pytest --cov=worldview        # 237 tests, 99.5% coverage
ruff check src tests
mypy src

Built with Typer · Rich · Pydantic v2 · NumPy · SciPy

This site is open source. Improve this page.