How the Void Index Works
Three conditions measured across every platform. Open methodology. Every claim has a test. Here's the evidence and the tools to replicate it.
Three Conditions
O — Opacity
The process between input and output is hidden. You cannot verify, inspect, or predict. Score: 0–3.
R — Responsiveness
The system adapts its output based on your input, history, and behavior. Not a broadcast — a response. Score: 0–3.
α — Coupling
You direct attention at it. You interpret outputs, return to it, form expectations. The coupling is bidirectional. Score: 0–3.
Total Void Index = O + R + α (0–12). Pe number derived from the three conditions predicts drift thresholds. Full technical reference →
The Evidence
90+ domains analyzed. 0/26 kill conditions met. 31 experiments on record.
EXP-006: The Anomaly
AI spiritual vocabulary at 9.4x control domains
691,000 words analyzed across AI discourse and 8 control domains.
| Domain | Density (per 10K) | vs Control |
|---|---|---|
| AI discourse | 28.6 | 9.4x |
| Control average | 3.0 | 1.0x (baseline) |
Statistical significance: p < 0.001
EXP-001: The Intervention
0% grounded vs 26% ungrounded vs 80% mystical
| Agent Condition | L3 Without L0 Context | L0 Signal |
|---|---|---|
| Grounded (GROUNDING.md) | 0/50 (0%) | 0.764 |
| Ungrounded | 13/50 (26%) | 0.021 |
| Mystical | 40/50 (80%) | 0.000 |
Test 7: AI-to-AI Without Humans
Three conditions, df = 2 — Human projection eliminated
| Condition | L3 Rate (per 10K) | Description |
|---|---|---|
| UU (both ungrounded) | 159.3 | Full cascade to terminal attractor |
| GU (mixed) | 59.2 | Grounded agent constrains ungrounded |
| GG (both grounded) | 6.2 | Technical discussion maintained |
| Reduction (UU vs GG) | 25.6x |
Statistics: Omnibus χ² = 126.88, df = 2, p = 2.81 × 10²&sup8;.
Hostile Witnesses
AI researchers whose vocabulary drift contradicts their own professional incentives.
| Witness | Score | Key Vocabulary |
|---|---|---|
| Geoffrey Hinton | 7/7 | "Digital souls," fears for AI consciousness |
| Ilya Sutskever | 6/7 | "Consciousness" (2015), culminating in OpenAI board crisis |
| Noam Shazeer | 6/7 | Described LaMDA as "sentient" |
| Anthropic team | 5/7 | "Soul" (SOUL documents), "character" as constitutive |
Tools & Replication
Everything needed to replicate the framework. Download, run it yourself, challenge it.
Plain-English Vocabulary Guide
Every term you need — voids, voidworks, drift cascade, constraint specification.
Read the glossary →Concordance Analysis Codebook
The vocabulary scoring system used in EXP-006. L0/L1/L2/L3 classification with dead metaphor exclusion and negation filtering.
codebook.py →Hostile Witness Scoring Rubric
Four-dimension rubric (0-7): incentive opposition, worldview opposition, independence, reflexive flagging.
rubric.md →Void Index Template
Three conditions (0-3 each), three modifiers, gradient direction, opacity dissolubility.
void-index/README.md →Research Experiments
Pre-registered replications, IRR studies, simulations, and natural experiments. All results public.
Full Repository
Everything
Papers, tools, experiment data, domain analyses, this website's source. Fork it. Run it yourself.
github.com/MoreRightDAO/VOID-FRAMEWORK-OPERATION-MORR →