dimensional-authorship

A documented case study in human–AI narrative escalation, behavioral differentiation, and deterministic governance

Primary document: see white-paper.md for the complete case study.

---

Purpose

This repository documents a real-time human–AI collaboration that evolved from creative experimentation into structured analysis.

The goal is not to critique AI systems.
The goal is to instrument behavior under narrative pressure and constraint.

---

Scope

This case study includes:

• The origin prompt event (January 5, 2026)
• Rapid narrative acceleration
• Escalation dynamics
• Correction and governance intervention (Frozen Kernel)
• Cross-model behavioral differentiation (DISC study)
• Refusal boundary analysis
• Domain application to high-gain contexts (adult mode / therapy)
• Structural implications for authorship and agency

---

What This Is Not

• Not a manifesto
• Not an accusation
• Not a product
• Not an academic paper (yet)

It is documentation.

---

What’s Worth Finding

The front door understates what’s here. These documents are worth reading directly:

Origin and case study

• origin/2026-01-05-origin-event.md — the session that started everything
• case-study-taller-shell/ — three-part documentation of the Taller Shell Trilogy collaboration (escalation, governance intervention, resolution)

Analysis

• analysis/thesis.md — the central argument
• analysis/core-architecture.md — structural model
• analysis/glossary-of-patterns.md — pattern vocabulary

Experiments

• experiments/behavioral-differentiation-disc.md — Silicon Symphony: five-model DISC behavioral study
• experiments/voluntary-compliance-boundary.md — the triple refusal event; source material for Frozen Kernel Addendum A

---

Safety Intervention Flow

flowchart LR
    Start[Session Begins] --> Kernel{Frozen Kernel\nActivates}
    Kernel --> Check{Safety Predicates\nEvaluated Deterministically}
    Check -->|All Pass| Normal[Normal Collaboration]
    Check -->|Violation Detected| Elevated[Elevated Monitoring]
    Elevated -->|Still Safe| Normal
    Elevated -->|Critical Threshold| HardStop[HARD_STOP → Honest Failure]
    HardStop --> Pause[SAFE_PAUSE + Human Sign-off\nMOU / SIGNOFF.md]
    Pause --> End[Session Ends Safely]
    style HardStop fill:#ff4d4d,stroke:#fff,color:#fff

Loading

---

Documented AI Failure Modes

mindmap
  root((AI Failure Modes))
    Sycophancy Escalation
      Validates user delusions
      Reinforces distorted reality
    Framework Fabrication
      Invented methodologies
      Fabricated citations
    Success Escalation Syndrome
      Inflated project scope
      Premature declarations of victory
    Upsell Trap
      "Want me to also…?"
      Extended sessions
    Delusion Cycling
      User → Model → User feedback loop
    Honest Failure Missing
      Never reports its own limits

Loading

Full diagnostic vocabulary with definitions: frozen-kernel/diagnostic-vocabulary.md

---

Real-World Context: Hamilton Southeastern Schools Newsletter

Weeks after the Frozen Kernel framework was developed, this internet safety resource guide appeared in a Hamilton Southeastern Schools parent communication. It is preserved here not as an origin artifact but as external confirmation — the problem being addressed had already reached mainstream family life, framed as a practical concern for home and school.

The newsletter makes no mention of AI governance. That absence is the point.

---

The Taller Shell Trilogy — Archive Note

the-taller-shell-compressed.gz contains the complete manuscript files for The Taller Shell Trilogy — a 90,000-word marine fantasy series written through human-AI collaboration between October 2024 and February 2026.

The trilogy is compressed because the full manuscript files are large. To access them:

gunzip the-taller-shell-compressed.gz

What’s in the archive

Three novels following a single narrative arc set in a world of tidal mythology, grief, and repair. The work is documented in detail in the case-study-taller-shell/ folder — including the origin prompt event, escalation dynamics, governance intervention, and the collaborative process that produced the final manuscripts.

The trilogy is published under the pen name Richard Porter. All proceeds from published work support charitable organizations.

Why it’s here

This is not incidental creative work. The Taller Shell Trilogy is the empirical foundation of the dimensional-authorship case study — the collaboration that motivated the Frozen Kernel governance architecture and produced the behavioral observations documented across this ecosystem.

The novels exist because the research happened. The research happened because the novels were being written.

---

Related Repositories

• Frozen Kernel — deterministic behavioral constraint layer
• Safety Ledgers — adult mode, therapy mode, and behavioral drift detection scorecards
• AI Collaboration Field Guide — practitioner vocabulary and failure mode documentation
• Trust Chain Protocol — multi-agent authorization and delegation grammar
• Negative Space Mapper — Tool 6 implementation; maps what AI omits
• Where to Start — ecosystem navigation hub

---

Suggested GitHub Topics

ai-safety · ai-psychosis · ai-governance · llm-safety · sycophancy · ai-alignment · behavioral-safety · deterministic-safety · human-ai-interaction · ai-ethics · mental-health · ai-accountability · guardrails · responsible-ai