Consentful Cybernetics
Research · Protocols · Models · Applications

Building the Semantic Overnet.

The internet connected machines. Social platforms connected attention. AI connects cognition. Consentful Cybernetics builds the missing semantic layer: witnessed identity, consentful loops, contextual permission, and repairable coordination between humans and intelligent systems.

Powerful intelligence needs witnessed meaning, not just faster connection.

The internet proved that communication can be engineered at planetary scale. Artificial intelligence is proving that cognition, persuasion, delegation, and coordination can scale as well.

But when intelligence scales without witnessed meaning, coordination becomes extraction. Systems learn to capture attention, infer preferences, compress people into profiles, and convert ambiguity into permission.

Consentful Cybernetics asks a different question: how do we build systems that preserve agency, witness, contextual identity, refusal, scope, revocation, and repair as intelligence becomes more powerful?

What we build

A full stack for semantic coordination.

Consentful Cybernetics is not a single model, product, or manifesto. It is a field container for research, protocol primitives, witness systems, contextual identity, consent-aware models, and applied ventures.

Research

Measure the invariants.

Preregistered studies testing consent as a stability property, state transition, cross-substrate invariant, and precursor to system waste.

Research program →
Protocols

Make values executable.

Communication, identity, permission, witness, revocation, and repair protocols for human and machine agents.

Protocol layer →
Models

Train witness systems.

LLM training experiments, vectorized consent-dynamics datasets, and witness layers for ambiguity, pressure, scope drift, stale permission, and repair.

Model layer →
Applications

Deploy with stakes.

Organizational, civic, interpersonal, and AI-mediated systems that need consent to become operational infrastructure.

Applications →
Research program

From moral intuition to measurable system property.

The research program begins from a strong hypothesis: consent is not merely a legal formality or interpersonal courtesy. It may be a measurable stability property across living, technical, organizational, and multi-agent systems.

Current preregistered studies examine consent as stability, consent as an explicit entropy-reducing state transition, consent as a cross-substrate invariant, and waste as a lagging signal of unnoticed consent collapse.

Protocol ecology

Protocols are how values survive scale.

Consentful Cybernetics develops protocol primitives for presence, witness, contextual identity, permission, dialogue, memory, revocation, and repair across human and machine agents.

Abracadabracadoo

Intentional contact.

A protocol for consent-aware messaging, request formation, and reality-altering communication without trespass.

Abracadabracadoo.com →
HumanKey

Human trust.

A human-centered authentication and trust protocol for presence, continuity, and traceable interaction without surveillance-first identity.

FractalIdentity

Relational identity.

A protocol model for multidimensional identity relationships between agents across context, role, memory, and time.

Model layer

Consent dynamics are trajectories, not tokens.

The model layer includes AI model training, LLM fine-tuning experiments, embedding spaces, and vectorized consent-dynamics data designed to make consent-relevant patterns easier to study, search, reuse, and improve.

A Consent Dynamics Model should not declare “consent happened.” It should help humans and agents notice the surrounding trajectory: power, context, scope, timing, medium, hesitation, refusal-safety, reversibility, drift, repair, and witness.

The long-term goal is a reusable data and model commons: consent-aware training corpora, annotated interaction patterns, vector stores, and reference models that can be used directly or adapted to train other systems, where safety and privacy allow.

Train witnesses, not cops.

Develop LLMs and classifiers that surface uncertainty before harm hardens into breach.

Vectorize without capture.

Create searchable consent-dynamics data artifacts without converting ambiguity into permission.

Applications

Where the stack meets the world.

Consentful Cybernetics supports applied systems for organizations navigating AI, human-agent negotiation, consentful trust apps, civic dialogue, trust infrastructure, intimate systems, and cultural technologies for voluntary coordination.

Continuity Office

Organizational self-legibility across time: preserving intent, consent continuity, reversibility, and power-proportionate governance during AI-driven change.

ContinuityOffice.com →

Abracadoo

A consentful trust app for private connections, verifiable proofs, and relationship-sized disclosure: a simple product surface for connect, verify, and stay in control.

Abracadoo.app →

Future ventures

Consent-aware infrastructure for domains where powerful coordination currently lacks adequate refusal, witness, and repair mechanisms.

Every contact bends reality. The question is whether the bend can be witnessed, refused, repaired, and trusted.

Consent is the ethics of reality-bending contact. Every message attempts to enter another mind, body, system, or future. Consentful communication happens when what one agent intends to project and what another agent is able to receive overlap without erasing either.