Measure the claim.
Preregistered studies testing consent as a stability property, state transition, cross-substrate invariant, and precursor to system waste.
Research program →
Research · Protocols · Models · Applications
Building the consent layer for voluntary intelligence.
Consentful Cybernetics develops the research, protocols, models, and applied systems required for powerful intelligence to coordinate without coercion, opacity, or capture.
Intelligence scales safely only when consent scales with it.
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 consent, 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, refusal, scope, revocation, context, and repair as intelligence becomes more powerful?
What we build
A full stack for consentful coordination.
Consentful Cybernetics is not a single model, product, or manifesto. It is a field container for empirical research, protocol primitives, consent-aware models, and applied ventures.
Make values executable.
Communication, identity, permission, witness, revocation, and repair protocols for human and machine agents.
Protocol layer →Train usable intelligence.
LLM training experiments, vectorized consent-dynamics datasets, and witness layers for ambiguity, pressure, scope drift, stale permission, and repair.
Model layer →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 identity, presence, permission, dialogue, memory, revocation, and repair across human and machine agents.
Intentional contact.
A protocol for consent-aware messaging, request formation, and reality-altering communication without trespass.
Abracadabracadoo.com →Human trust.
A human-centered authentication and trust protocol for presence, continuity, and traceable interaction without surveillance-first identity.
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, 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 →DeepTrust Dialogica
Protocol-native communication for auditable, structured dialogue between humans and autonomous agents.
Future ventures
Consent-aware infrastructure for domains where powerful coordination currently lacks adequate refusal, witness, and repair mechanisms.
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.