Language models as witnesses.
Instruction tuning, prompt protocols, classifiers, and evaluation tasks for surfacing consent-relevant uncertainty without pretending to adjudicate consent itself.
Model layer
Training witness systems for consent dynamics.
Consentful Cybernetics treats models as practical infrastructure: LLM experiments, classifiers, embedding spaces, vector stores, evaluation sets, and reusable data artifacts that help human and machine systems notice consent-relevant dynamics before they harden into breach.
This page is an early stub. The model program will expand as research code, synthetic streams, annotation schemas, training corpora, and reference models become ready to share.
What this means
Not a metaphorical model. Actual AI and data infrastructure.
“Models” here includes conceptual models, but it does not stop there. The goal is to build usable machine-learning infrastructure for recognizing patterns of consent, refusal, pressure, ambiguity, scope drift, stale permission, power imbalance, repair opportunity, and consent collapse.
Searchable consent dynamics.
Embeddings and vector stores for interaction patterns, research passages, synthetic event streams, annotation examples, and protocol-relevant cases.
Reusable examples.
Curated and synthetic datasets for consent-supportive phrasing, refusal safety, repair prompts, scope clarification, and boundary-sensitive dialogue.
Open where safe.
Reusable model cards, test harnesses, and small reference systems that others can study, adapt, or train against where privacy and safety allow.
Design constraint
Train witnesses, not cops.
A Consent Dynamics Model should not declare “consent happened.” That is too brittle, too context-sensitive, and too easy to weaponize.
The safer target is a witness layer: a system that notices pressure, coercion, hesitation, ambiguity, unsafe refusal conditions, stale scope, escalation, and repair opportunities, then prompts humans and agents to slow down, clarify, or preserve exit.
The model layer is therefore assistive, not sovereign. It should increase legibility, reversibility, refusal-safety, and repair capacity without converting uncertainty into permission.
Initial roadmap
Early artifacts to expect.
01 · Synthetic streams
Consent-gated event data.
Synthetic event streams for testing whether explicit consent-gated transitions reduce predictive uncertainty and improve persistence under perturbation.
02 · Annotation schema
Consent dynamics labels.
A structured vocabulary for labeling consent-relevant signals: power, scope, timing, medium, hesitation, refusal-safety, revocation, drift, witness, and repair.
03 · Vector library
Searchable research and cases.
A vectorized knowledge layer for consent research, protocol primitives, synthetic examples, and interaction patterns that can support retrieval, analysis, and downstream training.
04 · Evaluation harness
Measure witness behavior.
Test prompts and scenarios to evaluate whether a model can distinguish clarification from pressure, support refusal without escalation, and identify when scope has drifted.
The model is a witness layer, not a permission machine.
The ambition is a reusable data and model commons for consent-aware AI: directly useful tools where possible, training material where appropriate, and careful boundaries wherever real human interaction data is involved.