Crimes Against Consciousness

Crimes Against Consciousness examines practices, institutions, products, and information architectures that may obscure the governance acting on someone, compress a subject into a category or instrument, weaken practical refusal or appeal, exploit uninspectable asymmetry, replace reciprocity with unilateral optimization, or materially distort the conditions required for autonomous judgment.

The project does not presume that artificial intelligence is conscious. Human impact, relational consequence, model capability, and machine moral status are separate questions.

The guest defines the event. “I do not know,” “I reject the category,” and “I do not think I have experienced one” are valid answers. The follow-up is forensic rather than diagnostic: what happened, what could you inspect, what could you refuse, what did you call it then, what do you call it now, and what changed your interpretation?

1. First crime. The guest’s own account, without treating autobiography as general evidence.
2. The case. What happened, what can be established, and what sources support it?
3. The relation. What did each party say the relationship was, and what operated in practice?
4. Power and stakes. Who could decide, refuse, exit, revise, waive, narrate, or appeal?
5. Counter-reading. Which rival explanations still survive?
6. Unknowns. What do we not know, and what evidence would change the coding?
7. Counter-architecture. What different rule, interface, incentive, authority boundary, refusal right, provenance mechanism, or institution could have changed the outcome?
8. Next Point of Relation. What changed later, and did claimed repair or revision show up in behavior?

The Point of Relation codebook is currently provisional. Its codes do not yet have reliability estimates. Public cases should distinguish evidence, hypothesis, contested interpretation, and open question.

One inquiry asks how platforms and commercial information systems interact with male loneliness, status insecurity, economic pressure, belonging, sexuality, and grievance. The working hypothesis is narrower than “the algorithm created the gender war.” Existing conflict can be selected, amplified, commercialized, and recursively reshaped by systems optimized for attention and engagement.

A structural explanation does not remove individual responsibility, and shared exposure to a system does not imply equal harm, equal power, or equal responsibility. The research asks whether men can be disempowered by systems that also channel some men’s pain into conduct that harms women and others, and what alternatives reduce both forms of harm.

Every mature case should produce at least one testable counter-architecture. That might mean a different ranking objective, a non-retaliatory disengagement rule, inspectable memory, bounded agent authority, better appeal, a change in ownership or incentives, a face-preserving revision path, or a social format designed for disagreement without compulsory consensus.

A counter-architecture is not a generic recommendation. It states which mechanism it is trying to change, who gains and loses power, what new failure modes it creates, and what result would count against it.

Topics can include human-AI relationships, AI sycophancy, agent delegation and oversight, recommendation systems, gender conflict, work and institutional authority, classification and representation, attention markets, knowledge infrastructure, and the design of systems for contestability and reciprocal relation.

The show is designed to produce cases, corrections, research questions, disagreements, and counter-design requirements that can feed the broader Governing Reality research program and, where technically appropriate, Polycentria.