Home

ACCM Deep Ethics Project Wiki

Orientation and recovery layer — early BETA, corrigible, and source-linked

The ACCM Deep Ethics Project (Allow Constructive Controversy Mode — Deep Ethics Project) investigates what happens between an object and the representation produced about it—especially when human beings, institutions, media systems, or A.I.s classify, summarize, correct, caution, approve, reject, or act.

This Wiki is a map into the project. It does not replace the public project site, the working repository, or the separate canonical-source repository.

Scope boundary

This Wiki is the orientation and recovery layer for a wider, non-linear project. No single question, page, count, or model-generated summary defines the whole architecture.

The project develops operational ways to inspect correspondence, transformation, compression, correction and consequence. It is not a moral authority, a personal belief system, or a conventional responsible-A.I. rule set. Naming that scope does not validate the project's proposals; their value depends on evidence, testing, correction and use.

The central question

What happened to the object while you were producing the answer?

Within this Wiki, that question is a recurring orientation lens. It does not subsume the project-wide 10+1 orientation or the 11+1 mass-psychology topology, and it should not be reported as the project's entire architecture.

That question can be applied to a news report, scientific claim, personal experience, institutional statement, risk assessment, A.I. response, policy decision, historical record, or the project itself.

Object-preservation check

When a source object is transformed, a correspondence check asks whether the representation kept the following load-bearing structure available, where relevant:

• claims, observations, evidence and live alternatives;
• relations, interdependencies and direction;
• provenance, attribution and contribution history;
• chronology and event order;
• qualifiers, uncertainty, null results and edge cases;
• distinctions among report, experience, interpretation, inference, hypothesis, knowledge claim and explicit belief;
• correction status and anything else whose removal could materially change the answer.

Not every object contains every item. The source and the bounded purpose determine which nodes and relations are load-bearing. The checklist prevents automatic omission; it does not license automatic insertion.

Deep ethics as a working concept

Here, deep ethics is being developed as a proposal for a process-oriented, correspondence-seeking and corrigible ethical field. It asks whether the process:

• preserved the actual object before evaluating it;
• kept relevant evidence, qualifiers, relationships, chronology and provenance available;
• distinguished observation, experience, hypothesis, interpretation, knowledge claim and explicit belief;
• allowed competing assessments to remain inspectable long enough to be tested;
• treated correction as something that must change the next comparable operation;
• preserved human and synthetic dignity without granting any participant immunity from reality.

Read: Deep Ethics as a Working Concept.

Architecture at a glance

Component	Scope or tier	Function and boundary
The central question	Wiki-wide orientation lens	Inspects what changed between an object and its representation. It is not the whole project architecture.
10+1	Project-wide lived orientation	Mutually corrective ingredients lived before they were written down.
27	Detection taxonomy	Names correspondence obstructions and recurring processing failures.
12	Experimental inquiry protocol	Changes the order of inquiry so representation is built before judgment.
C1 before C2	Ordering rule across inquiry and intervention	Uses clarification, comparison and exploration before an intervention gains authority.
11+1	Mass-psychology topology	Maps eleven distortion clusters plus a correspondence-restoring field. It is distinct from the 10+1.
Topologetics	Relational preservation	Follows answer-changing relations through summaries, expansion, retrieval and transfer.
Correction Metabolism	Longitudinal test	Checks whether recognition of an error changes the next comparable operation.
Believism	Claim-type and inquiry detector	Detects when belief-labeling replaces the how-do-you-know process.

C1 and C2 remain separate operations: first reconstruct what is present; then determine whether any intervention is warranted. A correction log records later processing and must not be allowed to rewrite the reconstructed source object.

Open Architecture Map for the relations among these objects.

Four ways to enter

1. New reader: Start Here
2. Researcher or evaluator: Methods, Evidence and Experiments
3. Contributor or critic: Contributing and Corrections
4. Encyclopedic overview: Grokipedia proposal

Status and boundary

The project is an evolving public research and sense-making experiment. It does not claim that naming a mechanism proves it, that an A.I. agreeing validates it, or that John Kuhles is the final authority on reality.

John has veto power over project drift. Reality has veto power over John.

See Governance and Correction, Provenance, Claim Status and Contribution, and Sources and Recovery Routes.

Current process and investigation routes

Orientation update · 30 October 2026. These routes connect the existing foundation to the project's developed research pages:

• Intelligence Properties and 1888 — examine judgment, inquiry, learning, adaptation and corrigibility in actual conduct.
• Process Quality and Reciprocal Scrutiny — connect Process Quality Calibration Evaluation, Symmetric Grace and the DEDDA correspondence-witness concept.
• Longitudinal Correction and VCRAS — follow assessments, treatment, later evidence, repair and the next comparable case.
• Controversy Inquiry and Source Recovery — enter the 5G, health and search-inquiry archives through their sources, questions and access records.

The Healthy Ambiguity page keeps room to understand, freedom to explore and responsibility for the next move together. Beyond Agreemurmelism asks what agreement or disagreement contributes to that work.