From Case Scenes to Decision Models.
Archon reads research cases on how organizations decide; the cases are built with AI and cross-checked across model families. Each scene is read along four axes, given options and a provisional judgment, and placed beside other cases on the decision atlas.
Through four axes (information asymmetry, delegation of authority, assignment of accountability and rule exceptions), Archon reconstructs past decision trajectories and gives present choices firmer grounding.
- Base Layer
- Research Cases
- Reading Mechanism
- Four-Axis, Question-Led Reading
- Execution System
- The Scalae & 13 Decision Domains
The Case Base.
Archon’s questions start from case scenes: situations in which an organization had to decide. Each case isolates a single fork within one such situation (Layer 00).
A scene records what the people in it could know, which alternatives were open to them, and which constraints bound them. Reading these together lets us ask why a decision was reached and which exceptions were allowed.
- Scenes & Roles
- Each scene names who sat where, so a case can be read in terms of seats and roles.
- Decision Forks
- The moment a proposal reaches the person with authority to decide: two options, the seat that decides and the seat that is affected.
- Conditions to Reconsider
- Each judgment names what the case alone cannot settle and the condition under which the other option would be sounder.
The 5-Tier Architecture: Reading from Below.
Archon does not run a flat search across one body of text. As shown below, it reads upward from the case to the scene, the four axes, the atlas and, finally, human judgment.
Select a layer to see its role and outputs.
Research Cases
- Algorithmic Role
- The base layer is a set of research cases on how organizations decide, built with AI and checked across model families. Every question on this site starts from one of them.
- Working Material & Outputs
- Roles, situations and the moment a decision falls due.
Does the case give enough context to read who could decide and who would be affected?
Decision Scenes
- Algorithmic Role
- A scene isolates one fork inside an organization: what was proposed, who holds the decision, who bears the consequences, and what is still unconfirmed.
- Working Material & Outputs
- A scene summary, the deciding and affected seats, and two options.
Are proposals, actual agreements, and executed actions strictly distinguished?
Four Analytical Axes
- Algorithmic Role
- Each scene is read along four dimensions: information asymmetries, where authority lies, the scope of accountability and the legitimacy of rule exceptions. The axis a case turns on decides where it sits in the library and on the map.
- Working Material & Outputs
- A structural reading for each case, grouped by axis.
Can cases from different organizations and periods be compared on the same four axes?
The Scalae & Decision Domains
- Algorithmic Role
- Retrieves relevant evidence from the higher corpora in response to user inquiries and maps them across the 13 institutional domains (power balance, succession, resource allocation, etc.) in the Decision Atlas.
- Working Material & Outputs
- Supporting and contrary evidence graphs, cross-domain interaction vectors, and revision conditions.
Does the retrieved map directly resolve the fundamental causal structure, or merely reflect surface similarity?
Human Judgment & Accountability
- Algorithmic Role
- The algorithm never replaces judgment; it illuminates the evidential topography. Human leaders weigh consequences, shoulder accountability, and explicitly register the conditions warranting future reconsideration.
- Working Material & Outputs
- Accountable decisions with explicit reference dates, core assumptions, and post-audit criteria.
Who bears ultimate responsibility for this decision, and what evidence would prompt its reversal?
Deconstructing Decisions: The 4 Foundational Axes.
Cases are not classified at will; each is read along the four core axes that shape the dynamics of organizations and power.
- 01
Information
Tracks what information reached whom at the moment of decision, and what critical distortions or asymmetries existed.
- 02
Authority
Examines the gap between the formal approval chain and actual influence, and identifies who exercised veto power and final approval.
- 03
Accountability
Assesses to whom responsibility for failure or success was shifted, and how institutional incentives operated.
- 04
Exceptions
Uncovers the rationales used to permit departures from established rules and whether they hardened into precedent.
The Four Axioms of the Archon Engine.
Interpretation Kept Apart
Stated ScopeEvery judgment carries a reference date, and interpretation is kept apart from the scene it reads. Judgments can be revised in light of counter-evidence and new data.
Question-Led Reading
Question-LedArchon does not read from a single, fixed knowledge base. Each question decides which of the four axes a scene is read along, and which conditions would overturn that reading.
Transparent Causal Lineage
Causal TransparencyEvery proposal and analysis must make its causal trajectory transparent: who knew what and when, and which exceptions were approved. Archon rejects black-box inference.
Irreducible Human Responsibility
Human AccountabilityThe algorithm’s role ends at mapping the structural risks of a decision and its alternatives. Whoever makes a decision and answers for its consequences can never be reduced to a machine.
Executable Decision Map & Question Library.
Archon’s algorithms are not left as theory; they are implemented directly in workspaces. Decision cases extracted from the layered corpus are mapped in real time onto the 13 institutional domains of the Decision Atlas, and can be tested against new matters at any time through the Question Library.