Problem Classes

Overview

This document defines the formal Problem Classes addressed by PlainSight Lab.

Problem classes are structural, not domain-specific. They describe recurring failure modes that emerge across finance, AI systems, and social infrastructure when adversarial behavior, automation, and delayed correction interact.

A system may instantiate multiple problem classes simultaneously. Effective containment requires addressing all active classes, not only the most visible one.

This registry exists to prevent misdiagnosis.

PC-1: Unbounded Error Propagation

Definition
Systems in which errors propagate faster than humans can identify and correct them.

Structural Conditions

• High execution speed
• Delayed or indirect observability
• Costly or irreversible actions
• Human-bound correction pathways

Failure Mode
Local mistakes compound into systemic failures before intervention is possible.

Containment Strategy

• Early identification mechanisms
• Pre-execution constraints
• Invariant enforcement at decision boundaries

Derived From

• Systems Thesis — The Problem Class
• Truth & Governance Evolution — Phase 1

PC-2: Adversarial Incentive Exploitation

Definition
Systems where participants are structurally incentivized to probe, exploit, or manipulate rules for advantage.

Structural Conditions

• Competitive or zero-sum dynamics
• Rule complexity or ambiguity
• Observable reward asymmetry
• Weak or delayed enforcement

Failure Mode
Rules are optimized against rather than complied with. Good-faith assumptions collapse under sustained pressure.

Containment Strategy

• Assume adversarial behavior by default
• Remove ambiguity
• Make exploitation visible and costly

Derived From

• Systems Thesis — Adversarial Systems Are the Default

PC-3: AI-Accelerated Asymmetry

Definition
Systems where AI reduces the cost of attack, probing, or deception without equivalently reducing the cost of identification or correction.

Structural Conditions

• Automated strategy generation
• Scalable probing
• Human-only authority or correction
• No enforced execution constraints

Failure Mode
Error propagation and manipulation occur at machine speed while oversight remains human-bound.

Containment Strategy

• Deterministic execution layers
• Explicit authority separation
• Verification before action
• AI as advisory only

Derived From

• Systems Thesis — AI Reshapes the Adversarial Landscape
• Internal Constitution — Article VII

PC-4: Post-Hoc Governance Failure

Definition
Systems that attempt to govern behavior only after harm has occurred.

Structural Conditions

• Audit-based enforcement
• Reactive policy creation
• Punitive rather than preventive correction
• Slow institutional response

Failure Mode
Governance becomes symbolic. Damage is acknowledged but not meaningfully corrected.

Containment Strategy

• Governance before scale
• Constraints before execution
• Enforcement through refusal

Derived From

• Systems Thesis — Why Traditional Responses Fail
• Internal Constitution — Article V

PC-5: Authority Ambiguity

Definition
Systems where responsibility, authorship, or decision authority is unclear or diffused.

Structural Conditions

• Multiple overlapping actors
• Informal or implicit decision rights
• Weak attribution
• Blurred governance boundaries

Failure Mode
No one can be held accountable. Errors persist because responsibility cannot be assigned or enforced.

Containment Strategy

• Singular canonical truth
• Explicit authority boundaries
• Documented responsibility

Derived From

• Internal Constitution — Article II
• Internal Constitution — Article III

PC-6: Incentive Drift

Definition
Systems where local incentives (growth, usability, revenue) diverge from correctness or governance objectives.

Structural Conditions

• Product-driven optimization
• Market pressure
• Lack of binding constraints
• Informal exception handling

Failure Mode
Short-term success erodes long-term legitimacy and correctness.

Containment Strategy

• Incentive containment
• Mandatory escalation
• Correctness over convenience

Derived From

• Derivative Entity Charter Template — Article VI

PC-7: Premature Decentralization

Definition
Systems that distribute authority before invariants are stable or governance is bound.

Structural Conditions

• Early plurality of implementation
• Weak amendment rules
• Informal governance
• Ideological decentralization pressure

Failure Mode
Fragmentation, incompatibility, and legitimacy collapse.

Containment Strategy

• Phase-gated authority distribution
• Constitutionalization
• High amendment thresholds

Derived From

• Truth & Governance Evolution — Phase 1 & Phase 2.5

PC-8: Self-Perpetuating Authority

Definition
Systems where authority structures resist dissolution even after their original justification expires.

Structural Conditions

• Unbounded governance tenure
• No forced transition criteria
• Authority tied to individuals or entities
• Lack of neutral institutions

Failure Mode
Authority becomes extractive, illegitimate, or captured.

Containment Strategy

• Explicit obligation to distribute authority
• Designed end states
• Institutional handoff

Derived From

• Internal Constitution — Article IV & Article IX
• Truth & Governance Evolution — Phase 3

PC-9: Observability Collapse

Definition
Systems where identifying harm requires excessive social, political, or personal energy.

Structural Conditions

• Power asymmetry
• Social or institutional pressure
• Ambiguous evidence standards
• High false-positive or false-negative cost

Failure Mode
Harm persists because identification is too costly or dangerous.

Containment Strategy

• Mechanized detection
• Verifiable signals
• Early identification over late correction

Derived From

• Systems Thesis — Social Safety and Awareness

Closing Rule

These problem classes are authoritative classification primitives.

They may be referenced, combined, or instantiated by downstream documents and derivative entities, but they may not be redefined without formal amendment.