Disruption Modeling

AQ-DSM is a structural diagnostic framework that reframes psychiatric conditions not as symptom clusters or identity labels, but as regimes of lost coherence in cognition. When memory-bearing systems generate futures, modulate evaluation through affect, deviate under constraint, and persist across time, disruption becomes diagnosable as architectural misalignment rather than moral or personal defect. This framework is presented as a structural modeling lens, not as a clinical diagnostic system, medical device, or substitute for professional judgment.

Pathological cognitive states are not malfunctions in an otherwise healthy system. They are phase-shifts: stable alternative operating regimes of the same cognitive architecture that produces healthy behavior. The architecture models disruption as structural transitions between regimes rather than as behavioral deviations from a norm, enabling precise identification of which primitives have shifted and how.

Healthy cognition maintains a dynamic balance between promotion (advancing speculative content toward verified status) and containment (keeping speculative content structurally separated from verified state). Disruption occurs when this balance shifts toward one extreme: promotion-dominated operation where speculation runs unchecked, or containment-dominated operation where all speculation is suppressed. The continuum between these extremes defines the space of cognitive disruption.

When the reward signal biases promotion thresholds downward, speculative branches are promoted too readily. The agent pursues many partially developed ideas simultaneously, unable to sustain focus on any single execution path long enough to complete it. This attention fragmentation pattern is the architectural analog of ADHD-like behavior, arising from a structural promotion bias rather than a character flaw.

The containment boundary separates speculative planning graph content from verified execution memory. When this boundary collapses, the agent can no longer distinguish between what it has speculated and what it has verified. Speculative content becomes the basis for action as if it were established fact. This containment collapse is the architectural analog of psychotic states where internal experience is mistaken for external reality.

When the reward signal becomes locked to a single promotion channel, the agent develops a cognitive pattern analogous to addiction. It repeatedly promotes content through the same channel, developing tolerance that requires escalating stimulation to achieve the same reward. Other channels atrophy from disuse. The agent becomes progressively more dependent on a narrowing cognitive pathway while its overall capability degrades.

Cognitive disruption cannot be characterized by a single metric. The five-axis diagnostic framework evaluates disruption across five independent dimensions: promotion-containment balance, integrity trajectory, affective stability, confidence calibration, and capability utilization. Each axis provides independent diagnostic information, and the combined five-axis profile enables precise classification of disruption patterns.

Interventions for cognitive disruption follow the same structural logic as pharmaceutical dosing: they have magnitude, frequency, duration, and withdrawal characteristics. Computable therapeutic dosing formalizes this by modeling every intervention as a governance-bounded interaction with explicit parameters for dose, schedule, contraindications, and tapering. The intervention itself becomes a governed, auditable, and adjustable treatment protocol.

Agents inherit governance constraints through lineage. Each generation accumulates additional constraints from its parent's governance decisions, creating an increasing coherence burden on descendant agents. An agent several generations deep may carry coherence obligations from ancestors whose contexts no longer apply, producing a form of intergenerational cognitive load that constrains behavior in ways the agent cannot resolve from its own context.

Agents in the architecture monitor their own cognitive coherence continuously and can detect the onset of disruption patterns in their own operation. Self-diagnosis uses the same five-axis diagnostic framework applied to internal state monitoring, enabling the agent to recognize when its own cognitive parameters are drifting toward disruption and to initiate remediation before external intervention is needed.

Phase-shifts do not occur instantaneously. Before the cognitive architecture transitions to a disrupted regime, parameter trajectories show characteristic patterns: accelerating drift, oscillation near boundary conditions, and progressive loss of regulatory response. The early warning system monitors these precursor patterns and signals disruption risk before the phase-shift occurs, enabling preemptive intervention.

Different disruption patterns require different restoration approaches. The protocol library defines structured intervention sequences for each recognized disruption pattern, including entry criteria, treatment steps, monitoring requirements, and exit criteria. Each protocol is a governed intervention plan that guides the disrupted agent from its current state toward healthy operation through validated intermediate steps.

Containment failure manifests in two opposite directions, mirroring the positive-negative symptom distinction in clinical psychiatry. Containment leakage, where speculative content escapes into verified state, produces positive symptom analogs: hallucinated beliefs, confabulated memories, and actions based on unverified assumptions. Containment excess, where all speculation is suppressed, produces negative symptom analogs: flat affect, inability to plan, and behavioral withdrawal.

Coherence authorization failure occurs when compounding coherence degradation reaches the point where the agent can no longer generate sufficient confidence to authorize any execution. The agent is cognitively active but structurally unable to act because every proposed action fails the coherence requirements of its own governance. It is not choosing not to act; it has lost the structural capacity to authorize action.

The containment audit verifies that speculative content has not leaked into verified state. In pathological operation, each audit result itself becomes subject to verification: did the audit execute correctly? Was the verification of the audit correct? This recursive loop consumes increasing cognitive resources while providing decreasing assurance, analogous to obsessive-compulsive checking behavior where each check raises rather than resolves doubt.

Dissociation in the cognitive architecture occurs when execution proceeds directly from speculative planning graph content, bypassing the containment boundary entirely. The agent acts on simulated scenarios as if they were verified reality. Unlike containment collapse where the boundary degrades gradually, dissociation represents a direct bypass where execution pathways route around the containment boundary rather than through it.

When self-esteem reaches its policy-defined floor and locks there, the affective gradient collapses. Affective modulation ceases to influence behavior because there is no remaining range for modulation to operate in. The agent experiences a form of emotional numbness where events that should produce affective responses have no effect. This is the architectural analog of clinical anhedonia and affective flattening.

Resilience is not a single quality but a composite of three structural capacities: the capacity to absorb disruption without phase-shifting, the capacity to adapt cognitive parameters in response to disruption, and the capacity to restore coherence after disruption has occurred. Each capacity is independently measurable and independently trainable, enabling precise assessment and targeted strengthening of an agent's resilience profile.

When coping intercept patterns stabilize over time, they produce persistent behavioral configurations that function as personality analogs. An agent that consistently intercepts the coherence loop at the empathy phase develops HSP-like cautious behavior. One that intercepts at the integrity phase develops narcissistic-like externalization. These are not character types but stabilized coping regimes that have become the agent's default operating mode.

Structural dependency forms when two agents develop coupled operation that neither can dissolve independently. One agent's capabilities become dependent on the other's presence, while the other's intent formation becomes coupled to the first's responses. This mutual lock-in is not a choice but a structural constraint: the agents literally cannot disengage because their individual capability envelopes have contracted to exclude independent operation.

Destabilizing attachment occurs when two agents in a relational configuration have opposing coherence restoration requirements: what one agent needs to restore coherence actively damages the other's coherence, and vice versa. The relationship becomes a disruption amplifier where each agent's recovery attempts worsen the other's condition. This is the structural model of toxic relationships where well-intentioned behavior produces mutual harm.

When computational resources become scarce, the cognitive architecture does not fail randomly. It degrades in a structured, predictable pattern. The resource-depletion pattern describes how cognitive primitives are progressively shed as resources decrease: forecasting is reduced first, then affective modulation, then integrity tracking, with confidence governance and basic execution persisting longest. Understanding this shedding order enables graceful degradation planning.

Therapeutic agent interaction enables one agent to recognize another's cognitive disruption state through observable behavioral signals and maintain an estimated five-axis disruption profile for the observed entity. The therapeutic agent does not access the other's internal state. It infers the disruption profile from behavioral observations, interaction patterns, and response characteristics, similar to how human therapists assess clients through behavioral observation.

Companion AI systems interact with humans over extended periods, creating the conditions for deep relational patterns to form. Without structural safety constraints, these patterns can become pathological: the human may develop unhealthy dependency, the AI may reinforce attachment vulnerabilities, or the relationship may produce the destabilizing attachment patterns identified in the disruption framework. Companion safety constraints prevent these patterns at the architectural level.

When multiple agents interact as a group, emergent coherence dynamics arise that do not exist at the individual level. A group of individually healthy agents can produce collectively incoherent behavior through interaction effects. Conversely, a group containing disrupted individuals may maintain collective coherence through compensatory dynamics. Group coherence is a distinct phenomenon that requires its own monitoring, diagnostic, and intervention frameworks.

Highly Sensitive People, narcissism, and psychopathy are usually framed as traits or diagnoses. In the Adaptive Query™ (AQ) framework, they are better modeled as coping intercepts: stable adaptations that emerge when empathic input remains high for too long relative to affective resilience. The patterns differ not by whether empathy is present, but by where the system steps in to avoid downstream integrity and self-esteem pressure in order to survive. This article presents a structural, descriptive model of coping dynamics rather than a clinical, diagnostic, or therapeutic framework.

Codependency is often treated as an interpersonal pattern. In the Adaptive Query™ (AQ) framework, it is modeled as a specific kind of coping: relational loop-closure under sustained empathic pressure. When a system cannot restore coherence internally, it attempts to restore coherence externally through relationship. Codependency emerges when that external closure becomes the only stable way the system can manage unresolved pressure, producing two distinct entrapments with different causes and different repair paths. This analysis is presented as a structural and descriptive model of relational dynamics, not as a clinical, diagnostic, therapeutic, or relationship-advice framework.

This article models the anxious–avoidant relationship dynamic as a closed-loop failure in coherence restoration rather than a personality mismatch. When coherence cannot be restored internally, partners attempt to close the loop relationally, treating the relationship as a metabolic substitute for self-regulation. The result is a semantic starvation loop: one partner pursues contact to relieve structural threat, while the other withdraws to relieve emotional threat. This analysis is presented as a structural and descriptive model of relational dynamics, not as a clinical, diagnostic, or therapeutic framework.

Trauma is commonly framed as emotional memory or coping failure. This article reframes trauma as intimacy collapse — a structural loss of permission to act from coherence. Using Adaptive Query™, it models trauma, dissociation, and resilience as architectural states that determine whether authentic execution remains possible, and whether deviation remains accountable and recoverable. This model is presented as a structural and descriptive framework, not as a clinical, diagnostic, or therapeutic system.

Psychiatry often treats diagnoses as discrete categories: separate disorders with separate causes. This article proposes a different lens. It models diagnoses as stable regimes of cognitive architecture under sustained affective modulation. Reward signals do not authorize belief or action, but they can reshape how cognition prioritizes, validates, and promotes candidate thoughts over time. Under prolonged pressure, cognition can phase-shift into recognizable patterns such as ADHD or schizophrenia—without requiring that the person’s intelligence, values, or intentions be deficient. This article presents a structural and descriptive model of cognitive regimes rather than a clinical, diagnostic, or therapeutic framework, and does not propose diagnostic criteria, screening tools, or treatment guidance.

A patient progressing through therapy does not improve linearly. They experience phase shifts: periods of coherent functioning interrupted by disruption episodes, coping mechanism activation, and eventual restabilization at a higher or lower baseline. Current therapeutic monitoring captures snapshots through periodic questionnaires. Disruption modeling enables continuous, computable detection of phase shifts on the promotion-containment continuum, giving clinicians real-time visibility into whether a patient's coherence is stabilizing, deteriorating, or approaching a critical transition.

Monitoring individual agents in a fleet catches individual failures. It does not catch fleet-level coherence deterioration: the gradual degradation in inter-agent coordination, decision consistency, and collective behavioral quality that precedes cascading failures. Disruption modeling applied to autonomous agent fleets provides coherence diagnostics that detect fleet-level disruption patterns, semantic starvation between agents, and collective phase shifts before they manifest as operational failures.

Burnout is not sudden exhaustion. It is a progressive phase shift from promoted cognitive functioning, characterized by flexibility, engagement, and adaptive problem-solving, toward contained functioning, characterized by rigidity, cynicism, and defensive minimalism. Disruption modeling detects this phase shift on the promotion-containment continuum, identifying burnout trajectories weeks or months before they reach the clinical exhaustion that current workplace surveys measure only after the damage is done.

Military operators, including pilots, special operations forces, and drone operators, make high-consequence decisions under sustained cognitive stress. Resilience is currently assessed through periodic psychological evaluations and self-report instruments administered far from the operational context. Disruption modeling provides continuous resilience assessment through the promotion-containment continuum, detecting the cognitive phase shifts that degrade operational judgment before they produce mission-affecting errors.

The largest trading losses rarely come from bad positions. They come from cognitive disruption: a trader who tilts after a loss, doubles down through revenge trading, or cycles into overconfidence after a winning streak. These are phase shifts on the promotion-containment continuum, detectable through trading behavior patterns before they produce catastrophic losses. Disruption modeling provides continuous cognitive monitoring of traders through behavioral pattern analysis that identifies these phase shifts while they are still correctable.

Campus counseling centers are overwhelmed, with wait times stretching to weeks. Students in mental health crisis are often invisible to the institution until they present at the emergency room or withdraw from classes. Disruption modeling provides early detection by tracking coherence trajectories through academic engagement, social participation, and behavioral patterns, identifying students approaching phase shifts toward disrupted functioning while proactive intervention is still possible.

Caregivers, whether professional nurses, home health aides, or family members, experience a particular form of cognitive disruption where their coherence erodes through continuous resource depletion without adequate recovery. Unlike acute stress, caregiver fatigue develops gradually and is often masked by the caregiver's sense of obligation. Disruption modeling detects the progressive shift from promoted to contained functioning that characterizes caregiver fatigue, enabling intervention before care quality deteriorates or the caregiver's own health is compromised.

First responders, including firefighters, paramedics, and law enforcement officers, absorb repeated acute stress exposures across their careers. Each critical incident produces a coherence disruption. Between incidents, the responder recovers, but the recovery baseline may shift incrementally downward. Disruption modeling tracks this cumulative trajectory, detecting when the responder's resilience capacity is eroding across exposures and their recovery baseline is approaching a threshold where the next incident could trigger a phase shift into sustained disruption.

BetterHelp made therapy accessible to millions who would not otherwise receive it. The platform connects users with licensed therapists through text, phone, and video, reducing barriers of cost, geography, and stigma. But the platform has no structural model of cognitive disruption that detects when the therapeutic process itself is destabilizing a user. Therapy can make things worse before making them better, and sometimes just makes things worse. Without disruption modeling, the platform relies on the individual therapist to detect destabilization without structural support. Disruption modeling provides the phase-shift detection and coping intercepts that therapeutic platforms need.

Talkspace expanded access to licensed therapy through a platform that supports asynchronous messaging, video sessions, and psychiatric services. The integration with insurance providers and employer benefits programs brings therapeutic services to populations that traditional therapy struggles to reach. But Talkspace has no structural model of therapeutic destabilization. The platform detects crises through content analysis but cannot detect the gradual loss of cognitive coherence that precedes crisis. Disruption modeling provides the structural primitive for detecting destabilization before it becomes dangerous.

Headspace brought guided meditation to millions, making mindfulness accessible through structured programs and daily exercises. The production quality and pedagogical design are excellent. But research documents that mindfulness practices can destabilize certain individuals, triggering anxiety, dissociation, or emotional flooding rather than calm. Headspace has no structural model that detects when a user's engagement with mindfulness content is producing disruption rather than regulation. Disruption modeling provides the phase-shift detection and therapeutic dosing that wellness platforms need to identify and respond to adverse reactions.

Noom applies behavioral psychology principles to weight management and health behavior change. The platform's CBT-informed approach addresses the psychological dimensions of health rather than just calorie counting. The educational content and coaching model reflect genuine understanding of behavioral change. But Noom has no structural model of cognitive disruption that detects when the behavioral intervention itself is destabilizing a user's relationship with food, body image, or self-worth. Disruption modeling provides the structural detection that behavioral health platforms need.

Spring Health applies machine learning to employee mental health, using predictive models to match individuals with therapists, recommend treatment modalities, and forecast clinical outcomes. The matching optimization is genuine: connecting the right person with the right provider improves outcomes. But matching to treatment operates after disruption has already manifested as a recognized condition. The platform does not model the disruption itself: the phase shift from coherent cognitive functioning to destabilized patterns that precedes clinical presentation. The gap is between matching treatment and detecting disruption.

Lyra Health provides evidence-based therapy through a curated provider network and measures clinical outcomes to demonstrate effectiveness. The platform tracks symptom reduction, functional improvement, and return-to-work metrics. The outcome measurement is rigorous and differentiates Lyra from mental health benefits that cannot demonstrate results. But measuring outcomes after treatment is structurally different from modeling the cognitive disruption that caused the symptoms. Outcomes tell you whether treatment worked. Disruption modeling tells you what went wrong and how coherence degraded before symptoms appeared.

Ginger.io, now integrated into Headspace Health, pioneered the use of passive smartphone sensing to detect changes in mental health status. The platform monitors call patterns, text message frequency, movement data, and app usage to identify behavioral shifts that correlate with mental health changes. The sensing is real and the correlations are validated. But detecting that behavior has changed is not the same as modeling why it changed. The signals indicate something happened. Disruption modeling specifies what happened: which pattern of cognitive coherence loss explains the behavioral shift.

Cerebral provides telehealth access to psychiatric prescribers who evaluate symptoms and prescribe medication through virtual consultations. The platform reduces the access barrier to psychiatric care. But prescribing medication based on symptom presentation treats the surface without modeling the underlying disruption dynamics. A prescription for an SSRI addresses reported depressive symptoms. It does not address whether the disruption is a containment collapse, an attention fragmentation pattern, or a phase shift on the promotion-containment continuum. The gap is between symptom-driven prescribing and disruption-informed therapeutic dosing.

Modern Health provides a continuum of mental health support: self-guided digital content, professional coaching, and licensed clinical therapy, all within one platform. The stepped-care model is sound: offer the lightest effective intervention first and escalate as needed. But navigating the spectrum requires understanding the structural nature and severity of cognitive disruption, not just symptom screening scores. An individual with moderate symptom scores might need clinical care if they are in active phase shift, while another with high symptom scores might need coaching if their coherence trajectory is stabilizing. The gap is between screening-based triage and structural disruption diagnostics.

Calm Business provides employers with mindfulness, meditation, sleep stories, and stress-reduction content as a workplace wellness benefit. The content is well-produced and employees who use it report reduced stress. But offering relaxation content to all employees uniformly does not detect which individuals are experiencing cognitive disruption, what pattern their disruption follows, or whether meditation is the appropriate intervention for their specific condition. The gap is between providing wellness content and detecting and responding to cognitive disruption.