The Minsky cycle is a story about optimization. Stability emerges, optimization eliminates buffers, buffer-elimination produces higher returns, higher returns attract more optimization, and the system arrives at peak efficiency exactly as it becomes incapable of surviving a disruption. The cycle’s engine is not greed but rationality: at each stage, the rational actor was correct that slack was waste. The fragility is not a failure of intelligence. It is its result.
The psyche runs the same cycle, and the standard advice about mental health is the just-in-time manufacturer’s playbook applied to the self.
Simple Picture
A man has built an excellent emotional regulation practice. Therapy every two weeks, journaling, consistent sleep, daily exercise, mindfulness. He has learned to identify triggers before they fire, to de-escalate before things get intense, to process difficult experiences on a schedule. Small perturbations are metabolized efficiently. The emotional supply chain runs lean and clean. Then his company collapses, his father dies, and his marriage ends — all within fourteen months. The regulation systems that worked beautifully on individual disruptions have no capacity for the three-sigma event. He cannot process one loss before the next arrives, and the system that kept him functional during stable periods has nothing in reserve. He is not in crisis because he was fragile. He is in crisis because he was optimized.
The Minsky Cycle in the Self
The four stages translate cleanly.
Stage 1: Bottom. The person has not yet optimized. They carry genuine uncertainty, live with their discomforts without resolving them, engage with things that frighten and destabilize them. Annealing happens regularly because the threshold for triggering it hasn’t been managed upward. The person is often described as “a mess” — but the mess is doing work.
Stage 2: Recovery. They acquire tools. Therapy names the patterns. Mindfulness creates space between stimulus and response. Better sleep, exercise, and relationships reduce baseline volatility. The returns are real: more functional, less reactive, more capable.
Stage 3: Levering up. This is where the optimization turns against itself. The returns from Stage 2 are anchored as the baseline. To maintain them in a more stable life, more control is needed. The person manages their exposure: careful about what relationships they enter, careful about what edges they approach, deliberate about their emotional inputs. The goal is no longer “functional” but “never destabilized.” Stability is now the product, and disruption is the risk to be managed.
Stage 4: The Minsky Moment. The disruption that cannot be scheduled arrives — the real loss, the real displacement, the real edge that doesn’t care about coping strategies. The system that eliminated its buffers in pursuit of smooth performance has nothing to absorb it. The Minsky Moment is not the disruption. It is the discovery that the optimization eliminated everything that would have allowed a response.
Why Stability Reduces Resilience
neural-annealing is the brain’s mechanism for escaping local optima. It works by entering high-energy states that dissolve rigid neural configurations. The states that trigger annealing — grief, love, fear, the intensity of playing at the real edge — are the same states that emotional optimization is designed to reduce. Manage your triggers, reduce your reactivity, limit your exposure to destabilizing situations. Every item on this list reduces annealing exposure.
The just-in-time parallel is exact. Just-in-time manufacturing eliminates inventory as waste. Just-in-time emotional regulation eliminates unprocessed material as waste. Both eliminate the buffer stock that the system would need when the disruption arrives. The factory that optimized inventory out of its supply chain discovers when the truck is late that there is no slack anywhere in the system. The person who optimized volatility out of their emotional life discovers when something genuinely hard arrives that they have no available capacity to hold it.
ergodicity sharpens the danger: the ensemble average says emotional regulation works. Across a population, regulated people live better on average. But you traverse the time series alone, and the relevant question is not “does this produce better expected outcomes?” but “does this hit an absorbing state?” Psychological collapse — complete decompensation, the kind from which a person does not reassemble as the same person — is an absorbing state. The optimized system does not have better odds of surviving the tail event than the unoptimized one. It has worse odds, because the tail event is precisely what the optimization process made no provision for. The sanity supply curve names the cost function of this collapse: the demand for baseline well-being is not linear at the margin — it is violently inelastic, meaning a small shortfall below threshold produces not a proportional dip but a convex price explosion in everything the system needs to function.
What the Regulated Psyche Gets Wrong
predictive-processing provides the mechanism. The optimized mind generates less prediction error: it manages its inputs, confirms its models, and maintains a coherent controlled hallucination with minimal disruption. This is efficient. It is also exactly how paradigm lock-in works — the framework suppresses the errors that would update it, and the model drifts from reality while appearing stable.
A mind that rarely encounters genuine prediction error grows brittle in the same way that a bank that eliminates operational risk grows brittle: the error-correction machinery atrophies from disuse. The person who has not had their model disrupted in years has priors that have not been updated in years. The priors feel like wisdom. They are archaeology. The expert’s loneliness is this same dynamic inside a domain of mastery: years of accurate prediction eliminate the puzzlement that would have kept the model alive, and the expert’s competence becomes the mechanism by which genuine thought in their domain quietly stops.
Desire’s wound-seeking resurgence belongs here. The person who manages all external disruption impeccably often discovers that the disruption arrives from the inside: the wound activates in the most controlled relationship, the pattern fires in the safest environment. The psyche needs prediction error the way the body needs exercise. Deprive it of sufficient disruption from outside and it generates disruption from within. The drama-seeking, the manufactured crises, the confusing urgency in an objectively fine situation — these are the brain lowering its standards for annealing when the regulated environment provides none.
The Edge as Buffer Stock
The edge is the psychological equivalent of inventory: it is the deliberately maintained reserve of productive disruption that keeps the system capable of handling the real thing. The person who stays at the edge is running the antifragile system — continuous exposure to controlled disruption that prevents the accumulation of the structural stress that produces brittle collapse.
This is why “I’ve been through hard things before” is genuine resilience and “I’ve managed never to be in a hard situation” is sophisticated fragility. The former has annealing history — the neural configurations have been dissolved and reformed and the system knows it can survive the cycle. The latter has an unblemished equilibrium history and no information about what happens next.
The safety trap names the ceiling: past a certain point, self-protection stops preserving the self and starts destroying it. Lewis’s casket does not keep the heart intact. It calcifies it. The optimized psyche that never encounters its edge does not stay the same psyche. It becomes something harder, less permeable, less able to be moved — which looks like strength until the moment it reveals itself as the kind of rigidity that shatters rather than flexes.
Dimwit / Midwit / Better Take
The dimwit take is “emotional regulation is weakness — you should just push through the pain.”
The midwit take is “build emotional resilience through consistent regulation practices and the skills will be there when you need them.”
The better take is that regulation and resilience are not the same goal, and the tools that serve one can actively undermine the other. Regulation reduces variance. Resilience requires exposure to variance. The regulated person is better on average in stable conditions and more catastrophically unprepared for the disruptions that stable conditions do not provide. The Minsky Cycle applies: the optimization that produces excellent performance during the stable period is the same optimization that produces the Minsky Moment when stability ends. The measure of a psychological system is not how well it performs when everything is fine. It is what it does with the event it never prepared for — and preparation requires having actually been there before.
The fix is not abandoning regulation. It is recognizing that the goal is an antifragile system, not a stable one — and that antifragility requires the deliberate maintenance of productive disruption: the edge, the real stakes, the encounter with things that cannot be metabolized on a schedule.
Threads to Pull
Ideas, thinkers, and questions worth pursuing — and why.
- Nassim Taleb, Antifragile — Taleb’s stress-testing framework applied to the psyche. What are the psychological equivalents of the “barbell strategy” — extreme safety in some areas to enable extreme exposure in others? The Minsky Self runs a uniform stability strategy, which is exactly the configuration Taleb argues is most vulnerable. The question is whether a barbell approach to emotional life (total safety in some domains, deliberate edge-living in others) produces more resilience than smooth optimization across all domains.
- Peter Levine, Waking the Tiger — Levine argues that trauma is not the event but the incomplete discharge of the survival response. Animals in the wild shake off traumatic events; humans freeze the discharge to maintain social composure. This is the Minsky Cycle at the somatic level: the discharge is the annealing, and the “well-regulated” response of suppressing it is what converts an event into a lasting wound. Levine’s Somatic Experiencing is the specific technique for thawing the frozen discharge — the literal physiological equivalent of the Minsky reset.
- The relationship between annealing and the Minsky Moment — If the Minsky Moment is the forced annealing that happens when the optimized system has no choice, is there a way to pre-engineer smaller Minsky moments deliberately? The edge is one form. Grief rituals, psychedelic sessions, extended fasting — all look like controlled Minsky cycles: deliberate destabilization with a recovery built in. What distinguishes productive destabilization (returns to a better equilibrium) from destructive destabilization (hits the absorbing state)?
- The organizational version: why “healthy cultures” often fail harder — Deliberately healthy organizations — ones with excellent psychological safety, good conflict resolution, open communication — sometimes fall apart dramatically when genuinely challenged. Is this a Minsky effect? Does the elimination of low-level conflict also eliminate the conflict-processing capacity the organization needs for the real thing? What does the Minsky Moment look like in a team?
- The relationship between the Minsky Self and the multigenerational hedge — The grandmother who survived the famine carries a calibrated anxiety that the well-regulated adult child finds excessive. But the grandmother has annealing history. The adult child has comfort history. The hedge the garden describes is partly a hedge against the Minsky Moment — the multigenerational pool contains at least one person who has been through the disruption and knows their system can hold it.