Collapse and stabiliser are the same topology run in two directions. Collapse spends the optionality stock down; a good stabiliser deposits it back — across more than one component, without clicking a new ratchet.
Each stage transmits harm and quietly closes futures. The leading indicators say where the option-space is draining now.
Each stage names the platform to build. A stabiliser is high-leverage exactly when it runs the collapse in reverse — adding to the stock, not just absorbing one shock.
Grounded in Dark Matter Labs' own definition of optionality as the continuity of generative agency. "Recursive planetary risk" — the cascade's final stage — is simply the regime in which spend-down chronically outpaces rebuild. Keeping the stock above the level at which a system can still choose its own future is the literal objective, not a metaphor.
Three documents, one atlas — and the contradiction that had to be fixed.
The source set disagreed with itself. It used three different stage taxonomies for the same cascade, so a diagram box could not be matched to a paragraph. And its political fork contradicted itself: the flowchart forks into fragmentation vs. authoritarian consolidation — two failure modes, no exit — while the argumentation doc forks into authoritarian vs. inclusive renewal, its whole "urgency without fatalism" resting on that exit. The diagram omitted the paper's central point.
This atlas resolves both. It adopts one canonical eight-stage spine, and it draws the fork as it should be: a nested choice. The first branch is coercive vs. inclusive; the coercive branch then splits into fragmentation and consolidation. That keeps every distinction the sources drew — and restores the off-ramp the diagram had deleted.
The cascade, in eight stages, ending in a regime.
Deliberately conditional: climate is a risk multiplier, not a direct cause of war. Whether stress becomes violence depends on the social-vulnerability substrate. Evidence is strongest at the front (physical, food, water, displacement) and weakens by design toward the back (violence, autocratisation).
Biophysical tipping risk
Thresholds, compound hazards and irreversible feedbacks — ice sheets, AMOC, forests and carbon, permafrost, ocean and ecosystem stress. The shift from linear change to threshold dynamics.
Earth-system destabilisation
Heat, drought, flood, fire, water-timing shifts, ocean disruption and productivity collapse — threshold risk becoming lived, material disruption.
Critical dependency networks
Shocks propagate through food, water, energy, finance, insurance, logistics, information and care systems — the tight coupling that is efficient in calm and a transmission channel under stress.
Human & system stress / capability degradation
Displacement, hunger, disease, infrastructure failure, economic contraction — and the biological load (heat, pollution, toxic exposure, sleep loss) that erodes the very capacity to respond well.
Social-vulnerability substrate
Poverty, inequality, identity stress, information disorder, environmental injustice and low social capital decide whether a shock is absorbed or amplified — who is protected and who is abandoned.
Legitimacy failure
When survival pressure is distributed unfairly, crisis becomes political: the state is seen to sacrifice some to protect others. Below a threshold of trust, cooperative problem-solving stops. This is where a physical crisis becomes a political one.
The political fork — nested, not a line
Regional & planetary cascades
Displacement spillovers, proxy conflict, resource militarisation, weaponised interdependence, cooperation failure and economic fragmentation — the shock crossing borders and scales.
Recursive planetary risk
Not a single collapse but a regime: reduced adaptive capacity and greater exposure to the next shock. In optionality terms — the absorbing state, where interventions stop "taking".
Optionality, decomposed into six components.
"Option-space" is too coarse to act on. Decomposed, optionality has six components — each failing by a different mechanism, on a different clock. The decisive property is timing: the choice that would have avoided the damage is usually gone long before the damage is visible. Each is also the thing a stabiliser must put back.
Reachability
How many distinct, viable end-states remain accessible from here.
Foreclosed by: habitability loss, thresholds, lock-in.
Reversibility
Whether a move can be walked back if it proves wrong.
Foreclosed by: hysteresis — the master switch.
Response diversity
The number of independent ways to meet a need, so no single failure removes the function.
Foreclosed by: synchronisation of buffers assumed independent.
Slack (buffers)
Spare capacity that absorbs a shock without forcing an immediate, costly choice.
Foreclosed by: depletion + convergence until every shock is existential.
Lead time
The warning and runway to choose deliberately rather than react under duress.
Foreclosed by: compression — hazards outpacing adaptation.
Decision latitude
Whether choices are made through legitimate deliberation or extracted under coercion.
Foreclosed by: legitimacy collapse — futures chosen at gunpoint.
The stage that attacks the responder, not just the system.
The flowchart's most distinctive move is to isolate a stage where stress stops being external and becomes internal — a direct load on bodies and minds. It matters because it is a feedback: every later stage assumes a population able to sense, decide and act, and this stage erodes exactly that. It forecloses agency. Kept here, and evidenced with an honest gradient — the strong core separated from the speculative edge.
| Driver | What it degrades | Evidence |
|---|---|---|
| Extreme heat | Mortality, cognition, sleep, learning, labour capacity. | strong — IPCC AR6 WGII, Lancet Countdown, WHO. |
| Air pollution | Cardio-respiratory disease and death; emerging cognitive / neurodegenerative links. | strong — WHO (on the order of 7M premature deaths a year); neuro links maturing. |
| Undernutrition & food insecurity | Childhood development, cognition, immune function. | strong — FAO, WHO, GRFC. |
| Disease burden & trauma | Physical health; mental health under disaster, displacement, chronic insecurity. | strong–moderate — WHO, Lancet Countdown, UNDP. |
| Endocrine disruptors | Reproductive, metabolic, neurodevelopmental function. | emerging / contested — Endocrine Society, WHO–UNEP; sperm-count decline reported (Levine et al.), causation debated. |
| Sleep loss & "digital overload" | Attention, cognition, mental health. | weak / speculative — a hypothesis at population scale, not a finding. |
Read structurally, this stage matters because it is a feedback, not just an impact. A sick, exhausted, malnourished, traumatised population converts the same shock into more harm (deepening the vulnerability substrate), deliberates and cooperates less well exactly when coordination is needed (deepening legitimacy failure), and — the climate–conflict evidence finds temperature itself raises violence risk — feeds the fork. It also reveals something about optionality itself: most of the six components live in systems, but agency lives in the body. A society can hold a full option-space on paper and be unable to move into any of it if its people cannot act. And because early-life exposure to heat, pollution, undernutrition and toxicants has lifelong and plausibly heritable effects, capability degradation forecloses the next generation's agency too — the bodily route into irreversibility, and the reason the stabiliser model reserves a seat for a "future-generation agent".
Every foreclosure edge, and the platform that cuts it.
The centrepiece. Read the left three columns forward as the risk map — what forecloses, and the structural shape doing the foreclosing. Read the right two columns backward as the investment map — the stabiliser platform that reverses that edge, and the optionality components it restores. The two are the same rows.
| Stage | Topology | Forecloses | Stabiliser platform — the interruption | Restores |
|---|---|---|---|---|
| 1 · Tipping risk | threshold · ratchet | slack + reversibility at planetary scale — the margin that kept other thresholds "far away". | Arctic / methane / feedback stabilisation; governed climate-intervention research. Front-of-chain only — downstream cannot reopen this. | reversibility · lead time |
| 2 · Earth-system destabilisation | fan-out · depletion | response diversity and buffers in the physical base — heat, water, fire, coast fail together. | Watershed & aquifer stabilisation; fire-adapted landscapes; coastal living shields; bioregional multispecies farming; whole-city cooling. | slack · response diversity · reversibility |
| 3 · Dependency networks | fan-in · common-mode | response diversity — coupled buffers (food, water, energy, finance) fail together and jump domains. | Distributed energy microgrids; local manufacturing of essentials; circular construction & material banks; public spatial AI. | response diversity · slack |
| 4 · Human stress / capability | feedback | agency — the capacity to respond erodes as stress compounds. | Open-source health & care commons; resilience hubs; passive-survivability buildings; care-economy infrastructure. | slack · agency (capability) |
| 5 · Vulnerability substrate | common-mode | who is protected — the substrate sets the conversion rate from shock to harm. | Regenerative finance / insurance / procurement; managed-retreat & relocation finance; climate-receiving cities. | slack · reachability · response diversity |
| 6 · Legitimacy failure | loop · ratchet | decision latitude — the pivot: below a trust threshold, cooperation stops. | Participatory crisis governance & agentic democracy; public spatial AI as shared truth. | decision latitude · lead time |
| 7 · Political fork | bifurcation · ratchet | latitude locks (coercive) or reopens (inclusive). | Conflict-prevention & resource-sharing compacts; participatory crisis governance; self-sovereign land (diverts land conflict). | decision latitude · reversibility |
| 8 · Planetary cascades | contagion · bridge | reachability — futures drop off the map across whole regions. | Climate-receiving cities & humane mobility; conflict-prevention compacts; planetary-feedback cooperation. | reachability · response diversity |
What makes a platform a pass-through optionality stabiliser.
Not a product and not merely adaptation. A stabiliser is a population-scale, quasi-infrastructural platform that stabilises a critical domain while generating reusable capability that expands option-space in other domains. Its value is not just avoided harm — it is the capability stock it deposits.
The engine is not "intervention solves problem." It is "intervention creates an enabling stack that becomes useful across many future problems." Five links:
Systemic pressure
Land, heat, water, food, energy, health, migration or governance stress narrows options.
Stabiliser bundle
A coordinated platform deployed at quasi-infrastructure and population scale.
Enabling stack
It forces tools, rules, institutions, data, sensors, robotics, care or ecological design into existence.
Capability stock
Those capacities become reusable assets for other domains and future shocks.
Derivative optionality
New pathways open beyond the original problem — the source of high leverage.
Eight platforms — in effect, one per stage of the cascade.
Laid on the spine, the highest-leverage stabilisers resolve into a pattern: roughly one platform per major stage — which is why they are tier-one. Each cuts a whole stage and deposits capability downstream. The kicker on each names the components it restores.
Self-sovereign land & agentic spatial governance
Where food, water, housing, migration, energy and identity pressures converge. Diverts land conflict into coordination — stewardship sovereignty, not exclusion.
Restores: reachability · reversibility · latitude · diversity
Bioregional multispecies farming
De-correlates food; restores soil, water and biodiversity; cuts fertiliser and supply-chain dependence.
Restores: response diversity · slack · reversibility
Watershed & aquifer stabilisation
Buffers the water chokepoint; recharge and basin compacts restore reversibility and cooperation.
Restores: slack · reversibility · diversity · lead time
Whole-city cooling & thermal commons
Thermal slack plus the health and labour capacity that heat degrades directly.
Restores: slack · agency · lead time
Distributed energy microgrids
De-correlates power (inverts synchronisation); islandable grids are reversible; most other stabilisers depend on it.
Restores: response diversity · reversibility · slack
Open-source health & care commons
Rebuilds agency and distributed care so abandonment does not become grievance.
Restores: agency · response diversity · latitude
Public spatial AI & early warning
Manufactures lead time; shared situational truth coordinates land, water, fire, heat and migration.
Restores: lead time · decision latitude · reachability
Participatory crisis governance
Defends decision latitude — the pivot component. Fast without becoming authoritarian; sunset clauses and appeal rights.
Restores: decision latitude · reversibility · lead time
A proposed leverage rule — the ranking the source model lacks: a stabiliser is high-leverage in proportion to (edges cut × components restored × pass-through reach) ÷ negative-pass-through risk. And a sequencing rule: in any place, fund first where a component is closest to a ratchet or a closing loop — that is where reversal cost is about to jump. When several platforms' pass-throughs reinforce each other, you get the source model's flywheels — bioregional, urban-heat, water-security, human-capability.
The false-positive filter — not every stabiliser is high-leverage.
The strongest part of the source model is its self-criticism. Some interventions reduce harm now while creating deeper dependency, inequality or lock-in. Structurally, each stabilises one component while clicking a ratchet on another — failing the single testable bar: raise the stock across more than one component and keep the system reversible.
| Intervention | Stabilises now | Negative pass-through | Better design direction |
|---|---|---|---|
| Fossil-powered AC alone | Reduces heat exposure. | Raises emissions, grid stress, energy inequality. | Whole-city cooling, passive buildings, district cooling, microgrids. |
| Hard seawalls alone | Reduces local flood risk. | Ecological loss, false security, development lock-in. | Living shorelines, managed-retreat finance, adaptive coastal settlement. |
| Militarised borders | Manages migration pressure. | Violence, rights collapse, regional instability. | Climate-receiving cities, legal pathways, humane mobility compacts. |
| Surveillance-heavy crisis management | May reduce disorder. | Authoritarian lock-in, loss of trust. | Participatory crisis governance, privacy-preserving public data, appeal rights. |
| Centralised monoculture food | Maintains short-term output. | Input dependence, biodiversity loss, soil fragility. | Multispecies farming, robotics, nutrient circularity, seed commons. |
| Tier-one platform | Capture surface | Safeguard |
|---|---|---|
| Self-sovereign land | Neo-feudal enclave sovereignty; land grabs justified by "adaptation"; black-box algorithmic zoning; surveillance land management; speculative hoarding. | Public law + anti-enclosure rules + stewardship duties; tenure protection, Indigenous rights, consent, community land trusts; transparent, appealable models; data rights + local ownership; land-value capture + anti-speculation tax. |
| Bioregional farming | Ag-tech lock-in — proprietary seed / robotics / data platforms swap input dependence for a new one; carbon / greenwash capture of soil; displacement of smallholder & Indigenous practice. | Seed commons, open hardware, right-to-repair; producer-cooperative ownership of data & equipment; protect existing tenure & agronomy; un-gameable avoided-loss verification. |
| Watershed & aquifer | Water financialisation / markets that price out the poor; strong-riparian capture of basin compacts; "recharge" licensing new extraction. | Water as protected commons with a human-right floor before any tradability; downstream & ecological standing; additionality caps on recharge; transparent allocation. |
| Whole-city cooling | Green gentrification — canopy & district cooling raise rents and displace the heat-vulnerable they were meant to protect; energy-intensive cooling deepening the grid / emissions ratchet. | Target heat-vulnerable districts first; anti-displacement & rent protection tied to greening; passive-first + microgrid-backed to avoid the fossil-AC ratchet; public ownership of thermal infrastructure. |
| Energy microgrids | Utility / vendor enclosure of "community" grids (proprietary lock-in); energy haves-and-have-nots as grid defection strands the poor on a weakened shared grid. | Community / cooperative ownership + local maintenance skills; open interoperability standards; cross-subsidy & regulated fair access so defection doesn't strand others. |
| Health & care commons | Platform capture of health data (surveillance, insurer discrimination); unpaid (usually women's) labour dressed up as "community" care; biased AI triage. | Data rights, privacy & non-discrimination law; paid, valued care work not extraction; open protocols + clinical accountability + bias audits. |
| Public spatial AI | The sensors that warn can also watch (surveillance); model opacity decides whose risk is seen; privatised risk intelligence gives insurers / speculators the signal first. | Privacy-preserving design + data rights; public-utility status with open, equal access; transparency, contestability & audit of models; democratic oversight of what is sensed. |
| Participatory crisis governance | Participation-washing (consultation as legitimacy theatre); the emergency-powers ratchet it is meant to prevent; capture of "AI-supported deliberation" by whoever controls the model. | Binding (not advisory) mechanisms + hard sunset clauses + independent audit; rights floors emergency powers can't breach; transparent, contestable deliberation tooling under public governance. |
The operating rules, both directions.
Read the atlas down its length and the same structural truths recur — as instructions for where and when to act, and what to build.
Foreclosure precedes harm; so must the deposit.
Every stage closes its futures before the damage shows. The stabiliser that reopens them must arrive before the loss registers, not after.
Instruction: instrument leading indicators; build the stabiliser upstream of the lagging damage.
Reversibility is the bar a stabiliser must clear.
Hysteresis is the master switch between a stressor (recoverable) and a collapse (not). A stabiliser that clicks its own ratchet — a seawall, fossil AC — fails even as it helps.
Instruction: favour modular, removable, retreatable platforms; keep the pawl from clicking.
Leverage lives in chokepoints and diversity.
The largest losses come through synchronisation and at convergence nodes; so the largest gains come from de-correlating buffers and putting slack back at chokepoints — water, grid, fiscal capacity, public health.
Instruction: rank stabilisers by edges cut × components restored ÷ capture risk.
Deposits must be multiplicative.
Collapse forecloses several components at once, so partial defence of one fails if the others go. A high-leverage stabiliser adds to more than one component simultaneously — the definition of capability-throughput.
Instruction: reject single-component fixes; require pass-through into a second domain.
Sequence by proximity to the ratchet.
The recursive regime is an absorbing state, where interventions stop "taking". Order the portfolio by which component, in which place, is closest to irreversibility.
Instruction: fund first where reversal cost is about to jump; keep the system out of the absorbing state, where leverage still exists.
Sources for the cascade, the reading and the response.
Physical-risk and system-structure claims carry high confidence; conflict and legitimacy pathways are stated conditionally, mirroring the source documents. The stabiliser mappings are design logic, not demonstrated outcomes. Figures are order-of-magnitude, for orientation.
- [1]Dark Matter Labs — From Biophysical Tipping Risk to Cascading Violence (argumentation): the eight-stage spine and its institutional evidence base (IPCC, WMO, IPBES, World Bank, SIPRI, UCDP/PRIO, V-Dem). Climate as risk multiplier; prevention over late response.
- [2]Dark Matter Labs — Biophysical Tipping Risk, Social Vulnerability & Cascading Violence (systems flowchart): the one-page diagram; legitimacy as the pivot where biophysical crisis becomes political.
- [3]xCO analytical layer — Optionality Collapse Atlas: the six components, the eight-primitive topology grammar, the twelve-pathway atlas and the "foreclosure precedes harm" reading this atlas closes the loop on.
- [4]Dark Matter Labs — Pass-Through Optionality Stabilizers: A Logic Model (v2, May 2026): the stabiliser definition, the value equation, the pass-through engine, the 22-item catalogue and eight tier-one platforms, the flywheels and the negative-pass-through filter.
- [5]Dark Matter Labs — Defining Optionality (29 Mar 2026): optionality as the continuity of generative agency; foundational goods as the substrates of freedom.
- [6]Physical & human-system grounding cited in the sources: Global Tipping Points Report 2025; Planetary Health Check 2025 (seven of nine boundaries breached); Kornhuber et al. (synchronised breadbasket failure); UNHCR / IDMC (protracted displacement); V-Dem / Freedom House (autocratisation). Resilience-theory foundations: Walker, Holling, Folke.
The future closes quietly, component by component, before the damage is loud. The cascade names the route; optionality names the loss; the stabiliser names the deposit that buys it back.
integrated atlas
v1 · jul 2026
← all published notes · the earlier braid: tipping → collapse