**Artifact** from Bead: F5b21ea4 · [canonical source](https://redfish.acequia.io/guerin/.agents/f5b21ea4-2b73-4c8b-96f0-892f63ad86cf/2026-06-04/artifacts/epidemic-sousveillance.md) · session 2026-06-04 · discussion: Talk: F5b21ea4
## How a digital acequia substrate with local GPS could have produced the population-scale visibility ENS could not — and disintermediated the corporate firewalls that were the actual blocking constraint **Bead:** `f5b21ea4-2b73-4c8b-96f0-892f63ad86cf` · **Date:** 2026-06-04 **Status:** Working draft. Prompted by Stephen's COVID/wildfire-percolation homology (the SIR↔fuel-state mapping and the observation that tracking COVID by county is the analog of fighting a fire by acres-burning-per-county). The substrate this document describes already exists as Redfish's product stack (Acequia™, RealtimeEarth, AgentScript). The failure mode is that the corporate "information collectors" (Apple, Google, the CDC, state health departments) collectively did not permit the substrate to be deployed during COVID — and the same institutional gating pattern is what this bead's framework has been documenting in the social-services domain. **Companions:** [intelligence-blindness-and-the-digital-acequia.md](intelligence-blindness-and-the-digital-acequia.md), [wellness-structures-as-dissipative-structures.md](wellness-structures-as-dissipative-structures.md), [intermediation-and-selective-power.md](intermediation-and-selective-power.md), [santa-fe-rfei-data-platform.md](santa-fe-rfei-data-platform.md) (the 2019 Redfish proposal that named the architecture six years before COVID).
## The premise The information needed to produce a real-time, neighborhood-resolution heat map of the I compartment of SIR during COVID **existed**. Three streams together would have produced it: 1. **GPS location data** on every smartphone — held by Apple, Google, and the cellular carriers as the corporate gatekeepers of the "information collectors" stack. 2. **Confirmed-positive test results** — held by state health departments as institutional gatekeepers. 3. **Voluntary symptom reports, wastewater monitoring, syndromic surveillance** — held by various actors in fragmented form. What was missing was not data. What was missing was the *substrate that would let communities merge those streams into actionable heat maps without surrendering individual data to anyone*. Apple and Google shipped Exposure Notifications (ENS) — a privacy-preserving system that operated at bluetooth-proximity scale. Useful for individuals; useless for population-scale visibility. ENS chose **privacy-by-elimination**: don't collect the data, so it can't be misused. The acequia substrate would have chosen **privacy-by-architecture**: collect the data locally, aggregate it without ever centralizing it, produce community-scale visibility while individual locations never leave the phone unencrypted. That second path was already feasible in March 2020. It was not taken.
## What you would see The substrate is described from the perspective of *what becomes perceivable*, not how the underlying protocols work. Per Stephen's 2026-06-04 redirect, the framing stays at the level of direct affordances and information flow, not cryptographic primitives. Every phone runs a software agent on behalf of its user — a parciante in the agent-at-a-URI sense already developed in this bead's framework. The agent has access to the phone's GPS history (stored locally), the user's tested status if positive, the user's opt-in symptom reports. The agent does not send raw data anywhere. It exchanges aggregated, spatially- and temporally-binned signals with neighboring agents in a peer-to-peer mesh. What the user sees on their phone: - A **heat map of the I-front** at neighborhood resolution, updated continuously. The map shows where active transmission is clustered, the direction the front is moving, and the rate at which it is spreading or receding. - **Adjacent layers of context**: hospital capacity by region, vaccine availability by clinic, wastewater viral-load trends per sewershed, confirmed-test-positivity by ZIP code. Each layer is a different signal stream feeding the same composite picture. - **Personalized situational awareness** for vulnerable individuals: their neighborhood's current front position relative to their risk profile, the mutual-aid resources available in their area, the safest routes for essential trips. - **Institutional visibility** in the opposite direction: which hospital is at capacity, which county is under-testing, which state agency's response time is lagging. The watching flows from below to above as much as from above to below. What no one sees, anywhere in the system: - Any individual's GPS history - Any individual's tested status - Any individual's symptom reports The aggregation happens in such a way that population-scale patterns emerge without individual-scale exposure. This is the operational meaning of sousveillance applied to epidemic dynamics: the community sees the conditions and the institutional response; the individuals' raw data never leave their devices.
## What the corporate firewalls actually firewalled The substrate disintermediates several actors who currently hold gatekeeping power over pandemic information: **Apple and Google.** The phone OS vendors currently own the GPS data and gate access. ENS was their preferred answer because it did not require them to share that data with anyone — including the communities whose lives the data describes. A browser-based acequia substrate runs *on top of* the phone OS, treating the OS as commodity hardware. The GPS data stays on the phone (where it already is), but the agent that uses it is community-trusted rather than Apple-or-Google-controlled. This is the [distributed-origin architecture](https://redfish.acequia.io/) pattern — many local nodes serving the same logical paths, with corporate platforms reduced to commodity substrate. The disintermediation is *exactly what platform monopolies cannot permit*. ENS was the maximum disclosure Apple and Google were willing to make. A community-controlled substrate that aggregates GPS into population heat maps would have demonstrated that the corporate firewalls are not protecting users — they are protecting platform incumbency. **The CDC and state health departments.** The institutional gatekeepers currently hold confirmed-test data and choose what to publish, how to publish, and at what cadence. In the acequia model, the CDC's data becomes *one signal feed among many*, not the authoritative source. Communities aggregate the official feed alongside their own observations. The CDC's role shifts from gatekeeper to participant — exactly the Ostrom Principle 7 move (recognition of community rights to organize) applied to public-health intelligence. **The federal contractors who built the centralized dashboards.** Palantir, Deloitte, Accenture, and similar firms received hundreds of millions of dollars in 2020–21 federal contracts to centralize and visualize pandemic data. Their dashboards informed federal and state decision-making but were largely inaccessible to communities and to the actual sub-county scale at which mutual-aid networks operate. The acequia substrate makes those contractors largely unnecessary — communities run their own aggregation, on their own substrate, at the resolution they need. The federal investment in centralized pandemic-data contracting between 2020 and 2023 is estimated in the low billions of dollars. A community-substrate alternative built on open-source technology (AgentScript, Acequia, the existing decentralized-protocol stack) would have been in the low tens of millions. The intermediation is *expensive* and produces worse outcomes than the substrate would have.
## Sousveillance vs surveillance Surveillance and sousveillance are not "more vs less" of the same thing. They are structurally different: - **Surveillance**: a small number of actors (state, corporate) watch a large number of people from above. The information flows up to the watchers and is rarely reciprocated. Individual locations are exposed; institutional flows are opaque. - **Sousveillance**: the community watches *the conditions and the institutions* from below. Heat maps of the I-front are visible; aggregated infection dynamics are visible; institutional response is visible. Individual locations are never exposed. The acequia substrate inverts the COVID-era information topology. Under surveillance, the CDC, the state, and Apple/Google knew (in principle) who was positive, where they were, and who they were near. Under sousveillance, the community knows where the front is, how fast it is moving, which neighborhoods are vulnerable, and what the institutional response is doing — and nobody knows any individual's location. This is the operational meaning of Gibson's direct-perception framing from [wellness-structures-as-dissipative-structures.md](wellness-structures-as-dissipative-structures.md), applied to epidemic dynamics. The community **directly perceives the affordances** the front presents: where to act, where to shield, where mutual-aid networks need to form to protect vulnerable S sub-populations.
## The differential-protection corollary Stephen's Socratic probe from the prior turn: how does a "just sub-critical with frequent small events" regime hold for the genuinely vulnerable — the immunocompromised, the elder one fall from frailty, the household one shock from homelessness? The substrate's answer: **differential protection informed by population-level visibility.** A vulnerable individual sees the heat map. They see the front. They adjust their behavior — they shield, they ask for grocery delivery, they delay travel, they coordinate with their mutual-aid acequia for the resources that let them shelter — *because they can see the front coming*. This is the opposite of the universal-suppression / no-visibility regime we actually lived through during COVID. In the actual COVID, vulnerable individuals had no granular information. They got the same federal guidance everyone else got. They could either lock down maximally (with all the costs of isolation) or trust the public-health apparatus to keep transmission low enough (with all the costs of that bet failing). Under the acequia substrate, vulnerable individuals make *informed local decisions*. The mutual-aid networks (the dissipative structures from [wellness-structures-as-dissipative-structures.md](wellness-structures-as-dissipative-structures.md)) form around them. The population accepts low-level transmission to maintain criticality (avoiding the Bak-suppression antipattern from the prior Socratic turn — fuel build-up by another name). The vulnerable shield. The substrate makes both possible simultaneously, which the COVID-era infrastructure did not.
## What this would have looked like in March 2020 A counterfactual sketch, presented honestly as such: **Week 1 (early March 2020).** The CDC publishes daily confirmed-positive counts by state and selected counties. Wastewater monitoring is sparse but underway in a few cities. The Acequia substrate, already deployed via browser per the 2019 RFEI architecture, activates a public-health overlay. Citizens opt in via a simple browser visit to a community-trusted URL — no app install. Their phone's GPS history begins feeding the local agent; the agent contributes spatially-binned aggregates to the peer-to-peer mesh. **Week 2.** Heat maps render on every participating phone. Early signals emerge: certain neighborhoods show clustering; others show none. The maps are messy but actionable. Mutual-aid networks self-organize where vulnerable populations are adjacent to active clusters. Local school districts have data to inform closure decisions at ZIP-code resolution rather than statewide. **Week 3.** The CDC's data feeds into the substrate as one source. State health departments begin publishing more granular data because the substrate is rendering their existing aggregates anyway, and the bureaucratic incentive to provide *only* official numbers erodes when the substrate is producing better numbers. Apple and Google notice that ENS is not being adopted; they consider integrating their own GPS data into the substrate but are blocked by their commitment to ENS. The substrate continues to grow via browser-based adoption that bypasses the OS-level gatekeepers. **Month 2.** Population-level heat maps are standard. Vulnerable sub-populations have personalized situational awareness via opt-in alerting. Mutual-aid networks dissipate need gradients (grocery delivery, prescription pickup, child-care swaps) at neighborhood scale. The dissipative structures form because the gradients are visible and the substrate makes action possible. **Year 1.** Excess mortality is meaningfully lower than the historical actual. The vulnerable get protected by visibility; the population maintains adaptive behavior informed by real conditions; the centralized public-health apparatus continues to operate as one signal source among many. The intermediation pattern has been *partially* broken in one domain. The demonstration matters for what follows.
## Why it didn't happen The substrate was technically feasible. The substrate was already proposed (the 2019 RFEI). The failure mode was institutional: - **Apple and Google** chose privacy-by-elimination over privacy-by-architecture because the latter required them to *share architecture* with non-corporate actors. Sharing architecture is incompatible with platform-monopoly incentives. - **The CDC and state health departments** chose centralized control over distributed visibility because the latter required them to *cede informational authority* to communities. Ceding authority is incompatible with the bureaucratic gatekeeping pattern this bead has been documenting across domains. - **The federal contractors** who built the centralized dashboards had no incentive to propose the substrate alternative because their revenue model depends on the centralization being maintained. - **No institutional sponsor at the federal level** was willing to fund the substrate during the emergency window because doing so would have implicitly criticized the centralized apparatus that was simultaneously being funded. - The **2019 Redfish RFEI for exactly this kind of substrate**, submitted to the City of Santa Fe weeks before COVID arrived, was unfunded — and the same pattern played out at federal scale during the pandemic. The proposal that would have addressed the failure mode was blocked by the failure mode it addressed. The failure was not a technology failure. The failure was a *failure of institutional imagination by the gatekeepers*, compounded by a *failure of institutional sponsorship by anyone willing to break the gatekeeping*.
## What this means for the Groundworks proposal The epidemic-sousveillance frame is not a tangent. It is the strongest evidence in this bead's framework so far that: 1. **The intelligence substrate works at population scale.** The architecture exists, the demonstration was waiting, the gating constraint was institutional rather than technical. 2. **The intermediation pattern is consistent across domains.** Wildfire, social services, public-health emergencies all show the same gatekeeping failure mode. The Groundworks proposal is part of a family of attempts to break it. 3. **Communities self-organize when they can see.** The mutual-aid networks that formed during COVID did so blind. The substrate would have let them operate informed. The dissipative structures form regardless; the question is whether they form *fast enough and at the right scale* to relieve the gradient before catastrophe. 4. **The Bak/Prigogine synthesis is empirically grounded.** Populations did self-organize toward criticality in spite of, not because of, civic infrastructure. The infrastructure deficit is what produced the worst outcomes — not the absence of the structures, but the absence of the visibility the structures needed. For the published [acequia.org/groundworks/](https://acequia.org/groundworks/) proposal: the COVID case is the strongest single example that no Groundworks board member can dismiss as edge-case. The three-year blindness of COVID was *theirs*. They watched the centralized system fail in their own neighborhoods. A short paragraph in §The intervention or §Why this matters now naming this — *the substrate that would have helped during COVID is what we are now proposing for social services* — would do real work. The board's question becomes a direct one. The substrate exists. The intermediation pattern that blocked its deployment at federal scale is the same pattern blocking community self-organization in NM social services every day. The Groundworks proposal asks for a different institutional sponsor — the meta-nonprofit, the NM social-sector network — to do at state scale, for a different domain, what the federal government would not do at national scale for public-health emergency response. Do they want to be the institutional sponsor that breaks the pattern in NM, or another instance of the pattern?
## Notes - This document does not name cryptographic primitives (differential privacy, secure multi-party computation, federated learning) per Stephen's 2026-06-04 redirect. Those primitives are how the substrate's privacy guarantees would actually be implemented. The substrate framing here is at the level the board needs — what it does and why the alternative architecture is feasible. Technical detail belongs in a separate document for a different audience (probably a funders' technical-architecture brief). - The "Apple and Google as corporate firewalls" language is Stephen's; it should remain attributed if any of this prose moves into board-facing material. It is a sharper diagnosis than the standard "platform monopoly" framing because it names *what specifically is being firewalled*: the population-scale aggregation that would let communities see what is happening on the substrates they themselves are running on. - The counterfactual sketch in §What this would have looked like in March 2020 is presented honestly as a counterfactual. It is not a claim about what would have happened with certainty; it is a structured argument about what the substrate makes possible. Boards generally read counterfactuals as too speculative for board documents; this section is for the bead's thinking-in-public role, not for the published proposal. A condensed version could land in a footer or sidebar of the proposal but should not occupy main text. - Future work: this document is a *frame*, not an implementation plan. A companion technical-architecture document would describe the actual peer-to-peer protocol, the differential-privacy guarantees, the integration of public data feeds, the threat model, and the failure modes. That document would be useful for funders interested in the substrate as infrastructure investment. - The connection to indigenous cultural-burning practice (from the prior Socratic exchange — the parallel between acequia governance and Pueblo/Hispano cultural-burn fire ecology, both suppressed by centralized institutions, both rediscovered as durable solutions) is worth surfacing in a separate artifact. It is part of a *family* of continuous-low-level-dissipation traditions whose recovery is the subtext of this whole framework. Naming the family explicitly would strengthen the Groundworks proposal's claim that the acequia move is not isolated but part of a deeper recovery of indigenous-derived ecological intelligence.