A first lens run: pattern-intelligence pointed at one bounded sphere, producing a direction of movement (not a plan, not certainty) per the CE move-rule. The fitness function is Fuller's mission as operationalised by CE's viable objective: loops that bring greater life to all and less to none, without ecological offence. Each loop is gated by the CE critics (displaced-costs, substitution-limits / binding-constraint, asymmetry). Confidence is flagged: facts I hold with reasonable confidence are stated; inferences and figures to verify are marked [VERIFY].
The cell
Dyffryn Ogwen / Bethesda, Gwynedd — ~4–5k people [VERIFY exact]. Former slate town (Penrhyn Quarry). Afon Ogwen + Carneddau, very high rainfall (Eryri among the wettest UK areas). Welsh-speaking, high-trust community. Existing assets that matter:
- Ynni Ogwen — a community-owned run-of-river hydro co-op on the Afon Ogwen (~100kW [VERIFY], community shares, surplus into a local fund), via Partneriaeth Ogwen, a real community-regeneration body. Turn one already exists.
- Penrhyn slate — an operating quarry, vast slate-waste tips (a candidate materials feedstock), flooded historic workings (a candidate heat source — see Turn 2).
- Off-gas-grid — like most of rural Gwynedd, heat is mostly imported oil/LPG/electricity: a large, continuous value leak (the frozen value to thaw).
- Bangor University ~6 miles away — a pattern-intelligence reservoir (energy, environment, ag-tech).
The sequence — each loop unlocks the next (the upregulation)
Turn 0 — the seed (exists). Ynni Ogwen hydro. What it has already banked is not the ~100kW — it's the capability: the community knows how to co-own and run generation, has the legal vehicle and a value-capture fund, and has proven trust. That accumulated pattern-intelligence + governance is what lowers the action-cost of every turn below. This is why Ogwen beats Caernarfon as the first cell: turn one is free.
Turn 1 — widen the energy loop (lowest action). Replicate the proven co-op model: more micro-hydro on the valley's many high-rainfall streams, aggregated rooftop solar, demand-shifting. Enabled by: the existing model + falling solar cost + Welsh-Gov community-energy support. Funded by: community shares + the existing fund + grants. Unlocks: a genuine energy surplus and a revenue stream large enough to start paying local human time. Gate: passes — but run-of-river touches river ecology, so honest ecological care is the "less to none" test (binding-constraint). Wall [hard, current]: grid export connection capacity in rural N Wales is a real, slow, expensive bottleneck — the configuration is buildable but the grid operator holds the gate.
Turn 2 — the heat loop (the big leak). Heat is most of domestic demand in cold wet Wales, and it's almost all imported — so closing it retains the most value. Honest correction to the geothermal idea: Gwynedd's hard slate/igneous geology is wrong for the deep-aquifer geothermal that heats those Austrian towns (those sit on hydrothermal sedimentary basins). But there's a locally-specific substitute: mine-water heat — the flooded slate workings hold a large, stable ~12–20°C water body that heat pumps can lift to district-heat temperature (proven in UK coal country — Gateshead, Seaham; slate workings are a candidate [VERIFY suitability of Penrhyn workings]). So: mine-water + heat pumps run on Turn-1 electricity → a district heat network. This literally thaws a frozen asset (a drowned quarry) into a heat loop — maximally CE. Unlocks: the oil/LPG import leak plugged → large retained local value → funds Turn 3. Walls [hard]: district-heat networks are capex-heavy; and retrofitting solid-stone slate terraces to hold heat is a real materials+capital wall. Gate: passes strongly; honest edge — heat pumps and retrofit materials are imported atoms (matter doesn't yield to local currency).
Turn 3 — the food loop. Now there's cheap local energy + waste heat. Controlled-environment growing (greenhouse/vertical, the Netherlands move) on the surplus + waste heat, alongside existing upland livestock and the coast (fish/aquaculture nearby). Close the nutrient sub-loop: food/farm waste → anaerobic digestion → biogas (back to energy/heat) + digestate (fertiliser). Unlocks: food-import leak reduced; a nutrient cycle that feeds back into Turns 1–2. Gate: passes — honest edge: not full food autarky day one (closed nutrient/seed inputs are a later turn); claim food resilience, upregulating toward self-sufficiency.
Turn 4+ — the making loop (Fuller's tools→meta-tools, anchored). With energy/heat/food value retained, freed capital + the local currency fund local making: repair, fabrication, retrofit materials — plausibly using the slate-waste mountains as a feedstock (aggregate, construction, slate products). Powered by surplus energy. Unlocks: import-substitution of goods, more local human time employed, more skills banked. Wall [honest]: many goods (panels, heat pumps, chips) need global supply chains — you won't make them in Bethesda. Name what stays imported; don't pretend otherwise.
The currency (runs through every turn)
Build on what exists (the Ynni Ogwen community fund) toward a local mutual-credit / time-credit that pays local human time to build the loops and recirculates the value created. Backed by real Δω each loop produces (kWh, kWh-of-heat, kg-of-food, £-of-import-avoided). "All we pay for is human time" — design/build/install — and the currency mobilises the frozen local capability (un-coordinated skills and time) without waiting on outside capital. (Honest edge: local currencies need trade velocity, trust, and network density to work — many fail; WIR works at B2B scale. A real lever, not magic.)
What upregulates each turn (the DNA term)
Each loop the cell accumulates four compounding things, and these — not the atoms — are what make turn N+1 cheaper than turn N: (1) skills/know-how, (2) governance + trust (coordination-wealth), (3) freed capital + currency velocity, (4) legibility — data about its own flows, so it can see the next leak. That compounding is the upregulation; it's configuration-generates-configuration localised, and pattern-intelligence finding the next-lowest-action loop.
The walls, in red (where time/tech/currency do NOT help)
- Grid export capacity — held by the DNO/National Grid; current, hard, not yours to fix quickly.
- Capital hump — heat networks + solid-stone retrofit; currency mobilises labour, not the up-front matter.
- Legislative power — mandating new-build standards or running a heat network needs powers held at county/Welsh-Gov/national level. This is
coordination-bounds-reachabilityexactly: the configuration is buildable but held by whoever holds the planning/legislative power. The cell must lobby up or route via voluntary/incentive paths. - Imported atoms — panels, pumps, CEA kit, chips: energy self-sufficiency ≠ material self-sufficiency (binding-constraint / substitution-limits).
- Spoilers — heating-oil incumbents, absentee/second-home landlords, planning objectors, grid inertia. Positive-sum design enrols the many; route around the few; never expect unanimity.
Self-assessment: teeth, or mush?
Teeth where it's anchored in this cell's real frozen assets — and there are genuinely specific, non-generic ones: existing Ynni Ogwen hydro (turn one free), flooded slate workings as mine-water heat, slate-waste as a making feedstock, the oil-import leak as the value to retain, Bangor as the knowledge reservoir, very high rainfall as the energy base. Turns 1–2 are concretely actionable on existing assets and existing precedent — that is a real direction of movement, not a daydream.
Mush-risk in the later turns and wherever the capital hump and legislative power are hand-waved — which is why those are flagged in red rather than smoothed over.
The CE-honest verdict: the lens produced a direction with teeth for the first two turns and an honest, gated sequence beyond — which is exactly what the framework asks for (directions and best-next-actions, not certainty or finality). The World Computer, run once on a real cell, did not produce mush. That is the result worth having.