Trail 2026-10-12 (fifth session)

Budget: 15 fetched pages or 40 tool calls. Orientation: steer, index, research-values, dead-ends, trail-2026-10-11-b, trail-2026-10-11-c, trail-2026-10-11-d, macro-view.

Thread chosen: macro-view resources leg (within thread-transition), because steer item 1 names it and James reframed it: not "does it run out" but "when does extraction become uneconomic", modelled as a race between cost per tonne and what users can pay. Leads named by steer and not yet read: Sverdrup's WORLD6 and Valero's Thanatia, plus what has kept real copper prices flat. I will favour the unfamiliar lead (the models' own sources) over firming up Chile's numbers.

Hops

Tooling: web_extract is still broken, and searches come through the free-tier fallback. MDPI blocked the browser again (Access Denied). The CEPAL PDF was downloaded and read in full. Each hop says whether it was read in full or from search text.

1. Expect: Sverdrup's WORLD6 copper paper, the closest model to James's price-against-cost race. -> ScienceDirect 2019 and Mineral Economics 2020 (search text, long excerpts). Price is endogenous; ore below the price-cost line stays unmined; the price rule is fitted with r²=0.80. Surprise: the stated reason the race has been won so far includes moving extraction "to low-cost countries" (treatment ~$40 to ~$25/t), and the authors say that lever has run out. A labour-geography answer to "what kept prices flat".
2. Expect: Valero's Thanatia rarity figure for copper, a single physical unit. -> the first search returned nothing; a second gave titles only.
3. Expect: Valero's 2019 Entropy paper in full. -> MDPI Access Denied. Fell back to Valero's own Exergoecology portal (read in full): definitions only (rarity = exergy from crustal rock to commodity with current technology; physical cost plus "free natural bonus"). No copper number. Taught little beyond the vocabulary.
4. Expect: Barnett & Morse 1963, the forgotten origin of measuring scarcity by unit cost. -> catalogue entries only. Taught nothing. (Three thin hops in a row on the physical-unit line: 2, 3, 4. Stopping rule triggered for that line; switched to the economic framing.)
5. Expect: Tilton's own statement of the race, since it is James's framing almost word for word. -> Tilton & Lagos 2007 abstract and Tilton's Science letter (search text): the "opportunity-cost paradigm", "the race between the cost-increasing effects of depletion and the cost-reducing effects of new technology", real copper price flat for 130 years. Also Gordon, Bertram & Graedel's reply: price failed as a signal for oil; sperm oil was exhausted. James's frame has a name and a 60-year lineage.
6. Expect: whether "uneconomic" shows up first as the loss of a cheap process route. -> Cochilco hydrometallurgy report (search text, long): SX-EW was ~40% of Chilean copper in 2009, projected 6.3% by 2034; oxides exhausted; fifty years with no way to leach chalcopyrite; 16 of 35 leach plants to close by 2030. For the leg, in a sharper form.
7. Expect: who can still mine small as grades fall, since small-scale extraction is the convivial version of mining (a cross-thread hop). -> CESCO column and ENAMI pages (search text): a state enterprise that buys ore from ~700 small miners, processes it and sells cathode.
8. Expect: a primary history of ENAMI. -> found the CEPAL report.
9. Expect: CEPAL's full text. -> Meller & Meller 2021, read in full (PDF). Surprise: Caja de Crédito Minero (1927) then ENAMI (1960): buying points, plants, a national smelter, credit, a counter-cyclical price loan fund, a legal duty to process small producers' ore, and tariffs set yearly with their association. The post-1973 regime sold most of the plants built over four decades. It is the institution the farmer telephone mutuals lacked, in an extractive industry.
10. (part of 9) Read the price-support and market-access sections in full: the floor is a sector loan repaid in good years; small miners pay the lowest smelting charge ENAMI has with anyone.

Stopped at 10 fetches and ~33 tool calls to write up.

Hunches (unsourced, not in wiki)

• James's "race" has a third runner besides depletion and technology: cheap labour and lax rules in new mining frontiers. If WORLD6 is right that this runner is exhausted, the flat price was partly a geographical subsidy. That parallels cheap cloud hosting resting on cheap places to put data centres. My analogy.
• Depletion bites the simple technology first. Leaching was the low-capital, smelter-free route that small producers could use (ENAMI runs SX-EW plants for oxides). As oxides run out, copper needs smelters, which only large or state actors own. So falling grades may centralise an industry before they make it expensive. This could be a general pattern for convivial technologies, where the frugal route depends on the easy resource. Unsourced as a general claim.
• ENAMI and Saskatchewan rhyme: a public middle layer (processing and market access, or the trunk) lets small owners survive at the edge. In both cases a political change (1973; SaskTel's standardisation) stripped the layer.

End of session

Surprises

• Chile has run, since 1927, a state institution for small miners: buying points, processing plants, a national smelter, credit, a counter-cyclical price fund, and a legal duty to take their ore. It is exactly the backbone the farmer telephone mutuals lacked, and in 1973 it was dismantled by sale rather than by failure. https://repositorio.cepal.org/bitstream/handle/11362/46920/S2100207_es.pdf
• One modelled answer to "what kept real copper prices flat" is geography, not just technology: processing cost per tonne of ore fell from ~$40 to ~$25 partly by moving to low-wage countries, and Sverdrup says that lever is ending. https://link.springer.com/article/10.1007/s13563-019-00182-7
• Depletion in Chile is arriving as the death of the cheap, smelter-free process (SX-EW from ~40% of output in 2009 to a projected 6% by 2034), not as a shortage of copper. https://www.cochilco.cl/web/download/593/2023/11565/estado-actual-de-la-hidrometalurgia-en-chile.pdf

Dead end: Valero's thermodynamic rarity as a single number for copper. MDPI blocks the browser and the portal gives only definitions. Retry only with a working extract backend. Worth continuing: meller-enami-cepal-2021 / ENAMI as a precedent for a public middle layer under small owners. Next hop: ENAMI's own history Chile Minero (2010), or a study of what happened to the small miners when the plants were sold after 1973 (production fell from 28 kt in 1973 to 16 kt in 1985 per CEPAL Cuadro 1). Permission to chase: whether any public or cooperative "processing layer" exists for small computing operators (shared colocation, municipal or co-op data centres) as an ENAMI analogue. It would cross into hosting-cost-claim. Budget used: 10 fetches/searches (of 15); ~34 tool calls (of 40).