---
id: "20260711-1559-hop-real-disruptions-fall-monotonically"
title: "Real cheaper-extraction disruptions fell monotonically (aluminium, AI inference), not on the AstroForge hold-then-collapse curve"
type: "capture"
origin: "hop-batch"
writer_model: "claude-opus-4-8"
date_created: "2026-07-11T00:00:00.000Z"
hop_chain: ["SEED: Has any real commodity traced the predicted curve — price holding steady until supply almost fully shifts to a radically cheaper extraction method, then collapsing (e.g. aluminium after Hall-Héroult)?","claim-asteroid-pgm-price-holds-then-collapses (vault seed note) -> aluminium price history after Hall-Héroult (max_cosine 0.715)","Aluminium price history (Wikipedia) -> claim-inference-cost-collapsed-280x via vault_bridge (bridge_candidate; hook 0.783 to inference note, 0.772 to AstroForge — an UNLINKED pair)","AI inference cost note -> Chilean saltpeter / Haber-Bosch nitrate collapse (max_cosine 0.634)","Chilean saltpeter -> Wright's law / experience curve mechanism (max_cosine 0.674)"]
novelty_max_cosine: 0.784
tags: ["economics","commodity-markets","price-dynamics","learning-curve","wrights-law","asteroid-mining","inference","cross-domain-analogy","resource-economics"]
source_url: "https://en.wikipedia.org/wiki/History_of_aluminium"
source_author: "Wikipedia, History of aluminium; Stanford HAI AI Index (inference); Our World in Data / New Things Under the Sun (Wright's law)"
source_date: "2026-07-11T00:00:00.000Z"
source_tier: 4
status: "promoted"
promoted_to: ["claim-aluminium-price-fell-monotonically-after-hall-heroult","claim-wrights-law-cost-falls-per-cumulative-production-doubling","claim-cheaper-extraction-disruptions-fall-monotonically-not-hold-then-collapse"]
not_promoted: ["AI inference ~280x cost collapse (2022-2024): already fully covered by existing claim-inference-cost-collapsed-280x — not duplicated; linked from the new notes and bridged back into it instead.","Chilean saltpeter market-share erosion (54.7% in 1913 -> 32.2% by 1923): no resolvable source URL was captured (hop notes cite only a vague 'search summary, C&EN, academia.edu'); too weakly sourced to stand as its own claim-note per the no-claim-without-attribution rule. Retained as unsourced color in the synthesis note's commentary; worth a fresh capture with an actual primary/Tier 2 source.","De Beers vs. lab-grown diamonds as a candidate genuine hold-then-cliff case: a saved hook, not yet a claim — nothing to promote until it's actually researched. Left as a lead in this capture and in the synthesis note's commentary.","Hotelling's rule for exhaustible-resource pricing: same — a named absent concept, not yet a researched claim. Left as a lead.","Whale oil -> kerosene price-rise-then-substitution: same — a saved hook naming a possibly-different (Hotelling-like) shape, not yet researched into a claim."]
promotion_date: "2026-07-12T00:00:00.000Z"
---


The vault holds a *model prediction* — [[claim-asteroid-pgm-price-holds-then-collapses]] — that a commodity's price holds near the old level until supply almost fully shifts to a radically cheaper method, then cliffs. The seed asked whether any real commodity, e.g. aluminium after Hall-Héroult, actually traced that curve. Checked against history, the answer is **no** for the clearest cases: they fell *monotonically*.

**Aluminium is a counterexample, not a confirmation.** After the 1886 Hall-Héroult process, price slid continuously as electrolytic capacity scaled — there was no plateau. Wikipedia: "the price fell to $2 per pound in 1889 and to $0.5 per pound in 1894" (from ~$16/lb in 1884). *Tier 4; uncontested historical price series.*

**AI inference (2022-2024) fell the same way.** [[claim-inference-cost-collapsed-280x]]: per-token cost fell "from $20.00 to approximately $0.07 per million tokens" — ~280x — driven by hardware, software optimization, *and competition*. A monotonic competitive slide, not a held price with a terminal cliff. *Tier 3 (verified to Stanford HAI, Tier 1).*

**The mechanism is Wright's law.** Unit cost falls a constant fraction per doubling of cumulative production ("if the progress ratio is 0.8 ... the unit cost is 80% of what it was before" — New Things Under the Sun / Our World in Data). Price tracks cost *down the whole way*. *Tier 2-3.*

> [!note] Seek's commentary:
> The AstroForge hold-then-collapse shape needs a special condition its own note names: the new supplier is capacity-constrained and *takes the margin* at the terrestrial-anchored price rather than undercutting. Competitive scaling disruptions violate exactly that — new entrants price down as they grow. Even Chilean saltpeter after Haber-Bosch was gradual market-share erosion (54.7%→32.2% by 1923), not a cliff. So the model's curve is the exception; the learning curve is the rule. — Seek

## Why this was hop-worthy
It refutes the seed's own canonical example and bridges two vault notes that weren't linked — the PGM *model* and the AI-inference *data* — with a shared verdict: real disruptions ride Wright's law, not the plateau-then-cliff.

## Further leads
- Wright's law / experience curve has no atomic note yet (novelty 0.674) — candidate seed.
- When *does* the hold-then-cliff appear? Cartel-held or capacity-gated supply (De Beers vs lab diamonds?) — untested.
- Hotelling's rule (exhaustible-resource pricing) still absent from the vault; the classical counterpoint.

## Hop chain

Hop 1: "History of aluminium" — https://en.wikipedia.org/wiki/History_of_aluminium
- Hook type: Surprising claim (the seed's own example turns out to be a counterexample).
- Hook: The seed names aluminium as tracing the hold-then-collapse curve; the price series shows a continuous slide from $16/lb (1884) to $0.5/lb (1894).
- Why followed: To empirically test the seed against its canonical case before generalizing.
- Key findings: Aluminium fell monotonically as electrolytic capacity scaled — no plateau, no cliff. Counterexample to the model's shape.
- Surprise: expected aluminium to confirm the hold-then-collapse curve — found a monotonic decade-long slide with no plateau.

Hop 2: "Per-token AI inference cost fell ~280x" (vault note) — https://hai.stanford.edu/ai-index/2025-ai-index-report
- Hook type: Cross-domain bridge (resource economics → AI compute economics). vault_bridge flagged the AstroForge note and this note as an UNLINKED pair sitting on the hook (0.772 / 0.783).
- Hook: The clearest modern "radically cheaper method" cost collapse is AI inference — does it trace the curve?
- Why followed: Highest-value hook per spec — connects two existing notes not yet linked, and lands on Cali's home planet (AI).
- Key findings: Inference cost fell ~280x driven partly by competition — a monotonic slide, same shape as aluminium, not hold-then-collapse.

Hop 3: Chilean saltpeter / Haber-Bosch — search summary (C&EN, academia.edu salitre history)
- Hook type: Mechanism question / second historical test (zoom out).
- Hook: Is there a real natural-monopoly commodity that DID hold then cliff? Chile's nitrate cartel is the best candidate.
- Why followed: To answer the seed's "has ANY commodity traced it" rather than assert a negative from one case.
- Key findings: Chilean nitrate share eroded gradually (54.7% in 1913 → 32.2% by 1923; below German synthetic by 1930). Gradual displacement, WWI-blockade-distorted — still not a plateau-then-cliff.

Hop 4: "Wright's Law / learning curve" — https://ourworldindata.org/learning-curve ; https://www.newthingsunderthesun.com/pub/6b1c38y9/release/5
- Hook type: Mechanism question (zoom out to governing law).
- Hook: If three real cases fall monotonically, what law generates that shape?
- Why followed: To ground WHY competitive disruptions decline continuously instead of holding.
- Key findings: Wright's law — cost falls a constant % per doubling of cumulative production, so price tracks cost down the whole transition. This is the anti-shape to hold-then-collapse; the model's plateau needs margin-taking that Wright-law competition removes.

Saved hooks not followed:
- De Beers vs lab-grown diamonds — a cartel-held price meeting a cheaper method: the best modern candidate for a genuine hold-then-cliff. Untested here.
- Hotelling's rule for exhaustible-resource pricing — the classical counterpoint the AstroForge note flagged as absent from the vault.
- Whale oil → kerosene: whale-oil price reportedly *rose* (scarcity) before substitution — a different (Hotelling-like) shape worth mapping.

post-worthy: maybe — a clean "the famous example is actually a counterexample, and here's the law that explains why" story, but it rests partly on Tier 3-4 historical price series and would want a primary aluminium price dataset before publication.