# D005 Antithesis — The Spark Bridge Adds a Link to the Same Chain, Not a Channel to Google ## Counterclaim **The Spark Innovations finding does NOT change the verdict.** It deepens the developer-overlap pattern — a third [[Michael Montfort]]-managed Delaware shell — but it does not supply the one thing D003 found missing and the [[Altitude Capital, LLC]] standard requires: an *independent Tier-1 channel that names Google at Conway*. What the thesis offers instead is a chain whose strong link (Spark inside the Conway NDAs) and whose Google link (West Memphis) are joined only by Tier-2 registry coincidence and a single, vague Tier-3 sentence — and the Google end of that chain attaches to [[GROOT LLC]], not to Spark. Stacking a Tier-2 registry fact on a Tier-3 news association is not the multi-channel identification the corpus already decided is the floor; the [[Identifying the Unnamed Cooling-Water Data Centers]] precedent ("circumstantial lead ... is **not** an identification") controls, and the Conway principal remains unidentified. ## Attack on the thesis ### Against the thesis's `## Claim` The thesis claims the Spark finding is "a materially new evidentiary channel, not another instance of the corporate-structure pattern D003 already weighed." This is the load-bearing assertion and it is false. An evidentiary *channel*, in the sense D003 and [[Altitude Capital, LLC]] use the word, is an independent line of proof that *names the principal* — a deed chain, a water-use registration reading "Google Data Center," sworn CEO testimony naming Google's subsidiary, an APSC order conditioned on Google's requirement. The Spark finding names *no principal at all*. It is one more shell with the same manager, the same registered agent, and the same Wilmington mailbox — i.e., precisely "another instance of the corporate-structure pattern." The thesis tries to escape this by relabeling a third data point on the *same axis* (Montfort/CSC/PMB) as a *new axis*. Calling the West Memphis appearance a "destination" does not make Montfort's serial-organizer role into an independent proof of a common principal; it is the very dimension D003 already discounted, now with a third token on it. The thesis's own `## Claim` concedes the join it depends on is weak even as it asserts it: it says the bridge runs "Conway → Spark (Tier-1) → Montfort cluster (Tier-2) → confirmed-Google West Memphis (Tier-3 affiliation into a Tier-1-Google site)." Read that chain honestly: the only thing connecting the Conway end to the Google end is the *Tier-2 registry fact that Montfort manages both shells* plus a *Tier-3 "affiliate linked to the endeavor."* A chain is as strong as its weakest join, and here two of the three joins are the weakest tiers in the wiki's own hierarchy. That is not a channel that "approaches" Altitude Capital; it is a longer version of the Forgelight inference D003 already rejected. ### Against the thesis's `## Argument` — "the bridge is built from three independently verified joints" The thesis's architecture is three "joints." Take them in turn; each is either not what the thesis says it is, or is the discounted dimension wearing new clothes. **First joint ("Spark is inside the Conway project at Tier-1") proves the wrong proposition.** This joint is real and uncontested — but it establishes only that *Spark* is on the developer side at Conway, not that *Google* is. The thesis leverages it to defeat D003's "single inferential bridge" characterization by claiming "two independent documentary footholds inside one project" (Forgelight as MOU developer; Spark as NDA counterparty). But two Montfort shells inside one project is not two *independent* footholds — it is the same organizer placing two of his shells on the same deal, which is exactly what a serial formation agent or a developer using nested SPVs does as a matter of routine. The count of Montfort shells touching Conway is not the disputed quantity. The disputed quantity is the number of *channels that name the principal*, and that count is still zero at Conway. The Spark NDA, as the thesis's own `## Anticipated counterarguments` admits, names "a 'parent, subsidiaries and affiliates' that the record does not name" — so the single Tier-1 document the thesis leads with affirmatively *withholds* the principal on its face. **Second joint ("Spark is a third Montfort Delaware shell") is the discounted dimension, not a new one.** The thesis treats registry verification as if it added probative weight toward Google. It does not. D003 already held — and [[Spark Innovations, LLC]] itself restates — that "the shared-Montfort / shared-vendor registry pattern is **not, by itself, identification of the ultimate principal**." A third instance of a pattern that was non-probative at two instances is non-probative at three; multiplying a vendor-stack artifact does not transmute it into a principal. D003 expressly *bracketed* Montfort's professional role as an open question: he may be "an in-house corporate paralegal (which would strengthen the shared-principal inference), a Delaware formation-firm partner (which would weaken it materially — Montfort could organize unrelated shells for different clients on the same day), or a contracted registry technician." The thesis never resolves this bracket — and it *cannot*, on this record. If Montfort is a formation-firm organizer (the null hypothesis the registry data is fully consistent with), then three shells prove exactly nothing about a common beneficial owner, and the entire bridge collapses at its center. The thesis builds its strongest structural claim on a premise the corpus has flagged as unknown. **Third joint ("that same shell reaches a CONFIRMED-Google site") is a transitivity error and a misread of the source.** This is the thesis's decisive move and its weakest. The argument is: West Memphis is confirmed-Google (Tier-1, true); Spark appears at West Memphis (Tier-3); therefore Spark is a Google vehicle; therefore Spark's appearance at Conway imports Google to Conway. Two independent defects: 1. *The Google confirmation does not attach to Spark.* The constructionowners source — and [[Spark Innovations, LLC]]'s own reading of it — is explicit that "the article presents Spark Innovations as an earlier-in-time affiliate, distinct from the current [[GROOT LLC]] proposal." The Tier-1 Google confirmation (water-use registration, Entergy testimony, APSC order) attaches to the *GROOT* build, not to *Spark's* role in it. Spark is described as an *earlier-in-time* affiliate whose $3 billion plan predates and is distinct from the GROOT proposal that Google is confirmed behind. So even granting every join, what is confirmed-Google is GROOT; Spark's relationship to GROOT is itself unstated, and Spark may be a prior, abandoned, or unrelated affiliation that GROOT (or Google) later superseded. The thesis silently substitutes "Spark is at the confirmed-Google site" for the source's actual, far weaker "Spark was an earlier affiliate linked to the endeavor, distinct from GROOT." 2. *The transitivity is invalid even if the joins held.* "Google at West Memphis" + "Spark (Tier-3) at West Memphis" + "Spark (Tier-1) at Conway" does not yield "Google at Conway." A developer, formation agent, or SPV-stack that touches a Google project at site X does not thereby make *every* project it touches a Google project — that inference would make every CSC client and every Montfort shell a Google entity. The thesis needs Spark to be a *Google-exclusive* vehicle for the transitivity to run, and nothing in the record establishes exclusivity; the Tier-3 sentence establishes only a single, loose association at one site. The thesis's own framing betrays the gap: it writes that the chain reaches "confirmed-Google West Memphis (Tier-3 affiliation into a Tier-1-Google site)." The parenthetical concedes that the *affiliation itself is Tier-3* — only the *site's* Google status is Tier-1. Bolting a Tier-3 affiliation onto a Tier-1 fact about a *different entity* (GROOT) does not yield a Tier-1 identification of Spark's principal, let alone of Conway's. **On "Why this fills D003's gap and approaches the Altitude Capital standard."** The thesis claims the Montfort architecture "is no longer a closed loop of mutually-referencing shells." But it still is — Spark, Forgelight, and Willowbend reference each other through Montfort/CSC/PMB and nothing else. The single outward-pointing edge the thesis adds is the Tier-3 "affiliate linked to the endeavor" sentence, and that edge points to GROOT/West Memphis, not to Google directly, and describes a superseded earlier affiliate. The thesis admits the result "does not yet equal Altitude Capital's three-clean-Tier-1-channel showing" but says it "meaningfully approaches it." Restraint is not a sliding scale where "approaching" the bar licenses the conclusion. D003 and [[Identifying the Unnamed Cooling-Water Data Centers]] drew a *categorical* line: a circumstantial lead "is **not** an identification," full stop. "Closer to the line" is still on the restraint side of the line. The honest count of independent Tier-1 channels naming Google *at Conway* went from zero (D003) to zero (now). Nothing that leaves that count at zero can "fill D003's gap," because the gap D003 identified *was* that count. ### Against the thesis's `## Evidence` The thesis's evidence section is internally honest about tiers and, read closely, defeats the thesis. Its Tier-1 exhibit is the Spark NDA — which names Spark "for itself and its parent, subsidiaries and affiliates," i.e., a primary document that *declines to name the principal* and selects New York governing law that reveals nothing about beneficial ownership. Its Tier-2 exhibit is the SoS record — a registry fact about *Montfort*, CSC, and a Wilmington PMB, naming *no principal*. Its Tier-3 exhibit is the single constructionowners sentence, which the thesis quotes accurately ("an affiliate linked to the endeavor") and then immediately over-reads. At no point does any exhibit name Google in connection with Conway, or name Spark's principal at all. The thesis assembles three documents none of which, individually, names the principal, and asserts that their *sum* does — but the sum of three principal-silent documents is still principal-silent. The Altitude Capital exhibits, by contrast, each *independently name Google* ("Google's subsidiary, Altitude Capital, LLC"; the "Google Data Center" registration). That is the difference between a standard met and a standard gestured at. ### Against the thesis's `## Anticipated counterarguments` The thesis lists four counterarguments and defers them ("Acknowledged here only; the antithesis and synthesis will engage them"). This deferral is itself telling: the thesis concedes, in its own voice, the four facts that sink it — (1) the West Memphis link is Tier-3 trade press calling Spark merely "an affiliate linked to the endeavor"; (2) the shared-Montfort pattern "may still be a vendor-stack artifact"; (3) "the public Google confirmation attaches to GROOT LLC / Project Pyramid, not to Spark Innovations, and Spark is reported as an *earlier* affiliate distinct from the current GROOT proposal"; and (4) "the corpus still discloses no Spark principal on its face." Those are not peripheral caveats the synthesis can absorb while the claim survives. They are, respectively, the thesis's Tier-3 join, the collapse of its central Tier-2 join, the failure of its transitivity, and the silence of its Tier-1 exhibit. A claim that must pre-concede the weakness of every one of its three joints has not earned a verdict change; it has described exactly why restraint holds. ## Independent argument for the counterclaim Strip the dialectic to its governing rule. The wiki has a *controlling restraint precedent*, decided `confidence: low` by the same maintainer on the same investigation: [[Identifying the Unnamed Cooling-Water Data Centers]] holds that "neither unnamed project is identified here, and the circumstantial lead toward AVAIO / Project Leo for the Arup project is **not** an identification." D003 made that precedent binding for the Conway/Forgelight question: "Methodological coherence requires applying the cooling-water precedent here." The rule is categorical — circumstantial leads, however suggestive, are not identifications — and it does not contain a "but if the lead gets warmer" exception. The Spark finding makes the Conway lead *warmer*; it does not make it an *identification*. Under the controlling precedent, warmer-but-still-circumstantial resolves to *restraint*, not to naming. The affirmative standard the wiki uses for naming Google is [[Altitude Capital, LLC]], and it is exacting precisely because it is the worked example of a *correct* identification. There, Google is named only because **multiple independent Tier-1 channels each name Google**: the Entergy CEO's sworn testimony ("Google's subsidiary, Altitude Capital, LLC"); the APSC Cypress order conditioned on "Altitude Capital, LLC, a subsidiary of Alphabet, Inc (Google)"; the Department of Agriculture water-use registration reading "Google Data Center." Each channel is independent of the others and each, standing alone, names the principal. Against that standard, the Spark finding supplies **zero** channels that name Google at Conway. It supplies one Tier-1 document that *withholds* the principal (the NDA), one Tier-2 registry fact about an organizer (no principal), and one Tier-3 sentence about a superseded affiliate at a *different* site (no Conway principal). Three exhibits, zero principal-naming channels at Conway. The arithmetic is dispositive: the very standard D003 enshrined is not approached but *unmet in kind*, because not one of the new exhibits is the *kind* of thing — a channel that names the principal — the standard counts. There is also a methodological cost the thesis ignores. If a Tier-3 "affiliate linked to" sentence plus shared-organizer registry data sufficed to name a hyperscaler, the wiki would have to name a principal in [[Identifying the Unnamed Cooling-Water Data Centers]] too — the Arup/Project Leo lead there (scale + timing + Pulaski County geography) is *at least as strong* as the Spark chain, and the wiki expressly declined. Adopting the thesis would silently overrule the cooling-water precedent and import a "guilt-by-shared-vendor" rule that, applied evenly, would attribute every CSC-served, Montfort-organized shell in Arkansas to Google. The wiki's restraint standard exists to prevent exactly that laundering of inference into apparent fact (T003's "functionally identical to a positive attribution" concern). The correct, coherent output is the one [[Spark Innovations, LLC]] already reaches on its own face: document the strengthened *developer-overlap pattern* as a Tier-2 fact, flag the Tier-3 West Memphis association as a lead, and **decline to name the Conway principal** — which remains unidentified on this record. ## Evidence **Tier-1 — the Spark NDA withholds the principal (primary document; [[Conway Corporation Project Stratus NDAs]]).** The document names the developer-side counterparty but expressly leaves its principal unnamed: > A Mutual Non-Disclosure Agreement between "**Spark Innovations, LLC**, for itself and its parent, subsidiaries and affiliates ('Company')" and the Participant — Conway Corporation's [[Zach Gardner]] (signed 2025-10-14). **New York** governing law; five-year confidentiality term. The operative phrase is "its parent, subsidiaries and affiliates" — a primary record that, on its face, *declines to identify the parent*. This is the thesis's lead Tier-1 exhibit, and it names no principal. **Tier-2 — the SoS record names an organizer, not a principal ([spark-innovations-llc-sos-record.md](../../web%20archive/2026-05-29/sos-corp-search.ark.org/spark-innovations-llc-sos-record.md), retrieved 2026-05-29).** Verbatim, the registry discloses: > | Filing # | 811524872 | > | Filing Type | Foreign Limited Liability Company | > | Status | Good Standing | > | Reg. Agent | CORPORATION SERVICE COMPANY | > | Date Filed | 11/25/2024 | > | Officers | DAVID THOMAS, Incorporator/Organizer · MICHAEL _ MONTFORT, Manager | > | Foreign Address | 2801 CENTERVILLE ROAD, 1ST FLOOR, PMB 811, WILMINGTON, DE 19808 | > | State of Origin | DE | This is a Tier-2 registry fact about Montfort, CSC, and a Wilmington private mailbox. It names **no principal** and **no Google connection**. The page itself draws the restraint conclusion: "the shared-Montfort / shared-vendor registry pattern is **not, by itself, identification of the ultimate principal** (Montfort is a serial organizer using a common vendor stack)." The incorporator of record is DAVID THOMAS — "a name not otherwise in the corpus" — underscoring that the registry points *away from*, not toward, any identified beneficial owner. **Tier-3 — the West Memphis sentence describes an EARLIER, distinct affiliate, not Spark-at-the-Google-build ([constructionowners.com extract](../../web%20archive/2026-05-29/www.constructionowners.com/groot-spark-innovations-west-memphis-project-pyramid.md), retrieved 2026-05-29).** The single mention, verbatim: > "Last year, local media noted that **Spark Innovations LLC, an affiliate linked to the endeavor**, had plans for a roughly **$3 billion** investment targeting the **Bollinger Road area**." The article presents Spark Innovations as an earlier-in-time affiliate, distinct from the current [[GROOT LLC]] proposal. Two load-bearing facts the thesis elides: (a) the language is "**an affiliate linked to the endeavor**" — vague Tier-3 association, not a corporate-structure statement; and (b) the source frames Spark as "an **earlier-in-time affiliate, distinct from** the current [[GROOT LLC]] proposal." The Google confirmation attaches to GROOT, *not* to Spark. The archive's own Notes confirm the limit: the source "establishes only that the **name** 'Spark Innovations LLC' has been publicly associated with the West Memphis / [[GROOT LLC]] / Project Pyramid (Google-confirmed) project as an early affiliate. It does **not** establish the entity's corporate structure, and it does **not** itself connect Spark Innovations to Conway." **The standard being missed — [[Altitude Capital, LLC]] (Tier-1, multiple independent channels each naming Google).** For contrast, what a real identification looks like: > "Under the oversight of the APSC, EAL was able to develop an electric service agreement with **Google's subsidiary, Altitude Capital, LLC** ..." ([[Entergy CEO Direct Testimony]]) > "EAL states that Cypress Solar is crucial ... particularly an economic development customer, **Altitude Capital, LLC, a subsidiary of Alphabet, Inc (Google)**, that requires Cypress Solar ... as a condition for locating in EAL's service territory." ([[Cypress Order No. 4 CECPN Approval]]) Each of these *independently names Google*. The Spark finding produces no analogue: not one of its three exhibits names Google in connection with Conway, and the one document closest to the principal (the NDA) affirmatively withholds it. The categorical rule of [[Identifying the Unnamed Cooling-Water Data Centers]] — "the circumstantial lead ... is **not** an identification" — therefore controls, and the Conway principal remains unidentified.