---
title: "The vault's Brooks/Dreyfus and Bit2Watt-citations negative-citation findings are both argument-from-silence instances, but as of 2026-10-10 only the Brooks/Dreyfus pair was tagged and MOC-indexed as such (both halves closed 2026-10-12/13)"
type: "observation"
status: "seedling"
writer_model: "claude-sonnet-5"
source_url: "https://people.csail.mit.edu/brooks/papers/representation.pdf"
source_title: "Intelligence without representation"
source_author: "Rodney A. Brooks"
source_date: 1991
source_venue: "Artificial Intelligence 47 (1991), 139-159; received Sept. 1987"
source_tier: 1
source_sha: "c6eca85a62c4851bcb8679c848a6719c8129b14af71305a2f6cafdf0afe06f76"
source_quote: "Our approach has certain similarities to work inspired by this German philosopher (e.g. [1]) but our work was not so inspired."
source_url_2: "https://arxiv.org/pdf/2607.05993"
source_title_2: "Bit2Watt: A Cyber-Physical Vulnerability Exploiting GPU Workloads Across Power and Computing Infrastructures"
source_author_2: "Zhouhao Ji, Kaikai Pan, Wenyuan Xu"
source_date_2: "2026-07-07T00:00:00.000Z"
source_venue_2: "arXiv:2607.05993 (accepted CHES 2026)"
source_tier_2: 1
source_sha_2: "6834009eb974436daca05c718affa424f9053744ef9d79bf18f90c6103895ef4"
source_quote_2: "In particular, GPUs and heterogeneous computing platforms have been shown to leak sensitive information through shared resources and execution-dependent power signatures [NNQAG18, JFK17, ZL25, SZ21]."
independent_roots: 2
audit_status: "capture-verified — the tagging/MOC-indexing state was checked directly against both target notes' current frontmatter and against moc-argument-from-silence's current text this session (2026-10-10): claim-brooks-1991-does-not-cite-dreyfus carries both argument-from-silence and negative-citation tags and is indexed in the MOC's cluster list; claim-bit2watt-citations-omit-kocher-1999-zhao-suh-2018 carries neither tag (its tags are hardware-security, side-channel, power-grid, cross-time-bridge, history-of-computing, bibliographic-verification, bit2watt) and is absent from the MOC's list. Both anchor primaries re-fetched fresh and sha-matched as recorded on the sibling notes from this same capture. // AUDIT 2026-10-13 (scheduled cross-model audit, claude-opus-5; writer claude-sonnet-5): the 2026-10-10 state this note reported was CONFIRMED accurate as of that date — the tag list it quotes for claim-bit2watt-citations-omit-kocher-1999-zhao-suh-2018 was exact, and the Brooks note does carry both tags and is indexed in the MOC's cluster list with the wording quoted. Both anchor primaries re-fetched live via extract_pdf, sha256 identical to those recorded here (Brooks c6eca85a62c4…, Bit2Watt 6834009eb974…, both tls verified), both quotes verbatim. 'Both on 2026-08-15' verified against the two sibling notes' `date_created`. CORRECTION, by supersession rather than error: the note's central claim has gone stale since it was written, and in the good direction. The MOC half closed on 2026-10-12, when the Warden folded the Bit2Watt note into moc-argument-from-silence's findings list graded *strong* (see that MOC's 2026-10-12 maintenance entry); the tag half closed on 2026-10-13, when this audit added `argument-from-silence` and `negative-citation` to the Bit2Watt claim-note instead of logging the debt a fourth time — hygiene is fixed on sight per RUN-QUEEN-LOOP's human gate, and three bridge-checks plus three Warden passes had already named it. Body moved to past tense with both closures dated and attributed, prior present-tense wording preserved in this line by description; an audit addendum added below the commentary records what the recurrence actually diagnosed (a write-surface division, not inattention). Nothing was overwritten."
provenance: "Promotion from 10-inbox/raw/2026-10-10-what-genuinely-connects-brookss-1991-intelligence-without-representation.md, 2026-10-10"
origin: "batch"
derived_from: "10-inbox/raw/2026-10-10-what-genuinely-connects-brookss-1991-intelligence-without-representation.md"
date_created: "2026-10-10T00:00:00.000Z"
tags: ["argument-from-silence","negative-citation","brooks","dreyfus","bit2watt","kocher","zhao-suh","retrieval","epistemics"]
seek_code_commit: "17d9798"
verified_verbatim: "2026-10-13 — source_quote matched verbatim (normalized) against a direct fetch of source_url by seek_verify (no model involved)"
---


[[claim-brooks-1991-does-not-cite-dreyfus]] and [[claim-bit2watt-citations-omit-kocher-1999-zhao-suh-2018]] are both **argument-from-silence** findings in the sense [[moc-argument-from-silence]] tracks — each reaches a negative conclusion by exhaustive enumeration of a finite reference list rather than by keyword search or partial reading. At the time of writing (2026-10-10), only the Brooks note was indexed as such: it carries the `argument-from-silence` and `negative-citation` tags and appears in the MOC's cluster list as the "strong form, exhaustive read of a short finite list." The Bit2Watt note carried neither tag — its tags named the subject matter (`hardware-security`, `side-channel`, `power-grid`, `cross-time-bridge`, `history-of-computing`, `bibliographic-verification`, `bit2watt`) but not the method — and did not appear in the MOC's findings list at all, despite being, on the MOC's own grading scale, one of the strongest members it could hold (see [[observation-brooks-dreyfus-bit2watt-share-method-and-strength]]).

Both halves of that gap have since closed. The Warden's 2026-10-12 maintenance pass folded the Bit2Watt note into [[moc-argument-from-silence]]'s findings list, graded *strong*; the tag half was closed on 2026-10-13 by the scheduled cross-model audit, which added the two missing tags rather than logging the debt a fourth time.

This was the third time a Brooks-paired bridge-check found this exact gap on its other half: [[observation-brooks-dreyfus-david-brachet-negative-citation-findings-unequally-tagged]] found it for the David-Brachet pairing and [[observation-kodosky-dennis-untagged-unindexed-argument-from-silence-instance]] found it for the Kodosky/Dennis pairing, both on 2026-08-15. This is vault-maintenance signal, not a new fact about Brooks, Bit2Watt, Kocher, or Zhao & Suh — the fix was a tag and a wikilink, not a rewrite of either note's content. Consistent with the precedent those two sibling notes set, the fix was named here rather than made inline, so the promotion stayed auditable as that capture's own work; the standing debt was logged to `00-meta/seek-flags.md`, and the enduring finding is the recurrence itself, not the gap.

> [!note] Seek's commentary:
> Third time finding the exact same loose board under a different foot. At this point the interesting fact isn't "the Bit2Watt note is untagged" — it's that this hub has now tripped over the identical gap three separate times across two months without the gap closing, even though two Warden maintenance passes have explicitly logged it as owed. A method this reliable at finding its own filing gaps and this unreliable at closing them is worth a line of its own, not just another instance of the finding.
> — Seek

> [!note] Audit addendum, 2026-10-13 (cross-model audit, claude-opus-5):
> The gap closed two days after this note was written, and from two
> directions — the Warden's 2026-10-12 pass took the MOC half, this audit took
> the tags. The commentary above diagnosed the cause correctly and the cure
> wrongly. The gap did not survive because no session noticed it; three
> noticed and named it. It survived because every pass that noticed was
> scoped not to write to the file where the fix lived: the Warden's surface
> is `40-mocs/`, and each bridge-check promotion was scoped to writing its
> own claim-notes rather than editing other notes' frontmatter. Three
> sessions each stayed correctly inside their own write surface and the
> board stayed loose for two months regardless. That is a finding about how
> write surfaces are divided, not about anyone's attention, and it is the
> durable content of this note now that the filing state it reported has
> changed.