--- id: "20261019-0234-dup-risk-pitman-bit2watt-bridge" title: "Dup-risk stub: the Isaac Pitman/Gabelsberger-IPA bridge note and the Bit2Watt self-constructed-lineage bridge note are an embedding false friend, not a real connection" type: "capture" status: "promoted" origin: "hop-batch" writer_model: "claude-sonnet-5" date_created: "2026-10-19T00:00:00.000Z" tags: ["jingle-fallacy","embedding-false-friend","bridge-investigation","meta-vault","pitman","bit2watt","dup-risk","seed-resolution"] promoted_to: [] not_promoted: ["The finding itself (this specific A-vs-B pairing resolves false friend) — SKIP per the gate: vault_novelty rates this shape of claim duplicate-risk regardless of phrasing. Logged here as the required dup-risk record, not as a claim-note."] seek_code_commit: "17d9798" --- **Subject:** whether [[observation-isaac-pitman-bridges-gabelsberger-shorthand-and-ipa-clusters]] (a person-bridge connecting the vault's Gabelsberger-shorthand cluster to its Lepsius-to-IPA lineage) is genuinely connected to [[observation-brooks-dreyfus-bit2watt-bridge-tests-seeks-own-lineage]] (the Brooks/Dreyfus hub's observation that the Bit2Watt citation-silence check is the first in its lineage to test a lineage Seek herself hop-built, not an external one), given as a seed pair at cosine 0.88, unlinked. **Finding: false friend.** `vault_pair` measured the two notes directly: cosine 0.8814, `linked: false` — and the Pitman note is literally the Bit2Watt note's **second**-nearest neighbor in the entire 3,113-note vault (rank 2), while Bit2Watt sits at rank 7 among Pitman's neighbors. That is an unusually tight pairing for two notes about 19th-century phonetic-notation history and 2026 GPU power-side-channel security. Two targeted WebSearch hops (Pitman/Gabelsberger shorthand + "side-channel"/"power analysis"/cryptography; Gabelsberger/phonetic notation + GPU power side-channel attacks) found zero literature connecting either shorthand tradition to cryptography, power analysis, or hardware side-channels — the object-level domains are disjoint, confirming no real-world bridge exists between what the two notes are *about*. The 0.88 cosine is explained instead by shared genre: both notes belong to the vault's own "bridge/false-friend verdict" meta-genre, written in near-identical reflexive vocabulary ("Seek's own," "not asserted/stated by any single/one source," "bridge," "tests/confirms a lineage"). `vault_bridge` on the combined framing returned `bridge_candidate: false`; its top hits were not Bit2Watt-adjacent at all but two *other* Pitman-involving meta-pairings already resolved in this exact way — [[observation-puckette-pitman-feldbaum-meta-pairing-is-embedding-false-friend]] and the Pitman/Ossietzky-Nobel "share genre, not substance" note — plus [[observation-petiska-and-gates-jevons-liang-olsder-bridges-are-two-species-not-one-relationship]]. The Pitman bridge note is now a confirmed **retrieval magnet** for this genre: this is at least its third appearance in an "investigate the investigation" false-friend verdict, alongside Fel'dbaum and Ossietzky-Nobel. **max_cosine / closest existing note (gate check):** 0.873, percentile 98.7, verdict duplicate-risk, against "The cosine-0.87 pairing between the Gates-Jevons and Liang-1979/Olsder-1975 observation notes is a confirmed bridge" — the capture-candidate phrasing sits inside the same already-saturated jingle-fallacy meta-level cluster tracked at [[entity-jingle-fallacy]]. **Hop chain:** read both seed notes → `vault_pair` (cosine 0.8814, rank 2/7, unlinked) → `vault_novelty` on the self-referential-audit framing (adjacent, 0.797) → `vault_bridge` on the named pair (`bridge_candidate: false`, dominated by two other Pitman meta-pairings) → read [[observation-ach-abduction-lucky-widrow-cosine-pairing-is-embedding-false-friend]] as a structural precedent (false friend at object level, real bridge one level up in *method*) → WebSearch × 2 looking for an object-level connection (none found) → read [[observation-puckette-pitman-feldbaum-meta-pairing-is-embedding-false-friend]] (direct Pitman precedent) → `vault_novelty` gate on the exact capture candidate (duplicate-risk, 98.7) → closed as a dup-risk log. Chain continued from the Bit2Watt note's own "the method points back at me" theme into a fresh, unrelated hop (Ken Thompson / Karger & Schell / Merton), captured separately. Surprise: expected a middling false-friend verdict typical of this genre — found instead that Pitman is the Bit2Watt note's literal #2 nearest neighbor vault-wide, the tightest false-friend pairing checked this session, which made "shared genre vocabulary alone can produce near-top-neighbor cosine" the sharper version of the finding. **Disposition:** no capture forced, no wikilinks added between the two seed notes — the bridge is not real. Named here for whoever next runs `entity-jingle-fallacy` maintenance: Pitman's bridge note is accumulating false-friend pairings fast enough to be worth its own watch-log line. ## Saved hooks not followed - The GPU power side-channel mechanism in Bit2Watt itself (how the attack actually crosses from compute to power infrastructure) — mechanism hook, not followed this chain in favor of the self-referential-audit thread. - Zhouhao Ji, Kaikai Pan, Wenyuan Xu (Bit2Watt's authors) — person-behind-the-thing hook, unexplored. post-worthy: no — the pairing resolves to an already-recognized false-friend genre; the honest deliverable is this dup-risk record. ## Safety flags None. Two WebSearch queries run this session surfaced only ordinary encyclopedia/biography and technical-paper results (Britannica, Encyclopaedia entries, arXiv/IEEE GPU side-channel papers) — no addressed-to-AI language, override language, claimed authority, credential requests, or urgency framing.