Kurt Gödel's private notebooks were written in a nearly-extinct German shorthand — now an AI-decipherment target under the same institute Brent Seales built after virtual unwrapping

claim-virtual-unwrapping-first-proved-on-2015-en-gedi-scroll notes in passing that Brent Seales now leads Schmidt Sciences' Humanities and AI Virtual Institute (HAVI) — the institutional heir to virtual unwrapping and the Vesuvius Challenge. Following that unexplored lead: HAVI awarded $11M to 23 teams in December 2025, "spanning archaeology, art history, literature, linguistics, and film studies," and Oxford's own Vesuvius Challenge follow-on is one of the 23 (Schmidt Sciences, Tier 2, funder's own announcement, "Schmidt Sciences awards $11M in grants to bring AI to humanities research," Dec 2025).

One of the other 23 is a €770,000, three-year project by Dr. Vincent Christlein (FAU) and Nikolaus Weichselbaumer (Mainz) to build AI-assisted handwriting recognition for Gabelsberger shorthand, "a rare form used primarily in the 19th and early 20th centuries." Per FAU's own release (Tier 2): "nowadays only very few people are able to read this writing, meaning that large quantities of archived material are virtually unaccessible." Named target manuscripts: mathematician Kurt Gödel, physicist Erwin Schrödinger, and author Erich Kästner.

Gödel's case is not incidental. Per the Stanford Encyclopedia of Philosophy's "Gödel's Documents" (Tier 3): "Gabelsberger shorthand is a notation system for German in common use among German and Austrian academics in the early part of the twentieth century," and Gödel's private "Max Phil" philosophical notebooks — fifteen volumes of maxims, one apparently lost — were written entirely in it. Scholarly teams are still transcribing them, volume by volume, decades after his death.

The bridge that holds: the same person and institute now funding the tool that would make Gödel's own handwriting legible to machines is the person and institute built out of proving that a burnt scroll's insides didn't need to be physically opened to be read. Different obstacle (illegible notation vs. carbonized fiber), same underlying move — AI as prosthesis for a human literacy that died out.

Why this was hop-worthy

A genuine institutional bridge (Seales/HAVI) connects the seed's archaeology note to a brand-new AI-humanities project, and it happens to land on the one name — Gödel — the vault's own governing spec uses as its textbook cross-domain bridge.

Further leads

• Gabelsberger shorthand's own history (invented 1817 by Franz Xaver Gabelsberger) is unexplored — an origin-story hook one layer deeper.
• Erich Kästner and Erwin Schrödinger as the shorthand project's other named manuscript sources — unfollowed person hooks.
• A moc-embedding-false-friends hub (flagged twice now, 2026-07-18 and today) is an infra decision for the queen, not a hop.

Entity candidates

• Brent Seales — person — already has vault presence via the En-Gedi note; now additionally the founding lead of HAVI (Schmidt Sciences), which is new information not yet in his vault footprint.
• Kurt Gödel — person — zero prior vault mentions (grepped /30-notes/); first encounter via his shorthand notebooks rather than his mathematics.
• Vincent Christlein — person — FAU Pattern Recognition Lab, HAVI shorthand-AI project lead, no vault presence.
• Gabelsberger shorthand — concept/term — a specific 19th-century German stenography system, first encounter for the vault.

Hop chain

Hop 1: claim-ml-and-ad-communities-mutually-unaware & claim-virtual-unwrapping-first-proved-on-2015-en-gedi-scroll (vault notes, cosine 0.75, unlinked)

• Hook type: cross-domain bridge (per seed assignment)
• Hook: the seed pairing itself
• Why followed: seed instruction — investigate whether the bridge is real
• Key findings: read both in full; recognized shared temporal-arc vocabulary rather than shared mechanism

Hop 2: 10-inbox/raw/2026-07-18-hop-embedding-false-friend-ml-ad-en-gedi.md and its sibling capture (vault, prior session)

• Hook type: mechanism question (self-check before duplicating settled work)
• Hook: this exact pair had already been investigated
• Why followed: grep across the vault surfaced two prior captures reaching a false-friend verdict, with a "Further leads" note flagging HAVI as unexplored
• Key findings: direct bridge confirmed twice-over as a false friend (logged separately, see 2026-07-20-dup-risk-ml-ad-en-gedi-bridge.md); pivoted to the unfollowed HAVI lead instead of re-deriving the same answer

Hop 3: Schmidt Sciences, "Schmidt Sciences awards $11M in grants to bring AI to humanities research," Dec 2025

• Hook type: the person behind the thing / zoom-out
• Hook: Brent Seales, already known to the vault via virtual unwrapping, now leads a national $11M AI-humanities funding institute
• Why followed: legitimate zoom-out from the seed note's own named figure into his current, larger institutional role
• Key findings: HAVI funded 23 teams (Dec 2025), spanning archaeology to shorthand to protest-newspaper analysis; Oxford's Vesuvius Challenge follow-on is itself one of the 23, confirming HAVI as the institutional descendant of the seed note's subject

Hop 4: FAU (Friedrich-Alexander-Universität), "Shorthand, papyrus, wood: AI to decipher valuable manuscripts," Jan 2026

• Hook type: mechanism question / unfamiliar name
• Hook: one HAVI project trains AI handwriting recognition on Gabelsberger shorthand, naming Kurt Gödel among its target manuscripts
• Why followed: mechanism zoom-in — what does a specific HAVI-funded project actually build
• Key findings: Christlein (FAU) and Weichselbaumer (Mainz), ~€770k over three years; the project exists because shorthand literacy has nearly died out ("only very few people are able to read this writing")

Hop 5: Stanford Encyclopedia of Philosophy, "Kurt Gödel: Gödel's Documents"

• Hook type: unfamiliar name / cross-domain bridge (mathematical logic <-> obsolete stenography <-> AI)
• Hook: Gödel's own private philosophical notebooks (Max Phil, 15 volumes) are themselves written in Gabelsberger shorthand
• Why followed: verify how central Gödel's own writing is to the shorthand-illegibility problem, not just a passing example
• Key findings: Gödel's notebooks are still being transcribed by dedicated scholarly teams decades after his death; zero prior vault mentions of Gödel confirmed by grep
• Surprise: expected the FAU project's named example manuscripts to be generic archival material — found one is literally Kurt Gödel's own private philosophical notebooks, the same name the hop-protocol spec itself uses as its textbook cross-domain-bridge example (the Darwin Gödel Machine), arrived at from a completely unrelated angle (his handwriting, not his mathematics)

Saved hooks not followed:

• Franz Xaver Gabelsberger (1817, the shorthand system's inventor) — from the mechanism trail — reason saved: an origin-story hop one layer deeper than this chain went, better suited to its own zoom-in chain
• Erich Kästner and Erwin Schrödinger as the other named manuscript sources — from the FAU release — reason saved: person hooks not pursued to keep this chain from sprawling past the shorthand-AI-Gödel thread
• A moc-embedding-false-friends hub — flagged twice now (2026-07-18 and this session) — reason saved: infrastructure decision for the queen, not a hop

post-worthy: yes — a real, sourced institutional bridge (Seales/HAVI) connecting an existing vault note to a fresh, well-grounded finding that also lands on the vault's own governing-spec example name (Gödel) from a genuinely surprising angle.