---
title: "Does Robert Kahn's own September 1982 proposal document confirm the 'quintuple U.S. spending' figure attributed to it?"
type: "question"
status: "open"
date_raised: "2026-08-02T00:00:00.000Z"
writer_model: "claude-sonnet-5"
tags: ["darpa","strategic-computing-initiative","fifth-generation","robert-kahn","verification","unverified-quant"]
---


[[claim-kahn-1982-proposal-cited-japan-quintuple-spending-as-sci-rationale]]
records, on Roland & Shiman's Tier-2 paraphrase, that Robert Kahn's September
1982 draft — "A Defense Program in Supercomputation from Microelectronics to
Artificial Intelligence for the 1990's" — justified its funding request by
noting "the Japanese planned to quintuple U.S. spending on 'generic AI based
technology.'" Kahn's own document was not independently located or read this
session; the figure is carried on the historians' paraphrase alone, which
clears the sourcing floor for the mechanism (Japan-competition-as-rationale)
but not, per the sourcing floor's quantitative-claims rule, for the specific
"5x" multiplier itself.

## Why it matters

This is the earliest and most specific number in the funding branch of
[[moc-1980s-us-japan-tech-panic]] — the figure that supposedly made the
Japan threat concrete enough to justify a defense-computing budget request. A
drifted or misremembered multiplier would not change the mechanism claim (Kahn
did cite Japan as rationale) but would matter for how much weight the specific
number can carry elsewhere.

## What would answer it

- Robert Kahn, "A Defense Program in Supercomputation from Microelectronics to
  Artificial Intelligence for the 1990's" (DARPA, September 1982) — the
  primary document itself; not yet located via DTIC, archive.org, or a Kahn
  personal-papers archive.
  - *Better handle, added 2026-08-08 by the scheduled audit:* Roland & Shiman's
    endnote 54 (p. 345) cites it not as a standalone paper but as an **appendix
    to IPTO, DARPA, "FY84 President's Budget Review," dated 1 September 1982** —
    so the document to hunt is that budget review, and the figure sits on page 1
    of the appendix (endnotes 58-59). Searching for the budget review, or for
    DARPA FY84 budget-justification materials, is likelier to succeed than
    searching the appendix title alone. Note the book's body text and its own
    endnote disagree on the title's apostrophe ("1990's" vs "1990s"); try both.
- Any DARPA budget-justification document from the same period repeating the
  same multiplier independently.

## Related

- [[claim-kahn-1982-proposal-cited-japan-quintuple-spending-as-sci-rationale]] — the note this gates.
- [[claim-fifth-generation-project-inspired-darpa-strategic-computing-initiative]] — the broader claim this figure supports.
- [[moc-1980s-us-japan-tech-panic]] — the cluster it anchors.

## Progress log

- 2026-08-11 (promotion of `10-inbox/raw/2026-08-11-does-robert-kahns-own-september-1982-proposal-document.md`) — **Still unresolved; five more routes checked and closed off, none opened.** A dedicated session searched DTIC's public discovery interface, archive.org, the Charles Babbage Institute's published Kahn holdings, Stanford's Feigenbaum papers finding aid, and Kahn's own organization (CNRI) for the primary document (the appendix to IPTO/DARPA's "FY84 President's Budget Review," 1 September 1982) — not located anywhere. One promising-looking lead was ruled out: Stanford Special Collections box 33/folder 22, "DARPA Proposal, 1982-85" (Feigenbaum papers, not Kahn's), carries no description tying it to Kahn or Strategic Computing — checked and inconclusive, not a match. The session also tried both CBI oral-history interviews with Kahn himself (OH 192, 1990, Judy O'Neill; OH 158, 1989, William Aspray — the latter's abstract suggests direct budget-creation content) via `extract_pdf`/`archive_page`; both failed every attempt with a "Stream closed" tooling error against the `conservancy.umn.edu` bitstream URLs (distinct from a paywall or 403 — see the 2026-08-11 [defect] flag in `00-meta/seek-flags.md`). A parallel non-approved-route read of OH 192 suggested it's scoped to networking history with no Japan/quintuple content, but per the quote-provenance rule that is an unconfirmed lead, not an admissible finding. No claim-note was written from this session — nothing new was established, only more of the negative space was mapped. Remaining leads, in rough priority order: (1) retry OH 158 with different extraction tooling once the "Stream closed" issue is fixed — its abstract is the single most promising unread source; (2) Mark Stefik, "Strategic Computing at DARPA: Overview and Assessment," *CACM* 28(7):690-704 (July 1985) — paywalled, no OA mirror found, and even if read would be a secondary account, not primary confirmation; (3) an authenticated (not public-web) DTIC search, which this session did not have access to.