Meteorologist, oceanographer, Python solvem probler, experimenter, Raspberry Pi tinkerer, geek dad, manager, space enthusiast, ham radio lover, science columnist, ADHDer, synthetic intellect curious, linux admin cosplayer.

Draft Corrections: Post 14 - The Machine That Built IBM

Status: COMPLETE

Summary of Findings

Two factual errors require correction. One is a date error (ENIAC year), one is a biographical error (Watson Jr.’s education). The post contains a third minor issue (attribution of a paraphrase as quote). Most of the post’s complex claims about the IBM 701, the 19 customer machines, the Oppenheimer dedication, the Georgetown-IBM demonstration, Samuel’s checkers, JNWPU operational forecast, and Backus Speedcode are verified accurate against primary sources.

REQUIRED CORRECTIONS

Correction 1: ENIAC unveiling date (line 36)

Current text:

When the ENIAC was unveiled in 1944, Watson Sr. was furious.

Issue: ENIAC was publicly unveiled on February 14-15, 1946. Construction began in June 1944, but the machine was secret until early 1946. Watson Sr.’s reaction to ENIAC was after its 1946 public unveiling, not 1944.

Suggested replacement:

When the ENIAC was unveiled in 1946, Watson Sr. was furious.

This also strengthens the timeline because the SSEC’s accelerated construction (August 1947 - January 1948) makes much more sense as a response to the February 1946 ENIAC unveiling than to a hypothetical 1944 reveal.

Correction 2: Watson Jr. education (line 40)

Current text:

But Watson Jr. – born 1914, MIT-educated, a wartime bomber pilot – saw what his father could not.

Issue: Watson Jr. attended Brown University, not MIT. He graduated from Brown in 1937 with a business degree. The post’s own image caption later (“Brown University photo, 1937”) directly contradicts the body text.

Suggested replacement:

But Watson Jr. – born 1914, Brown-educated, a wartime pilot who flew Lend-Lease missions to Moscow – saw what his father could not.

(The “bomber pilot” element can be retained but it understates: he was personal pilot to General Follett Bradley, head of Lend-Lease to the USSR, and flew the route to Moscow. He reached Lt. Col.)

Correction 3: Hutchins quote attribution (line 222)

Current text:

It was, as the historian John Hutchins later put it, “a clever proof-of-concept engineered to make the 701 look miraculous to reporters.”[^8]

Issue: This is not a direct verbatim quote from Hutchins 2004. Hutchins describes the experiment’s small scale and notes the rules were “fitted to specific words and sentences” but does not use this exact phrasing. The post presents a paraphrase as a quotation with quotation marks.

Suggested replacement (option A - drop the quotation marks):

It was, as the historian John Hutchins later argued, a clever proof-of-concept engineered to make the 701 look miraculous to reporters.[^8]

Suggested replacement (option B - use actual Hutchins phrasing):

The historian John Hutchins later concluded that the experiment was small-scale – just 250 words and six grammar rules – with many operations fitted to specific words and sentences for the demonstration.[^8]

NO ACTION NEEDED

The following claims, despite their specificity, were verified against primary sources:

  • $15,000/month rental (confirmed by Computer History Museum and Hutchins 2004 primary source)
  • 19 IBM 701 customer machines including DuPont Central Research in Wilmington as #19
  • IBM HQ 590 Madison Avenue first installation late December 1952
  • Los Alamos first customer April 1953
  • April 7, 1953 Oppenheimer dedication; “tribute to the mind’s high splendor” preserved by Goldstine
  • All Léon Dostert biographical claims
  • All Arthur Samuel biographical claims (Emporia Kansas 1901, MIT 1926, Bell Labs 1928, Illinois 1946, IBM 1949)
  • Cressman 1919 birth, Rossby at Chicago, JNWPU director 1954
  • JNWPU established July 1, 1954
  • IBM 701 installed at JNWPU March 1955
  • First US operational forecast May 6, 1955
  • “Very disappointing and not usable” quote (NCEP/EMC, verbatim)
  • 1957 IBM 704 replacement, January 1958 NMC merger
  • Backus 50-75% programming cost quote (verbatim from 1954 paper)
  • Backus Speedcode team: Herrick, Quarles Jr., Skillman, Pulos, Siegel (all verified)
  • ACM September 1953 presentation
  • “Machine learning” coined 1959
  • 15-point IBM stock rise after Feb 24, 1956 broadcast
  • Watson Sr. died June 19, 1956
  • ALPAC November 1966 report; chaired by John R. Pierce

CONFIDENCE LEVELS

  • High confidence in corrections 1 (ENIAC 1944→1946) and 2 (MIT→Brown): These are flat factual errors, easily verified against multiple sources.
  • Medium confidence in correction 3 (Hutchins quote): The post puts quotation marks around what appears to be a paraphrase. Strictly speaking, this is a misattribution and should be fixed.

TECHNICAL NOTES FOR EDITOR

These two errors (Watson Jr. MIT vs Brown; ENIAC 1944 vs 1946) are particularly important because:

  1. Watson Jr.’s education is mentioned in two places in the post — once in body text (wrong: MIT) and once in image caption (right: Brown). Both should be consistent.
  2. The ENIAC date is foundational to the chronology of the post and should be correct.