Newsgroups: soc.history.science,comp.ai.philosophy,alt.cyberspace,soc.history.moderated
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!news.bluesky.net!news.sprintlink.net!noc.netcom.net!netcom.com!lhb
From: jmc@SAIL.Stanford.EDU (John McCarthy)
Subject: Re: What if x had lived longer?
Message-ID: <lhbDA5AD7.E6L@netcom.com>
X-Submission-Address: soc-history-mod@bcm.tmc.edu
Sender: lhb@netcom23.netcom.com
Reply-To: jmc@cs.Stanford.EDU
Organization: Computer Science Department, Stanford University
References: <3r8mn1$ohg@nntp5.u.washington.edu> <3ra7us$kcb@shell1.best.com>
X-Admin-Address: history-mod@best.com
Date: Wed, 14 Jun 1995 04:12:43 GMT
Approved: lhb@netcom.com
Lines: 30

Von Neumann did not dominate American computer design after the very
first years.  Many of his opinions were over-run.  He did not like
programming languages, floating point or large memories.  In fact,
according to Paul Armer, von Neumann wrote a letter to the Air Force
saying that RAND Corporation ordering a 32K word memory for their IBM
704s was a waste of Government money.  RAND got the money for the
memories.  I'm sure von Neumann would have changed his mind about many
of these questions.  I hoped that he would come to the Dartmouth
Summer Research Project on artificial intelligence in summer 1956, but
he was already dying at the time.

Turing's ideas about computer design were over-run by events also.
The "Big Ace" was finished at the National Physical Laboratory in
Teddington in December 1958.  Finishing his old design must have been
an act of piety.  It used mercury delay lines arranged very cleverly
so that the information a program wanted would often be coming off a
delay line just when it was wanted.  However, the machine was finished
just as IBM came out with 32K memories for the 704.

I think Turing would have made a significant contribution to
artificial intelligence if he had lived longer.

-- 
John McCarthy, Computer Science Department, Stanford, CA 94305
*
He who refuses to do arithmetic is doomed to talk nonsense.

http://www-formal.stanford.edu/jmc/


