human-nets@ucbvax.ARPA (01/20/85)
From: Charles McGrew (The Moderator) <Human-Nets-Request@Rutgers>
HUMAN-NETS Digest Sunday, 20 Jan 1985 Volume 8 : Issue 2
Today's Topics:
Query - Use of Computer Networks for Impaired People,
Responses to Queries Floppy Disk Storage (3 msgs) &
A Rural Net of Micros,
Computers and People - AI People &
CDs vs. Books (2 msgs)
Information: Online Technical Reports
----------------------------------------------------------------------
Date: Fri, 11 Jan 85 11:00 IST
From: Henry Nussbacher <VSHANK%weizmann.BITNET@WISCVM.ARPA>
To: <human-nets-request@rutgers.arpa>
I would like to know if any articles or digests appear about the
following topics:
1) Use of the network for mute and hearing impaired people.
2) Use of the network for paralyzed and stroke victims (severely
immobolized).
The reason I am asking is because a person here in Israel has a group
of people that she is "head" of. They are (I think) called "Society
of Hearing Impaired People In Israel" and she knows a little about
networks and she felt that it would be great for these people to
communicate with friends and people in a similar situation throughout
the world.
Stroke victims and immobolized people feel limited in their activities
and their scope of outside contact. Being able to sit at a terminal
and have electronic conversations with people who are also immobolized
might be extremely theraputic.
I am wondering if any medical journal or researcher has explored this
side of "human-nets".
Hank
Weizmann Institute of Science
Rechovot, Israel
------------------------------
From: Norm Shapiro <norm@rand-unix>
Date: 15 Jan 85 09:41:45 PST (Tue)
To: INFO-MICRO@brl-vgr
Subject: Floppy Disk Storage: Steel or Plastic
Almost all of the floppy disk storage gadgets I have seen sold are
made out of plastic. It struck me that it would be better to store
floppy disks in steel, or some other iron alloy, so that when they are
kept near computers, terminals, modems etc they will not be subject to
damaging stray magnetic impulses.
I finally somebody, (Mead-Hatcher, Buffalo N.Y), who makes steel
floppy disk housings. But a collegue remarked that steel, which
although it will shield the disks from magnetic radiation, can itself
become magnitized and thereby damage the disks it stores.
What's the best way to store floppy disks: in steel or plastic?
In my case, the disks are 5.25 inch, double sided, quad-denisty.
Norman $Shapiro
norm@rand-unix
...!decvax!randvax!norm
213 393 0411
The Rand Corporation
1700 Main Street
Santa Monica CA 90406
------------------------------
Date: 15 Jan 1985 16:22:00-EST
From: mlsmith@NADC
To: INFO-MICRO@BRL-VGR.ARPA, norm@rand-unix
Subject: Floppy Disk Storage: Steel or Plastic
Cc: kushnier@NADC
We have some very attractive _wood_ floppy boxes that are much
stronger than plastic, and since they look good are located on the end
of the table.
mlsmith@nadc.ARPA
------------------------------
From: kyle.wbst@XEROX.ARPA
Date: 16 Jan 85 12:25:14 EST
Subject: Re: Floppy Disk Storage: Steel or Plastic
To: Norm Shapiro <norm@RAND-UNIX.ARPA>
Cc: INFO-MICRO@BRL-VGR.ARPA
I agree with the comment that steel could become magnetized and thus
cause more harm in the long run. Your question raises another issue,
though. Plastic can accumulate a static electric charge. I have seen
some ads recently (selling anti static devices for computer rooms)
that suggest that static discharges can cause damage to floppy disks.
My first reaction to such ads was Hog Wash... more ad hype. But then I
thought, if the discharge was strong enough it might create a
localized electromagnetic field of sufficient strength to cause an
error on very high density floppy disks. Does any one know if this
could be a real problem?
Earle.
------------------------------
Date: 11 January 1985 04:58-EST
From: Jerry E. Pournelle <POURNE @ MIT-MC>
Subject: query re rural net of micros
To: dual!islenet!bob @ UCB-VAX
It isn't just rural nets that have those characteristics. Think
about the computer phenomenon in the US, and what's happening
this year...
------------------------------
Date: 11 Jan 1985 20:33 EST (Fri)
From: Wayne McGuire <MDC.WAYNE%MIT-OZ@MIT-MC.ARPA>
To: ihnp4!utzoo!henry@UCB-VAX.ARPA
Subject: AI, laser disks, etc.
CLAIMS ABOUT AI
Spencer:
"Ah, I wondered what sort of response I would get from
the AI folks to that flame."
Let me hasten to point out that (1) I am not an expert on AI, but only
someone with a strong interest in it, (2) the AI community is
extremely diverse, and comprises many different, often contradictory
viewpoints, and (3) I didn't imagine the idea processing software I
described for optical disk-based personal assistants to be a product
of especially advanced AI research. I do think your brand of
skepticism is valuable: it will help keep the field of AI honest.
MEMORY & INTELLIGENCE
Spencer:
"Exercising heroic willpower, I will refrain from describing
in detail the things I have seen done on an 8K PDP8.
Substituting memory for thought
is often a worthwhile
tradeoff, but there is a mythology growing up that you can't
do anything useful in less than (your favorite number here)
megabytes. Nonsense. Guano. Utterly false."
Much memory alone does not for intelligence make, but it is a
necessary (if not sufficient) ingredient of intelligence, as oxygen is
for living mammals. Intelligence is probably to an important extent a
function of the number of conceptual connections that exist among the
number of conceptual objects in an organism or machine (conceptual
connections are themselves, of course, conceptual objects):
"... Experts differ from novices in science, chess, and other
fields that have been studied not only in having more information in
permanent memory but also, and more significantly, in being able to
process it efficiently. Among experts, for example, items of
information are more thoroughly indexed and thus can be rapidly
brought to conscious memory. The items, moreover, are elaborately
associated or linked with one another. Two consequences of these
associations--the ability to recover information by alternative links,
even when parts of the direct indexing are lost, and the capacity for
extensive means-ends, or trial-and-error, searches--are the essential
processes called into play in all problem-solving, from the most
elementary scientific discovery to the most advanced....
"The greatest advantage of the expert--and, conversely, the
biggest problem for the novice--attempting to gain literacy in
cognitively demanding fields is 'chunking,' the representation of
abstract groups of items as linked clusters that can be efficiently
processed. Such chunks may underlie mental processes ranging from the
childhood stages of cognitive development identified by Jean Piaget,
to the themata found by Gerald Holton in the history of scientific
discovery. Simon estimates that 50,000 chunks, about the same
magnitude as the recognition vocabulary of college-educated readers,
may be required for the expert mastery of a special field. The
highest achievements in various disciplines, however, may require a
memory store of one million chunks, which may take even the talented
about seventy hours of concentrated effort per week for a decade to
acquire...."
Herbert J. Walberg. Scientific literacy and economic
productivity in international perspective. Daedalus 112 (2), Spring
1983, pp. 1-28.
All this indexing, linking, associating, and chunking that Walberg
refers to is very memory intensive, both in human beings and machines.
In the future really interesting AI programs, especially in the domain
of natural language, will probably require many terabytes of memory or
more stored in densely connected parallel processing networks. An 8K
PDP8 won't quite suffice.
SELF-KNOWLEDGE
"I already know rather more about myself than the feds do;
don't you?"
In principle it might be possible for a large organization (like the
government) or any outside agent to know much more about you than you
know about yourself. Just because you inhabit your body doesn't mean
that you possess the information, or intelligence, or analytical
skills to understand in great depth your own personality, behavior
patterns, and social situation.
Imagine an organization with software that integrates the best
analytic knowledge and methods from economics, sociology, psychology,
literary analysis, history, anthropology and other disciplines, and
with the power to apply that software in an automatic multidimensional
analysis of the full set of information about any person (all the
kinds of things I suggested systematically archiving on optical
disks). In this set of personal data we could also include all the
relevant contextual information surrounding someone (especially
behavior by other agents) which that person might not be aware of or
have access to. I think you see my point. I don't know that such
organizations or software exist, but they could in theory. (On a much
less ambitious scale, programs probably already exist for
understanding social networks, and the positions of individuals within
them, on the basis of analyzing communications traffic.) Personal
biodisks would at least give one in machine readable form some of the
raw material with which to analyze one's own life with intelligent
programs.
-- Wayne McGuire <wayne%mit-oz@mit-mc>
------------------------------
Date: Tue, 8 Jan 85 11:33:28 pst
From: hpda!hptabu!dclaar@Berkeley (Doug Claar)
Subject: CDs obsoleting books
Re: CDs obsoleting books
The biggest factor I see holding back this "revolution" (didn't they
say TV would replace books, too?) is the display medium. I'm surprised
that no one else has mentioned it. The example that everyone uses is
the Encyclopedia Brittanica, but what most people are refering to is
its WORDS. Removing the pictures removes much of the usefulness.
Another example: I looked at one of my old text books, and every page
had either different fonts, or diagrams, or a picture, or colored
letters. All of these things made the text easier to understand. The
current displays, besides being hard to read, lack the color and
graphics capabilities to present this information, let alone the
equivalent of "4 color" pictures. Finally, there is the matter of
capacity. Most books present far more characters per page than a
terminal (and you can see 2 pages at once). Here are some examples:
what width x length #characters visible factor
----------------------------------------------------------------------
Terminal 80x24 1920 1920 1
Fiction (Sherlock Holmes) 55x36 1980 3960 2
Engineering text 85x46 3910 7820 4
Macro Economics text 78x60 4680 9360 4.9
Byte magazine 105x59 6195 12390 6.4
This becomes even more obvious is you include pictures. A "picture may
be worth 1000 words", but it costs many more (computer) words.
Clearly, the "human interface" has a ways to go before books die.
Doug Claar
HP Computer Systems Division
UUCP: { ihnp4 | mcvax!decvax }!hplabs!hpda!dclaar
-or- ucbvax!hpda!claar
ARPA: hpda!dclaar@ucb-vax.ARPA
------------------------------
Date: Thu, 10 Jan 85 12:35:46 pst
From: amdcad!phil@Berkeley (Phil Ngai)
Subject: use for CDs
When will AAA come out with a CD having all their maps on it?
And when will GM have a reader available in their cars?
I can't wait...
------------------------------
Date: Fri, 11 Jan 85 20:27:50 est
From: krovetz@nlm-mcs (Bob Krovetz)
Subject: online technical reports
The following is a list of people who can be contacted at various
sites on the net for ordering technical reports. I've tried to
determine who is the site contact, whether they have an on line
bibliography, if they have a mailing list for notification of new
TR's, and if the TR's themselves are available on line. If anyone
knows of this information for any sites I haven't mentioned, please
send me a message and I will post a followup to the net. Note that
the mailing lists mentioned are U.S. mail, not electronic! Online
bibliographies at the various sites may be FTP'd by logging in with
id: ANONYMOUS and password GUEST (this only applies if you are on the
ARPANET)
Yale: Donna Mauri (MAURI@YALE) is the contact person for AI or
cognitive science reports. There is no online list of those
reports, but she can send a hard copy list. For
non-AI/cognitive science reports the contact person is Kim
Washington (WASHINGTON@YALE).
CMU: No online list, however they do have a mailing list for
notification of recent TR's. TR's can be ordered over the
net. The contact person is Sylvia Hoy (HOY@CMU-CS-A).
MIT: There is an online list, but the publications office is
undergoing a restructuring, so it isn't available at the
moment. A contact for ordering the TR's will be established
at some future time.
SRI: No online line of just the report names, but there is a
list of the reports plus abstracts. Tonita Walker
(TWalker@SRI-AI) is the contact person. Many of the reports
are available for FTPing.
UTEXAS: A list of current reports is in {UTEXAS}<cs.tech>TRLIST. A
master list of reports still in print is under MASTER.TR.
Many of the current reports themselves are also available in
the above directory, but they contain text formatting
commands. The directory contains a file READ.ME which tells
which text formatter was used for which reports (SCRIBE vs.
NROFF). Reports may be ordered by sending mail to
CS.TECH@UTEXAS-20.
BBN: No reports or list online (no list even in hard copy).
Contact author directly about getting a copy of the TR.
PARC: Maia Pindar (PINDAR@XEROX) is the contact person. An online
copy of the bibliography is not available at the moment, but
Ms. Pindar may be contacted to obtain a hardcopy.
Rutgers: Contact Christine Loungo (LOUNGO@RUTGERS) or Carol Petty
(PETTY@RUTGERS) to obtain reports. They maintain a mailing
list to distribute notices of the TR's and the abstracts.
The abstracts of recent reports are online and under:
{RUTGERS}<library>tecrpts-online.doc.
ISI: Lisa Trentham is the contact (LTRENTHAM@ISIB). There is a
list of the available reports under
{ISIB}<BBOARD>ISI-PUBLICATIONS.DOC
Stanford: Stanford reports are issued by four sources: the HPP
(Heuristic Programming Project), the AI lab, the Center for
the Study of Language and Information (CSLI), and the
Computer Science Department. HPP reports are available
without charge by contacting Paula Edmisten
(EDMISTEN@SUMEX). Please be reasonable with your requests;
no more than 15 at a time! There is no online bibliography
available, but a hard copy may be requested. There is an
online bibliograpy of AI lab reports in AIMLST in
[BIB,DOC]@SU-AI. Some of the reports are available online
and are so indicated in the bibliography. Reports from CSLI
may be requested from Dikran Karagueuzian (DIKRAN@SU-CSLI).
A bibliography of the reports is stored under
{SU-CSLI}<CSLI>CATALOG.REPORTS. CSLI will also be issuing
lecture notes, and a bibliography of these will be under
{SU-CSLI}<CSLI>CATALOG.LECTURE-NOTES. The reports are
available without charge, but there is a charge for the
lecture notes. There is also a charge for reports published
by either the AI lab or the Computer Science Department, but
information as to cost and/or availablity may be sent to
Kathy Berg (BERG@SU-SCORE). A bibliography of CSD reports
from 1963 to 1984 is available for $5.00. The department
maintains a mailing list for notification of new TR's. You
can be added to it by contacting Kathy Berg. Updates are
sent out about five or six times per year.
------------------------------
End of HUMAN-NETS Digest
************************