taylor@hplabsc.UUCP (06/20/86)
Computers and Society Digest, Number 6
Friday, December 13th 1985
Topics of discussion in this issue...
The Cost versus the Freedom of Information
More on the Value of Information
Computers and the Law
Feedback on the last issue of CAS
A Worm Example
Heavy topics
More thoughts on information
----------------------------------------------------------------------
Date: Mon, 25 Nov 85 11:50:13 PST
From: ihnp4!ucbvax!harry@ucbarpa.berkeley.edu (Harry I. Rubin)
Subject: The Cost versus the Freedom of Information
Dave, I like your notion of exploring cost vs. freedom of information,
but I think you have mixed together two issues which need to be recognized
as seperate, even if you want to consider them together. One issue is
the volume of information people are confronted with these days, and
in particular the volume of electronic mail/news/notes.
The other issue concerns the organizations (businesses, universities, etc.)
which pay to support computer systems, and the rights of such organizations
and how they can ensure that the computers are used to their benefit.
For the individual reader, the cost vs. benefit question is time-to-read
vs. usefulness or interest of a piece. For the organization the cost side
of cost/benefit includes both computer resources to support the mail/news/etc
and the time of the people who read it. The benefit is usually rather
vague, on the order of "better informed staff."
Screening programs could be employed in both cases (if only we knew how to
build good ones). There is little technical difference between
John Random Reader using a filter on his incoming mail and administrators
filtering all mail or netnews coming into the the system. There is
however an important moral or ethical difference; the former is a person
using a tool to help him, the latter is a bit of heavy handedness.
It may be justified by the fact that the administrator, acting for the
organization, pays for the computer, but it still seems rather unpleasant
to me.
Harry Rubin
------------------------------
From: hplabs!aurora!eugene (Eugene miya)
Date: 25 Nov 1985 2031-PST (Monday)
Subject: More on the Value of Information
> Date: Tue, 5 Nov 85 15:51:52 MST
> From: hpcnou!dat (Dave Taylor)
> Subject: The Value of Information
>
> I'm starting to wonder if our society is placing the
> right emphasis on information...
> There is a crucial problem with estimating the 'value' of
> information that isn't being addressed.
> The problem I think is that with the electronic media, there is
> no way to judge the VALUE of the information without actually having read
> the particular item.
>
> the value of the information
> is going to have to be evaluated because it isn't reasonable to expect
> companies to subsidize junk mail and miscellaneous trash...
actually in some ways its being devalued or deflated.
>
> Which brings us to the real question here - how can we evaluate,
> on an ad hoc basis, the information being sent about the system?
>
> One solution is to write 'intelligent' software that does some
> sort of filtering for us.
>
> Another alternative is to somehow limit and/or police the
> system so that only information that was useful to some sort of
> governing body would be forwarded (the "mod.*" approach to news).
>
> A third method is to sort of 'blend' the previous two approaches,
> and have system 'auditors' that would peruse the available information
> and attach a "value" rating to each piece.
>
> Seems like perhaps we've reached two issues:
>
> 1. The "rights" of information
> versus
> 2. The "cost" of information.
>
> Any thoughts?
>
> -- Dave Taylor
Yes there are interesting tools which have been proposed and
other interesting tools I've seen at places like PARC. The NLS
(Englbart) and the Tioga (Teitelman) editors offer ways of structuring
ideas and acting at Tables of Contents (Mac's ThinkTank is a crude
shell of these system as is MacDraw to the PARC drawing software).
I know several people who have thought of active, network discriminators,
ala say the movie Tron, or Alan Kays recently posted thoughts to the net
(captured by some one else). Yes, you can make or measure value, the
real problem comes from the Used Car Saleman down the street, who just so
happens to be your brother-in-law, so he's an exception.....
On the subject of cost, Knuth at SU wrote a paper on this for the American
Math Monthly, and I have to locate it for someone else. So......
------------------------------
> Date: 8 Nov 1985 1609-EST (Friday)
> From: Marc Mengel <ihnp4!pur-ee!pucc-j!aaz>
> Subject: A little bit of information ...
>
> I think one of the major problems associated with databases
> like the NCIC, (or the various "bad tenant", "bad check passers"
> etc. databases) is that they don't keep *enough* information.
> . . . . . greatly shortened but an interesting contrast with the above.
> So what can be done to keep people from acting on too little
> information? The first choice is to provide more information.
> The second choice is to keep people from acting,
> with such things as disclaimers, telling when the information was
> last updated, and where one should go for more information before
> taking any action.
> Any other ideas out there?
>
> Marc Mengel
Yes, we are being flooded. No, I don't think more information will help.
Yes, updating is extremely important. I fear this is a social rather than
a technological problem. My cop out. there is nothing to prevent a
country like say South Africa from buying an IBM computer and using it
for "evil purposes" except other socially conscious people.
In the AI digest I was contemplating "Artificial Intelligence".....
substitute any variety of words for "intelligence" which make up
intelligence. It's very easy to build systems for "discrimination"
in the behavioristic sense, but consider trying to implement systems
or "artificial compassion" and its much harder. Forget emotion for a second
and consider a social worker expert system to decide who gets welfare.
It's not done purposely on the basis of economic need. Sure deceiption
is possible. Anyway, the analogy is vague, I hope you can consider it,
I have friends at Intellicorp and other AI places who have a hard time with it,
more if it comes to me.
Too much time on the net.
--eugene miya
------------------------------
From: hplabs!aurora!eugene (Eugene miya)
Date: 25 Nov 1985 2007-PST (Monday)
Subject: Computers and the Law
Re: Reference for computers and law
Law (yecch, personally), but you asked:
The person best known in the computer field to go into law is probably
Dr. Susan Nycum who is part of a law firm in Palo Alto. There is
a lot of interest in this field under a variety of subtopics. I'm
sorry, I don't know the firm she is with, but she recently spoke
at a CPSR meeting on the legal limits of AI [I was unable to make].
--eugene miya
------------------------------
Date: 9 Dec 85 11:26 EST (Mon)
From: _Bob <hplabs!Carter@RUTGERS>
Subject: Re: Computers and Society Digest #5
The Gary North stuff is pretty thin soup. Can we leave this kind
of journalism to Geraldo Rivera and get back to serious discussion?
_B
------------------------------
Date: Thu, 12 Dec 85 12:19:48 MST
From: hpcnou!dat (Dave Taylor)
Subject: A Worm Example
[Since discussing risks to the public of computers is the topic of another
group (Risks@SRI-NIC) I'm going to request people to try to keep the
conversation to information that's more directly related to our topics
of discussion. Before we do, though, I can't resist this message...]
Dave,
I do know of a worm deliberately installed in commercial software.
A few months back I got a very worried call from my home builder, who
explained that their firm had purchased an IBM PC and an accounting package
from a Colorado company to run on it. On Jan. 1 of the new year, suddenly
all their records were inaccessable. They made a phone call to their software
vendor, who explained that (for a hefty price) a new software package could
be obtained which would allow the records to be read and processing to continue.
The building company had not been informed of this "feature" when they
purchased the original software package. I read their purchase agreements
and manuals and did not find any mention of the worm.
As far as I can tell, this is extortion. I advised the builder to roll over
and buy the new software (they had a business to run).
Marc Clarke, Loveland Technology Center
------------------------------
Date: Thu, 12 Dec 85 12:26:22 MST
From: hpcnou!dat (Dave Taylor)
Subject: Heavy topics
Just when I thought it was okay to continue with our normal topics
of discussion in this group...
I just received a call from someone at the National Computer Security
Center in Virginia (who wishes to keep a low profile) discussing the
points raised in the note from Gary North.
Unfortunately, this person couldn't offer many reassuring words and
rather had comments like "If it's true [worms/viruses in bank software]
*pause* then there's some [reason for] concern..."
Another comment made was "If there's collusion [at the places
writing the software] there's nothing we can do about it..."
The more interesting question raised by all this, and (amazingly)
relevent to the group, is - How would people react if they knew
about the potential dangers of our newly computerized society?
What are the social/economic and political ramifications of the
computer industry informing people that machines are NOT omniscient?
The person at NCSC promised to have someone from the US Treasury
department contact me about the potential threats due to the
lack of reliability (perceived or otherwise) so one can only assume
that this is pretty serious stuff!!
-- Dave Taylor
------------------------------
Date: Fri, 13 Dec 85 16:41:48 MST
From: hpcnou!dat (Dave Taylor)
Subject: More thoughts on information
Harry Rubin has some interesting points about my oversimplified
approach to discussing the difference between the value and the cost of
information. As he points out, there are (at least) two issues that
deserve seperate discussion that were lumped into one - the volume of
information and the rights of the organizations transmitting the information.
For an individual there are a number of factors that determine
the 'value' of an item of information, including:
o timeliness (something we haven't touched on at all yet)
o brevity (ie the succintness of the information presentation)
o truth (another issue we haven't discussed yet)
o relevence to their personal interests
o relevence to their professional interests/job
o interest (these last three are indeed mutually exclusive
sometimes)
o agreement or disagreement with the bias of the presenter
(especially true on the anonymous medium of electronic
mail and notes)
I'd like to begin a further discussion of these factors -
1. Timeliness
The area where the computer communications systems can really
outdo more conventional information transmission sources is in the speed
that the messages can be propagated throughout the professional com-
munity. Imagine a network of people working on IC fabrication where
they can all compare notes and discuss breakthroughs NOW, not over the
course of months or years...it'd be wonderful!
Somehow, though, this all has to tie in to reality, and the
most noticable areas where timeliness is affected are in the willing-
ness of companies to spend money to propagate information that can be
either irrelevent or actually counter-beneficial to their business in
a timely (expensive) manner. For example, most Unix sites transmit
messages during off-hours only due to the phone rate structure and the
work load during the day. This means that the timeliness of each
individual message is adversely impacted for profit reasons.
The second area where reality steps in is in the free transmission
of information at all - why would my company want me to tell fellow IC
designers about my wonderful breakthrough? We'd want to hoard the
knowledge and exploit it for profit (leading to 'knowledge merchants',
but that's another topic entirely).
Suffice to say that the timeliness of the electronic information
is a significant part of it's relative value to the individual.
2. Brevity
As E.B. White stated an astounding number of times - "Be Succinct!"
Unfortunately, a lot of the information that is recieved via the electronic
medium (like this digest :-) tends to the overly windy and long. Part of
the problem is that people still have a hard time expressing themselves
clearly on phosphor, but part of it is that any random twit can send
lots of information to the world at large (a power play, one presumes).
On the USENET there are all too many examples of this behaviour. Without
going to heavily into the psychological reasons or prevention techniques,
it is clear that the shorter something is (given the same information
presented clearly) the better.
3. Truth
One of the last great issues to be discussed about the
electronic information transmission systems is the security problem
of misinformation (deliberate or otherwise) of the public. There is,
unfortunately, no easy way to rate the truth of an iota of information.
Any thoughts or ideas?
4. Relevence/Interest
This is probably the most obvious area where the individual
rates the value of the information they recieve - summed up as "Do
I care?"
5. The Presentation Bias
As a great newscaster once said "EVERYone's biased....everyone
but me.".
Let's be honest - due to the informal and anonymous nature of
electronic communications, people are much more opinionated than they
would normally be in person. This gives lead to some pretty astounding
biases to information and rather vicious personal attacks on people
who present information with other biases. This is a problem for
the individual receiving information, however. It ties in with the
problem of determining the truth of something - how can one know that
they shouldn't discount some information due to the bias of the presenter,
or perhaps that they should seriously ponder something presented in a
lackadaisical manner?
This isn't any different from printed media, I suppose, except
for the fact that there are a LOT more sources of information on a
network. It's harder to know 'where someone is coming from' if you
can't associate them with a group or anything.
I encourage further discussion of these issues...
-- Dave Taylor
-----------------------------------
To have your item included in this digest, please mail it to any
of the addresses; ihnp4!hplabs!taylor, or taylor@HPLABS.CSNET.
You can also simply respond to this mailing.
-----------------------------------
End of Computers and Society Digest
-----------------------------------