The Risks Digest

The RISKS Digest

Forum on Risks to the Public in Computers and Related Systems

ACM Committee on Computers and Public Policy, Peter G. Neumann, moderator

Volume 2 Issue 55

Wednesday, 28 May 1986

Contents

o Culling through RISKS headers; SDI
Jim Horning
o Blind Faith in Technology, and Caspar Weinberger
Herb Lin
o Risks of doing software quality assurance too diligently
PGN from Chris Shaw and the Torrance Daily Breeze
o Collegiate jungle
Mike McLaughlin
o Decease and Desist -- Death by Computer
Deborah L. Estrin
o The Death of the Gossamer Time Traveler
Peter G. Neumann
o Computer Ethics
Bruce A. Sesnovich
o Info on RISKS (comp.risks)

SDI; Culling through RISKS headers [Message entirely edited]

Jim Horning <horning@src.DEC.COM>
Tue, 27 May 86 11:51:06 pdt
    [[Jim and several others called my attention to an article in the
      NYTimes of 27 May 86, page 9.  I have excerpted from the article, as
      follows.  PGN ]


          "Feasible Computer Control For Missile Shield Doubted"
              by Charles Mohr (Special to the New York Times)

  "An expert [Jim Horning] in computer programs who was asked to advise on
  research into defense against long-range nuclear missiles says he is
  skeptical that a reliable computer system to control such a defense can
  ever be devised."

The article quotes from a letter from Jim Horning to Douglas Waller (on the
staff of Senator William Proxmire):

  "To date no system of this complexity has performed as expected (or
  hoped) in its first full-scale operational test; no one has advanced
  any reason to expect that an S.D.I. would either.  A huge system that
  is intended to be used at most once, and cannot be realistically tested
  in advance of use, simply cannot be trusted."

The article also quotes a statement signed by 36 of the 61 experts who
attended a workshop on computing March 16-19 at Pacific Grove CA:

  "The effective defense from nuclear annihilation of the lives, homes and
  property of the American people, as embodied by the Strategic Defense
  Initiative (Star Wars), requires highly reliable computer systems of
  unprecedented complexity.  As experts in reliable computing, we strongly
  believe that a system meeting these requirements is technologically
  infeasible."

The article notes Dave Parnas' role in the ongoing discussions, and also

  "Lieut. Gen. James A. Abrahamson, the director of the missile defense
  organization, has said that computer programming was probably the most
  difficult technical problem faced by his group.  But he stresses an
  optimistic view that it can be solved and argues that Mr. Parnas has
  applied "unrealistically high criteria"."

  Mr. Horning, who like Mr. Parnas has written computer programs for
  weapons systems, is supportive of Mr. Parnas, observing that "there has
  been a movement toward Parnas' position" among those knowledgeable about
  technology.

The article also quotes from Jim Horning's "trip report" to participate in a
meeting of the Strategic Defense Initiative Organization (see RISKS-1.2, 28
August 85).   END of PGN excerpting.]

     [Wow, it is 9 months to the day since RISKS-1.2, and we've had 99
      issues (not counting the "pilot issue", RISKS-1.1, on 1 Aug 85).
      I hope we are not overwhelming you, but I also hope we can keep
      up the generally good quality of contributions.  PGN]


Blind Faith in Technology, and Caspar Weinberger

<LIN@XX.LCS.MIT.EDU>
Sun, 25 May 1986 17:45 EDT
On the blind faith in technology, it is interesting to note that, when
initial reports came in after the bombing of Libya that U.S. bombers had
hit the French Embassy, Weinberger said,

        "That's impossible.  They weren't ordered to do that."


Risks of doing software quality assurance too diligently

Peter G. Neumann <Neumann@SRI-CSL.ARPA>
Wed 28 May 86 21:02:44-PDT
From the Torrance Daily Breeze, 19 May 1986, page 1, courtesy of Chris Shaw:

              Death threats dog fired whistleblower
                (by James Hart, Aerospace writer)

  Finding a new job after getting fired can be hard enough, but Edward F.
  Wilson never expected the death threats.  Wilson, a computer software
  engineer fired narly a year ago from a small Hawthorne-based aerospace
  company, says he's paying the price for speaking out against government
  contracting abuses.  The threats -- anonymous, of course -- have come over
  the telephone twice in recent weeks at his Long Beach home...
  "Whistleblowing, I'm afraid, is not very popular," he said with a sigh.

  He said that soon after being asked ... to draw up software quality-assurance
  programs required by the government, he realized that Amex Systems officials
  were doing it strictly for show.  "They said to me on several occasions that
  they had no intention of implementing them," he said.

The article goes on to document Wilson's memo to his employer, his being
fired for "being a troublemaker", his filing a wrongful discharge suit, the
ensuing criminal investigation currently underway on unnamed government
programs, various denials, etc.  Dina Rasor, director of the Project on
Military Procurement, a self-styled watchdog agency in Washington D.C. spoke
about the situation:

  "I've heard of whistleblowers being blackballed from the industry and of
  government whistleblowers put in 'do-nothing' jobs, but in five years of
  working with these people I've never had anyone receive a death threat.
  ...  What I've found is so unusual about Ed Wilson is that he made his
  complaints known to the company well before he was fired.  He hasn't
  brought all this up later as sour grapes."

  Wilson said he remains optimistic he will eventually find a job, but
  admits his "faith in the system is diminishing."  "I did what I thought
  was in the best interests of the country," he said.


Collegiate jungle

Mike McLaughlin <mikemcl@nrl-csr>
Tue, 27 May 86 08:42:00 edt
Darwinian selection will solve the backup problem on campus.  Them that
backs up will survive, them that don't, won't.

Permission is granted to delete "campus" and insert any other sphere of
computer-supported activity presently known or yet to be discovered.

    Mike McLaughlin <mikemcl@nrl-csr.arpa>


Decease and Desist -- Death by Computer

Deborah L. Estrin <estrin%usc-cseb@usc-cse.usc.edu>
Mon, 26 May 86 18:43:45 pdt
An editorial appeared in yesterday's (Saturday's) LA Times.  It is written
by Forman Brown, on the subject of computer error.

Following are a few exerpts:

  "I first became aware of my death last May when my checks began to bounce.
  Never having experienced bouncing checks before, and knowing that I had
  quite a respectable balance at the bank, I was both shocked and angry.  When
  I examined the returned checks and found, stamped over my signature on each
  of them, in red ink, "Deceased", I was mystified. Then, when one of the
  recipients of my checks, a utility company, demanded that I appear in
  person, cash in hand, plus $10 for their trouble--their trouble--I was
  shocked, angry and mystified. I wondered just how they expected us deceased
  to acquiesce."

Well, to paraphrase, Brown went to the bank, the series of tellers could not
believe such a thing had happened and said it was probably the computer's
fault and sent him home to write new checks and explanations--including one
to a friend who thought he was dead due to the "deceased" notice on the
bounced check.

Then the next month he found that his social security payment was not
credited to his account. On investigation he found that whatever troubled
the computers "had spread to those of the Social Security system as well."
This went on for a couple of months despite visits to Social Security.  Then
finally the bank agreed to credit the amount to his account until Social
Security started payment again--which they did several months later.

Brown thought the story was over until his physician contacted him recently
to say that Medicare had refused to accept his bill for services rendered
becuase the date of the service was six months later than the date of the
patient's decease...

He concludes by saying that if he were 20, all this might merely be
irritating, but since he is 85 the prospect of death is too near to be
treated lightly.


The Death of the Gossamer Time Traveler

Peter G. Neumann <Neumann@SRI-CSL.ARPA>
Wed 28 May 86 22:08:47-PDT
Dr. Paul MacCready has had some marvelous successes, including the first and
only human-powered flight across the English Channel in 1979 on his Gossamer
Condor.  His Time Traveler, a short-winged model of the prehistoric
Quetzalcoatlus northropi from 65 million years ago, had made something like
43 consecutive safe flights and starred in a film, "On the Wing",
replicating the original appearance and flying style of QN.  Weighing in at
44 pounds, it includes battery-operated motors, a computerized autopilot,
and ground-based radio controls.  Unfortunately, the bird chose the day of
its first public appearance, 17 May 86 at Andrews Air Force Base, to have
its head break off.  Computer archaeologists of the future will of course
try to ascertain whether the accident was due to human error in overtaxing
the creature, or to a computer program bug in the safety controls that might
have otherwise have prevented flight instability, or some other cause.  We
hope that the head crash can be repaired.  The construction cost, variously
reported as $500,000 and $700,000, was funded by the National Air and Space
Museum and the Johnson Wax Company.  [Maybe this was inspired by its more
modern precursor, the "one-SEATER WAX-WING".]

Your roving [raving or raven'?] reporter, PGN


Computer Ethics

<rti-sel!dg_rtp!rtp41!dg_rama!bruces%mcnc.csnet@CSNET-RELAY.ARPA>
Tue, 27 May 86 13:36:39 edt
The following is a copy of a review I wrote for a recent newsletter of the
Boston chapter of Computer Professionals for Social Responsibility (CPSR).
Readers of RISKS may be interested, as well.

METAPHILOSOPHY is a British journal published three times yearly which is
dedicated to considerations about particular schools, fields, and methods of
philosophy.  The October 1985 issue, Computers & Ethics (Volume No. 16, Issue
No. 4), is recommended reading [...].

This issue's articles attempt to define and delimit the scope of Computer
Ethics, and examine several emerging and current concerns within the field.

One current concern is responsibility for computer-based errors.  In his
article on the subject, John W. Snapper asks:  "...whether it is advisable to
...write the law so that a machine is held legally liable for harm." The author
invokes Aristotle's "Nichomachean Ethics" (!) in an analysis of how computers
make decisions, and what is meant by "decision" in this context.

On the same subject, William Bechtel goes one step further, considering the
possibility that computers could one day bear not only legal, but moral
responsibility for decision-making:  "When we have computer systems that ...can
be embedded in an environment and adapt their responses to that environment,
then it would seem that we have captured all those features of human beings
that we take into account when we hold them responsible."

Deborah G. Johnson discusses another concern:  ownership of computer programs.
In "Should Computer Programs Be Owned?," Ms. Johnson criticizes utilitarian
arguments for ownership, as well as arguments based upon Locke's labor theory
of property. The proper limits to extant legal protections, including
copyrights, patents, and trade secrecy laws, are called into question.

Other emerging concerns include the need to educate the public on the dangers
and abuses of computers, and the role of computers in education.  To this end,
Philip A. Pecorino and Walter Maner present a proposal for a college level
course in Computer Ethics, and Marvin J. Croy addresses the ethics of
computer-assisted instruction.

Dan Lloyd, in his provocative but highly speculative article, "Frankenstein's
Children," envisions a world where cognitive simulation AI succeeds in
producing machine consciousness, resulting in a possible ethical clash of the
rights of artificial minds with human values.

The introductory article, James H. Moor's "What is Computer Ethics," is an
ambitious attempt to define Computer Ethics, and to explain its importance.
According to Moor, the development and proliferation of computers can rightly
be termed "revolutionary":  "The revolutionary feature of computers is their
logical malleability.  Logical malleability assures the enormous application of
computer technology." Moor goes on to assert that the Computer Revolution, like
the Industrial Revolution, will transform "many of our human activities and
social institutions," and will "leave us with policy and conceptual vacuums
about how to use computer technology."

An important danger inherent in computers is what Moor calls "the invisibility
factor." In his own words:  "One may be quite knowledgeable about the inputs
and outputs of a computer and only dimly aware of the internal processing."
These hidden internal operations can be intentionally employed for unethical
purposes; what Moor calls "Invisible abuse,"  or can contain "Invisible
programming values":  value judgments of the programmer that reside, insidious
and unseen, in the program.

Finally, in the appendix, "Artificial Intelligence, Biology, and Intentional
States," editor Terrell Ward Bynum argues against the concept that "intentional
states" (i.e. belief, desire, expectation) are causally dependent upon
biochemistry, and thus cannot exist within a machine.

If you're at all like me, you probably find reading philosophy can be "tough
going," and METAPHILOSOPHY is no exception.  References to unfamiliar works,
and the use of unfamiliar terms occasionally necessitated my reading
passages several times before extracting any meaning from them.  The topics,
however, are quite relevant and their treatment is, for the most part,
lively and interesting.  With its well-written introductory article, diverse
survey of current concerns, and fairly extensive bibliography, this issue of
METAPHILOSOPHY is an excellent first source for those new to the field of
Computer Ethics.

[METAPHILOSOPHY, c/o Expediters of the Printed Word Ltd., 515 Madison Avenue,
Suite 1217, New York, NY  10022]

Bruce A. Sesnovich         mcnc!rti-sel!dg_rtp!sesnovich
Data General Corp.         rti-sel!dg_rtp!sesnovich%mcnc@csnet-relay.arpa
Westboro, MA               "Problems worthy of attack
                            prove their worth by hitting back"

Please report problems with the web pages to the maintainer

Top