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 3 Issue 15

Sunday, 29 June 1986

Contents

o A Personal View on SDI from Harlan Mills
Herb Lin
o Having an influence from "within the system"
Herb Lin
o Re: Research programs that pay for themselves
Rich Cowan
o Text Scanners
Fred Hapgood
o Info on RISKS (comp.risks)

A Personal View on SDI from Harlan Mills

<LIN@XX.LCS.MIT.EDU>
Sun, 29 Jun 1986 06:12 EDT
On the whole, I am touched by Harlan Mills' remarks.  But I am bothered by
two things.  He says that 

    I [Mills] regard SDI as a political question that will be ultimately
    settled in our political system by the 525 members of our Congress.
    I trust them to make the *WISEST* disposition possible of this question.

    I depend on the Congress
    to make the final, collective, decisions, in how to *BEST* reflect that
    strength for peace in political, diplomatic, and military matters.
    [Emphasis added by me]

These comments reflect a trust in a rational process of government
that I wish I could share; it almost sounds as though he believes that
whatever decision the Congress makes will be right *by definition*.  I
have seen too many instances in which Congress manifestly did NOT do
the right thing to believe in their collective wisdom.  The nature of
a democratic system forces me to *abide* by their decisions, but that
is not the same thing as approving of them or believing in their
wisdom. (On the other hand, I would not trade democracy for anything
else.)

At a somewhat more fundamental level, he states that 

    .. it is somewhat misleading to convert
    the problem of SDI feasibility into the question of software perfection.
    ... The best man can do in
    any physical system is to reduce the probability of failure to low levels,
    not to zero.

The latter statement is a position with which all TECHNICAL analysts
agree: a perfect system is impossible.  But the POLITICAL debate has
been cast in terms of "Do you want to defend yourself or not?",
"eliminating (NOT reducing) the threat of nuclear ballistic missiles"
and "the immorality of threats to kill innocent civilians".  

The technical analysis of the political questions posed above is
absolutely clear, and is that it is impossible to develop technology
that will allow us to get rid of offensive nuclear weapons and shrug
off nuclear missiles should they happen to be launched our way).
Technical analysts then debate the technically more interesting
question of what CAN be done, in which case Mills' comment that

    ... the intent
    of most scientists and engineers working on SDI is to explore the technical
    side intelligently enough to provide the widest range of options possible
    for the political and diplomatic side.

makes a great deal of sense.

But SDI supporters in the political arena find THIS question much less
interesting.  The support that SDI garners from the population at
large, and indeed from those that push it arises from the fact that
defense against ballistic missiles is a truly revolutionary
possibility, that will result in a military posture that is
qualitatively different from that which exists at present.  It won't,
as SDI supporters admit when pushed; they say defenses will enhance
deterrence, and that we will still have to accept societal
vulnerability and to rely on the threat of retaliation to deter Soviet
attack.

Looking at the question from another side, all technical analysts
agree that it is possible to build SOMETHING that sometimes does some
fraction of what you want it to do, and the interesting technical
questions are what is the nature of this something, what will it be
able to do, and how often can it do it.  But the political debate is
cast against the backdrop of technology that is capable of meeting a
certain absolute level of performance, and a rather high one at that.
The technology to do THAT is much more demanding -- if the level of
performance is societal perfection, then it's not reachable at all.
The political proponents try to have it both ways; they want the
political support that comes from belief in the feasibility of this
very demanding technology, and they try to deflect technical criticism
of this political position by saying the question is one of
discovering what technology can do.

Thus, until the broader political debate can be recast in terms of the
desirability of IMPERFECT defenses, and SDI supporters concede
POLITICALLY that defenses will not do what is being claimed for it,
technical analysts, in my view, are fully justified in pointing out
that perfection is not possible.  When SDI supporters make this
concession, the perfect defense issue will become a dead horse
politically as well as technically, and we can all go on to talk about
more interesting things.


Having an influence from "within the system"

<LIN@XX.LCS.MIT.EDU>
Sat, 28 Jun 1986 17:52 EDT
    From: Richard A. Cowan <COWAN>

    You have here touched upon what I believe is -- more often than not -- a
    delusion:  that it is more effective to work within the system to change
    it than to protest it from without.  

Without addressing the specific merits of doing SDI work at this time,
I think this statement needs qualification.  

There is a role for people outside the system.  There is also one for
people inside the system.  Activists are necessary to bring political
pressure.  But they have to have some technical credibility.  As bad
as things are in government now (with people believing in the Tooth
Fairy,.. excuse me, I meant perfect ballistic missile defense), there
is only minimal support for other things that other people would also
like to have -- teaching creationism in the schools for one.  The
reason is that there is NO serious scientific opinion that creationism
has any literal validity at all.  I can assure you that if there were,
the battle to keep creationism out of the textbooks would be a lot
more difficult to fight.

Technical credibility is not the same thing as being "inside the
system".  But "the system" does many things, some of which are
probably right, and others wrong.  But should that mean that people
should give up on the whole thing?  Some of the most effective critics
of the system are those who have extensive experience in it -- Richard
Garwin comes to mind as a prime example.  His effectiveness comes
about because he knows what he is talking about, and it is hard to
imagine that he could have developed his expertise had he remained
forever outside the system.  By contrast, Kosta Tsipis -- while he has
made a rather significant name for himself in the public domain -- has
been identified in most of the public debate that I have heard as a
flake who instinctively knee-jerks against US defense; Tsipis has
never been part of "the system".  (This is not to make a judgement
about the quality of Tsipis' work.)

Then why doesn't the system stop doing silly things?  I guess the
answer has to take the form -- if you think things are bad now, just
imagine how much worse they would be without the likes of Garwin.
While being technically right doesn't necessarily mean that your
position will win, being technically wrong is often the kiss of death.


Re: Research programs that pay for themselves

Matthew P. Wiener <weemba@brahms.berkeley.edu>
Sun, 29 Jun 86 02:47:49 pdt
I'd like to add a small comment to Richard Cowan's remarks.

One concern about SDI spinoffs is that DoD gets to choose some of
them.  I wonder if, for example, we are going to see more incidents
like the ASATing of Solar Max--a fully working scientific satellite
whose routine operating grant renewal was turned down last summer
to provide a suitable test target.

ucbvax!brahms!weemba    Matthew P Wiener/UCB Math Dept/Berkeley CA 94720


Text Scanners

"Fred Hapgood" <SIDNEY.G.HAPGOOD%OZ.AI.MIT.EDU@XX.LCS.MIT.EDU>
Sat 28 Jun 86 06:33:34-EDT
The archetypal computer risk is of course unemployment. With regard to this
issue, does anyone know what sort of inroads page and form scanners are or
are not making into the data entry industry, and what features are pacing or
retarding penetration into that market?  Or would anyone have any
suggestions of whom I might call to find out more?
            [Please respond privately to Fred unless your 
             response has RISKS-related implications.  PGN]

Please report problems with the web pages to the maintainer

Top