exploit the possibilities
Home Files News &[SERVICES_TAB]About Contact Add New

ap.htm

ap.htm
Posted Dec 21, 1999

ap.htm

tags | encryption
SHA-256 | ff5ef9917d003276848f6d7bbd6704a630a4543cbf416f8170f20073e23b50df

ap.htm

Change Mirror Download
<!--

Author: Infowar.Com

-->
<HTML>
<HEAD>
<TITLE>Assassination Politics</TITLE>
</HEAD>
<BODY BGCOLOR="#ffffff">
<P ALIGN=Center>
3 April 1997<BR>
Source:
<A HREF="http://www.infowar.com/class_1/BELL1.html-ssi">http://www.infowar.com/class_1/BELL1.html-ssi</A>
<P>
<HR>
<H1 ALIGN=Center>
Assassination Politics
</H1>
<P ALIGN=Center>
<B>by Jim Bell </B><BR>
<H2>
<B>Part 1</B>
</H2>
<P>
I've been following the concepts of digital cash and encryption since I read
the article in the August 1992 issue of <I>Scientific American</I> on"encrypted
signatures." While I've only followed the Digitaliberty area for a few weeks,
I can already see a number of points that do (and should!) strongly concern
the average savvy individual:
<P>
1. How can we translate the freedom afforded by the Internet to ordinary
life?
<P>
2. How can we keep the government from banning encryption, digital cash,
and other systems that will improve our freedom?
<P>
A few months ago, I had a truly and quite literally "revolutionary" idea,
and I jokingly called it "Assassination Politics": I speculated on the question
of whether an organization could be set up to legally announce that it would
be awarding a cash prize to somebody who correctly "predicted" the death
of one of a list of violators of rights, usually either government employees,
officeholders, or appointees. It could ask for anonymous contributions from
the public, and individuals would be able send those contributions using
digital cash.
<P>
I also speculated that using modern methods of public-key encryption and
anonymous "digital cash," it would be possible to make such awards in such
a way so that nobody knows who is getting awarded the money, only that the
award is being given. Even the organization itself would have no information
that could help the authorities find the person responsible for the prediction,
let alone the one who caused the death.
<P>
It was not my intention to provide such a "tough nut to crack" by arguing
the general case, claiming that a person who hires a hit man is not guilty
of murder under libertarian principles. Obviously, the problem with the general
case is that the victim may be totally innocent under libertarian principles,
which would make the killing a crime, leading to the question of whether
the person offering the money was himself guilty.
<P>
On the contrary; my speculation assumed that the "victim" is a government
employee, presumably one who is not merely taking a paycheck of stolen tax
dollars, but also is guilty of extra violations of rights beyond this.
(Government agents responsible for the Ruby Ridge incident and Waco come
to mind.) In receiving such money and in his various acts, he violates the
"Non-aggression Principle" (NAP) and thus, presumably, any acts against him
are not the initiation of force under libertarian principles.
<P>
The organization set up to manage such a system could, presumably, make up
a list of people who had seriously violated the NAP, but who would not see
justice in our courts due to the fact that their actions were done at the
behest of the government. Associated with each name would be a dollar figure,
the total amount of money the organization has received as a contribution,
which is the amount they would give for correctly "predicting" the person's
death, presumably naming the exact date. "Guessers" would formulate their
"guess" into a file, encrypt it with the organization's public key, then
transmit it to the organization, possibly using methods as untraceable as
putting a floppy disk in an envelope and tossing it into a mailbox, but more
likely either a cascade of encrypted anonymous remailers, or possibly
public-access Internet locations, such as terminals at a local library, etc.
<P>
In order to prevent such a system from becoming simply a random unpaid lottery,
in which people can randomly guess a name and date (hoping that lightning
would strike, as it occasionally does), it would be necessary to deter such
random guessing by requiring the "guessers" to include with their "guess"
encrypted and untraceable "digital cash," in an amount sufficiently high
to make random guessing impractical.
<P>
For example, if the target was, say, 50 years old and had a life expectancy
of 30 years, or about 10,000 days, the amount of money required to register
a guess must be at least 1/10,000th of the amount of the award. In practice,
the amount required should be far higher, perhaps as much as 1/1000 of the
amount, since you can assume that anybody making a guess would feel sufficiently
confident of that guess to risk 1/1000th of his potential reward.
<P>
The digital cash would be placed inside the outer "encryption envelope,"
and could be decrypted using the organization's public key. The prediction
itself (including name and date) would be itself in another encryption envelope
inside the first one, but it would be encrypted using a key that is only
known to the predictor himself. In this way, the organization could decrypt
the outer envelope and find the digital cash, but they would have no idea
what is being predicted in the innermost envelope, either the name or the
date.
<P>
If, later, the "prediction" came true, the predictor would presumably send
yet another encrypted "envelope" to the organization, containing the decryption
key for the previous "prediction" envelope, plus a public key (despite its
name, to be used only once!) to be used for encryption of digital cash used
as payment for the award. The organization would apply the decryption key
to the prediction envelope, discover that it works, then notice that the
prediction included was fulfilled on the date stated. The predictor would
be, therefore, entitled to the award. Nevertheless, even then nobody would
actually know WHO he is!
<P>
It doesn't even know if the predictor had anything to do with the outcome
of the prediction. If it received these files in the mail, in physical envelopes
which had no return address, it would have burned the envelopes before it
studied their contents. The result is that even the active cooperation of
the organization could not possibly help anyone, including the police, to
locate the predictor.
<P>
Also included within this "prediction-fulfilled" encryption envelope would
be unsigned (not-yet-valid) "digital cash," which would then be blindly signed
by the organization's bank and subsequently encrypted using the public key
included. (The public key could also be publicized, to allow members of the
public to securely send their comments and, possibly, further grateful
remuneration to the predictor, securely.) The resulting encrypted file could
be published openly on the Internet, and it could then be decrypted by only
one entity: The person who had made that original, accurate prediction. The
result is that the recipient would be absolutely untraceable.
<P>
The digital cash is then processed by the recipient by "unbinding" it, a
principle which is explained in far greater detail by the article in the
August 1992 issue of Scientific American. The resulting digital cash is
absolutely untraceable to its source.
<P>
This overall system achieves a number of goals. First, it totally hides the
identity of the predictor to the organization, which makes it unnecessary
for any potential predictor to "trust" them to not reveal his name or location.
Second, it allows the predictor to make his prediction without revealing
the actual contents of that prediction until later, when he chooses to, assuring
him that his "target" cannot possibly get early warning of his intent (and
"failed" predictions need never be revealed). In fact, he needs never reveal
his prediction unless he wants the award. Third, it allows the predictor
to anonymously grant his award to anyone else he chooses, since he may give
this digital cash to anyone without fear that it will be traced.
<P>
For the organization, this system also provides a number of advantages .By
hiding the identity of the predictor from even it, the organization cannot
be forced to reveal it, in either civil or criminal court. This should also
shield the organization from liability, since it will not know the contents
of any "prediction" until after it comes true. (Even so, the organization
would be deliberately kept "poor" so that it would be judgment-proof.) Since
presumably most of the laws the organization might be accused of violating
would require that the violator have specific or prior knowledge, keeping
itself ignorant of as many facts as possible, for as long as possible, would
presumably make it very difficult to prosecute.
<P>
<HR>
<H2>
Part 2
</H2>
<P>
"At the Village Pizza shop, as they were sitting down to consume a pepperoni,
Dorothy asked Jim, 'So what other inventions are you working on?" Jim replied,
'I've got a new idea, but it's really evolutionary. Literally REVOLUTIONARY.'
'Okay, Jim, which government are you planning to overthrow?,' she asked,
playing along.
<P>
'All of them,' answered Jim."
<P>
<B>Political Implications </B>
<P>
Imagine for a moment that as ordinary citizens were watching the evening
news, they see an act by a government employee or officeholder that they
feel violates their rights, abuses the public's trust, or misuses the powers
that they feel should be limited. A person whose actions are so abusive or
improper that the citizenry shouldn't have to tolerate it.
<P>
What if they could go to their computers, type in the miscreant's name, and
select a dollar amount: The amount they, themselves, would be willing to
pay to anyone who "predicts" that officeholder's death. That donation would
be sent, encrypted and anonymously, to a central registry organization, and
be totaled, with the total amount available within seconds to any interested
individual. If only 0.1% of the population, or one person in a thousand,
was willing to pay $1 to see some government slimeball dead, that would be,
in effect, a $250,000 bounty on his head.
<P>
Further, imagine that anyone considering collecting that bounty could do
so with the mathematical certainty that he could not be identified, and could
collect the reward without meeting, or even talking to, anybody who could
later identify him. Perfect anonymity, perfect secrecy, and perfect security.
And that, combined with the ease and security with which these contributions
could be collected, would make being an abusive government employee an extremely
risky proposition. Chances are good that nobody above the level of county
commissioner would even risk staying in office.
<P>
Just how would this change politics in America? It would take far less time
to answer, "What would remain the same?" No longer would we be electing people
who will turn around and tax us to death, regulate us to death, or for that
matter sent hired thugs to kill us when we oppose their wishes.
<P>
No military?
<P>
One of the attractive potential implications of such a system would be that
we might not even need a military to protect the country. Any threatening
or abusive foreign leader would be subject to the same
contribution/assassination/reward system, and it would operate just as
effectively over borders as it does domestically.
<P>
This country has learned, in numerous examples subsequent to many wars, that
once the political disputes between leaders has ceased, we (ordinary citizens)
are able to get along pretty well with the citizens of other countries. Classic
examples are post-WWII Germany, Japan, and Italy, and post-Soviet Russia,
the Eastern bloc, Albania, and many others.
<P>
Contrary examples are those in which the political dispute remains, such
as North Korea, Vietnam, Iraq, Cuba, Red China, and a few others. In all
of these examples, the opposing leadership was NOT defeated, either in war
or in an internal power struggle. Clearly, it is not the PEOPLE who maintain
the dispute, but the leadership.
<P>
Consider how history might have changed if we'd been able to "bump off" Lenin,
Stalin, Hitler, Mussolini, Tojo, Kim Il Sung, Ho Chi Minh, Ayatollah Khomeini,
Saddam Hussein, Moammar Khadafi, and various others, along with all of their
replacements if necessary, all for a measly few million dollars, rather than
the billions of dollars and millions of lives that subsequent wars cost.
<P>
But that raises an interesting question, with an even more interesting answer.
"If all this is so easy, why hasn't this been done before?" I mean, wars
are destructive, costly, and dangerous, so why hasn't some smart politician
figured out that instead of fighting the entire country, we could just 'zero'
the few bad guys on the top?
<P>
The answer is quite revealing, and strikingly "logical": If we can kill THEIR
leaders, they can kill OUR leaders too. That would avoid the war, but the
leadership on both sides would be dead, and guess who is making the decisions
about what to do? That's right, the LEADERS!
<P>
And the leaders (both theirs and ours!) would rather see 30,000,000 ordinary
people die in WWII than lose their own lives, if they can get away with it.
Same in Korea, Vietnam, the Gulf War, and numerous other disputes around
the globe. You can see that as long as we continue to allow leaders, both
"ours" and "theirs," to decide who should die, they will ALWAYS choose the
ordinary people of each country.
<P>
One reason the leaders have been able to avoid this solution is simple: While
it's comparatively easy to "get away with murder," it's a lot harder to reward
the person who does it, and that person is definitely taking a serious risk.
(Most murders are solved based on some prior relationship between the murder
and victim, or observations of witnesses who know either the murderer or
the victim.)
<P>
Historically, it has been essentially impossible to adequately motivate an
assassin, ensuring his safety and anonymity as well, if only because it has
been impossible to PAY him in a form that nobody can trace, and to ensure
the silence of all potential witnesses. Even if a person was willing to die
in the act, he would want to know that the people he chooses would get the
reward, but if they themselves were identified they'd be targets of revenge.
<P>
All that's changed with the advent of public-key encryption and digital cash.
Now, it should be possible to announce a standing offer to all comers that
a large sum of digital cash will be sent to him in an untraceable fashion
should he meet certain "conditions," conditions which don't even have to
include proving (or, for that matter, even claiming) that he was somehow
responsible for a death.
<P>
I believe that such a system has tremendous implications for the future of
freedom. Libertarians in particular (and I'm a libertarian) should pay particular
attention to the fact that this system "encourages" if not an anarchist outcome,
at least a minarchist (minimal government) system, because no large governmental
structure could survive in its current form.
<P>
In fact, I would argue that this system would solve a potential problem,
occasionally postulated, with the adoption of libertarianism in one country,
surrounded by non-libertarian states. It could have reasonably been suspected
that in a gradual shift to a libertarian political and economic system, remnants
of a non-libertarian system such as a military would have to survive, to
protect society against the threats represented by foreign states. While
certainly plausible, it would have been hard for an average naive person
to imagine how the country would maintain a $250 billion military budget,
based on voluntary contributions.
<P>
The easy answer, of course, is that military budgets of that size would simply
not happen in a libertarian society. More problematic is the question of
how a country would defend itself, if it had to raise its defenses by voluntary
contribution. An equally simplistic answer is that this country could probably
be defended just fine on a budget 1/2 to 1/3 of the current budget. True,
but that misses the point.
<P>
The real answer is even simpler. Large armies are only necessary to fight
the other large armies organized by the leadership of other, non-libertarian
states, presumably against the will of their citizenry. Once the problem
posed by <I>their</I> leadership is solved (as well as ours; either by their
own citizenry by similar anonymous contributions, or by ours), there will
be no large armies to oppose.
<P>
<HR>
<H2>
Part 3
</H2>
<P>
In the 1960's movie, "The Thomas Crown Affair," actor Steve McQueen plays
a bored multi-millionaire who fights tedium by arranging well-planned high-yield
bank robberies. He hires each of the robbers separately and anonymously,
so that they can neither identify him nor each other. They arrive at the
bank on schedule, separately but simultaneously, complete the robbery, then
separate forever. He pays each robber out of his own funds, so that the money
cannot be traced, and he keeps the proceeds of each robbery.
<P>
In my recent essay generally titled "Digitaliberty," or earlier "Assassination
politics," I hypothesized that it should be possible to LEGALLY set up an
organization which collects perfectly anonymous donations sent by members
of the public, donations which instruct the organization to pay the amount
to any person who correctly guesses the date of death of some named person,
for example some un-favorite government employee or officeholder. The
organization would total the amounts of the donations for each different
named person, and publish that list (presumably on the Internet) on a daily
or perhaps even an hourly basis, telling the public exactly how much a person
would get for "predicting" the death of that particular target.
<P>
Moreover, that organization would accept perfectly anonymous, untraceable,
encrypted "predictions" by various means, such as the Internet (probably
through chains of encrypted anonymous remailers), U.S. mail, courier, or
any number of other means. Those predictions would contain two parts: A small
amount of untraceable "digital cash," inside the outer "digital envelope,"
to ensure that the "predictor" can't economically just randomly choose dates
and names, and an inner encrypted data packet which is encrypted so that
even the organization itself cannot decrypt it. That data packet would contain
the name of the person whose death is predicted, and the date it is to happen.
<P>
This encrypted packet could also be published, still encrypted, on the Internet,
so as to be able to prove to the world, later, that SOMEBODY made that prediction
before it happened, and was willing to "put money on it" by including it
outside the inner encrypted "envelope." The "predictor" would always lose
the outer digital cash; he would only earn the reward if his (still-secret)
prediction later became true. If, later on, that prediction came true, the
"lucky" predictor would transmit the decrypt key to the organization,
untraceably, which would apply it to the encrypted packet, and discover that
it works, and read the prediction made hours, days, weeks, or even months
earlier. Only then would the organization, or for that matter anyone else
except the predictor, know the person or the date named.
<P>
Also included in that inner encrypted digital "envelope" would be a public
key, generated by the predictor for only this particular purpose: It would
not be his "normal" public key, obviously, because that public key would
be traceable to him. Also present in this packet the predictor has earned.
(This presentation could be done indirectly, by anintermediary, to prevent
a bank from being able to refuse to deal with the organization.)
<P>
Those "digital cash" codes will then be encrypted using the public key included
with the original prediction, and published in a number of locations, perhaps
on the Internet in a number of areas, and available by FTP to anyone who's
interested. (It is assumed that this data will somehow get to the original
predictor. Since it will get to "everyone" on the Internet, it will presumably
be impossible to know where the predictor is.) Note, however, that only the
person who sent the prediction (or somebody he's given the secret key to
in the interim) can decrypt that message, and in any case only he, the person
who prepared the digital cash blanks, can fully "unbind" the digital cash
to make it spendable, yet absolutely untraceable. (For a much more complete
explanation of how so-called "digital cash" works, I refer you to the August
1992 issue of Scientific American.)
<P>
This process sounds intricate, but it (and even some more detail I haven't
described above) is all necessary to:
<P>
1. Keep the donors, as well as the predictors, absolutely anonymous, not
only to the public and each other, but also to the organization itself, either
before or after the prediction comes true.
<P>
2. Ensure that neither the organization, nor the donors, nor the public,
is aware of the contents of the "prediction" unless and until it later becomes
true. (This ensures that none of the other participants can be "guilty" of
knowing this, before it happens.)
<P>
3. Prove to the donors (including potential future predictors), the organization,
and the public that indeed, somebody predicted a particular death on a particular
date, before it actually happened.
<P>
4. Prove to the donors and the public (including potential future predictors)
that the amount of money promised was actually paid to whoever made the
prediction that later came true. This is important, obviously, because you
don't want any potential predictor to doubt whether he'll get the money if
he makes a successful prediction, and you don't want any potential donor
to doubt that his money is actually going to go to a successful predictor.
<P>
5. Prevent the organization and the donors and the public from knowing, for
sure, whether the predictor actually had anything to do with the death predicted.
This is true even if (hypothetically) somebody is later caught and convicted
of a murder, which was the subject of a successful "prediction": Even after
identifying the murderer through other means, it will be impossible for anyone
to know if the murderer and the predictor were the same person.
<P>
6. Allow the predictor, if he so chooses, to "gift" the reward (possibly
quite anonymously) to any other person, one perhaps totally unaware of the
source of the money, without anyone else knowing of this.
<P>
Even the named "target" (the "victim") is also assured of something: He his
best "friend," could collect the reward, absolutely anonymously, should they
"predict" his death correctly. At that point, he will have no friends.
<P>
This may represent the ultimate in compartmentalization of information: Nobody
knows more than he needs to, to play his part in the whole arrangement. Nobody
can turn anyone else in, or make a mistake that identifies the other
participants. Yet everyone can verify that the "game" is played "fairly":
The predictor gets his money, as the donors desire. Potential future predictors
are satisfied (in a mathematically provable fashion) that all previous successful
predictors were paid their full rewards, in a manner that can't possibly
be traced. The members of the public are assured that, if they choose to
make a donation, it will be used as promised. This leads me to a bold assertion:
I claim that, aside from the practical difficulty and perhaps, theoretical
impossibility of identifying either the donors or the predictor, it is very
likely that none of the participants, with the (understandable) hypothetical
exception of a "predictor" who happens to know that he is also a murderer,
could actually be considered "guilty" of any violation of black-letter law.
Furthermore, none of the participants, including the central organization,
is aware, either before or after the "prediction" comes true, that any other
participant was actually in violation of any law, or for that matter would
even know (except by watching the news) that any crime had actually been
committed.
<P>
After all, the donors are merely offering gifts to a person who makes a
successful prediction, not for any presumed responsibility in a killing,
and the payment would occur even if no crime occurred. The organization is
merely coordinating it all, but again isolating itself so that it cannot
know from whom the money comes, or to whom the money eventually is given,
or whether a crime was even committed. (Hypothetically, the "predictor" could
actually be the "victim," who decides to kill himself and "predict" this,
giving the proceeds of the reward to his chosen beneficiary, perhaps a relative
or friend. Ironically, this might be the best revenge he can muster, "cheating
the hangman," as it were.)
<P>
In fact, the organization could further shield itself by adopting a stated
policy that no convicted (or, for that matter, even SUSPECTED) killers could
receive the payment of a reward. However, since the recipient of the reward
is by definition unidentified and untraceable even in theory, this would
be a rather hollow assurance since it has no way to prevent such a payment
from being made to someone responsible.
<P>
<HR>
<H2>
Part 4
</H2>
<P>
In part 3, I claimed that an organization could quite legally operate, assisted
by encryption, international data networking, and untraceable digital cash,
in a way that would (indirectly) hasten the death of named people, for instance
hated government employees and officeholders. I won't attempt to "prove"
this, for reasons that I think will be obvious. First, even if such an operation
were indeed "legal," that fact alone would not stop its opponents from wanting
to shut it down. However, there is also another way of looking at it: If
this system works as I expect it would, even its claimed "illegality" would
be irrelevant, because it could operate over international borders and beyond
the legal reach of any law-abiding government.
<P>
Perhaps the most telling fact, however, is that if this system was as effective
as it appears it would be, no prosecutor would dare file charges against
any participant, and no judge would hear the case, because no matter how
long the existing list of "targets," there would always be room for one or
two more. Any potential user of this system would recognize that an assault
on this system represents a threat to its future availability, and would
act accordingly by donating money to target anyone trying to shut it down.
<P>
Even so, I think I should address two charges which have been made, apparently
quite simplistically, claiming that an implementation of this idea would
violate the law. Specifically: "Conspiracy to commit murder" and "misprision
of felony."
<P>
As I understand it, in order to have a "conspiracy" from a criminal standpoint,
it is necessary to have at least two people agree to commit a crime, and
have some overt act in furtherance of that crime.
<P>
Well, this charge already "strikes out" because in the plan I described,
none of the participants <I>agrees</I> with ANYONE to commit a crime. None
of the participants even informs anyone else that he will be committing a
crime, whether before or after the fact. In fact, the only crime appears
(hypothetically; this assumes that a crime was actually committed) to be
a murder committed by a single individual, a crime unknown to the other
participants, with his identity similarly unknown.
<P>
Remember, the "prediction" originally sent in by the predictor was fully
encrypted, so that the organization (or anyone else, for that matter) would
be unable to figure out the identity of the person whose death was predicted,
or the date on which it was predicted to occur. Thus, the organization is
incapable of "agreeing" with such a thing, and likewise the donors as well.
Only if the prediction later came true would the decrypt key arrive, and
only then would the organization (and the public) be made aware of the contents.
Even then, it's only a "prediction," so even then, nobody is actually aware
of any crime which can be associated with the predictor.
<P>
"Misprision of Felony"
<P>
This crime, sort of a diluted form of "accessory before and/or after the
fact," was claimed to qualify by "Tim of Angle," who subsequent to my answer
to him on this subject has totally failed to support his initial claim. (A
recent curiosity is that this crime is one that has been charged against
Michael Fortier, the person who claims he helped OKC bombing suspect Tim
McVeigh "case the joint" at the Federal building.)
<P>
I include it here, nevertheless, because his simplistic (and un-careful)
reading of my idea led him to perhaps the "closest" law that one might allege
that the participants would have broken. Tim claimed: No. That's called
"misprision of felony" and makes you an accessory before the fact. Arguably,
under the felony murder rule you could get TOA> capital punishment in
a state that has such.
<P>
However, I did a little library research, checking <I>Black's Law
Dictionary</I>. Here is the entry for this item: "Misprision of felony. The
offense of concealing a felony committed by another, but without such previous
concert with or subsequent assistance to the felon as would make the party
concealing an accessory before or after the fact. <I>United States v.
Perlstein</I>, C.C.A.N.J., 126 F.2d 789, 798. Elements of the crime are that
the principal committed and completed the felony alleged, that the defendant
had full knowledge of that fact, that the defendant failed to notify the
authorities, and that defendant took an affirmative step to conceal the crime.
<I>U.S. v. Ciambrone</I>, C.A. Nev., 750 F.2d 1416, 1417. Whoever, having
knowledge of the actual commission of a felony recognizable by a court of
the United States, conceals and does not as soon as possible make known the
same to some judge or other person in civil or military authority under the
United States, is guilty of the federal crime of misprision of felony. 18
U.S.C.A 4." See also Obstructing Justice in <I>Black's Law Dictionary</I>.
<P>
The only "element" of this crime which is arguably satisfied is the first:
Some person other than the defendant for "misprision of felony") committed
a crime. The second element fails miserably: "...that the defendant had full
knowledge of that fact... " My previous commentary makes it clear that far
from "full knowledge of that fact," other participants are carefully prevented
from having ANY "knowledge of that fact." The third element, "..that the
defendant failed to notify the authorities..." is also essentially non-existent:
No other participants have any information as to the identity of a predictor,
or his location, or for that matter whether he has had any involvement in
any sort of crime. In fact, it would be possible for each of the other
participants to deliver (anonymously, presumably) copies of all correspondence
they have sent, to the police or other agency, and that correspondence would
not help the authorities even slightly to identify a criminal or even necessarily
a crime.
<P>
In fact, normal operation of this organization would be to publicize "all"
correspondence it receives, in order to provide feedback to the public to
assure them that all participants are fulfilling their promises and receiving
their rewards. This publication would presumably find its way to the police,
or it could even be mailed to them on a "fail[ing] to notify authorities."
Nevertheless, none of this material could help any authorities with their
investigations, to their dismay.
<P>
The fourth and last element of the crime of "misprision of felony", "...and
that defendant took an affirmative step to conceal the crime," would totally
fail. The organization would not " conceal" the crime. In fact, it will have
no ability to do anything to the contrary, if for no other reason that it
<I>has</I> no knowledge of the crime! And as described above, it would carefully
avoid having access to any information that could help solve the crime, and
thus it would escape any obligations along these lines.
<P>
Summary:
<P>
In hindsight, it is not surprising that such an organization could operate
legally within the U.S., although at least initially not without political
opposition. First, this is at least nominally supposed to be a "free country,"
which should mean that police and other authorities aren't able to punish
behavior just because they don't like it.
<P>
Secondly, it is obvious that most laws today were originally written during
an era in which laws assumed that "conspirators" at least knew each other,
had met each other, could identify each other, or had (at least!) talked
to each other. On the contrary, in my scenario none of the participants even
know on what continent any of the others reside, let alone their country,
city, or street. They don't know what they look like, sound like, or for
that matter even "type like": None of their prose, save a few sparse
"predictions," ever gets communicated to anyone else, so even text-comparison
programs would fail to "target" anyone.
<P>
Equally surprising (to those who originally wrote the laws against "conspiracy")
would be "Person A's" ability to satisfy himself that "Person B" deserves
the award, without knowing that "Person B" is (or is not) actually responsible
for a particular death.
<P>
<HR>
<H2>
Part 5
</H2>
<P>
In the previous four notes on the subject of Digitaliberty, I've suggested
that this concept (collecting anonymous donations to, in effect, "purchase"
the death of an un-favorite government employee) would force a dramatic reduction
of the size of government at all levels, as well as achieving what will probably
be a "minarchist" (minimal government) state at a very rapid rate. Furthermore,
I pointed out that I thought that this effect would not merely affect a single
country or continent, but might in fact spread through all countries essentially
simultaneously.
<P>
But in addition to such (apparently) grandiose claims, it occurs to me that
there must be other changes to society that would simultaneously occur with
the adoption of such a system. After all, a simplistic view of my idea might
lead one to the conclusion that there would be almost no governmental structure
left after society had been transformed. Since our current "criminal justice
system" today is based totally on the concept of "big government," this would
lead a naive person to wonder how concepts such as "justice," "fairness,"
"order," and for that matter protection of individual rights can be accomplished
in such a society.
<P>
Indeed, one common theme I've seen in criticisms of my idea is the fear that
this system would lead to "anarchy." The funny thing about this objection
is that, technically, this could easily be true. But "anarchy" in real life
may not resemble anything like the "anarchy" these people claim to fear,
which leads me to respond with a quote whose origin I don't quite remember:
<P>
"Anarchy is not lack of order. Anarchy is lack of ORDERS."
<P>
People presumably will continue to live their lives in a calm, ordered manner.
Or, at least as calm and ordered as they WANT to. It won't be "wild in the
streets," and they won't bring cannibalism back as a national sport, or anything
like that.
<P>
It occurs to me that probably one of the best ways to demonstrate that my
idea, "assassination politics" (perhaps ineptly named, in view of the fact
that its application is far greater than mere politics), would not result
in "lack of order" is to show that most if not all of the DESIRABLE functions
of the current so-called "criminal justice system" will be performed after
its adoption. This is true even if they will be accomplished through wholly
different methods and, conceivably, in entirely different ways than the current
system does.
<P>
I should probably first point out that it is not my intention to re-write
the book of minarchist theory. I would imagine that over the years, there
has been much written about how individuals and societies would function
absent a strong central government, and much of that writing is probably
far more detailed and well-thought-out than anything I'll describe here.
<P>
One reason that ALMOST ANY "criminal justice system" would be better and
more effective than the one we currently possess is that, contrary to the
image that officialdom would try to push, anyone whose job depends on "crime"
has a strong vested interest in <I>maintaining</I> a high level of crime,
not eliminating it. After all, a terrorized society is one that is willing
to hire many cops and jailers and judges and lawyers, and to pay them high
salaries. A safe, secure society is not willing to put up with that. The
"ideal" situation, from the limited and self-interested standpoint of the
police and jailers, is one that maximizes the number of people in prison,
yet leaves most of the really dangerous criminals out in the streets, in
order to maintain justification for the system. That seems to be exactly
the situation we have today, which is not surprising when you consider that
the police have had an unusually high level of input into the "system" for
many decades.
<P>
The first effect of my idea would be, I think, to generally eliminate
prohibitions against acts which have no victims, or "victimless crimes."
Classic examples are laws against drug sales and use, gambling, prostitution,
pornography, etc. That's because the average (unpropagandized) individual
will have very little concern or sympathy for punishing an act which does
not have a clear victim. Without a large, central government to push the
propaganda, the public will view these acts as certainly not "criminal,"
even if still regarded as generally undesirable by a substantial minority
for a few years. Once you get rid of such laws, the price of currently illegal
drugs would drop dramatically, probably by a factor of 100. Crime caused
by the need to get money to pay for these drugs would drop drastically, even
if you assume that drug usage increased due to the lowering of the price.
<P>
Despite this massive reduction in crime, perhaps as much as 90%, the average
person is still going to want to know what "my system" would do about the
residual, "real" crime rate. You know, murder, rape, robbery, burglary, and
all that. Well, in the spirit of the idea, a simplistic interpretation would
suggest that an individual could target the criminal who victimizes him,
which would put an end to that criminal career.
<P>
Some might object, pointing out that the criminal is only identified in a
minority of crimes. That objection is technically correct, but it's also
a bit misleading. The truth is that the vast majority of "victim"-type crime
is committed by a relatively tiny fraction of the population who are repeat
criminals. It isn't necessary to identify For example, even if the probability
of a car thief getting caught, per theft, is only 5%, there is at least a
40% probability of getting caught after 10 thefts, and a 65% chance after
20 thefts. A smart car-theft victim would be happy to donate money targeting
ANY discovered car-thief, not necessarily just the one who victimized him.
<P>
The average car-owner would be wise to offer such donations occasionally,
as "insurance" against the possibility of his being victimized someday: An
average donation of 1 cent per day per car would constitute $10,000 per day
for a typical city of 1 million cars. Assuming that amount is far more than
enough to get a typical car thief's "friends" to "off" him, there is simply
no way that a substantial car-theft subculture could possibly be maintained.
<P>
Another alternative is that insurance companies would probably get into the
act: Since they are going to be the financial victims of thefts of their
insured's property, it is reasonable to suppose that they would be particularly
inclined to deter such theft. It is conceivable that current-day insurance
companies would transmogrify themselves into investigation/deterrence agencies,
while maintaining their insurance role, in view of the fact that they have
the most to lose. This is particularly true because if "assassination politics"
(as applied to criminals and crime) comes about, they could then actually
DO SOMETHING about the problem, rather than merely reporting on the statistics
to their customers and stockholders.
<P>
Such companies would also have a strong motivation to provide a workable
system of rewards for solving crimes and identifying criminals, rewards that
(naturally enough!) can be given out totally anonymously.
<P>
While I would like to talk about the other advantage of this new kind of
justice, the fact that politicians and other government employees would no
longer have de-facto immunity in most cases, the reality is that since we
would no longer HAVE "politicians and other government employees," to mention
that advantage would be redundant.
<P>
The principle is valid, however: In today's system, you can have people known
to be guilty of crimes, but not prosecuted because they are part of "the
system." Classic examples would be heroes of the right (Oliver North) and
heroes of the left (Jim Wright) who either escape prosecution or conviction
for "political" or "bureaucratic" reasons. With "assassination politics"
that would simply never happen.
<P>
<HR>
<H2>
Part 6
</H2>
<P>
A frequent initial belief among people who have recently heard of my
"assassination politics" idea is the fear that this system will somehow be
"out of control": It would end up causing the death of ordinary, "undeserving"
people.
<P>
This system, however, will not be without its own kind of "control. "Not
a centralized control, decidable by a single individual, but a decentralized
system in which everyone gets an implicit "vote." A good analogy might be
to consider a society in which everyone's house thermostat is controlled
to operate at a temperature which is set for the entire country. Each person's
control input is taken as a "vote," whether to get hotter, colder, or to
stay the same temperature. The central control computer adjusts the national
setpoint temperature in order to equalize the number of people who want the
temperature colder and hotter. Each house is at the same, nationally set
temperature, however. Clearly, no one individual is in control of the setting.
Nevertheless, I think it would be generally agreed that this system would
never produce a REALLY "off the wall" temperature setting, simply because
so many people's inputs are used to determine the output. Sure, if a group
of 10,000 kids decided (assisted by the Internet) together to screw with
the system, and they all set their houses' thermostat inputs to "hotter,"
they could SLIGHTLY increase the overall setting, but since there are probably
about 100 million separate dwellings in the U.S., their fiddlings will be
drowned out by the vast majority of the population's desires. Is this system
"out of control"? True, it is out of the "control" of any single individual,
but nevertheless it is well within the control of the population as a whole.
<P>
It turns out that "assassination politics" actually has a rather similar
control mechanism to the one I've described above. First, I've pointed out
that if I were to operate a centralized system such as this, I'd only accept
donations naming people who are in violation of the "Non-Initiation Of Force
Principle" (NIOFP), well known to libertarians. By this standard, government
employees (who have accepted paychecks paid for with funds stolen from citizenry
by taxes) and criminals whose crimes actually had a victim would be included.
Let's call this hypothetical organization "Organization A," or OrgA for short.
<P>
True, somebody else might be a little less scrupulous, accepting donations
for the termination of ANYBODY regardless of whether he "deserves" his fate
(call them "Organization B," or OrgB, for short.) Most potential donors (who,
I suggest, would have "typical" levels of scruples) would see that if they
patronize OrgB, their interests wouldn't be protected. For example, OrgB
(if it survives and thrives) might later come back to target them, because
of some other donor. OrgA would not. Naturally, our "ethical" donors don't
want this, so they would choose to give their donation to the most "ethical"
organization that will accept it. This maximizes the donors' benefit, and
minimizes the potential harm.
<P>
Since BOTH organizations will accept donations for "deserving" victims, while
only OrgB will accept them for "just anybody," it is reasonable to conclude
that (capitalism being what it is) OrgB's rates (the percentage of the price
it keeps as profit) can be and will be higher for its donations (that's because
there is less competition in its area of specialization.) Thus, it would
be more economical to target "deserving" people through OrgA , and thus donors
will be drawn to it. In addition, OrgA will become larger, more credible,
believable and trustworthy, and more potential "guessers" (assassins?) will
"work" its system, and for lower average potential payments (all else being
equal.) Even so, and ironically, the average donation level for people listed
by OrgA would likely be higher, since (if we assume these are "deserving"
people) more people will be contributing towards their demise.
<P>
After all, if a potential donor wants to "hit" some government bigwig, there
will be PLENTY of other donors to share the cost with. Millions of donations
of $1 to $10 each would be common and quite economical. On the other hand,
if you just selected a target out of the telephone directory, an "undeserving"
target, you'll probably be the only person wanting to see him dead, which
means that you'll probably have to foot the whole bill of perhaps $5K to
$10K if you want to see any "action. " Add to that OrgB 's "cut," which will
probably be 50%, and you're talking $10K to $20K. I contend that the likelihood
of this kind of thing actually happening will be quite low, for "undeserving
victims."
<P>
Now, the die-hards among you will probably object to the fact that even this
tiny residual possibility is left. But consider: Even <I>today</I> it would
be quite "possible" for you to pick a name randomly out of a list, find him
and kill him yourself. Does this frequently happen? Apparently not. For just
one thing, there's no real motive. Unless you can show that the application
of "assassination politics" would dramatically increase the likelihood of
such incidents, I suggest that this "problem" will likely not be a problem
after all.
<P>
For a while, I thought that the "lack of a motive" protection was momentarily
overturned by a hypothetical: I thought, suppose a person used this system
as part of a sophisticated extortion scheme, in which he sends an anonymous
message to some rich character, saying something like "pay me a zillion dollars
anonymously, or I put out a digital contract on you." For a while, this one
had me stumped. Then, I realized that an essential element in this whole
play was missing: If this could be done ONCE, it could be done a dozen times.
And the victim of such an extortion scheme has no assurance that it won't
happen again, even if he pays off, so ironically he has no motivation to
pay off the extortion. Think about it: The only reason to make the payment
is to remove the threat. If making the payment can't guarantee to the target
that the threat is removed, he has no reason to make the payment. And if
the target has no reason to make the payment, the extortionist has no reason
to make the threat!
<P>
Another, related (and equally simplistic) fear is that political minorities
will be preferentially targeted. For example, when I pointed out that
"establishment" political leaders would probably "go" quite quickly, one
wag suggested to me that "libertarian leaders" could likewise be targeted.
Such a suggestion reflects a serious misunderstanding of political philosophy,
and libertarians in particular: I consider it obvious (to me, at least) that
libertarians NEED no leaders. (You don't need leaders if you don't want to
control a population, or achieve political power. The only reason libertarians
"need" leaders today is to take places in the government and (then) to shut
it down.) And if my idea is implemented, "libertarian leaders" represent
no more of a threat to anyone than the average libertarian citizen.
<P>
Fully recognizing this, another (and far more credible) person thought a
while, and in a proud revelation suggested that one way that the establishment
would "fight back" is to convert to a government that is based on fully
decentralized authority, as opposed to the leader-centric system we have
today. Such a system could not be attacked by killing individual people,
any more than you can kill a tree by pulling off a single leaf. His "solution"
was, in effect, to totally disband the current government and turn it over
to the public at large, where it highly de-centralized system that is not
controlled by a tiny fraction of the population in a structure called a
"government," essentially identical to his idea. So in effect, the only way
the government can survive is to totally surrender. And once it surrenders,
the people win. And in practice, it will have no alternative.
<P>
Will this idea be "out of control"? To a great extent, that depends on what
your definition of the word "control." I have come to believe that "assassination
politics" is a political Rorschach (ink-blot) test: What you think of it
is strongly related to your political philosophy.
<P>
<HR>
<H2>
Part 7
</H2>
<P>
Dear libertarian Friend,
<P>
I very much understand the concerns you voiced about my idea which I call,
"Assassination Politics," because this essay is nothing if it is not radical
and extreme. I wrote it, in the middle of last year, partly because I think
libertarianism and libertarians in particular need to address what is, if
not a contradiction," is at least an intolerable reality: On the one hand,
we are told not to initiate aggression, but on the other we are aggressed
against by the government every time it collects a tax.
<P>
I much appreciate the way some people I know have "dropped out" of the system,
and the guts that such a tactic requires. But that's the problem, I think:
Only those with the "guts" do it, which gives the government fewer targets
so that it can spend more time attacking the few who oppose it. The reality
is that the government STILL collects taxes, and it STILL uses that money
to violate our rights. We all know that's wrong.
<P>
My position is quite simple: If tax collection constitutes aggression, then
anyone doing it or assisting in the effort or benefiting from the proceeds
thereof is a criminal. This is quite analogous to current law which prosecutes
co-conspirators. While I am not holding out "current law" as some sort of
gold-standard of reasonableness that we must always accept, on the other
hand I think it's plausible to use it to show that once we have come to the
conclusion that taxation is theft, the prescription follows directly by a
form of reasoning allegedly acceptable to society: It is reasonable to "attack
the attackers" and their co-conspirators, and everyone who is employed by
the government is thus a co-conspirator, even if he is not directly involved
in the collection of those taxes. That's because he IS involved in
<I>benefiting</I> from the proceeds of these taxes, and he presumably provides
a certain level of "backup" to the young thugs that governmental organizations
often hire.
<P>
I realize, and you should too, that the "non-aggression principle" says nothing
about the EXTENT of the self-defense/retaliation that one might reasonably
employ in defending one's own rights: In a sense, that sounds like an omission
because it at least suggests that a person might "unreasonably" defend himself
with lethal force when far less drastic means might normally be called for.
For what it's worth, I think most people will behave responsibly. But I think
it is pretty straightforward to argue that whatever means are necessary to
stop the attack, are reasonable given the terms of the non-aggression principle:
If a given means are known to be inadequate to actually stop the attack,
then further and more serious means are reasonable and called-for.
<P>
To set up a reasonable analogy, if I'm walking down the canonical "dark alley"
and am accosted by a man wielding a knife threatening me with it, it is
presumably reasonable for me to pull a gun and threaten back, or possibly
take the encounter to the final conclusion of gunfire. Even if I should choose
to hold my fire and test to determine whether my actions deterred him, I
can't see that this possibility binds me morally. And should he advance,
despite the gun, as if to attack, I should feel no remorse in shooting him
and taking myself out of danger. If you accept the premises so far, you
apparently accept the principle that escalation of the self-defense/retaliation
is reasonable as long as if the current level of returned counter-threat
is inadequate to stop the aggression initiated by the other party. To believe
otherwise is to believe that ultimately, you are obligated to accept a certain
high level of aggression simply because you do not have the resources (yet)
to resist it. I totally reject this concept, as I hope you would.
<P>
So if, hypothetically, I could have an anonymous conversation with a hard-nosed
government employee, and asked him, "If I killed one of your agents, would
you stop trying to collect that tax from me," his predictable reaction would
be, "no, we would continue to try to collect that tax." In fact, he would
probably hasten to add that he would try to have me prosecuted for murder,
as well! If I were to ask if killing ten agents would stop them, again they
would presumably say that this would not change their actions.
<P>
The conclusion is, to me, obvious: Clearly, there is no practical limit to
the amount of self-defense that I would need to protect my assets from the
government tax collector, and to actually stop the theft, so I suggest that
logic requires that I be morally and ethically allowed (under libertarian
principles) to use whatever level of self-defense I choose.
<P>
You raised another objection, that quite frankly I believe is invalid. I
believe you implied that until a specific level of escalation is reached
( such as the Feds showing up on your doorstep, etc) then it is not legitimate
to defend oneself. Delicately, I must disagree. As we all well know, government
ultimately operates primarily not on actual, applied force, but simply the
threat of future force if you do not comply. True, there are people who have
decided to call the government's bluff and simply drop out, but the reality
is that this is not practical for most individuals today. This is no accident:
The government makes it difficult to drop out, because they extort the
cooperation of banks and potential employers and others with which you would
otherwise be able to freely contract. In any case, I fail to see how not
"dropping out" makes one somehow morally obligated to pay a tax (or tolerate
the collection of one). I trust you did not inadvertently mean to suggest
this.
<P>
The reason, morally, we are entitled to shoot the mugger if he waves the
knife in our face is that he has threatened us with harm, in this case to
our lives, but the threat the government represents to the average citizen
(loss of one's entire assets) is just as real, albeit somewhat different.
Since government is a past reality, and a present reality, and has the immediate
prospects of being a future reality as well, I sincerely believe that the
average citizen can legitimately consider himself CONTINUOUSLY threatened.
The aggression has already occurred, in continuously occurring, and has every
prospect of continuing to occur. If anything would justify fighting back,
this would.
<P>
To continue the analogy, if you've been repeatedly mugged by the same guy
down the same dark alley for each day of last month, that DOES NOT mean that
you've somehow consented to the situation, or that your rights to your assets
have somehow been waived. With my "Assassination Politics" essay, I simply
proposed that we (as libertarians as well as being ordinary citizens) begin
to treat aggression by government as being essentially equivalent to aggression
by muggers, rapists, robbers, and murderers, and view their acts as a continuing
series of aggressions. Seen this way, it should not be necessary to wait
for their NEXT aggression; they will have always have been aggressing and
they will always BE aggressing, again and again, until they are stopped for
good.
<P>
At that point, the question shifted to one of practicality: Sure, theoretically
we might morally have the "right" to protect ourselves with lethal force,
but if they have any reputation at all, government agents have a habit of
showing up in large numbers when they actually apply direct force. To take
a position that you can only defend yourself when <I>they've</I> chosen the
"where" and "when" of the confrontation is downright suicidal, and I hope
you understand that I would consider any such restriction to be highly unfair
and totally impractical. Understand, too, that the reason we're still stuck
under the thumb of the government is that to the extent it's true, "we've"
been playing by THEIR rules, not by our own. By our own rules, THEY are the
aggressors and we should be able to treat them accordingly, on our own terms,
at our own convenience, whenever we choose, especially when we feel the odds
are on our side.
<P>
I understand, obviously, that the "no initiation of aggression" principle
is still valid, but please recognize that I simply don't consider it to be
a valid counter-argument to "Assassination Politics," at least as applied
to targets who happen to be government agents. They've "pre-aggressed," and
I don't see any limit to the defenses I should be able to muster to stop
that aggression completely and permanently. Not that I don't see a difference
between different levels of guilt: I fully recognize that some of them are
far worse than others, and I would certainly not treat a lowly Forest Service
grunt in the same fashion as an ATF sniper.
<P>
Now, there is one more thing that I would hope we could get straight: As
I originally "invented" this system, it occurred to me that there could be
certain arguments that it needed to be "regulated" somehow; "unworthy" targets
shouldn't be killed, etc. The "problem" is, what I've "invented" may (as
I now believe it to be) actually a "discovery," in a sense: I now believe
this kind of system was always inevitable, merely waiting for the triad of
the Internet, digital cash, and good encryption in order to provide the technical
underpinnings for the entire system. If that is genuinely the case, then
there is no real way to control it, except by free-market principles.
<P>
It would be impossible, for example, to set up some sort of "Assassination
Politics Dictator," who decides who will live and who will die, because
competition in the system will always rise to supply every demand, albeit
at possibly a very high price. And if you believe the maxim that "absolute
power corrupts absolutely," you wouldn't want to accept any form of centralized
control (even, perhaps, that of your own!), because any such control would
eventually be corrupted. Most rational people recognize this, and I do too.
I would not have invented a system where "Jim Bell" gets to make "all the
decisions." Quite the contrary, the system I've described absolutely prevents
such centralization. That, quite frankly, is the novelty and dare I say it,
the beauty of this idea. I believe that it simply cannot be hijacked by
centralized political control.
<P>
As I pointed out in the essay, if <I>I</I> were running one of the organizations
accepting those donations and offering those prizes, I would selectively
list only those targets who I am genuinely satisfied are guilty of the violation
of the "non-aggression principle." But as a practical matter, there is no
way that I could stop a DIFFERENT organization from being set up and operating
under DIFFERENT moral and ethical principles, especially if it operated
anonymously, as I anticipate the "Assassination Politics"-type systems will
be. Thus, I'm forced to accept the reality that I can't dictate a "strongly
limited" system that would "guarantee" no "unjustified" deaths: I can merely
control my little piece of the earth and not assist in the abuse of others.
I genuinely believe, however, that the operation of this system would be
a vast improvement over the status quo.
<P>
This, I argue, is somewhat analogous to an argument that we should be entitled
to own firearms, despite the fact that SOME people will use them
wrongly/immorally/illegally. The ownership is a right even though it may
ultimately allow or enable an abuse that you consider wrong and punishable.
I consider the truth of such an argument to be obvious and correct, and I
know you would too.
<P>
I realize that this lacks the crisp certitude of safety which would be reassuring
to the average, "pre-libertarian" individual. But you are not the "average
individual" and I trust that as long-time libertarians you will recognize
rights must exist even given the hypothetical possibility that somebody may
eventually abuse them.
<P>
I do not know whether I "invented" or "discovered" this system; perhaps it's
a little of both. I do genuinely believe that this system, or one like it,
is as close to being technologically inevitable as was the invention of firearms
once the material we now know as "gunpowder" was invented. I think it's on
the way, regardless of what we do to stop it. Perhaps more than anyone else
on the face of this planet, this notion has filled me, sequentially and then
simultaneously, with awe, astonishment, joy, terror, and finally, relief.
<P>
Awe, that a system could be produced by a handful of people that would rid
the world of the scourge of war, nuclear weapons, governments, and taxes.
Astonishment, at my realization that once started, it would cover the entire
globe inexorably, erasing dictatorships both fascistic and communistic,
monarchies, and even so-called "democracies," which as a general rule today
are really just the facade of government by the special interests. Joy, that
it would eliminate all war, and force the dismantling not only of all nuclear
weapons, but also all militaries, making them not merely redundant but also
considered universally dangerous, leaving their "owners" no choice but to
dismantle them, and in fact no reason to KEEP them!
<P>
Terror, too, because this system may just change almost EVERYTHING how we
think about our current society, and even more for myself personally, the
knowledge that there may some day be a large body of wealthy people who are
thrown off their current positions of control of the world's governments,
and the very-real possibility that they may look for a "villain" to blame
for their downfall. They will find one, in me, and at that time they will
have the money and (thanks to me, at least partially) the means to see their
revenge. But I would not have published this essay if I had been unwilling
to accept the risk.
<P>
Finally, relief. Maybe I'm a bit premature to say it, but I'm satisfied we
<I>will</I> be free. I'm convinced there is no alternative. It may feel like
a roller-coaster ride on the way there, but as of today I think our destination
is certain. Please understand, we <I>will</I> be free.
<P>
Your libertarian friend,
<P>
Jim Bell
<P>
jimbell@pacifier.com
<P>
Something is going to happen... Something... Wonderful!
<P>
<HR>
<H2>
Part 8
</H2>
<P>
The following article appeared in the Sunday, February 4, 1996 issue of
<I>Asahi Evening News</I>, in an article written by columnist Paul Maxwell,
page 6. He writes a regular column about the Internet for this newspaper.
<P>
"Networks: Paul Maxwell"
<P>
"Dial Internet for murder"
<P>
'The first thing we do, let's kill all the lawyers." (Shakespeare, Henry
VI).
<P>
A startling and controversial idea has surfaced on the Internet recently--fear
with me for a moment while I explain it. It is based on two technological
developments: digital cash and encryption software.
<P>
Briefly, digital cash is a system for transferring funds from one person
to another on the Net. For this system to be as good as cash, the transactions
must be capable of being conducted anonymously, just like in real life. (You
go into the Seven-Eleven, buy a Cafe Latte, and nobody knows your name or
your credit history. The purchase is not recorded in a database of your consumer
preferences.)
<P>
Several competing schemes for digital cash have been launched, but the one
that eventually gains universal acceptance will surely have this anonymity
feature.
<P>
The second innovation is a kind of software called public-key encryption.
It allows you to send a file or an email message that is "locked" in such
a way that it can only be opened by the intended recipient. The recipient,
however, cannot open it until given a "key." This "key" may then be used
to encrypt a return message that can only be opened by the original sender.
<P>
Freelance visionary and tinkerer Jim Bell has been following both of these
developments for the past few years. Recently, he asked himself a couple
of tough questions: "How can we translate the freedom afforded by the Internet
to ordinary life?" How can we keep government from banning encryption, digital
cash, and other systems that will improve our freedom?"
<P>
Suddenly, Bell had a revolutionary idea. ("Revolutionary" is the word he
uses, and it fits.) You and me--the little guys, the ordinary working people
of the world--could get together, all pitch in, and pay to have every rotten
scoundrel in politics assassinated. And we could do it legally. Sort of.
Bell imagined an organization that would award "a cash prize to somebody
who correctly 'predicted' the death of one of a list of violators of rights,
usually either government employees, officeholders, or appointees. It could
ask for anonymous contributions from the public, and individuals would be
able to send those contributions using digital cash."
<P>
He explains that "using modern methods of public-key encryption and anonymous
digital cash, it would be possible to make such awards in such a way so that
nobody knows who is getting awarded the money, only that the award is being
given. Even the organization itself would have no information that could
help the authorities find the person responsible for the prediction, let
alone the one who caused the death. "Are you following this? Let's say that
we, the public, decide we've finally had enough of [insert name of villain].
Ten dollars from me, ten from you--suddenly there's a million dollars in
a fund. The money will go to the first person who can "predict" the date,
time, and circumstances of the villain's death. Obviously, this information
is only known in advance by the assassin.
<P>
He sends an anonymous, "locked" message. He kills the villain. He sends the
"key" to the message. He has, without ever revealing his identity, "correctly
predicted" the murder. The "key" that he has provided is then used to "lock
the award money in a file that is then publicly posted on the Internet. Only
the person who originated the key may open the file and claim the digital
cash.
<P>
In other words, public anger could finance cash awards for assassinations.
The organization that collected the money and announced a list of possible
targets would never know about a crime in advance, and would never know the
identity or whereabouts of a criminal. It would not technically be guilty
of conspiracy or complicity.
<P>
Jim Bell has thought about this a lot, and feels that the idea is technically
feasible, practical, even foolproof. Suppose for a moment he's right? What
are the implications?
<P>
World leaders live with the threat of assassination every day of their lives.
But at the local level, this could really have an impact. And the "target"
list wouldn't necessarily to politicians--any offensive public personality
would be fair game. Picture yourself a year from now, sitting around with
friends. Somebody says, "Remember when Juice Newton got whacked?" And you
say, "Yeah--best ten bucks I ever spent."
<P>
Satisfying as it might be to declare war on asinine pop singers, Bell has
a more civic-minded suggestion: Let's kill all the car thieves. He reasons
that a very small number of career criminals are responsible for nearly all
car thefts. If one million car owners in a given metropolitan area contributed
just four dollars a year, it would create $10,000 a day in "prize money"
for the "predictor" of any car thief's death.
<P>
"Assuming that amount is far more than enough to get a typical car thief's
'friends' to 'off' him," he writes, "there is simply no way that a substantial
car-theft subculture could possibly be maintained."
<P>
Jim as high hopes for his plan--he thinks it could eventually lead to the
end of political tyranny. But if you don't like this idea, he has others.
In a recent email exchange, I asked what he was doing now.
<P>
"I recommend that you rent the movie, "The Day the Earth Stood Still.," he
answered. "I'm working on a similar project."
<P>
<HR>
<H2>
Part 9
</H2>
<P>
by Jim Bell, February 27, 1996
<P>
For about a year I have been considering the implications of "Assassination
Politics," and for more than six months I've been sharing the subject and
my musings with you, the interested reader. I've also been debating the issue
with all comers, a self-selected bunch who range from enthusiastic proponents
to clueless critics. Ironically, some of you have even chided me for "wasting
time" with some of the less perceptive among my numerous "opponents." In
defense, my response has always been that when I respond to a person, I do
it not primarily for his benefit, but for others who might be fence-sitting
and are waiting to see if my idea will break down anywhere.
<P>
If there is anything which has fascinated me as much as the original idea,
it is this vast and dramatic disparity between these various responses. It's
been called everything from "a work of genius" to "atrocious," and probably
much worse! Clearly, there must be a fundamental, social issue here that
needs to be resolved.
<P>
While nobody has quite yet said it in those terms, I'm sure that more than
one of you have probably wanted to react to my prose with the line, "See
a shrink!" [American slang for a psychiatrist, for the international readers
out there.] Well, in a sense that's exactly what I did, but the "shrink"
I "saw" had been dead for over five decades: Sigmund Freud. Much to my surprise,
I was handed a copy of a book, <I>Introduction to Great Books</I> (ISBN
0-945159-97-8) which contained (page 7) a letter from Freud to Albert Einstein.
On page 6, there is an introduction, describing the reason for this
communication. It says:
<P>
"In 1932, the League of Nations asked Albert Einstein to choose a problem
of interest to him and to exchange views with someone about it. Einstein
chose "Is there any way of delivering mankind from the menace of war?" as
his problem and Sigmund Freud as his correspondent. In his letter to Freud,
Einstein said that one way of eliminating war was to establish a supranational
organization with the authority to settle disputes between nation as and
power to enforce its decisions. But Einstein acknowledged that this solution
dealt only with the administrative aspect of the problem, and that international
security could never be achieved until more was known about human psychology.
Must right always be supported by might? Was everyone susceptible to feelings
of hate and destructiveness? It was to these questions Freud addressed himself
in his reply."
<P>
Interestingly enough, when I first started thinking about the idea that I
would later term "Assassination Politics," I was not intending to design
a system that had the capability to eliminate war and militaries. What I
was targeting, primarily, was political tyranny. By my standards, that included
not merely totalitarian governments but also ones that many of us would consider
far more benign, in particular the Federal government of the United States
of America, "my" country. Only after I had thought of the fundamental principle
of allowing large numbers of citizens to do away with unwanted politicians
was I "forced," by my work up to that point, to address the issue of the
logical consequences of the operation of that system, which (by "traditional"
ways of thinking) would leave this country without leaders, or a government,
or a military, in a world with many threats. I was left with the same fundamental
problem that's plagued the libertarian analysis of forming a country in a
world dominated by non-libertarian states: It was not clear how such a country
could defend itself from aggression if it could not force its citizens to
fight.
<P>
Only then did I realize that if this system could work within a single country,
it could also work worldwide, eliminating threats from outside the country
as well as corrupt politicians within. And shortly thereafter, I realized
that not only could this occur, such a spread was absolutely inevitable,
by the very nature of modern communications across the Internet, or older
technologies such as the telephone, fax, or even letters written on paper.
In short, no war need ever occur again, because no dispute would country
he intended to war with, obviously, but he would also draw the ire of citizens
within his own country who either didn't want to pay the taxes to support
a wasteful war, or lose their sons and daughters in pointless battles, or
for that matter were simply opposed to participating in the aggression. Together,
all these potentially-affected peoples would unite (albeit quite anonymously,
even from each other) and destroy the tyrant before he had the opportunity
to make the war.
<P>
I was utterly astonished. Seemingly, and without intending to do so, I had
provided a solution for the "war" problem that has plagued mankind for millennia.
But had I? I really don't know. I do know, however, that very few people
have challenged me on this particular claim, despite what would normally
appear to be its vast improbability. While some of the less perceptive critics
of "Assassination Politics" have accused me of eliminating war and replace
it with something that will end up being worse, it is truly amazing that
more people haven't berated me for not only believing in the impossible,
but also believing that the impossible is now actually inevitable!
<P>
A little more than a week ago, I was handed this book, and asked to read
Freud's letter, by a person who was aware of my "little" philosophical quandary.
I began to read Freud's letter in response to Einstein, having never read
any other word Freud had written, and having read essentially none of the
works of the giants of Philosophy. (Now, of course, I feel tremendously guilty
at the omission in my education, but I've always been attracted more to the
"hard sciences," like chemistry, physics, mathematics, electronics, and
computers.) Since this letter was specifically on war, and the question of
whether man could ever avoid it, I felt perhaps it would contain some fact
or argument that would correct what was simply a might end up being right,
but alternatively hoped that if wrong, I would be soon corrected. I was fearful
that I was wrong, but also fearful that there would be nothing in this essay
that would assist me in my analysis of the situation.
<P>
About a third of the way through Freud's letter, I had my answer. Below,
I show a segment of Freud's reply, perhaps saving the whole letter for inclusion
into a later part of this ongoing essay. While I could drastically oversimplify
the situation and state, "Freud was wrong!," it turns out that this brief
conclusion is at best highly misleading and at worst flirting with dishonesty.
By far the greater part of Freud's analysis makes a great deal of sense to
me, and I would say he's probably correct. But it is at one point that I
believe he goes just a bit wrong, although for reasons which are entirely
understandable and even predictable, given the age in which he lived. It
must be remembered, for example, that Freud was born into an era where the
telephone was a new invention, broadcast radio was non-existent, and newspapers
were the primary means that news was communicated to the public. It would
be highly unreasonable for us to have expected Freud to have anticipated
developments such as the Internet, anonymous digital cash, and good public-key
encryption.
<P>
In some sense, at that point, my biggest regret was that I couldn't discuss
the issue with either of these two communicants, Freud having died in 1939,
and Einstein in 1955, after having helped initiate research that led to the
development of the atomic bomb, the weapon that for decades and even now,
makes it absolutely, vitally important to eliminate the possibility of war
from the world.
<P>
But I'll let Dr. Freud speak, as he spoke over sixty years ago, because he
has much to say:
<BLOCKQUOTE>
"Such then, was the original state of things: domination by whoever had the
greater might--domination by brute violence or by violence supported by
intellect. As we know, this regime was altered in the course of evolution.
There was a path that led from violence to right or law. What was that path?
It is my belief that there was only one: the path which led by way of the
fact that the superior strength of a single individual could be rivaled by
the union of several weak ones. "L'union fait la force." [French; In union
there is strength.] Violence could be broken by union, and the power of those
who were united now represented law in contrast to the violence of the single
individual. Thus we see that right is the might of a community. It is still
violence, ready to be directed against any individual who resists it; it
works by the same methods and follows the same purposes. The only real difference
lies in the fact that what prevails is no longer the violence of an individual
but that of a community."
</BLOCKQUOTE>
<P>
[But below is where I think Freud falls into a certain degree of error, perhaps
not by the standards and realities of <I>his</I> day, but those of ours.
My comments are in square brackets,], and Freud's comments are quoted "".
Freud continues: ]
<BLOCKQUOTE>
"But in order that the transition from violence to this new right or justice
may be effected, one psychological condition must be fulfilled. The union
of the majority must be a stable and lasting one. If it were only brought
about for the purpose of combating a single dominant individual and were
dissolved after his defeat, nothing would be accomplished. The next person
who though himself superior in strength would once more seek to set up a
dominion by violence and the game would be repeated ad infinitum. The community
must be maintained permanently, must be organized, must draw up regulations
to anticipate the risk of rebellion and must institute authorities to see
that those regulations--the laws-- are respected and to superintend the execution
of legal acts of violence. The recognition of a community of interests such
as these leads to the growth of emotional ties between the members of a united
group of people--communal feelings which are the true source of its strength."
[end of Freud's quote]
</BLOCKQUOTE>
<P>
[Those of you who truly comprehend the idea of "Assassination Politics" will,
I'm confident, understand exactly why I considered this segment of Freud's
letter to be important enough to include, and will probably also recognize
why I consider Freud's analysis to go wrong, albeit for comparatively minor
and understandable reasons. I will address the last paragraph in greater
detail, to explain what I mean. I will repeat Freud's words, and address
each of his points from the standpoint of today's situation and technology.]
<BLOCKQUOTE>
"But in order that the transition from violence to this new right or justice
may be effected, one psychological condition must be fulfilled. The union
of the majority must be a stable and lasting one." [In a sense, Freud is
absolutely correct: Whatever system is chosen to "govern" a society, it must
continue to operate "forever." ] Freud continues:
<P>
" If it were only brought about for the purpose of combating a single dominant
individual and were dissolved after his defeat, nothing would be accomplished."
</BLOCKQUOTE>
<P>
[This is where the problem begins to creep in. Freud is leading up to justifying
the existence of a formal government as he knew them in the 1930's, based
on the continuing need for keeping the peace. The first, and I think, the
most obvious problem is that Freud seems to implicitly assume that the purpose
of the union will actually be fulfilled by the formation of a government.
Freud, who died in 1939, didn't see what his survivors saw, a "legitimate"
government in Germany having killed millions of people in the Holocaust,
or many other incidents subsequent to that. And Freud, whose letter was written
in 1932, was probably not aware of the slaughter of the Russian Kulaks in
the late 1920's and early 1930's, or the purges which followed. Freud could
have felt, generally, that the problems with a country's governance were
caused either by inadequate government or simply a rare example of government
gone bad. We know, to the contrary, that governments very frequently "go
bad," in the sense of violating citizen's rights and abusing the power entrusted
to them. Few may end up killing millions, but to assume that we must continue
to tolerate governments just because they don't go quite as far as Nazi Germany
would be foolish in the extreme.]
<P>
[The second problem is the implicit assumption that the long-term control
he (correctly) sees MUST come from an organization like a traditional government.
True, in the era in which Freud lived, that conclusion made a great deal
of sense, because a well-functioning government appeared superior to none
at all. And it was at least plausible that such control COULD come from a
government. But as the old saying goes, "Power corrupts, and absolute power
corrupts absolutely."]
<P>
[To use a house's thermostat as an analogy, but differently than I did in
"Assassination Politics part 6," a person who lived in an era before automatic
furnace thermostats would always conclude that a person's efforts would have
to be continually directed towards maintaining an even temperature in his
house, by adding fuel or limiting it, by adding more air or restricting,
etc. To the extent that this manual control constitutes a "government," he
will believe that this hands-on control will always be necessary. But we
now live in a time where a person's time is rarely directed towards this
effort, the function having been taken over by automatic thermostats which
are cheap, reliable, and accurate. They are also, incidentally, essentially
"uncorruptible," in the sense that they don't fail except for "understandable"
reasons, and repair is cheap and easy. (And a thermostat can never be bribed,
or get tired, or have its own interests at heart and begin to subvert your
own commands.) Quite simply, the progress of technology has put control of
temperature in the hands of an automatic, error-free system that is so reliable
as to be ignorable most of the time.]
<P>
[I argue that likewise, the progress of technology would allow an automatic
system to be set up, which I called "Assassination Politics" (but could probably
use a more apt name, since its application extends far beyond the issue of
politics) different from traditional government, a difference somewhat analogous
to the difference between a person's full-time efforts and an automatic
thermostat. Aside from the dramatic reduction in effort involved, an automatic
system would eliminate the errors caused by inattention by the operator,
such as leaving, falling asleep, or other temporary lack of concentration.
These failures are somewhat analogous to the failure or misbehavior of a
corruptible or indifferent or even a malicious government.]
<P>
[This makes a government like Freud saw totally unnecessary. Of course, Freud
could not have anticipated the technological developments that would make
an "automatic" replacement for government even possible, and thus he followed
his contemporary paradigms and sought to justify the governments as they
then existed.] Freud continues:
<BLOCKQUOTE>
"The next person who thought himself superior in strength would once more
seek to set up a dominion by violence and the game would be repeated ad
infinitum."
</BLOCKQUOTE>
<P>
[This statement is correct, but I think it misses the point: Many functions
of individuals and machines are never "completed", and must "be repeated
ad infinitum." (The most basic example: If we are optimistic about the future
of the human race, by definition reproduction and survival must be "repeated
ad infinitum.") That does not mean that the mechanism which handles that
need must be any more complicated that the minimum necessary to achieve the
control needed. I agree that a system of long-term control is necessary;
where I disagree with Freud is simply that I believe that a vastly better
method of control now can potentially exist than the traditional governments
that he knew. To the extent that he couldn't have anticipated the Internet,
anonymous digital cash, and good encryption, he had no reason to believe
that government could be "automated" and taken out of the hands of a tiny
fraction of the population, a fraction which is corruptible, malicious, and
self-interested. Also, by not being aware of modern technology, he is unaware
how easy it has become, conceptually, for people to come together for their
self-defense, if that self-defense required only a few kilobytes be sent
over fiber-optic cables to a central registry. Freud's objection to an "endlessly
repeating" system breaks down in this case, so his conclusion need not be
considered valid.]
<P>
Freud continues:
<BLOCKQUOTE>
"The community must be maintained permanently, must be organized, must draw
up regulations to anticipate the risk of rebellion and must institute authorities
to see that those regulations--the laws-- are respected and to superintend
the execution of legal acts of violence."
</BLOCKQUOTE>
<P>
[Again, I think Freud misses the point. He refers to "the risk of rebellion,"
but I think he forgets that the main reason for "rebellion" is the abuse
by the government then in control. (Naturally, it looks differently from
the standpoint of that government!) If the latter problem could be eliminated,
"rebellion" would simply never occur, for there would be no reason for it.
If those that were "rebelling" were in the wrong, violating somebody's rights,
then my "Assassination Politics" system would be able to take care of it.
This, presumably and understandably, Freud could never have foreseen. Also,
Freud does not address the question of whether or not the government which
promulgates those laws is doing so in a way primarily for the benefit of
the public, or those who populate the government itself. Graft was well known
if Freud's time; it seems to me that he should have addressed the question
of whether or not an entity called a "government" could actually achieve
the benefits he claims justify the government, without being subverted by
those who control it, for their own interests. If not, then there is certainly
a issue to be addressed: At what point do the depredations of a parasitic
government exceed its benefits? And can we find a way to do without it?]
Freud continues:
<BLOCKQUOTE>
"The recognition of a community of interests such as these leads to the growth
of emotional ties between the members of a united group of people--communal
feelings which are the true source of its strength." [this is end of the
portion of Freud's letter which I quote here.]
</BLOCKQUOTE>
<P>
One of the interesting things about this statement is that it is the development
of tools such as the Internet which will be eliminating the very concept
of "foreign" and "foreigner." They will become artificial distinctions. There
is clearly much precedent for this, from the country in which I live, America.
When formed, it contained people whose primary loyalty was to their
<I>state,</I> not to the Federal government as a whole. Even our civil war,
from 1861 to 1865, was based on loyalty to states or regions, rather than
the country as a whole. To cite just one example, myself, while I reside
in the state called Washington, I've lived in a number of other states, but
I don't consider myself loyal to any particular state. (Perhaps using myself
as an example is misleading, because at this point I don't consider myself
"loyal" to any government at all!)
<P>
In fact, later in Freud's letter, he says, "Anything that encourages the
growth of emotional ties between men must operate against war." Sadly, Freud
did not live to see the development of the Internet, and the massive
international communication which it has already begun to foster. In
<I>his</I> day, the ordinary people of one country and another rarely
communicated, except perhaps for letters with relatives from "the old country"
that emigrated. The idea of going to war with people from whom you get email
on a daily basis is, in itself, a "foreign concept" to me, and I hope it
will remain so! In that sense, Freud was very right: "Assassination Politics"
active or not, it will be much harder for governments to whip up their citizens
into a frenzy to kill the enemy if they can type to them every day. Frustratingly
left unanswered is a question whose answer I'd like to know: Could I have
convinced Freud, or Einstein, that "Assassination Politics" is not only a
necessary or even an unavoidable system, but also a GOOD one? Could I convince
them today, had they miraculously survived until today, aware of the last
64 years of history subsequent to their correspondence?
<P>
Jim Bell jimbell@pacifier.com
<P>
Klaatu Burada Nikto
<P>
Something is going to happen... Something...Wonderful!
<P>
<HR>
<H2>
Part 10: "Non-Euclidean Thinking"
</H2>
<P>
by Jim Bell
<P>
An interesting communication I had recently on the subject of "Assassination
Politics." My commentary is preceded with >> or nothing; the other
person's commentary starts with a ">". The subject is how to actually
implement this system, and my first comment notices the fact that despite
my efforts, the government has not attempted to use this issue to justify
some sort of crackdown on net rights, or anything like that.
<P>
I think they're actually afraid to start the debate,
<P>
I think they don't believe you're a threat.
<P>
You're probably right about this. I guess I'll have to think of something
to change their minds, huh?
<P>
Remember, they have incredible >amounts of money with which to hire bright
but greedy people. All they have to do is find the people running the "Guess
the Death Date" lottery. They would have great incentive to apply their
considerable resources to this end.
<P>
Your logic is excellent. But as strange as it may seem, there may be a different
way... Let's see, how do I explain? First, a little diversion that may or
may not be relevant to this subject, but initially won't appear to be so.
<P>
Somewhere around 20-25 years ago, I read some item concerning Howard Hughes,
the late billionaire. It described the history of his business ventures,
in fields such as aircraft ("Spruce Goose" is a well-known example) but also
mentioned that Hughes Tool was (originally?) into oil-well drilling equipment.
<P>
I don't know how much you know about oil well drilling and drill bits, but
they look nothing like the classic fluted drill bits common in hardware stores.
Oil well drill bits consist of multiple ultra-hard carbide points mounted
on rotating shafts mounted at the end of the drill "string," and these shafts
must be connected to the main shaft with bearings. They roll around on the
rock, not sliding, and they "spall" off pieces of rock due to enormous applied
pressure.
<P>
Oil well drilling is done by lubricating the drilling operation with what
is called "drilling mud," which is actually a slurry of solids in water,
which is primarily used to cool the cutter and wash away the rock chips and
dust produced in the operation. Now, since the rotating cutter wheels must
spin on their axis, that means they have to be run on shafts with bearings
installed. These bearings cannot be perfectly sealed and thus protected against
rock and mud dust, and their useful lifetime is strongly limited by their
quality.
<P>
And since every time they wear out the whole drill string has to be pulled
from the well, that's an EXTREMELY expensive proposition for well-drillers.
So it should not be surprising that these guys considered bearing quality
to be very, very important. A little improvement was worth a lot of money.
<P>
"Quality", to a bearing manufacturer, is strongly related to surface hardness,
and traditionally, the best bearings were (and, mostly, still are) the hardest.
But there's a problem: Ultimately, a very hard circular bearing rotating
on a very hard flat surface (especially if its heavily loaded) applies nearly
all its for on a single point (for ball bearings) or on a single line (for
roller bearings) and that eventually causes bearing failure. So there was
an upper limit, generally, on how good you could get in bearings. And the
hardest won. Until Hughes.
<P>
[don't go to sleep yet... it gets relevant real soon]
<P>
According to the source I read, what Hughes Tool did that made them really
rich was quite simple and counter-intuitive: Rather than trying to make his
bearings as HARD as you can get, he made them SOFT, very soft, "almost as
soft as lead." (Which, if you know anything about metals, is very soft indeed.)
The bearings deformed on their raceways, spreading out the load over a far
larger area, and the resulting bearings were the best in the business. (He
probably also applied a lot of research into how to avoid "metal fatigue,"
but that's quite another story.)
<P>
Very counter-intuitive, but he "won" precisely because he did exactly the
opposite of what everyone "knew" was the proper way to go. Okay, so that
explains a genius who later became a billionaire who later turned into a
neurotic, or worse. "What," you will ask, "does this all have to do with
Assassination Politics?"
<P>
Well, to draw an observation originally posited in an essay titled the "Libertech
Project," about 7 years ago, libertarians (of all people) are "non-Euclidean
thinkers." Basically, this means that we recognize that the best way to go
from "point A" to "point B" is NOT NECESSARILY a straight line. And like
Columbus, who sailed west in order to go east, sometimes it is necessary
to sit down, and totally re-think your strategy if you're trying to accomplish
some goal.
<P>
By "classical" thinking, "Assassination Politics" would have to be the best,
tightest-security, more protected organization that has ever existed on the
face of this planet. Just about EVERY powerful person would want to kill
anybody who had anything to do with such a system. The codes would have to
be unbreakable, the remailers would have to be certain, but most importantly,
each and every participant would have to be perfectly anonymous to even have
a prayer of pulling it off. Especially the operators of such a system. Especially
them.
<P>
That's classical thinking. And that's what I thought a few months ago. I
thought, "it's do-able, but it's gonna be a lot of work!"
<P>
But let's suppose, for a moment, that somebody "pulls a Hughes." Rather than
trying to make the hardest bearings in the world, why doesn't somebody try
to make the softest? Rather than trying their darndest to stay anonymous,
or wait and let somebody else implement this system, why not just "let it
all hang out," (as the saying went in the 1960's) and publicly announce that
they're implementing this system, come hell or high water, and invite anyone
who wants to participate to help form what will be the LAST revolution on
earth, the one that'll take down ALL the governments.
<P>
This sounds crazy, right? I mean, who wants to die? Who wants to commit suicide
just to... just to... just to... make an ENTIRE WORLD FREE FOREVER? Free
from wars, militaries, governments, taxes, political oppression. Free from
the kind of totalitarian governments that existed and currently exist. Free
from the Holocausts that have killed Jews, Cambodians, Armenians, Russian
Kulaks, Iraqi Kurds, Chinese dissidents, Native Americans, and oh so many
others? "Who, exactly, would be stupid enough to risk death to make the world
free???"
<P>
Everyone who volunteered to fight to fight Hitler, to name just one example.
Remember, or have we forgotten so soon, that occasionally people die to keep
the rest of us free. That's the way it's been for hundreds of years. The
United States of America was founded by people who risked death to shake
off the yoke of a government that was, by the standards of the day, not
particularly bad.
<P>
Think about it. Somebody had to be the first one to start banging on the
Berlin Wall, with a sledgehammer, in 1989. Somebody had to be the first to
walk through. Somebody had to be the first to stand up and say, "Enough!"
And the ironic thing is, the most strangely unusual thing, is that the entire
Eastern Bloc fell, almost bloodlessly, in a couple weeks, because one by
one everybody realized that all that's sometimes required is to finally stand
up and be counted, and to just say no to the government. When the time was
right, all it took was a slight push and the dominoes tumbled down.
<P>
Now, don't get me wrong. I'm not suggesting that EVERYONE would be identified.
The "donors" to the system would remain perfectly anonymous, and the "guessers"
would likewise be perfectly anonymous, but the organization itself would
be made up of real people, who have published addresses, who have simply
decided that they have had enough of the current system and are going to
participate in a PERFECTLY LEGAL enterprise by the laws of the country, and
just DARE the government to try to stop them. The organization wouldn't have
to buy ads; the publicity firestorm would be enormous. Suddenly, all the
politicians would be put on the spot! Instead of being asked by the reporters
for their position on the economy, pollution, the budget deficit, or some
other thing, they'll ask, "Why should the public NOT want to see you dead?"
<P>
When would be the best time to do it? Why, during a major political campaign!
When Congress is out of session, and they can't pass legislation without
calling some sort of emergency session. But it won't matter anyway, for a
few weeks the organization doesn't actually have to take bets or make payments,
they'll merely publicize their efforts for all to see. To reassure the public,
they could announce that they'll only take bets on elected and appointed
political officeholders...and anyone who tries to stop the system. And the
politicians will be scurrying around, looking for political cover, trying
to figure out how to NOT look scared, but at the same time each is wondering
if he'll be the first to go. And all the while, the public will be loving
it, laughing at the efforts of the politicos to cover their collective asses,
and taking private bets among themselves on who will be the first one to
die.
<P>
Prosecute the participants? On what charge? "Conspiracy to commit gambling"?
Which prosecutor would risk appearing to be impeding the progress of a useful
system? At that point, the organization's members will just be publicly
exercising their first-amendment rights. Which judge would take the case?
Now THEY'RE on the spot, THEY have to decide what to do. I contend that in
an election year, before the election, there would be mass resignations from
Congress, or members deciding "it's just not fun anymore" and decline to
return even if re-elected, as well as the complete loss of whatever residual
confidence the public has in the government. Whew! Is this all just wishful
thinking? I really don't know!
<P>
<HR>
<P>
<I>[End]</I>
</BODY></HTML>
Login or Register to add favorites

File Archive:

June 2022

  • Su
  • Mo
  • Tu
  • We
  • Th
  • Fr
  • Sa
  • 1
    Jun 1st
    19 Files
  • 2
    Jun 2nd
    16 Files
  • 3
    Jun 3rd
    28 Files
  • 4
    Jun 4th
    0 Files
  • 5
    Jun 5th
    0 Files
  • 6
    Jun 6th
    19 Files
  • 7
    Jun 7th
    23 Files
  • 8
    Jun 8th
    11 Files
  • 9
    Jun 9th
    10 Files
  • 10
    Jun 10th
    4 Files
  • 11
    Jun 11th
    0 Files
  • 12
    Jun 12th
    0 Files
  • 13
    Jun 13th
    0 Files
  • 14
    Jun 14th
    0 Files
  • 15
    Jun 15th
    0 Files
  • 16
    Jun 16th
    0 Files
  • 17
    Jun 17th
    0 Files
  • 18
    Jun 18th
    0 Files
  • 19
    Jun 19th
    27 Files
  • 20
    Jun 20th
    65 Files
  • 21
    Jun 21st
    10 Files
  • 22
    Jun 22nd
    8 Files
  • 23
    Jun 23rd
    6 Files
  • 24
    Jun 24th
    6 Files
  • 25
    Jun 25th
    0 Files
  • 26
    Jun 26th
    0 Files
  • 27
    Jun 27th
    0 Files
  • 28
    Jun 28th
    0 Files
  • 29
    Jun 29th
    0 Files
  • 30
    Jun 30th
    0 Files

Top Authors In Last 30 Days

File Tags

Systems

packet storm

© 2022 Packet Storm. All rights reserved.

Hosting By
Rokasec
close