1998-01-09 - Re: Jim Bell… lives… on… in… Hollywood!

Header Data

From: Ryan Lackey <rdl@mit.edu>
To: David Honig <honig@otc.net>
Message Hash: 53192a8fd4810065b66a1d7d1867643cb612a1ac0dc4a728b198e0b067a3bcf6
Message ID: <tw74t3dixo1.fsf@the-great-machine.mit.edu>
Reply To: N/A
UTC Datetime: 1998-01-09 13:42:54 UTC
Raw Date: Fri, 9 Jan 1998 21:42:54 +0800

Raw message

From: Ryan Lackey <rdl@mit.edu>
Date: Fri, 9 Jan 1998 21:42:54 +0800
To: David Honig <honig@otc.net>
Subject: Re: Jim Bell... lives... on... in... Hollywood!
Message-ID: <tw74t3dixo1.fsf@the-great-machine.mit.edu>
MIME-Version: 1.0
Content-Type: text/plain



<daemon@ATHENA.MIT.EDU> (David Honig) writes:

> At 10:48 PM 1/7/98 -0800, Sergey Goldgaber wrote:
> >
> >1 - Anonymity is technically feasable.
> >
> >2 - This requirement is a legal necessity.  Otherwise, the organization
> >    may be seen as advocating murder.
> >
> >Obviously, if the "Death Pool" was fully anonymous, there would be
> >no way to tell if the winner had contributed in any way to the death.
> >
> >Thus, I think we may be well on our way to Assasination Politics.
> >
> > - Sergey Goldgaber
> 
> I agree, but "contribute to death" needs to be operationalized.  Here's a
> proposal:
> If a homicide suspect is arrested within N months, they will be isolated
> from the net
> and the owner of the winning ID will have to perform a challenge-response.
> Since
> the suspect couldn't have replied, they are different; if a pair
> collaborated, well,
> when a hit man is caught, his payoff matrix will usually make him turn in
> the client.

Given strong cryptography and something like my current Eternity DDS
almost prototype (a reliable distributed way of selling 
storage-compute-bandwidth being the relevant part), why couldn't the 
incarcerated person have left an agent out on the net to handle the 
challenge for him, and hold the money in anonymous trust for him until
he gets out?  I can't think of any anonymity-preserving system which 
contains an "is-a-person" predicate -- even if you asked an AI-hard
question, you could blind the question and post it to usenet or CNN
or something and quote one of those responses (which would be wise to
do anyway for styleometry prevention).

The other option is having a non-anonymous system, or one that is
anonymous until someone tries to collect the prize, but in that case,
it's not all that interesting a problem.

Ryan the Nightshifted
-- 
Ryan Lackey
rdl@mit.edu
http://mit.edu/rdl/		






Thread