MECHANICAL INTEGRITY

     Einstein liked to do 'thought experiments' which tended to annoy
the dimwits around him who couldn't do them.

     So here is a thought experiment, lets see how many dimwit answers
we get back from it.

     Phase 1 of The Proof spends most of its time analyzing the guy in
the army tank who can only see the world via a TV screen connected to
a video camera on the outside of the tank.

     The integrity of the data link between the video camera and the
TV set is of paramount importance, because we want the guy in the tank
to have a 100 percent accurate picture of the world within the limits
of the connection and the technology of the time.

     It was understood by many scientists that if an enemy agent could
get at the circuitry and software connecting the video camera to the
TV set it would be possible to alter the image on the TV.  The enemy
agent could then make the guy in the tank see things that weren't
actually out there, or not see things that were.

     The problem therefore of verifying the integrity of the data flow
from the outside world to the TV set became an all consuming problem
to the theoreticians working on the problem.

     They were particularly interested in situations where there were
no outside observers to help verify that the tank had not been messed
with.

     They wanted to know if it was theoretically possible to verify
the integrity of the connection between the outside world and the TV
set USING ONLY THAT CONNECTION to look at and verify itself.

     For the sake of the thought experiment, they assumed a few
things.

     One the video camera outside the tank was actually infinitely
mobile and could look at any part of itself or the circuitry all
the way back to the TV set to find breaks in integrity or signs
of espionage.

     It was assumed that any espionage that might take place would
take place after the tank left the factory in known good working
order.  It was also assumed that sometimes tanks are 'left alone' not
under observation which might then make a good time for such espionage
to take place.

     At the time the tank was built, its circuitry had a complete and
accurate built in diagram of what it was supposed to find and so it
could compare what it saw with what was expected.

     Second they allowed for the existence of two independent video
cameras, each connected by their own circuitry to two independent TV
sets in the tank.

     Each one of the video camera could not only study its own
circuitry in detail, but also that of its brother.

     Given such a setup, and a number of years down the road, would it
be possible to state with 100 percent certainty that one or both of
the data links had not been corrupted by espionage?

     If it is not possible to verify such a data link with perfect
certainty, is it possible to prove THAT with perfect certainty?

     Homer