Subject: america is a virus
From: Edward Cranswick <e_cranswick@yahoo.com>
Date: Sun, 23 May 2004 03:34:40 -0700 (PDT)
To: Michael Dwyer <m.dwyer1@bigpond.net.au>

michael-
i finished "prey" last tuesday -- basically read it
within a day after i decided to do so, it's been a
while since i made such an obsessive attack upon a
book -- and i feel that it is another one of
critchon's leaps from harvard medical school to
hollywood via what is nominally fiction, or perhaps
that in reverse order ... huh? what? huh? there was a
famous grand old man of seismology, jack healy, who
put together a very large USGS (US Geological
Survey)seismic refraction field investigation in saudi
arabia in the 1970's (the US has always had a special
relationship with SA ever since FDR made a deal with
King Ibn Saud in 1945). jack reasoned that since the
final product would be plots, graphs, showing the
structure of the Earth, first he would find a good
printer/plotter that would work in the saudi desert
and then work his way from there back to the computers
that would process the data and then to the
seismographs that would record the explosions, rather
than the reverse planning sequence. similarly, i see
that crichton sees a series of movie scenes, with the
emphasis on the graphics & special effects, and works
his way back to the text that nominally connects those
to an empirical "scientific" reality.

when i was about 16, i built a robot that avoided
walls. it consisted of a three-wheel cart with a
unpowered rear-center wheel and front right & left
wheels, each with its respective and separate drive
motor, and mounted on top of the cart was a "cyclops
eye". The eye was one photocell in a box (designed and
sold for a beam-interrupt doorbell) surmounted by a
light whose beam was directed forward in the direction
of cart motion, in the direction the photocell "saw".
by means of a motor rotating an arm with an attached
lever that swivelled the box, the eye scanned
left&right, and a switch controlled by the rotation
position distinguished these two directons. in
ordinary operation, both drive motors would be on and
the robot would advance forward with the eye scanning.
when light was reflected back from some object in the
robot's path, the photocell would trigger and turn off
the scanning motor, leaving the "eye" looking in the
direction of the object, and also turn off the drive
motor on the opposite side, causing the robot to turn
in the direction, away from the object. after the eye
could no longer "see" the object, the drive motor that
had been turned off would be turned on, the robot
would start moving forward in the new direction and
scanning would recommence. the robot dragged an
extension cord attached to mains power (US: 120 V AC
60 Hz) and what with its motors, relays, lamp, etc.,
consumed 50-100 Watts. with the appropriate lighting
conditions and setting of its photocell sensitivity
which controlled how far forward it could see, it was
able to circumnavigate a room with bare walls
containing a few large simple objects. other than a
cart controlled by a paper tape that i built about
four years later when i was 20 and back home at my
mother's recovering from a broken ankle, the cyclops
robot was my last big "hardware" project.

in 1967, about the same time i built the cyclops,  i
started working as a FORTRAN computer programmer on an
IBM 360/30 computer at the local family looney bin by
virtue of nepotism. by 1969, our office had a 360/44
and a 360/50 that had HALF A MEGABYTE OF RAM (in those
days, there was virtually no operating system overhead
and the user had access to all memory) and i had begun
drinking and correspondingly misbehaving in an attempt
to escape my nerdy/geeky tendencies. i had a friend --
one of the group with whom i went to Woodstock 1969 --
who got me to find and get the cyclops out of the
closet so he could borrow it for a project in driver's
education class, telling that the teacher he had built
it to experiment with robot-controlled cars that
avoided collisions, i.e., would permit drunk driving
(as they call drink driving in the States -- actually
i don't think he made that suggestion).  i was a
half-assed competent FORTRAN programmer and had worked
with GPSS (Generalized Purpose Simulation System(?)) a
bit and done some simple heuristic research by that
time, and i became intrigued by the thought of
modeling or simulating the behaviour of the cyclops.
Accordingly, because i was at university and worked
all night friday nights and had access to the vast 0.5
MB of RAM after midnight when the job-stream had
completed, i wrote a FORTRAN program that had a
digital entity -- the cyclops -- that could move about
in an array-space of 128 x 64 elements. it would leave
a designated character at each position it reached so
as to form a "track" of its movements in that space
that could be graphically displayed, i.e., printed
out, after some given number of iterations, i.e.,
movements (133 x 66 was standard dimensions of the
output line-printer).

originally, i was thinking in terms of optimizing the
"seeing distance" of the entity with respect to the
complexity of the objects in its array-space
"environment". however, because i did not want to
waste an inordinate amount of printer paper in
displaying the results of these simulations, it
occurred to me that i could run several different
entities simultaneously, each having different
behavior parameters and distinguished by a different
tracking character ... and suddenly it occurred to me
that i was simulating the respective fitnesses of
different digital organisms to navigate an environment
... darwin, here we come ... 

i was raised as an athiest with darwin and newton
sustituting for other traditional dieties. the gospel
according to my father -- who could speak 5-6 several
languages from having translated from the bible, many
passages of which he knew by heart -- was that
everything in the universe could be explained in terms
of the revelations of these two gods -- only my father
himself was exempt from the corresponding natural laws
(his death at the age of 46 from a heart attack
induced by shooting speed and eating potato chips
provided some empirical evidence to the contrary ...
the contrary of something ... ), consequently, "the
origin of the species" (which i have yet to read) or
rather the principle of the survival of the fittest
has always been a concept close to my heart.

within a few friday nights, cyclops had been buried in
the digital dust as a cyber fossil of dead code, and
my program had evolved into a little closed ecological
system containing simulated autotrophs (photosynthetic
plants), heterotrophs (animals), and protists/bacteria
(to clean up the shit, i.e., clear array-space). the
different organisms fed upon one another or otherwise
interacted and had various "life rules", i.e.,
parameters, that determined what minimum input of
"food" they needed to survive and what to reproduce
themselves. i had a somewhat jury-rigged scheme for
random mutation that would produce a random variation
in the parameters of the offspring but it was all
constrained to stay within the pre-wired or rather
coded structures that i had written -- there was no
true mutation of the code itself on a byte-level
analgous to mutations of genes (e.g., see Tom Ray's
work with Tierra
<http://www.his.atr.jp/~ray/tierra/whatis.html>). so i
would start the program running, and every several
thousand iterations, i.e., time steps, or so it would
print out another view of the digital environment,
showing the variation of the numbers and
characteristics of the different "species" (hard to
define because i did not include sex). on what was to
be the last run, there was a major ecological
catastrophe sometime after time step 57000 and
everything died, the whole ecosystem became extinct --
i never ran the program again and spent the next
decade drowning myself in booze.

in my subsequent more & less neurotic moments, i
sometimes wondered what had brought about this digital
demise, and about a year later, in autumn (oct-nov)
1970, i woke up with an awful hangover in the back
seat of someone's car in which i'd passed out the
night before and suddenly had this vision of digital
doom, SOAs, self-organizing algorithms -- but i did
not anticipate that digital viruses would be
deterministically created, intentionally written, two
decades later.

the fundamental assertion of the gaia hypothesis is
that Earth appears as if it were a living organism --
one of the chief criticisms of this notion is that
"life" has evolved in response to the process of
natural selection, the survival of the fittest of many
competing individual organisms, but there is only one
Earth, i.e., one gaia. an objection to this criticism
is that it implicitly and fallaciously defines an
organism as a distinct individual with a finite
lifetime and immutable characteristics constrained by
its single set of DNA. This definition excludes
colonial organisms that have potentially unlimited
lifetimes and which exhibit a range of genetic
information within the colony as a whole. for example,
a mangrove swamp contains many apparently separate
mangrove trees which may exhibit some genetic
variation but they actually share one root system
which continues to live and function despite the death
of an individual tree.

so the Internet is an entity that appears to be one
organism that evolves and propagates itself around the
Earth.

in conclusion, similar to your identification of
knowlege as a virus, hollywood special effects may be
the virus that escapes from the screen and destroys
the world.

regards to lex.
-edward