trainoff.pasa@XEROX.ARPA (08/15/84)
Matter transmitters do not necessarily have to be matter duplicators or
matter transmutters. The common every day examples of information
transmission, such as radio and television broadcasts, Xerox machines,
stereo recordings, computer storage, etc., have the property that the
copying is not a destructive process. That is to say, you can record
music and make a copy without destroying the original. The same may not
be true for a matter transmitter. It may be necessary to destroy the
object to record its data. As far as the matter transmuting goes, does
a Xerox machine have to transmute air into ink?
Simply accumulating the data would be a monumental task. It might not
be necessary to make an EXACT, identical copy of something to retain its
essential charateristics. I define essential characteristics to
include, form, total mass, viablity of living creatures, memory and
personality of sentient creatures. It is very unlikely that a single
electron having its spin flipped will effect the overall personality of
a duplicated person (Who knows? Would that make a person flip out?
(sorry)) Now, to just get an idea of how much information we are
talking about let us ignore things like electron spins, excited
electronic and nuclear states, etc. and try just to duplicate the
correct atoms in a human. I am going to make a few unrealistic
assumptions to do a top of the head calculation (an order of magnitude
in the order of magnitude will be fine (think about that for a moment)).
An average person masses around 65 kilograms, round off to 1e5 g. A
human is mostly composed of hydrogen, oxygen, carbon, and nitrogen. Let
us assume that they are in roughly equal amounts by number (they aren't,
but it should be close enough). They have an average atomic weight of
around 10 g/mole. Avogadro's number (6.02e23 atoms/mole) can be rounded
to 1e24 atoms/mole. So we have on the order of 1e28 atoms. For a given
person, assume that each atom is one of the four I have mentioned. In
terms of the binary storage, each position requires 2 bits for each atom
so we have 2e24 bits. Now recall that this is a terribly gross
underestimate of the total amount of information stored in the
wavefunction of a person. The true wave function would also have to
include information on the positioning, velocity, etc. of each subatomic
particle. The true quantum state must include interations between all
of the particles, so it probable scales as the factorial of the number
of the particles. So in some sense my order of magnitude is probably
too small by many, many of orders of magnitude. As, just a guess, it
doesn't seem too unreasonable to say a human might require more
than1e100 bits of total storage. I just picked that to defy anyone
trying to envision a computer storage that could store 2e24 bits. 2e24
bits is almost conceivable, 1e100 is not.
Now it is quite conceivable that a device that attempted to duplicate
the exact quantum state of human would not be able to store all of the
information that it is transmitting to a receiving station. A good
example of this type of information device is a television camera. It
is quite believeable that a handheld video camera could take more data
than it could store. I know, you could always buy a video tape
recorder, but conceptually, it is ridiculus to assume that any device be
able to store as much information as it can aquire. In addition, since
it appears very unlikely that we could ever make enough computer storage
to store an object, it seems even more unlikely that we would ever be
able to process the data. (Have you ever tried to take a FFT of a big
array? It scales as n*log(n). 1e102 is a mighty big number. (Well
just take 1e102 processors in parallel and....))
Then there is the problem of the data transmission. It is going to take
a while. Assume that a matter transmitter is built which solves the
storage problem. How would you like to hop into it, press the button
("Press the button, Max!"), and arrive just in time to watch the
universe fold up in the Big Crunch? Sounds like a neat one way time
machine.
This entire diatribe has also ingored one minor detail; the Heisenberg
uncertainty principle. I addessed the data storage problem. This point
is acedemic since we can't accumulate the data in the first place.
From the fallible, flailing fingers of
Steve "I just couldn't resist any longer" Trainoff
Xerox Special Information Systems
101 N. Halstead, Pasadena CA. 91106