giles@ucf-cs.UUCP (Bruce Giles) (11/27/83)
I, unfortunately, thought that that was a very realistic portrayal
of a physically advanced civilization. I say unfortunately because I
expect that we will soon be to that point ourselves.
Here is a possible future history which illustrates what I am talk-
ing about. (Anyone wishing to base a story on this fh is welcome to do
so, incidently). First, some background information:
<1>: We have 'expert knowledge' systems in physical science today.
The ones I have heard about will determine molecular structure
from x-ray diffraction patterns, and perform geological prospect-
ing from a data base.
<2>: The emphasis in modern test equipment is on simple, rugged
*phyical* measurement devices with advanced mathematical
analysis by an on-board computer. This is in contrast to
complex, delicate equipment providing direct measurements.
<3>: We are on the threshold of a 'robotics revolution' where far
more sophisticated interactions between the items mentioned in
parts (1) and (2) will be commonplace.
Based on this information, a very possible future history is as follows:
(The number to the left represents the year I expect this to happen by.)
2000: Intelligent automated test equipment is commonplace in
pharmacutical (?) companies, research & development companies,
the military, and major universities. This equipment is
used to (a) ensure quality, (b) suggest possible sources of
contamination, (c) and do basic physical measurements required
for further work.
2020: Several major intelligent probes are launched to all other
planets in the system, a number of major moons, possibly circa-
solar and interstellar space, and almost certainly the deep
ocean on earth. These probes return not only the raw data
they aquire, but also possible theories to explain their obser-
vations; and most importantly, if they encounter something unex-
pected they have the ability to drop less important measure-
ments, reallocate equipment, design the experiment, perform same,
and analysis the results, ALL WITHOUT COMMANDS FROM EARTH!
As a result of these probes, nearly all physical knowledge of the
universe after the launch of these probes is due to machines,
not humans.
2050: The meta-expert systems will be developing increasingly more
general expert systems. As a result, more interdisciplinary
experiments will be performed by computers. Additionally,
social science 'experiments' will be commonplace using the
census data as the data base.
2070: Most of the Nobel Prizes go to computer systems. Most programs
are written by computer. Most program-generating programs are
generated by computers. Most ....
2100: At least one probe has been sent into a previously inaccessible
location, after being designed, built, and operated entirely by
computers, and being designed entirely on principles discovered
by computers. Possible probes include: a stellar probe (into the
sun's outer atmosphere; a Jovian probe (into the inner atmosphere
of Jupiter); a earth probe (desending into the lower crust/upper
mantle of the earth); a interstellar probe (using high acceler-
ation drives, going into *deep* space).
At this point my crystal ball becomes too cloudly to be usable, but the point
I am trying to make is that even as soon as a single century from now, it may
very well be that most up-to-date knowledge is found by machine, not human,
intelligence. After that, how long will it be before it is considered
improper for a graduate student in a physical science to propose *actually*
doing an experiment for his thesis/dissertation? After all, can't a computer
do it faster, cheaper, better, etc. if it has not already been done?
Then in a 'mere' millenia, we will have a culture which can approach that in
*Startide Rising*. What would it be like after 7.5 billion years, and thou-
sands of sentient species have come and gone?
Cravet(?): I personally would want to continue duplicating experiments from
now until the big crunch. This would change portions of the above
culture, in that it would be acceptable to suggest you measure 'c'
for the 241,642,998,235,243rd time in recorded history, *so long
as* a computer system does the actual work. This is in contrast
to the culture in *SR* where it is not acceptable to perform the
experiment at all, but rather the value must be taken blindly
from the *Library*.
Any counter future histories?
ave discordia
-------------
Bruce Giles
decvax!ucf-cs!giles (UUCP)
UCF, Dept of Math, Orlando Fl 32816 (Snail)rf@wu1.UUCP (12/01/83)
Given the current human tendencies, I would guess that even the
effort to study *The Library* would be too much for most people.
Even the effort of teaching the Libraries enough to be useful in
a given culture might prove too much. Some cautious races would
probably suspect the Library of spying on them--and in light of
events in *Startide Rising* might be correct in their
suspicions.
I suspect that in dealing with a Library that old and big:
- The problem of learning the Library would be about as
difficult as the problem of studying the physical universe.
- Some substantial portion of the Library would be in error,
either through failure of checks made before the data was
entered or as a result of obsolesence. Even supposedly
fixed physical laws might change in billions of years.
Given the Galactic reverence for the Library, correcting
mistakes would be quite difficult.
It may be that, at some level, *thinking* is a process which
itself is limited. Even if not, wise sentients might not wish
to construct sentients who could outthink them. So perhaps the
Galactics never got about a millenium ahead of us.
Randolph Fritzciaraldi@rochester.UUCP (Mike Ciaraldi) (12/04/83)
Don't forget--the ancient Greeks felt that the right way to find out about the nature of the universe was to think and reason about it, NOT to do experiments. Thus, they thought heavier bodies fell faster, etc. Our modern "scientific method" says the opposite. Can we tie this in with the idea of looking something up in the Library? I suspect the important thing, from the point of view of the intelligence and creativity of the civilization, is how the information in the library is perceived. There is bound to be a combination of raw observational data, statistical and other data which comes from analysis of the raw data, and theories (and "laws") explaining them. This applies to the physical sciences as well as the social sciences, and could even spill over into art, literature, religion, etc. Example--one of the arguments about capital punishment is whether it deters crime. Given several billion years of data on crime rates in various civilizations and species, you could do a huge regression analysis on things like penal codes, affluence levels, etc., and find the (statistical) causes of crime under various circumstances for various types of beings. How valid are these results in predicting future behavior? The upshot is--do the library users ask the library computers for alternative hypotheses and their likelihood of being correct, or do they ask for the most likely answer, or do they ask for "the ANSWER" and assume it is absolutely and forever true? Excuse me if some of this has already been addressed in the books in question. I haven't read them yet. Mike Ciaraldi ciaraldi@rochester