[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Random results vs low probability results



While I am not now, nor ever likely to become, a proponent of chaos theory,
the analogy to coding "random" events doesn't apply. In programming, the
greater  challenge in programming is not the actual construction of code,
but rather the recognition of all possible events during the execution of
that code. You have accounted for the genetic history for the species
involved in a mass-kill event (for example..." there is some bias when
choosing a starting point for the array") but you cannot account for the
non-random nature of the natural disaster. I am not sure if you are familiar
with this technique....a couple of buddies are mine employ this technique
for hard-core security (& w/o too much detail)...a random # is generated
using sub-molecular particles colliding into a measurement device. The
particles are from an unstable isotope. This to them, a random event. I
pointed out to my friends, that this would be a hackable defense. All you
have to do is figure out the physics. The only problem with that is that we
currently can't. So for all practical purposes, this method of cryptology is
sound.  It sure would be nice if those weather guys could get  it right, but
the fact remains that far to many variables exist for a biological entity to
account for all potential threats in its finite lifetime.  So a bolide
impact, volcano, poison gas, is in effect, a random event., no matter the
differences in sensory systems. A deer might detect a "poison gas" explosion
at an earlier time, than a ground sloth, but that does not change the
"randomness" of the event itself. In order to remove the randomness from the
event, a deer would have to understand all the events that directly led up
to the explosion (and it might  be argued that all events leading to the
explosion might need explanation). I couldnt even begin to count how may
studies plot morphospace vs extinction patterns. But that doesn't change the
random act, it merely helps to identify the cause of extinction or the
converse, why some survived and some didnt. Which leaves us right where we
were before.

Analogy: you code in an autosave feature into a spreadsheet you created.
This should minimize the effects of a poorly written, over-bloated operating
system like Micro$oft  WinDung. A hillbilly terrorist, angry at the US Govt,
sets off a bomb made of fertilizer outside the building your network resides
in. Result, a random event from your perspective, made your autosave feature
not so adaptive. Why didn't you account for that? Say you did, you have
disaster recovery procedures in place....i.e. your data is spread out like a
population of rats. Negating the trouble of spreading out gentic data
without creating a "hard-copy" , natural disasters that can cause
extinctions, are usually a HECK of lot more powerful that a  truck bomb.

-Dave Lessin
-----Original Message-----


>   I agree with Joseph Christopher Beamon in that too many events labeled
as
>random are really what I would call low probability results(LPR). Events
>such as speciation and extinction are governed very little by random events
>and much more by LPR.
>   In programming, we work with psedo-random numbers and call them random
>with the understanding that they are not truly random.Most are based on the
>computers internal clock to generate an array of "seed" numbers to start
>from. The computer clock is considered "accurate" in increments of 1/10 of
a
>second. There is bias though in determining a starting point for generating
>the random number array. It can be improved(made more random) by adding
>sub-routines, but it will not be random in a mathematical sense.
>   To end, I would like to know if any studies have been done in post
>extinction event fossils to indicate morphological trends across all
>vertebrates. Such as adaptations for burrowing or digging which could aid
in
>surviving extinctions in mammals, reptiles, amphibians. Thanks, now back to
>lurking and programming and sports analysis.
>                                                            Tim Bollier

>From: Joseph Christopher Beamon <c992@scatcat.fhsu.edu>
>>chaos is the mutations and mating combinations formed in each
>>generation, as well as random kill-offs (the afore-mentioned poison gas
>>clouds) and random survivals (those critters in the afore-mentioned
>>valley).
5