A revolutionary OS/Programming Idea

[Lynn H. Maxson]

John Newman writes:
"...Sometimes we think that when evolution theory refers to 
'chance,' we mean randomness.  What we really mean is 
'unpredictable,' relative to present human standards.  But for 
one to talk rationally about any concept, beit evolution or 
whatever, one must apply cause and effect.  Any other 
method employs the 'god of the gaps.'  We don't need to be 
creating unnecessary black boxes (however tantalizing that 
may sound)."

As I suggested earlier you need a good dose of "Language, 
Thought, and Reality: Selected Writings of Benjamin Lee 
Whorf" and just to crowd it a little more Korzybski's "Science 
and Sanity".  I could continue with a litany of others perused 
over my longtime involvement in the study of general 
semantics.

In software the basic building block, our universal black box, 
is the IPO (Input/Process/Output).  We connect them to make 
systems.  We could say that the input represents cause and 
the output effect.  But the minute we put two of them 
together we find that the effect of one is the cause of the 
next.  If we put the two of them as a closed cycle as a system 
with feedback, cause and effect become indistinguishable.

We do wierd things things in a language like english with our 
subject-verb map we place on reality.  We say, "The lightning 
flashed" when no separation is possible: the lightning and the 
flashing represent a single phenomenon.

In english we have a concept of "private property" or 
"ownership" due to the presence of "possessive pronouns".  
Languages without them have no such concept.  In short we 
use language to create maps of reality.  Sometimes we 
irrationally try to make reality fit (conform) to our maps 
instead of making changes to our maps to better fit the 
territory.

The general semanticist would say, "Whatever you say 
something is it is not.": the map is not the territory.  We poor 
software people might offer (as I have in the past) that you 
cannot sit in the shade of a simulated tree.

I tried to do a google search on a different book by E. F. 
Schumacher, author of "Small is Beautiful" in which what we 
can seek to know of reality falls into four categories, only 
one of which includes the scientific method, the predictive 
science of cause and effect.  Cause and effect are two sides 
of the same coin.  From any cause (the effect of some earlier 
cause) the effect is what occurs next regardless of its 
predictability.  Cause and effect, this sequence, represents  
arbitrary "cuts" in an otherwise continuous process.  Reality is 
ongoing, never hesitating regardless of the arbitrary "stops" 
we engage with our language, our map of the process.  In 
reality no "god of the gaps" occurs as no "gap" exists.

"... An assertion not founded in evidence.  Von Neumann 
machines are based on the concept of universality--It can 
emulate any other machine.  The human brain is Universal 
machine precisely because the brain can abstract a conception 
of any given 'possible' machine.  The catch is, one's brain must 
be at least as complex as any machine it can emulate. ..."

Von Neumanm machines can emulate other von Neumann 
machines.  To say otherwise to quote you is "an assertion not 
founded in evidence".<g>  We have no evidence that the brain 
is a "universal machine".  In fact we have evidence to the 
contrary.  It cannot conceive every possible machine, of which 
the "given" ones remain only a subset.

"...Many agree that computers are universal, because a mac 
can emulate a pc and vice versa, etc. ..."

Well, it's always somewhat disturbing when so many exist in 
denial, using the rules of mathematical induction 
inappropriately.  Computers are not universal machines, for 
the simple reason they are based entirely on formal logic and 
it implementation in logic circuits, "ands", "ors", "nots", etc..  
They may be universal within the realm of logic machines, one 
capable of expressing any other.  They are not universal with 
respect to non-logic-circuit-based systems, systems not 
subject to logical decomposition.

You see you are making an assertion here, not something 
which is either true or false, but unproven.  That proof 
process continues.  A RISC architecture can emulate a CISC 
one because it was designed with that purpose specifically in 
mind.

"...Statement based on unfounded assertion (see above).  
Matter is the machine that exhibits, or one could say, emulates 
life and intelligence.  We can emulate matter in 
computers(when we figure it out).  ..."

"Whatever you say something is it is not."  Matter does not 
exist except as a word, machine does not exist except as a 
word.  Life does not exist except as a word.  Intelligence does 
not exist except as a word.  These words are "map" terms 
different from the territory.  They are our "inventions".  
Unfortunately we can work with them in ways that we cannot 
in reality.  They are for our convenience.  Reality does very 
well without them.<g>

We cannot emulate matter in computers.  To do so implies 
that it is decomposable, that reality (the name we give to it) 
can somehow be broken up, stopped, separated, and yet 
remain the same.  Whenever you do this you deny its very 
essence.   If you want to fly in the face of Heisenberg's 
Principle of Uncertainty, be my guest.  You will not end up 
with reality.

"...So, first, you must formulate an argument against a 
mechanistic nature.  Then you have the problem of proving 
that this 'non-mechanical' black box can interface in a 
mechanistic way with the mechanisms that are evidenced by 
nature (like evolution and sentience--both prominent black 
boxes in creationism). ..."

Again "mechanisms", "evolution", "sentience", and "black 
boxes" exist only in our maps and not necessarily in the 
territory.  In fact they may inhibit our ability to understand it.  
We can engage in fantasies with our words.  Reality has no 
use for it.

"...What of a function like phi/Golden-Ratio?  Is its 
inexolerably resultant infinite sequence an intention of 
someone who implements the function?  How can it be when 
it is impossible for any programmer to know the entirety of 
the outputting sequence?  No one can influence the evolution 
of phi or pi.  For all intents and purposes, they evolve on their 
own. ..."

I've hear this illogical argument before from supposedly logical 
people.  The fact that we cannot predict all possible outcomes 
does not mean they are not predictable.  As they are 
software based we know they are predictable.  We know that 
if we had a infinitely long life we could emulate the same 
progression with the same results.

"... So, evolution likely harnesses search functions which 
exhibit this type of unboundedly complex output, thus 
maximizing the search space.  From what I know about 
complexity science, that's my take on it. ..."

Evolution doesn't exist in reality.  It is our "crutch" that we 
employ in our maps of it.  Do not confuse the map with the 
territory.  We need to understand that all these words we 
have created may or may not actually represent what occurs 
in reality.

The only distinction I make here lies in the assertion of a 
non-programmable machine, one whose program does not lie 
in an instruction set.  It doesn't program itself (as it doesn't 
have to) and you can't program it without interfering with 
what it would do on its own.  With an instruction set you may 
emulate but not replicate it.  In doing so you will have created 
something else.  I am not arguing for or against the usefulness 
of such  an approach, only that whatever results will not will 
not be the "same" as what you emulate.

If that's not your concern, then at least have the courtesy to 
remind others as well as yourself that the emulated results 
work "as if" they were the real thing, consciously reminding 
yourself that they are "not" the real thing.  The difference 
continues to exist.  When you get away from the implications 
of "as if" that's when you confuse the map with the 
territory.<g>