John Herbster writes:
Quote
When it was started, this thread was about the progression
from universal tools (like knives, sticks, and fire) to
more specialized tools and possible analogies to the
progression of computer programming tools and technology.
One can depict the current progression in many ways. For example, how to
"interpret" knives and sticks in more abstract terms, in order to avoid
comparing apples with peaches?
One attempt would be to say that knives and sticks are tools in the form
of "structures" applied to other structures, in order to (re)shape those
structures.
Now, today we have added to it other "levels of abstraction" in that we
do not only produce "structures" (hardware tools), we also equip the
tools with Behaviour (as a complement to structure). And not only that,
we can "teach" our tools to optimize a given task, with more or less
guidance when the actual work begin.
From biology we can learn how adaptive concepts can very efficiently
optimize itself for a given task, for example the immune system applies
a form of biochemical "genetic algorithms" in super-quickly trying out
different biochemical structures fit to lock up foreign intruders, etc.
In all this we clearly see distinguishable corrective, and protective
concept in action.
At last, if you compare some very significant foundational
characteristics of modern tools with the old knives and sticks, you'll
soon notice that in the old days the tools were both Guided and
Integrated (if several sub-systems were involved) by manual guidance,
that is, the human brain was the Real Time Integrator guiding the work
processes.
Today we have developed both adaptive and self optimizing control
systems for such things.
The next step, the big challenge for today and tomorrow's software
systems, seems to me to be stable solutions for Scalability and
Integration of the many "sub-technologies" we master today.
The conceptual solutions for Scalability and Integration already exist
(although not being well known or understood). Some of the basic
principles required for such have already been tried using Delphi, but
probably any modern programming language will do. The trick lays not in
the language, just like in the DNA the conceptual solutions it comprises
lays NOT embedded in the amino acids, instead the specifications of the
concepts which the amino acids embeds lays in how the chemical words are
ORGANIZED.
That is, I could have written this post in any natural language, using
entirely different words and letters, while still retaining the same
semantics.
And now Semantics, for Organizing and controlling structure, and
Behavior, and for directing Optimization and most important, Integrating
all this in Scalable systems, on different levels of abstraction (we can
simulate even "abstract principles" today, not only run physical
models), that is some of the aspects I find most eye-catching when
comparing with knives and sticks, and untamed fire.
But in G K Chestertons heyday they still didn't very much about
biological information systems.
Don't forget though that "biological information systems" are by no
means "modern". They have only started to slowly become KNOWN to us, and
to some extent also understood, in our days.
Even the hidden secret of computational performance lays embedded in
molecular biochemistry. I have dubbed it "implogic", that which is
illustrated by the specially designed IBM super computer "blue gene"
(2005) which were planned to fold a fairly simple protein. The plan was
to perform the folding task within a year 2005 of intense computation.
In our bodies millions of such (parallel) folding processes are
performed in less than a second, using the principle of Implogic.
One could say that we still have some ways to go before we can say that
we master the art of processing and integrating scalable and stable
dynamic information systems.
Regards,
// Rolf Lampa
