Dear Reading Diary 4

22.01.13
THE NONTRIVIAL MACHINE

Okay, I tried to avoid this enty as long as possible, as it´s going to be the most complicated one. Complicater to explain and a bit harder to understand than previous entries. As I´m not the best in logical thinking i had to spend some time before I wrapped my mind about that matter. But to understand it opened my mind´s horizont again a bit more. So pondering about this post might be rewarded with an astonishing insight on your side as well !

All previous entries about my partly summarize of Heinz von Foersters biography "Part of the World"
kind of build up and correspondend with this essential post.
So let´s deal with the trivial machine in comparison to the nontrivial machine.

McCulloch and Walter Pitts were two smart guys who wrote an article called " A Logical Calculus of Ideas Immanent in Nervous Activity".
This article got pretty famous amongst science because of an amazing discovery they made and describe in it.
It´s centered about neurons, the nerve cells. A neuron can give and receive an elecrificial signal. The neurons are conncected to each other and can send messages in electrific ways.
McCulloch and Pitts found out that everything, what can be described, can be transferred in a nervous network.
This opens the doors to Artificial Intelligence. Because according to that system a machine now can do everything what can be described. You just have to build an according artificial kind of nervous network.
If you say then "but machines can´t do THAT..." , you just have to describe what that THAT is; and another machine will be build which exactly can do THAT.

Now we come to trivial machines. An example for such a machine is an Anagrammor. That thing turns one letter into another letter, A to B, B to C, C to D and D to A.
Our trivial Anagrammor now has just one transformation rule: A to B, B to C and so forth.
If we add more transformation rules, the Anagrammor ends up to be nontrivial: An A can now become as well a C. And the next time another A gets a D. because now different decision options are available.

The transformation rule of a trivial machine can be found out. An analysis is possible.
to find out the transformation rules of nontrivial machines is pretty impossible, as they constantly change their way of the working progress.

The trivial machine is predictable.
 Aha, our trivial Anagrammor always takes a B after an A, a C after a B etc....
Nontrivial machines are unpreditable.
 I will never know if my nontrivial Anagrammor chooses a C, B or D after the A.

people changed their way of thinking over the course of history. Back in the past, in the 18th century, the world was imagined as a trivial machine. it was believed that the course of the world, the whole future can be predicted, according to the believe that the principles of how the world ticks were as precise and systematical as a clockwork. if the world´s structures are like a clockwork, it would be possible to find out how the world ticks.

That assumption can be falsified. It can be refuted like this:
We became aware how we decide over undecidable questions. Humans are nontrivial machines. How we act and decide is unpredictable.

If we try to decide now if the world is a trivial or nontrivial machine we have to ask ourselves: Are we part of the world we try to analyze?
because if we are part of the world, a nontrivial element comes into the world. Then the world can´t be trivial anymore, as the course of happening isn´t predictable anymore. So we can´t foresee the future.


There are undecidable questions. We have the freedom to choose between alternatives when it comes to those questions.
With freedom comes responsibility. We are responsible for the decisions we made.