Re: random, was Predictions & duplications

From: Marchal <marchal.domain.name.hidden>
Date: Thu Oct 18 05:34:05 2001

John Mikes wrote:


>Bruno, I appreciate your choice of incompressibility - as far as
>mathematical views are concerned.


And you know that with comp a case is made there is nothing
outside mathematics (even outside arithmetics) so that's ok for me.
But you know also that the determinist self-dup entails the
possibility that we are confront with such form of indeterminism.


>How about a "random" choice of a color
>from a hundred others?


If the colors are clearly distinguishable, you can use an incompressible
string of O1-symbols translated into base 100.
It is not practical, because you will not be able to prove it is
random, but the point is that the UD generate all such sequences, so
comp "nature" makes such random "choice" all the "time".


>can this be algorithmic and incomressible?


Yes. But it cannot be known for sure.


>Or a choice "at random" from available several routes, how to defend an
>innocent accused in court?


I don't see the relationship. To defend a innocent accused in court,
you better should refute the evidences given by those in charge.


>I admire you, physicists, ...


I'm not a physicist. I am, let us say, an hesitator between
biology and chemistry, who after discovering "godel's result"
realise that an abstract biology exists as a branche of mathematics.
So I decide (in the early 70) to study mathematics for showing
that it is not the cells who obeys to the laws of chemistry (like
James Watson said in "molecular biology of the gene") but it is
the laws of chemistry which obeys (in a deeper way, sure) to the
laws of cells (self-replication).


>...for writing an equation to everything.
>I cannot
>find this applicable with infinite variables and infinite levels of
>influences among applicable factors (=natural systems).


If the equation of everything is supposed to give a complete
explanation, or even just a complete description, I certainly
do not believe in it. Not because of some natural systems (I
don't believe that exists with comp) but because there is no
complete theory of numbers and/or machines.


>I do not find Russell's "choice without a cause" applicable without
>interjecting "known" before 'cause'.


Instead of "choice without a cause" I would say "selection of
possibility" without any mean to influence the selection, like in
the self duplication experiment in the comp frame. Here we can know
even God cannot know the result of the selection.


>Which does not mean deterministic
>extremism, the choices are close and fractalously propagating into quite
>diverse routes, so it is beyond designability how a later situation will
>look.


Yes.

Bruno
Received on Thu Oct 18 2001 - 05:34:05 PDT