(urth) 5HC : Skinner, Turing, Laplace

maru marudubshinki at gmail.com
Sat Mar 19 23:45:31 PST 2005

Chris wrote:

> ...
> Bring the demon inside the universe and close the system. Now he can 
> make predictions, and tell them to you. But these predictions require 
> a full knowledge of the state of the universe - which now includes 
> himself. The demon must "know" the disposition of his own brain states 
> - he must know the totality of what he knows. So for each item X in 
> his knowledge, the fact that he knows X requires him to know (that he 
> knows X), and this in turn requires him to know (that he knows (that 
> he knows X)), and so on in an infinite regress. This isn't just a 
> trivial abstraction. The problem is analogous to the problem I 
> mentioned of making a precise map of the universe. If the map is 
> inside the universe, then it must include itself, which must include 
> itself, which must include itself.
> [A curious aside... An abstract but almost entirely meaningless 
> possible solution to the map problem is to simply call the universe a 
> scale map of itself. This requires a somewhat... flexible... concept 
> of "representation". If you were George Berkeley you might apply such 
> a solution and say that the universe *is* the demon's mind. But such a 
> demon would no longer be Laplace's, it would not be predicting 
> anything, it would be making the future (and all of being) by thinking 
> it. But in any event, this wanders just a bit off from the purpose of 
> the thought problem.]
> Civet
Have you ever heard of a replicating von Neumann machine?  I mention 
them because they  at first sight
to suffer the same reflectivity problem as LaPlace's demon.  But the 
solution is similar to Berkeley's
but different: at one time the machines is storing its data, but this 
data is also simultaneously processed as a program
which copies that data (in the software version).  No physical ones have 
been built (the closure problem:
getting a machine to derive
and develop all its resources and manufacture all its parts and also to 
be robust enough to actually work.)
I think isomorphisms are the key here.

But if you do not like that explanation, consider a demon equivalent to 
an angel: it can be any size it wants.
And since by default this universe is continuous, the demon can make as 
many brains as he wants and fit each in
an infinitely small space. Now, since the demon has a visible head and 
presumably brain for me to converse with
and receive predictions from, it follows that it posses infinite amount 
of brain; I suspect an infinite brain could
contain and hold that 'infinite regress'. In fact, an infinite number of 
infinite regresses.

Dat kwazy set theory! So counterintuitive.

More information about the Urth mailing list