This is the crazy logic that pervades AGI -

“just because all algorithms – numbering many billions – have totally failed to 
produce anything new whatsoever – doesn’t mean that they couldn’t still do so”..

“just because we have no evidence that matter is doing any computing of its 
behaviour, doesn’t mean that it *couldn’t* be doing some computing”

Well, no, and the laws – i.e. regularities of behaviour - of the universe 
*could* be upended tomorrow – just because the earth has always gone one way 
round the sun, doesn’t mean it couldn’t opt for a new course, say the opposite 
direction, tomorrow.

It’s true nothing is absolutely guaranteed.

But to qualify as more than another AGI crazy type, you have to offer some 
*reason* – some empirical reason why – a) algorithms could suddenly produce new 
elements or b) matter, and presumably some parts of matter, are doing some 
computing.

You have to produce real world reasoning about the nature of 
things/algorithms/matter – not just logical legalistic wordplay -  and that is 
totally lacking here.

From: Jim Bromer 
Sent: Friday, August 03, 2012 8:47 PM
To: AGI 
Subject: Re: [agi] Fw: Pre-pub Offer: A Computable Universe

Intuitively, I would agree that the proposition that a natural occurrence is 
"computing" the effects it has seems like a stretch of the term. However, the 
idea is interesting.  And I can use reason about to find some interesting 
validations of the conjecture. First of all, any model is imperfect.  So 
imperfection is not in itself a good reason to reject a model.  And models are 
based on abstractions and simplifications of the event being modelled so the 
idea that there are limitations on the application of a event of the universe 
does not show that the event cannot be likened to a computation.  Next, while 
someone may set up a computation, that does not mean that he controls every 
aspect of it.  And we can observe computations that other people have started.  
So the recognition that an observer of a natural event is not necessarily 
controlling the event is not, in itself, a substantial reason to reject the 
conjecture that the event can be likened to a computation.  Furthermore, we can 
interact with natural events and so we can make conclusions like that which 
states that the orbit of a satellite is "computing" the orbit of an object.  
Why not?  Some natural events have not been reduced to pure systems of 
regularity, but we can use mathematics to model systems that exhibit some kinds 
of regularity without requiring that the totality of the system be regular.

So while the conjecture is based on a stretch of the term computation, it can 
still be enlightening to consider it.  We might learn something and we may 
sharpen our thoughts about such things.

However, there is no way to establish that all things and events in the 
universe are governed only by regular and computable methods.  It is just an 
interesting idea which is really useful when it is done well.

Jim Bromer


On Thu, Aug 2, 2012 at 4:33 PM, Mike Tintner <[email protected]> wrote:

  Perhaps this is a useful insight into where people may be going wrong.

  The “laws of physics” are computable – IOW it is true that humans can use 
computers to express and explore mathematically defined regularities of 
behaviour in matter.

  But it doesn’t follow that matter itself is doing the computing, or that some 
deus ex machina behind the scenes is doing the computing.

  Just because planets follow regular, mathematically definable orbits, doesn’t 
mean that they are computing their orbits – or that there’s a mathematical 
God/divine principle who/which use mathematics to calculate and arrange their 
orbits.

  This looks like another example of the logicomathematician’s extreme 
solipsism – “if I’m doing it, everyone and everything must be doing it  - the 
world is just an expression of my mind.”

  Er no it isn’t.


  From: Jim Bromer 
  Sent: Thursday, August 02, 2012 9:13 PM
  To: AGI 
  Subject: Re: [agi] Fw: Pre-pub Offer: A Computable Universe



  On Thu, Aug 2, 2012 at 8:55 AM, Mike Archbold <[email protected]> wrote:
  >
  >
  > It seems like the goal posts keep getting moved around as far as what does
  > and doesn't constitute computing, now seemingly becoming so broad that just
  > about any interaction is "computing" complete with a $99 book on sale.  It
  > looks interesting anyway... 


  The (so called) laws of physics are based on computable functions. That is an 
old idea. So, the argument goes, perhaps all the laws of physics may be 
expressed using computable relations.  Even non-mathematical relations might be 
expressed as a kind of logical relation and a logical relation may be expressed 
as a mathematical relation. The thing that makes this argument interesting is 
that the laws of physics have played such an important role in modern science.

  There is a difference between the concept of computation and an effect of a 
pysical reaction because a computation can be used to model a wider range of 
computational systems than some particular reaction.

  However, there is a chance that the computable laws of physics have acted as 
a powerhouse not because nature is just a system of computable effects but 
because the arithemtic of the modernn positional notation system has an 
effective power that is still a bit hazy.  Indeed, you can even see that lack 
of clarity in this group.

  The modern numbering system represents an impressive method of "compression" 
where a value may be represented in a highly efficient manner.  (Imagine if you 
had to use ten million sticks to represent 10,000,000.  It is not as efficient 
as using ten -or eleven or eight- characters of the alphabet.  Ten million 
things vs a selection of eight marks drawn from a selecton of twelve variations 
or characters.  That alone is one of the most amazing things that human beings 
have ever accomplished.  But another, even more unexpected feat has been 
accomplished using the modern numbering system.  By using addition or 
multiplication we are able to "compress" the number of steps that we need to 
take to calculate a transformation of a method on two such numbers.  We don't 
have to add ten million representations of 10,000,000 in order to calculate 
what 10,000,000 X 10,000,000 equals, instead we can do it in around 64 
representation-steps.  (I am not going to check that but it is in the ball 
park. Of course we don't have to do all of the steps for a multiplication where 
the two decimal numbers has long strings of zeros in them.  I knew that.)

  Saying that the underlying reality is a computable universe is taking a major 
leap.  Saying that the underlying reality of the major advances that have 
emanated out of physics and computer programming is the power to "compress" 
representations and "compress" transformations of such representations is not 
that much of a step.  

  By expressing this I am starting to see that particular kinds of 
relationships might be expressed in order to compress certain kinds of 
transformations of representations of ideological objects.  However, I don't 
see anything that would come anywhere near the compressed transformations of 
arithmetic. Party like it's 30 BC.
  http://en.wikipedia.org/wiki/Cuneiform

  Jim Bromer


        AGI | Archives  | Modify Your Subscription   

        AGI | Archives  | Modify Your Subscription   


      AGI | Archives  | Modify Your Subscription   



-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968
Powered by Listbox: http://www.listbox.com

Reply via email to