(Note: any definition of “non-linear” you claim you meant should cover
the large class of AGI approaches you claim are bound to fail because of
Richard-complexity.)
Ed Porter
-----Original Message-----
From: Richard Loosemore [mailto:[EMAIL PROTECTED]
Sent: Thursday, April 24, 2008 12:38 PM
To: agi@v2.listbox.com
Subject: Re: **SPAM** RE: [agi] Adding to the extended essay on the
complex systems problem
Russell Wallace wrote:
On Thu, Apr 24, 2008 at 2:14 AM, Mark Waser <[EMAIL PROTECTED]> wrote:
> The real gotcha, though is the "Are the functions describing
> the behavior deeply nonlinear". You're just not going to find that
with the
> first three.
Actually, it's true of every program significantly more complex than
"Hello World" that the functions describing the behavior are deeply
nonlinear. This shouldn't be too surprising, because it's also true of
every electronic device significantly more complex than a length of
wire, every chemical system other than a hydrogen atom, every
mechanical device with more than one or two moving parts, every fluid
dynamics system that involves turbulence - and every neural network
more complex than a one-layer perceptron, so if complexity made
systems undevelopable, not only could ANNs not operate, but organic
brains could neither develop in the individual nor evolve in the
species.
The notions of nonlinearity and complexity simply don't do what
Richard wants them to do.
You have fallen into two traps that I have to deal with over and over.
1) The claims are meant to be applied at a chosen level of description -
jumping down to other levels of description is not relevant.
For example, if a physicist that "below the elastic limit, a spring is a
linear system", would you drop down a level and call the physicist a
fool, because "every chemical system other than a hydrogen atom [is
deeply nonlinear]" .... the exact statement that you just made above?
If you would not do such a thing to the physicist, why do you try that
trick on the argument I just presented?
2) There is a widely accepted, broader sense of 'nonlinear' than merely
"not described by an equation in which x is proportional to y". That
broader sense is roughly equivalent to "cannot separate the variables".
You then go flying off on a sequence of non-sequiteurs ("so if
complexity made systems undevelopable ...") which imply that I made
arguments that I did not.
Richard Loosemore
-------------------------------------------
agi
Archives: http://www.listbox.com/member/archive/303/=now
RSS Feed: http://www.listbox.com/member/archive/rss/303/
Modify Your Subscription: http://www.listbox.com/member/?&
Powered by Listbox: http://www.listbox.com
------------------------------------------------------------------------
*agi* | Archives <http://www.listbox.com/member/archive/303/=now>
<http://www.listbox.com/member/archive/rss/303/> | Modify
<http://www.listbox.com/member/?&>
Your Subscription [Powered by Listbox] <http://www.listbox.com>