An AI (AGI) program has to have Input-Output capabilities.  This seems so
obvious that it is.  An AI program must be compnonential.  I do not know
how to say this in a more sophisticated fashion.  Even if the AGI data is
represented as a weighted method or if the program was written for an
analog computer, the basis would still be componential.

It seems obvious that pure logic is insufficient to represent AGI.
Although it might be possible, it would be much too inefficient for us to
use with contemporary computers.  There are few people who argue that pure
deductive logic should be the basis for AGI.

And it seems obvious that because of the number of components that would be
needed to refer to the objects, relations and the nature of those objects
and relations that would exist in many IO Data Environments that we are
able to deal with, there are going to be combinatorial problems based on
the number of combinations of components (of idea-like data) that might be
used to refer to some aspect of the IO Data Environment.

Because this combinatorial complexity problem is exhibited in every AI
endeavor that goes beyond anything simple, it is pretty easy to guess that
the contemporary problem with genuine AGI is due to combinatorial
complexity.

Jim Bromer

On Sat, Nov 10, 2012 at 8:02 AM, Jim Bromer <[email protected]> wrote:

> On Thu, Nov 8, 2012 at 10:30 PM, Todor Arnaudov <[email protected]> wrote:
> I agree with Aaron's overall position regarding the absurdity/uselessness
> of the logical paradoxes - bugs of data and combinatorial games when the
> sensori-motor roots are forgotten or lost.
> ---------------------------------------
> I have no idea why you would even bother to write something like this. If
> you are claiming that the use of sensori-motor data is what would make AGI
> feasible, whereas those of us who feel differently are clueless then you
> would, in your estimation, have everything that you needed to create an
> actual AGI program. Even though robotics is a little difficult, visual and
> audio is cheap and easy to input and it would not be impossible to create a
> very simple robot that had sophisticated visual and audio capabilities.
> On the other hand if you did agree that even with a sensori-motor data an
> AGI program is probably beyond your current capabilities then what basis do
> you have to say that sensory-motor data is necessary for AGI. Let's see,
> you are saying that robots with sensori-motor data do not have
> bugs???????????? Of course that is not what you are saying. So you are
> really just saying that "combinatorial games" are the problem of
> logic-based AGI right? Again, do you honestly believe that combinatorial
> problems is not a problem with robotic based AGI? If combinatorial problems
> are not a problem with robotic-based AGI then what is the contemporary
> problem? Why hasn't genuine AGI emerged from the field of robotics? Are you
> working in robotics? If so, explain how your achievements have solved the
> problems that AGI have encountered.
> You did provide reasons for your belief, but as I have tried to explain, I
> do not feel that the two reasons that you used in your introduction were
> substantial.
> Jim Bromer
>



-------------------------------------------
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