There is a problem of saying that a categorical type can be used as an
element of a pattern derived from the groupings of various instances
of that type.  Suppose that you ask the question of how many patterns
are in the entire collection, like the one that I offered before
(found below). Not only were the 5 patterns made of different
characters but if you claim that the type (ie a character in this
case) can be used to refer to 5 patterns of a type (or of 25
representations of that type) then there would be an infinite number
of patterns in the collection because you could say those characters
belong to an infinite number of types (composed of any definition
offered that would include the characters.)  So by this reasoning you
could claim that any collection was a pattern.

I have two things to say about this criticism.  One is that a group of
characters from the alphabet (or from a greater collection of
typographical characters) is something that is understood by many
people and has strong recognition as an important example of
commonality.  (The word "type" even comes from that system).  But
secondly, even more relevant to our conversation, is that one of the
most important problems that AGI has to overcome is the problem of
finding hidden patterns.  I disagree with people who think that we can
find important patterns by examining Input thoroughly enough. For many
cases one has to infer the relationships (that may have been derived
from experiences interacting with the IO data environment) in order to
find the hidden patterns that can be critically important to making
AGI programs feasible.

So although there are some valid criticisms of going too far with the
extension of this definition of what a pattern is, at the same time,
the discovery of hidden patterns, including those that can only be
inferred by reasoning that can only come from many many sources of
information, is what AGI is all about.

Jim Bromer


On Sun, Aug 26, 2012 at 6:27 AM, Jim Bromer <[email protected]> wrote:
>
>
> 9 repeated typographical characters (like "l l l l l l l l l" is the most 
> primitive form of patterns and the referent of the cardinal number does -of 
> course- refer to the pattern).  However, it is so useful to think of 
> something that is a type that can be drawn from a pattern (like any 
> typographical character) that we may start thinking of collections of 
> patterns as patterns.  Name everything that is repeated in this collection:
>
> l l l l l
>
> e e e e e
>
> g g g g g
>
> h h h h h
>
> a a a a a
>
> Not only is "l", "e", "g", "h" and "a" repeated 5 times but there are also 5 
> groups of the 5 typographical characters.  Problems like this are useful to 
> help people who are capable of dealing with unconventional insights to do so. 
> Being able to work with ideas like this is a sign of intelligence and 
> child-level maturity.  So the dull conventionalist (who is aware of one of 
> the narrowest definition of the idea of pattern) might only see 5 separate 
> patterns but the more intelligent person will be able to deal with the less 
> conventional insight that there are also 5 different examples.  If you can 
> see that different typographical characters might be taken as being -of a 
> type- then you should be able to understand what we are talking about.  Being 
> able to understand stuff like this is very important in computer programming 
> since you may want to treat a type as a pure variable representation.  If you 
> can't accept that then you are not talking about the same thing the rest of 
> us are talking about.
>
> 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