Perhaps I should call it Asymmetric-Symmetric Functions or Relationships or something. I thought that it really only applied to mathematics, like combinatorics, so I started to wonder what some of the non-mathematical AGIers might think: what does this have to do with AGI? So I tried to see if I could apply it to language as a way of saying - see it is really more general than just mathematics. I came up with this:
Say you have a sentence-concept like, "Jennie gave Bob a book." The symmetric sentence might be, "Bob gave Jennie a book." But that symmetric sentence does not retain the original meaning of the sentence. Chomsky was fond of the transformation which would produce, "Bob was given a book by Jennie," but notice that this is not perfectly symmetrical and it could be confusing in this particular case. Working from my ideas about math, I tried to force a symmetric sentence which retained the meaning of the sentence. I finally came up with, "Bob received a book from Jennie." This is a 'symmetric' sentence, it is close to retaining the meaning and I think it works better than, "Bob was given a book by Jennie." However, at the same time it is asymmetric. (I guess it is asymmetric in a few ways.) But never the less it is a form of asymmetric symmetry in a very true sense. If you wanted to set some simple computational order to store (or create) sentence-ideas in an efficient way, you could run into complications with sentence-concepts that have a different meaning when the Person Nouns are exchanged. Jennie Gave Bob a Book. Bob Was Given a Book by Jennie. Bob Received a Book from Jennie So I ended up with these two very symmetrical forms of an Asymmetrical Symmetric Relationship that might exist, for example, at a deep transformational level. Jennie Gave a Book to Bob. Bob Received a Book from Jennie Jim Bromer On Tue, Oct 13, 2015 at 5:05 PM, Jim Bromer <[email protected]> wrote: > I became annoyed about the discussion of a computational model which > was called non-algorithmic, but then I started thinking of one of my > pet ideas, asymmetric symmetry. A derivation taken from some kinds of > initial systems may exhibit a great deal of symmetry. The symmetry of > the derived system may be somewhat abstruse or abstract. But if the > parts of the initial state of the system are then changed, (imagine > points that are moved or deleted) then the symmetry of the derived > system may be skewed past recognizable form. What I am trying to say > is that the output of methods which gained efficiency or traction due > to the abstract symmetry of the derivation methods (acting on the > right kind of initial system) may still be found by assuming that > symmetry exists in some hidden (or more hidden) form for the modified > initial state. > > So even though the term, "Asymmetric Symmetry" may look like a dim > contradiction of terms, it can describe a valuable way to look at > certain kinds of systems. > Jim Bromer ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657 Powered by Listbox: http://www.listbox.com
