It seems that language is one of the basis problems with all fields of science. When individuals use language in unusual ways they conceal the underlying process. In programming the proper naming of a library can make it accessible and bad names can make it less useful. If I name a graphic primitive Kill_All_Running_Process() and it simply displayed a triangle with three points and a surface, then it would be hideously irritating to use. In the same way physics establishes names and related concepts with no sense as to the ability to be used outside of a selected group that speak the manufactured dialect.
The underlying method or concept is unchanged by the association of a symbol, however the communication and thus its common utility is destroyed by assigning a vocabulary that obfuscates the relationships. It seems to me that if I am to create a new extension to C, that I should deal with the interface problem to the culture. C is chosen because it can be compiled to practically any machine code and has a reasonably well defined behavior in practice.
It would seem that if all objects and methods were named from a common category, that it would be coherent and extensible. If a set of rules determined how a new object was to be named, the discovery of the relationship of its symbolic representation and actual nature could easily be determined.
The computer represents everything as numbers and so it could be said that a unique number is as good as any language to represent an object. I have discovered many things in the field of physics and I don't waste time giving them cute names. It is a disability and not a method that fosters communication. I discovered something today in quantum physics that I had devised by other means and in order to get to the concept it was necessary to travel the sphere of knowledge all around the globe to find a concept which is easily accessible by turning around.
So I need to deal with language in an AI so that what is said is what is known. Thus the relationships of matter prescribe the relationships of its use. I do not care to translate a concept to a language and system that does not contain a symbol for constructing the concepts. Dancing about the camp fire.
In cladistics the relationships of organisms is presented as a tree or categories and it may seem to demonstrate a relationship, but the underlying physics is such that it is possible to move from point A to point B in an infinite number of ways. To say that B is derived from A is simplifying the relationship and throwing away some aspects. Is it a toad or a frog? Is it a quark or a neutrino? It is a system of communication without any underlying rational basis. Why would object A and object B, which differ by some small variation be called toad and frog? I could see that if I were using UUIDs to represent objects in an ENUM like fashion that the number would have no relationship to the object or its relationship to other objects. That is the crux of the matter, The language is arbitrary and the attempt to resolve it by using a mathematical system that has missing information simply compounds the failure to communicate the identity.
It has a very significant meaning when it is necessary to communicate with an alien race. If they are as silly as we, then it would be a nightmare. Communication must begin with symbolism that represents the common mode between two agents. In the computer world we have ASCII, and many other standards that serve to translate one form of energy to another to represent a common object. I am simply considering this in the hopes of devising a method for universal communication. I am certain that life exists elsewhere in the universe, the problem is how to establish meaningful communication. It is difficult enough with people of a common origin.
0 comments:
Post a Comment