Friday, October 15, 2010

GEB 1

Ok - First off, I am looking at restarting this, and I'll explain all when I get there.

But for now, gotta get this idea out:

So, GEB is talking about strange loops, self-referential statements, etc. There's a better way to put that that I think might come out the next couple of thoughts, but:

Branch 1: So, imagine you're learning a new language from someone you have no common language with. They can reference things around them, utter a sound, and from context, you can reasonably assume that the sound means the referenced thing. The sounds of "apple" and the context of the fruit connect to give the symbol "apple" meaning.

Branch 2: One of the problems of AI (AFIAK), is that, while we can make systems that can figure out relations between symbols, it can be hard to argue that the system then knows what a symbol means. Indeed; I think we can see this in certain mathematics, where we can do symbol manipulation on a set of symbols and rules, with no care as to what the symbols correspond to. They're meaningless in their lack of reference.

So the problem is then - how do we get meaning into symbols? More - how do we get meaning into symbols without an outside reference - because was can build a system of symbols and rules with no outside; ie, the Universe.

But! And here, I think, might be a crux point of GEB: A Godel strange loop is a symbol manipulation that references the system it is in, perhaps in the way that the word "apple" references a fruit, or the word "English" references the very system it belongs too.

Basically, with that example, without a reality "apple" will forever remain a meaningless symbol, which "English", because it references that which it is a part of, will always have meaning.

I'm thinking this might give new meaning to "I think, therefore I am", but I'm not sure yet.

No comments:

Post a Comment