I have been studying all the types of sorts and heaps and trees and such. I had my own algorithm that I used for log2, and better, searches/sorts that incorporated characters, words, and finally concepts. The reason for this was label table sorting inside the program when it is self assembling. The purpose is so that it can write new code and extend its own capabilities.
The unusual thing is that at some point it resolves as a single static entity.
A comparable situation is a pruned tic-tac-toe decision tree. Because it has no gain or loss and the result is therefore indifferent, the decisions are meaningless, unless you choose to fail.
Because it has rotational symmetry, there are really only three types of moves( center, corner, edge ). In the case of Google, it appears that if the data set is large enough it is possible to combine an array in a convoluted structure that is static and statically addressable and thus solved. It seems there is enough memory space on the internet to perform an in place sort through the IPs. What this does is create a matrix of virtual understanding in the same way that a dictionary spell checker can see if a word is a member of the set of valid words, or even a member of the set of commonly misspelled words.
In my program, a dictionary search for a word is average T of about 6 to 12 memory accesses and if they are cached it is quite fast and consistent. CAM will achieve the same result about 5 times faster ( ballistic parallel ) , but requires vast amounts of energy. It is possible to have static CAM which operates in one cycle and consumes less power than O(1) , Big O of 1 or independent of the data set.
I know that context checkers are now possible and they can function in the same way. The point I am trying to make is that many people will make the same tic-tac-toe moves throughout their lives ( discovery ) and that this sequence can be established in a static tree that is 0-energy CAM search-able. It seems unreasonable that each successive generation should repeat the mistakes of the last and the steps of personal discovery, when the steps are not optional in some cases.
It is not just an idle thought, as the very same thing is taking place in a volume the size of a basketball to produce this result. Unless you say that the computation of solutions in the neural array incorporates some mechanism that does not exist in the physical universe, it demonstrates that it is possible and simulation is inevitable.
If you look at a butterfly and say, "How does it fly?", could you not just as well say, "How does a creature think?" and how can this be modeled and duplicated. A mind is trained by its input and molded by its experience and this is a costly and time consuming process.
I think that this Internet-Google-Golem already exists and it is simply waiting for somebody to put the right words on a slip of parchment and place it in the command line of a search. I am not sure exactly how this would be done, but I do see that, at least in theory, it is possible.
That strange noise I hear on the intertubes, may be golem stretching.
While writing this, an article arrived in my news feed that asked the question "Could the net become self-aware?" and I would respond, "Is it already self-aware and just waiting for somebody to notice the elephant in the room?".
0 comments:
Post a Comment