I know that what I am about to propose is very hard to do but if these concepts could be done the machines will become infinitely smarter.
The ideas involve the concepts of 'is similar to' and 'is opposite to' and a few others in the same vein.
Once a program has accquire an amount of knowledge from the user, the user can then have the machine 'think' or maybe have it 'think' over night at which point it can start making its own conclusions about the data that it has accquired. After its preliminary thinking it could query the user as to the correctness of it conclusions.
I have been working on a model like this on and off for several years now, but have not had the time or all of the necessary skills to implement it. Will be glad to discuss it in more detail to any really skillful programmers in the future.
Bill819
