One problem with a well defined "AI Programming Language" is that these attempts tend to falter on the simplest of problems. For instance, if you define a negation but do not define the oppositions (or alternatives) for the negation then the program will not be able to use the negation at all well. The real problem is that the possibilities are so great that it is impossible for an "AI Programmer" to define the entire background of possibilities that might be called to analyze a problem (like the one to find if a negation should invoke some kind of consideration of an opposing idea) since the whole idea is to design the program to react to novel situations. I feel that this sort of problem might be reduced with a reference to some tertiary associations to an idea-object that the program could consider when dealing with an unexpected use of the idea (like learning that it should be negated.) With the AGi database annotation programming language this situation could be studied in a controlled situation (if the annotation programming language proves feasible).
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
