Plans can be made by using components of classes (or categories) of types of events that could happen. So I can make some predictions about unusual events and plan for them even though I can't really predict every significant aspect for an event before hand. I had realized that an AGI program had to use generalizations of kinds of situations (and the like) and I realized that an AGI program has to reason using collections of possible events but for some reason I had never synthesized these two plans for my project in any formal way before. Jim Bromer
On Sat, Jan 31, 2015 at 10:34 AM, Jim Bromer <[email protected]> wrote: > Of course the future actions of an adaptive object are unpredictable > so the mapping of the future is essentially non-deterministic except > possibly for the near future. The set of every possible string of > actions is at least exponentially complex. Since an AGI object cannot > possibly know everything and since it does not have adequate memory to > store everything about a complicated world space that means that any > computational mapping of future actions in that world is not only > non-deterministic but really indeterministic as well. ------------------------------------------- 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
