An AGI program has to have some means to verify the conjectures that it
creates.  This verification method is not going to be perfect but there has
to be some means for the program to qualify the theories that it tries to
use.  A lot of guys have talked about prediction. So I challenged them to
try using prediction in their own lives to see how it worked.  I then
decided to try it myself. And the use of prediction did help me crystalize
some insights around the predictions but they tended to serve as temporal
benchmarks.  I did however, find that I was able to use those benchmarks to
start to analyze why I did or did not meet my original predictions.  I may
be just talking to myself right now, but that is not an invalidation of
what I am saying.

If you feel that if a function in an AGI program works then it must be a
validation of the effectiveness of the function then I would say that you
are seriously delusional about this.  The problem is what does "it works"
means. So a theory has to be validified through a structural integration
process.  So I will try to start integrating some sort of structural
integration model of verification into my reports on my prediction about my
AGi project.  However, I am not exactly sure how to do that yet.

So if you have some method of verification that you think would work in an
AGI program then I am saying try it against your own problem solving in
real life.

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

Reply via email to