Ben, et al,

I had a prospective high-level thought-way past some of the impasses here.

I think I can see generally how an intelligent machine might function, with
some presently unbuildable boxes like a "self-organizing potential-reality
filter". I think I could diagram (probably with lots of bugs to chase out
later) how these boxes might potentially fit together, that on first glance
would look a bit like the architectural diagram of a complex CPU. I
suspect/presume that you may have already drawn such architectural diagrams
in your continuing quest to make OpenCog intelligent, as you would seem to
need SOME sort of high-level architecture in mind to ever hope to develop
such a thing. If such a diagram exists, could you please post it?

I suspect that a "competition" of architectures might bear some fruit. For
one thing, I suspect that the connections between boxes would have to be of
the "equation balancing" type that I have discussed, for such things work
together. Alternatively, perhaps you have already found prospective ways
past such apparent impediments?

At minimum, this should clarify the "grand challenges" in mathematics,
wet-lab science, etc., that must be met before an AGI can be built, short
of "simply" copying-without-understanding brains (which neither of us think
is even possible).

Steve



-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968
Powered by Listbox: http://www.listbox.com

Reply via email to