Ben,

> My perspective on grounding is partially summarized here
> www.goertzel.org/papers/PostEmbodiedAI_June7.htm
>"Clearly, embodiment makes the task of teaching a proto-AGI system a heck of a 
>lot easier – to such a great extent that trying to create a totally unembodied 
>AGI would be a foolish thing."

I wouldn't be so sure about that. Your embodiment approach might be
better for dealing with the implicit knowledge problem than trying to
specify every single fact using the predicate logic or by a
conversation-based teaching. But there is another way, and that's
teaching through submitted stories [initially written in a formal
language] = a solution I'm trying to implement when I (once a while)
get to my AGI development. Stories (and the formal language) provide
important contextual data and collections of those stories can supply
a decent amount of semantic knowledge useful for generating (&
clarifying) the implicit knowledge / grounding particular concepts.
Writing stories is [I guess] generally easier than setting up
scenarios-to-learn-from in simulated 3D world. And all the data
processing and attention allocation etc you need in order to handle
the 3D world is IMO an unnecessary overkill. But, I have to admit, my
AGI is not well functional yet (+ I definitely have AGI stuff to
learn), so - just sharing my current opinion. ;-)

Regards,
Jiri Jelinek


-------------------------------------------
agi
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=8660244&id_secret=108809214-a0d121
Powered by Listbox: http://www.listbox.com

Reply via email to