On Wed, Aug 6, 2008 at 8:25 PM, Brad Paulsen <[EMAIL PROTECTED]> wrote:
> Jiri,
> I'd really like to hear more about your approach.  Sounds bang-on!  Have you
> written a paper (or worked from papers written by others) to which you could
> point us?

Brad,

I'm not aware of anyone else using or describing this approach (the
story-based learning using FL). And when I discussed it with some AGI
developers/specialists, it was typically kind of discouraging. They
were like "Nah, theoretically ok, but practically dead end!". But,
"AGI specialist", what do they know.. ;-)) I figured it's well worth a
shot + it's fun to play with. Unfortunately I just kept very busy with
non-AI priorities in last 2 years. No, I didn't write a white paper on
the topic (yet). As a hardcore developer, I'm much better in writing
code & cryptic notes on napkins than in writing official docs and
white papers (though, once a while, I'm forced to do it all). Some
brief/basic high-level info about that project (one of many I work on)
is here:
http://www.busycoder.com/index.php?link=28&lng=1
But again, I don't want to be too loud about it because of the limited
progress so far. Anyway, with the formal language (/wizard-based
communication format), I'm trying to provide a semantic baseline.
Basically a set of the key semantic building blocks (with a built-in
understanding) which could be used to generate rich stories for the
AGI to understand and learn from. Our brains also have various
semantics built-in. There are simply certain concepts for which the
AGI's attempt to somehow learn it "on its owns" would [IMO] be a
complete waste of resources + it would require senses similar to those
we have.

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