On Thu, Feb 20, 2014 at 6:46 PM, Ben Goertzel <[email protected]> wrote:

> Do you mean: Given two numbers x and and y drawn from a specific sample S
> of numbers (or a specific probability distribution D over the set of
> numbers)?
> Without this background S or D, the question is meaningless...
> Given a distribution D, one can draw a sample S, of course; so the case
> where one has a sample S is sufficient to deal with
> One sensible measure would be: What ratio of the numbers in the sample S
> are greater than max(x,y) or less than min(x,y) , as opposed to lying
> between x and y?
>


But even this definition is (relatively) trivial relative to  the problem
of AGI.

Ben Goertzel wrote:

> An advantage of ranking based approaches like these, is that they are
> robust with respect to the wide variety of different probability
> distributions one encounters in the real world...
>


You don't "encounter" different probability distributions in the "real
world", you derive them from "observations" of the real world. Just as the
question would be meaningless unless he was talking about a specific sample
or a specific probability distribution, the question would be trivial (and
meaningless) in AGI without a much greater development of the potential
relations of the background knowledge of those observed distributions. What
(I think) we are really interested in is how can different distributions be
combined and how can subtle features be discovered in distributions (even
when those features may not correspond easily to a general numerical
process or a general normalization principle).  I can't express that
clearly enough for everyone to appreciate but a relevant conclusion is to
wonder whether "meaning" can be derived from a general principle of
numerical similarity? If it can't, then the question is whether the
question is significantly relevant to the problem of solving
the contemporary challenges of AGI.

If you had answered the question using examples of numerical relations
(with simple but powerful examples) and everyone understood those examples
then I'd have to conclude that the underlying principle of numerical
relations must have a great deal of relevance to the problem of
discovering, representing and conveying meaning. But instead, when you
realized that you had a chance to teach something that could be useful to a
lot of people, you started out using words. There is something in using
words that allows us to combine and distill meaning in such ways that can
make them useful to convey ideas. Of course mathematics can be used to
convey ideas, but words are the most general way that we can convey the
greatest range of ideas.



-------------------------------------------
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