From: Ben Goertzel (ben@goertzel.org)
Date: Wed May 03 2006 - 16:11:40 MDT
On 5/3/06, Damien Broderick <thespike@satx.rr.com> wrote:
> At 04:59 PM 5/3/2006 -0400, Ben wrote:
>
> >the exemplar approach to categorization may be a useful
> >heuristic in many cases, but it definitely does lead to bad errors in
> >many cases.
>
> Not sure if you mean it inaccurately predicts human cognition,
>or it
> correctly predicts logically erroneous human cognition.
In that sentence I meant that it correctly predicts logically
erroneous human cognition.
But the other is also true: in some cases it inaccurately predicts
human cognition
> My point, and
> I think Richard's, is that such artificial and contrived challenges,
> while of course valid in their own terms, miss the complexity and
> everyday utility of extrapolating from exemplars, especially when an
> abundance of apparently salient information is available.
Agreed... we are tuned for situations so complex and messy that
precise inference is very difficult, and so we screw up in many cases
where precise inference is easy...
ben
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT