|
Authored by: Anonymous on Tuesday, April 30 2013 @ 08:10 PM EDT |
Welcome to the Great Levelling Machine?
I thought RK is attempting to understand, and thus reconstruct,
intelligence through its dependence on language as a means of
both storage and communication. My concern is not about the
accuracy of the information in Wikipedia. It is that Wikipedia
contains a non-negligible amount of language which is not
constructed according to accepted norms due to its editing
by non-fluent speakers. I observe that this is not limited to
the English version.
So RK has a problem, set an artificial norm and reject non-conforming
samples; or accept all and build an understanding of language
that is less than optimal. To overcome the hypothesis that some
human intelligence will always be superior to machines, the
latter action will require the human to teach the machine to
recognise sub-standard language. Perhaps this is the secret
he was unwilling to reveal in the bool.
[ Reply to This | Parent | # ]
|
|
|
|
|