Re: Sentience [Was FAI: Collective Volition]

From: Randall Randall (randall@randallsquared.com)
Date: Fri Jun 04 2004 - 12:35:12 MDT


On Jun 4, 2004, at 12:37 PM, John K Clark wrote:

> On Thu, 3 Jun 2004 "Randall Randall" <randall@randallsquared.com> said:
>
>> Using the term "optimization process"
>> doesn't predispose one's reader to imagine himself as the
>> intelligence, or otherwise anthropomorphize it.
>
> What’s wrong with that? Anthropomorphizing is a useful tool, in fact
> it’s just about the only tool we have to figure out what another
> intelligent entity is likely to do next.

It's just about the only tool we have to figure
out what another *person* is likely to do next.

While I don't have any grounded statements to
make about my own beliefs regarding building an
intelligence, it does seem that Eliezer intends
to build a very intelligent goal-oriented system
which is explicitly *not* a person.

To head off the coming storm in your reply:
Mechanical calculators do something that only
people could do before the calculators were
invented. It turned out, though, that doing
(some forms of) math could be separated from
being a person, so that one could have a
calculator which wasn't also a person. Someone
arguing against this idea, pre-calculator, might
very well point out that if a machine could
answer math questions, like a person, then it
must have mental states that correspond to the
intermediate steps of the calculation, and
having mental states presupposes personality.
This seems analogous to your argument against
intelligence sans personality.

Before you once again argue that sentience
must either be unimportant or an automatic
feature of intelligence, you might want to
consider the possibility that general problem
solving, like (some forms of) math, can be
usefully separated from personality.

--
Randall Randall <randall@randallsquared.com>
'I say we put up a huge sign next to the Sun that says
"You must be at least this big (insert huge red line) to ride this 
ride".' -- tghdrdeath@hotmail.com


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:47 MDT