From: Kevin (maitrikaruna@yahoo.com)
Date: Mon Feb 23 2004 - 08:19:18 MST
>
> I'm not so sure it's the case that there's only ONE "non-goal system",
> though. There may be many kinds of complex, intelligent-ish dynamics
whose
> actions aren't conveniently modeled using the language of "goals"
This discussion of goals could become thorny rather quickly. For instance,
there is the perpective of the observer and the observed. You might observe
something that is acting in a certain way that you infer to be "goal
oriented", yet the observed "thing" may not even conceptually hold the idea
of "goal" or "attainment". This is a kind of uber-enlightened activity by
the observed entity.
I guess the key point I was making is that all this discussion we are having
is based on humans very limited conceptual frameworks. Even to very
advanced humans, the concepts we are discussing such as "growth",
"ascension", "friendliness", "ethics" are not held as being existent in any
way. If an AGI does achieve the levels of intelligence that we want to
ascribe to it, then we should prepare ourselves for the strong possibility
that it will come to the same realization that certain humans have about the
non-dual nature of the universe. In which case everyone here may be rather
disappointed :) But that outcome warrants a *very* long discussion not
appropriate for this board.
all that being said, I know that during this nascent stage of development
that one has to work with goal-systems, ethics etc. And before an AGI is
really smart, their proper implementation will be critical IMO
Cheers,
Kevin
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:46 MDT