From: Ben Goertzel (ben@goertzel.org)
Date: Fri Dec 28 2001 - 07:46:53 MST
About "faith" and whether Eli's gotta have it or not... (I really love these
endless threads dissecting Eliezer's psychology. We haven't gotten onto his
potty training yet for some reason; I guess there's a paucity of Freudians
on the list...)
Anyway, I partly agree with Samantha here, actually.
But only partly!
Of course, as Hume observed back when I was in diapers, the problem of
induction is
unsolvable. *Purely* rational inference about the universe is not possible.
Even highly-refined probabilistic inference engines like your own brain have
to make some a priori assumption about the background probability
distribution
(to cast Hume in Bayesian terms).
Call it faith, call it intuition, or call it Caledonia Mahogany's Elbows,
there's something in addition rational calculation going on when we make
judgments, particularly in such data-poor domains as the end-state of
the cosmos!
I also agree with Eli's point that our physical models of the universe are
very unstable and not to be taken that seriously in the "big picture."
However,
I think the problem of induction show itself rather clearly in this
observation.
The time series of physical models of the universe is really not very long,
so
there's a lot of faith/intuition/assumption in extrapolating it *or* even in
estimating its variance as Eliezer is implicitly doing...
BUT -- I guess the problem with the word "faith", Samantha, is that it often
goes along
with the idea that *some things should not be rationally questioned.* If
you have
faith in life after death, you should not even consider rational arguments
against
it.
This is not the same as the kind of a-rational intuition that must underly
the
assumption of background probability distributions. The mind must make some
underlying, biased assumptions in order to reason; but it can be flexible
about what
these assumptions are. "Faith" implies a forced lack of flexibility in this
regard,
which is surely NOT optimal in terms of maximizing cognitive power (though
it may
well be optimal as a mechanism of spreading a certain assumption throughout
society).
-- Ben G
> "Eliezer S. Yudkowsky" wrote:
>
> > Perhaps. Understand that I do not have "faith" that immortality is
> > possible. I am simply stating that before we get all emotional
> about this
> > issue - that is, before we begin making value judgements or
> philosophical
> > assumptions based on it - we should remember that the model the
> prediction
> > is based on is a model which historically has often changed and
> currently
> > is still in flux.
>
> Faith may be a word that has fallen into disrepute in some
> camps. However, if I did not have "faith" that X (SI,
> singularity, immortality) was possible I wouldn't spend any time
> at all working towards actually producing such. In short, some
> questions are not effectively tackled by armchair cogitations.
> They can only be resolved, to the extent they are resolvable, by
> taking a working hypotheses and proceeding to check whether it
> is or can be made real.
>
> We know so very little about what "the universe" is and includes
> that it is extremely hubristic to claim that the universe simply
> does not allow X or that there is no way for anything that was
> part of the universe to survive the "death" of said universe or
> certain types of transformation thereof.
>
> - samantha
>
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:37 MDT