From: Ben Goertzel (ben@goertzel.org)
Date: Sun Feb 01 2004 - 17:05:54 MST
> > I don't think there can be an objective morality, but I wonder if there
> can
> > be an "objective" (in some sense) set of system-theoretic criteria for
> > assessing ethical systems (which would not pertain to the actual content
> of
> > these systems).
>
> Isn't it the same thing? Having an objective way of determining which
> subjective theories of morality are good and which are not, would
> amount to
> being able to choose one objectively good morality.
Not necessarily. Maybe it would just give you a "partial ordering" on
moralities, i.e. a tool that would allow you to -- in some cases -- declare
that one ethical system is better than another. In other cases, the tool
might tell you "no comparison is possible, the systems are too
incommensurable."
Also, the kind of criteria I'm thinking of wouldn't give a quantitative
assessment of the quality of an ethical system, so it wouldn't be able to
make fine distinctions between roughly equally high-quality ethical systems.
And two roughly equally high-quality-overall ethical systems might make
drastically different conclusions about specific situations.
> I suppose if you can
> prove that an objective morality does not exist, then you would also have
> proved that an objective way to choose good moralities also does not
> exist....
That really depends on the details of how you define "objective morality"
> It comes down to exactly this. I have no clue how to build a
> quale detector,
> since we still don't know how qualia are generated. So _if_ this
> conversation comes down to "I agree in principle, but what to do in
> practice?"
I don't know if I agree in principle, because I don't know enough about
qualia.
I am not even sure that "degrees of positivity" of qualia are the sort of
thing that can be summed up, and the sum of one set of positive qualia
compared to the sum of another set of positive qualia.....
This seems a little dubious to me. Summing of degrees does not exist on the
level of qualia (Peircean First); it exists on the level of mathematical
relationships (Peircean Third). Mapping qualia into summable qualities
seems to require some substantial theoretical assumptions ... so the quality
of an ethical system based on qualia-summation seems to rely on the quality
of the scientific research programme containing these theoretical
assumptions... no?
> My suggestion has been from the
> beginning to have the AI have "figure out qualia" as an intermediate goal
> before it starts doing big changes in the universe which will affect many
> sentients...
Yes, there's a germ of an interesting idea there.. ;-)
I guess that a Universal Mind Simulator AI would have a lot of useful data
for "figuring out qualia", if it was trying to simulate the experiences of
every mind in the universe...
ben g
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:45 MDT