From: Mikko Särelä (email@example.com)
Date: Sat Jan 29 2005 - 08:41:23 MST
On Fri, 28 Jan 2005, Ben Goertzel wrote:
> > So I was continuing to try to come up with ideas to make objective
> > morality work. And eventually it seemed to me that the only possible
> > way it can work is to directly tie the laws of morality to the laws of
> > physics in order to define an 'ultimate goal' (telos) that is truly
> > universal (holds at all times and places). The goal 'move towards the
> > Omega Point in the optimal way' is the only possible one I know that
> > could fit the bill.
> True, but of course this argument could easily wind up being a lengthy
> and artistic "reductio ad absurdum" of the concept of objective morality
Of course it could. But if I understood Marc correctly, he does not try to
argue that FAI and going toward omega point is equivalent. He is saying
that he has a hunch that they are equivalent and that he is trying to see,
what it leads to.
Anyways, I think that we should divide the problem into certain categories
to make it more understandable.
First we can make proposals as potential FAI designs. These may be in the
form of required component, or something that must not be in the FAI.
Then we can create methods for analyzing where each of these designs will
And finally we can create better methods for analyzing both parts of the
What is important at this stage, is not necessarily to create the one true
FAI design (IMO). The more important thing is to develop the necessary
tools to do all the required analysis.
I'll just end with a note that unfortunately I haven't had the time to
read Eliezer's current work (since my own PhD research eats my time pretty
well and is not quite close enough to the subject at hand). This means I
do not know if he's already working on the problem from this angle.
This problem of FAI is such that it does require - it deserves - such
attention that proper methods and tools are created for scientific enquiry
-- Mikko Särelä "I too don't really find Monty Python all that exciting, but don't tell anyone I said that." Anonymous
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:50 MDT