From: Ben Goertzel (firstname.lastname@example.org)
Date: Sun Nov 24 2002 - 11:18:49 MST
> > Could you give me an example of the empirical verification of
> > an ethical/moral system, and then an example of the empirical
> refutation of
> > one?
> Sure. It's nothing fancy.
> First, one has to define the goal. Here are some examples of goals:
> To get to heaven.
> To be considered a good person by oneself and others.
> To perpetuate the human species.
> To increase the quantity and quality of sentience in this Galaxy.
> To increase in complexity of the universe.
> Second, one has to define the system that is intended to reach (or
> move towards) a goal. Here are some examples of systems:
> Friendly AI.
> Third, you compare how well actions take within the system move
> toward (or reach) the goal. If actions within the system do in fact
> achieve the goal, then the system is verified. Otherwise it is
> This can be summed up with the short phrases like: Does the system
> achieve its goals? Does it work?
This much is straightforward. Because the goal is given, and one is then
asking how well the goal is achieved by a given system.
The difficulty comes in comparing various goals against one another. I.e.,
the hard part is choosing a goal!
For example is
A = "increase the quantity and quality of sentience in this Galaxy."
a better goal than
B = "increase in complexity of the universe."
.5 * A + .5*B
.3 * A + .7 * B
> I ask myself "what is the most Right goal"? or "what goal encompasses
> all the other Right goals?". Then I ask myself "what has the
> universe been doing up to now? (that we know of)". And one answer
> appeared to be 'Increasing in complexity'. This seems to be the
> answer the Universe is giving us when we ask "what are you doing?".
> This answer correlates well with many of the human moral systems that
> have grown and prospered over the centuries. Even now we, as humans,
> are attempting to push the complexity of our environment and
> ourselves to ever greater heights. This seems to be what the
> universe *does*, and as such, I suspect it is the Right thing to do.
> My definition of Rightness makes a first stab at measuring sentient
> actions against-and-within this background of what the universe is
> moving toward. Although we are the universe measuring itself, and it
> seems circular, it is not. It is just self-contained.
> Actually, Ben, I have you to thank for some of these ideas. Please
> pull them apart if you can, because I would love to get it Right ;>
Now you are positing a concrete meta-goal:
** Find out the universe's implicit goal system, and adopt a goal system as
close as possible to the universe's implicit goal system **
This is an interesting meta-goal.... And indeed, one that a superhuman
intelligence may be better able to execute than us mere humans. Since our
brains seem to be having a pretty tough time finding out the universe's
implicit goal system !!
But why is this meta-goal better than some other one? There is still,
ultimately, a basic choice being made...
One might wonder why the universe needs help in achieving its goals. By
definition, it's going to do so anyway ;-) Are we really helping it by
helping it "achieve its goals even better"??? Paradoxes of free will seem
to lurk inside your meta-goal....
Personally, I do have a goal of increasing the amount of complexity &
pattern in the universe, AND a goal of preserving human & sentient life....
And I'm not so sure about your meta-goal.
My son Zeb asked my dad once "Ted, do you like me more than the universe?"
Ted said ... "Hmmm. yeah, I'll have to choose you Zeb. When you think
about it, the universe is kind of a mixed bag..." ;-)
-- Ben G
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:41 MDT