Shane Legg wrote:
On 5/27/07, *Richard Loosemore* <[EMAIL PROTECTED]
<mailto:[EMAIL PROTECTED]>> wrote:
What possible reason do we have for assuming that the "badness" of
killing a creature is a linear, or even a monotonic, function of the
intelligence/complexity/consciousness of that creature?
You produced two data points on the graph, and two inequalities:
B(Human) > B(Mouse)
I/C/C(Human) > I/C/C(Mouse)
I thought two was enough to illustrate the argument, but if
you really need more... let's do 6:
Human, Chimpanzee, Goat, Mouse, Ant, Bacteria.
I presume most people would rate the things in this list as
going from most complex/intelligent/conscious to least?
I also guess that most people would rate the killing of these
living things as going from most bad to least bad?
Obviously individual cases will vary, for example killing a
person who is about to blow up a whole city of people is
not going to be so bad. But, such external factors aside,
there seems to be a general principle.
If one accepts that there is, then the question becomes:
Where should we put a super human intelligent machine
on the list? If it's not at the top, then where is it and why?
I don't claim to have answers to any of these questions,
I'm just wondering what other people's thoughts are.
Extrapolating from any number of datapoints is unreliable when all you
have is a rank ordering.
Your rank ordering is extremely culture-dependent: are you asking
someone who lives in India, for example?
And one final lesson from experimental design: have you ascertained
that the relevant factor in the suggested rank ordering, above, is
degree of complex/intelligent/conscious, rather than, say, the more
likely empathic response (which has more to do with cuteness than
anything else)? For example, try:
Human, Chimpanzee, Jackal, Gerbil, Ant, Bacteria.
I suggest that most people would rate Gerbil more worthy than Jackal.
More generally, I think the question is silly: I have only responded
from a narrowly technical, data-fitting perspective.
Richard Loosemore
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=4007604&user_secret=7d7fb4d8
