RE: group-based judgement

From: Smigrodzki, Rafal (SmigrodzkiR@msx.upmc.edu)
Date: Sat May 25 2002 - 12:37:09 MDT


                Eliezer S. Yudkowsky [mailto:sentience@pobox.com]

                However, there are ethical implications to altering our
treatment of people
                based on prior probabilities rather than observed behavior.
Bayes's theorem
                necessarily influences our estimate of the prior
probabilities, but does not
                necessarily control whether we choose to alter our treatment
based on those
                probabilities. Arguably the situation is viewed as an
iterated Prisoner's
                Dilemna in which we agree not to prejudge in exchange for
not being
                prejudged by others. To put it another way, being judged
only by your
                actual, personal actions and not your statistical
associations with the
                actions of others is a public good; it helps preserve the
ethical structure
                of reciprocal altruism, which requires that you adjust your
treatment of an
                agent based on those actions that are subject to the judged
agent's control.

                ### A very cogent argument. Yet, one could extent some of
the reasoning further.

                Optimal use of scarce shared resources (as in tax monies for
security purposes) is important as a public good, since it helps bring about
the benefit of all (as in maintaining our ability to cheaply travel by air,
despite terrorist threat). An optimization could put a higher burden on
some participants in our long-term prisoners' interaction. On the other
hand, insistence on equal distribution of burdens (in the interest of
maintaining a morally symmetric structure) could impair the functioning of
that structure, harming all.

                Perhaps the solution would be to allow statistically-based
group judgments if the burden on the group members is not great (as in
somewhat increased risk of being searched at the airport) but the benefit
for all of us is significant. After all, each of us is a member of many
groups which in many contexts might have to pay more for the better
functioning of the whole, and the combination of added burdens affecting any
particular person might be quite equitably spread throughout the population,
thus ensuring moral symmetry at a higher reference level. Additionally,
moral symmetry might be maintained by offering compensation commensurate
with the added burden (as in giving an amount of money to the innocent
members of targeted groups who are actually searched).

                Rafal



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:14:22 MST