From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Sun Aug 10 2003 - 01:10:32 MDT
Lee Corbin wrote:
>
> Here are the first few paragraphs after the abstract. If
> this doesn't acidly comment on our recent Robert-centered
> discussion, I don't know what does.
From Tetlock, P.E. (1999). Coping with trade-offs: Psychological
constraints and political implications. In S. Lupia, M. McCubbins, & S.
Popkin (eds.), Political reasoning and choice. Berkeley: University of
California Press:
"The value pluralism model also warns us not to assume that explicit
trade-off reasoning is always cognitively or morally superior to
categorical rejection of trade-offs. It is not difficult to identify
historical contexts within which contemporary sympathies overwhelmingly
favor those factions that engaged in vociferous trade-off denial:
Churchillian opponents of British appeasement of Nazi Germany in the
1930's who denounced Chamberlain’s effort to strike a subtle balance
between deterrence and reassurance (Tetlock & Tyler, 1996) and
abolitionists in the slavery debates of antebellum America who denounced
moderates who sought integratively complex compromises that would pressure
the Union, avert war and indefinitely preserve slavery in certain states
(Tetlock, Armor, & Peterson, 1994). It is also not difficult to identify
historical contexts within which contemporary observers deplore one or
both of the values that complex trade-off reasoners attempted to balance
against each other. Pragmatic Nazis were quick to recognize a gruesome
trade-off between their goals of mobilizing military resources to win the
war and devoting resources to the extermination of Jews. There is nothing
intrinsically meritorious about trade-off reasoning."
My argument was not that Robert was posing a taboo tradeoff, but that he
was posing a false tradeoff - genocide would have no beneficial
consequences, despite the attempt to frame it as a taboo tradeoff which
rationalism requires considering.
I don't think outrage should always be impermissible. I think outrage can
always be questioned. Sometimes outrage survives questioning and the
proposal of random acts of genocide is one of those cases. I acknowledge
that I could be wrong in the outrage or even the judgment of
undesirability, but the judgment of undesirability and the associated
outrage represent my best guesses. What I object to is not that Robert
analyzed the issue, but that he got the analysis wildly wrong. It is mere
rationalism to assume that any issue worth analyzing, or any issue that
provokes deep emotional reactions, must have a complex, weighty,
counterintuitive answer with many pros and cons where the conclusion
ultimately negates all the emotional reactions involved. Sometimes
rationalists arrive at complex taboo-violating analyses of what seem like
simple issues, but it is a profound error to generalize and start thinking
that whoever offers the most complex analysis and violates the most
taboos, sounds the most rational.
The undesirability of genocide is a very complex issue in terms of ethics
under uncertainty and the evolutionary psychology of opposing tribes, but
what the analysis ends up saying only affirms the verdict of history and,
yes, the 21st-century taboo: the power of genocide to improve the human
condition is far less than what is frequently imagined. Robert's proposal
is not a taboo tradeoff, it is a false tradeoff.
I believe that a hospital should not spend a million dollars to save the
life of a five-year-old child, because if you take a million dollars away
from a hospital, people are going to die. I would applaud the moral
courage of a hospital administrator who made that decision, and the
willingness to confront taboo tradeoffs and risk community outrage to make
the right choice. It doesn't mean that I'll be any less outraged if you
propose killing a five-year-old child to save a dollar, and this is
roughly what I think Robert did.
-- Eliezer S. Yudkowsky http://singinst.org/ Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Sun Aug 10 2003 - 01:20:09 MDT