From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Mon Oct 01 2001 - 21:09:20 MDT
Mike Lorrey wrote:
>
> "Eliezer S. Yudkowsky" wrote:
> >
> > Mike Lorrey wrote:
> > >
> > > This conflict reflects the problems moral philosophers have with the
> > > trolley death paradox: given a choice between 5 people dying on one
> > > track, or flipping a switch that will only kill one person, versus 5
> > > people dying in the trolley accident if you refuse to shove one person
> > > in front of the train, most people will look at flipping the switch as a
> > > more moral choice than shoving someone in front of the trolley, despite
> > > the equal cost in lives lost versus lives saved.
> >
> > Excuse me, but when our moral intuitions say that two things are
> > different, they usually are. If you shove a person in front of a trolley,
> > and he dies, and then it turns out that it was a false threat... well,
> > oops. If the switch is a false threat then flipping it does no damage.
> > Uncertainty about the outcome of actions is one of the foundations of
> > ethics, one of the environmental conditions to which moral intuitions are
> > an adaptation, and quite relevant here.
>
> Uh, you don't understand the scenario. In both situations, you have a
> choice of five people definitely dying versus one person definitely
> dying. No 'uncertainty'.
Of course there's uncertainty. If you're dealing with 100% probabilities
then you have left the universe humanity grew up in, and your moral
intuitions are going to be totally out of sync. Telling people to believe
in a hypothetical scenario like that proves nothing except that the
philosophers or cognitive scientists fail to understand the problem they
are exploring. We are hardwired not to be that certain. It would take an
act of deliberate mental self-control to switch off the perception of
uncertainty - it can't just be done on a philosopher's instruction. It
would take an evolutionary psychologist to furthermore work out which of
the moral intuitions are related to the presence of uncertainty and apply
a compensating counterbias so as to work out an abstract response which
would be suited to this alternate universe in which it is possible to be
certain of things.
Let's suppose that you have the choice of either six people dying, or six
people or six people dying. Let us furthermore suppose that the laws of
arithmetic are altered so that all numbers are equal to six, but not all
sixes are equal to each other. Do you throw the switch?
-- -- -- -- --
Eliezer S. Yudkowsky http://singinst.org/
Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:11:06 MST