From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Mon Oct 15 2001 - 07:03:03 MDT
Samantha Atkins wrote:
>
> "Eliezer S. Yudkowsky" wrote:
> >
> > Eugene Leitl wrote:
> > >
> > > On Sat, 13 Oct 2001, Alex F. Bokov wrote:
> > >
> > > > If you were a terrorist, and you had several kilos of stolen Russian
> > > > anthrax those mean FBI guys wouldn't let you load onto a crop duster,
> > >
> > > We have zero evidence that they have Biopreparat stuff stockpiled.
> >
> > Are we discussing the latest in string theory? Are we preparing a grant
> > proposal for peer review? Hypotheses that predict thousands of dead
> > people do not require experimental confirmation, just a significant
> > Bayesian prior.
>
> When you are talking theories that result in seeming
> justification for bombing the hell out of an unspecified number
> of countries and killing an unspecified number of people and
> costing the lives of unspecified thousands of our own military
> and shutting down a lot of our freedoms you damn well better
> have something a hell of a lot better than a "Bayseian prior".
I'm not talking about that; I'm talking about, say, buying a gas mask. Or
moving out of urban areas. Obviously you need a much higher level of
proof to take actions that impact others than you do to make your own
choices.
-- -- -- -- --
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 - 08:11:23 MST