From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Thu Sep 01 2005 - 16:12:08 MDT
Peter de Blanc wrote:
> On Thu, 2005-09-01 at 12:05 -0600, brannen wrote:
>
>>It has been decades, literally, since I read Jaynes but, IIRC, the
>>argument starts with the inferential fallacy 'absence of evidence
>>being evidence of absence' and goes downhill from there.
>
> Just a quick note: for a Bayesian, absence of evidence *is* evidence of
> absence.
I wondered if anyone would catch that. But if one is going to make that
point, one really should do three other things:
1) Phrase it mathematically: say that P(A|~B) < P(A) iff P(A|B) > P(A).
2) Post a link to an explanation, for the benefit of those bewildered
by a seeming violation of logic.
E.g: http://yudkowsky.net/bayes/bayes.html
3) Note the qualitative fact that absence of evidence may be *extremely
weak* evidence of absence.
PS: Let's call the author "Julian Jaynes", so as not to confuse with
the late Bayesian Master E.T. Jaynes.
-- Eliezer S. Yudkowsky http://intelligence.org/ Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:52 MDT