From: Eliezer Yudkowsky (sentience@pobox.com)
Date: Wed Jun 02 2004 - 02:08:50 MDT
Damien Broderick wrote:
> At 04:31 PM 6/1/2004 -0400, Eliezer wrote interestingly, ending with:
>
>> Regarding optimization processes as compressing outcomes into narrow
>> regions of configuration space gets rid of a lot of the dreck about
>> "intentionality" and all the wonderful properties that people keep
>> wanting to attribute to "intelligence". Natural selection produces
>> information in DNA by compressing the sequences into regions of
>> DNA-space with high fitness.
>
> The dangers of this approach are evident when we ask whether there are
> such entities as `regions of DNA-space with high fitness'. Phenotypes
> have empirically-tested fitness within given eliciting and sustaining
> environments that include other phenotypes similar to and different from
> themselves. Genes don't (except, say, in the countervailing example of
> homeoboxen, which are spectacularly promiscuous and don't care *what*
> phenotypes they help build). See, e.g., critiques by Gabriel Dover and
> Barbara Herrnstein-Smith among many others. Configuration spaces are
> human constructs produced, amusingly enough, by such dreck as
> intentionality. (Just a note of caution.)
Yes, I'm well aware of the dangers of 'fitness' rather than 'correlations
between allele substitution rates and alleles, given statistical properties
of an ecological and genetic environment'. But it's still where the
information comes from.
-- Eliezer S. Yudkowsky http://intelligence.org/ Research Fellow, Singularity Institute for Artificial Intelligence
This archive was generated by hypermail 2.1.5 : Tue Feb 21 2006 - 04:22:38 MST