From: Eliezer S. Yudkowsky (sentience@pobox.com)
Date: Sat Aug 25 2001 - 13:43:45 MDT
Chris Hibbert wrote:
>
> http://singinst.org/CFAI/info/glossary.html#gloss_sexual_selection
>
> Eliezer's gloss on sexual selection is much too negative. It implies
> that it's usually subject to hijacking, rather than usually based on
> something that is not a "cheap signal".
[...]
> I think all the examples referred to in the gloss (antlers, peacock's tail,
> intelligence) have been justified reasonably well as expensive signals that
> are good fitness markers. If they were not so, it's likely that the
> process Eli described would have been able to hijack the signals, and
> end up driving the species extinct. Doesn't the fact that these species
> continue to survive give us some reason to believe that sexual selection, while
> arguably stupid, isn't destructive in the observed cases?
[...]
> Eliezer, what
> made you decide to present this topic in such a negative light?
What a strange notion. First off, I have no particular emotion about
sexual selection, in case you were wondering. I care about the structure
of evolutionary forces because evolution is the source of human cognition,
plus a case of wry humor for the possibility that human intelligence
happened not because it was so useful and powerful but because a bunch of
apes wound up with the self-propagating delusion that intelligence was
sexy.
The reason why the possible dysgenic nature of sexual selection is
emphasized is because it's so important for evolutionary theorists to
realize that evolution does not always operate to produce greater
fitness. As an exception to the surface rule, pondering it helps us to
understand the real, deeper rules that truly underly evolution.
It doesn't matter whether sexual selection focuses on signals that are
cheap or expensive. The point is that sexual selection will give more
weighting to those signals than their environmental fitness would warrant
because of the extra weighting for reproductive fitness. If you view
evolution as a flawed version of a normative goal system with
environmental fitness at the top, then "sexual selection" is a systemic
flaw that tends to greatly increase the strength of certain subgoals for
no particular reason. To the extent that there is only so much selection
pressure to go around, then sexual selection is usually dysgenic - unless,
by pure unplanned coincidence, sexual selection drives a species out of a
local minimum and into previously unexplored but fruitful areas of the
fitness gradient. This is probably what happened with humanity and
intelligence.
-- -- -- -- --
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 - 08:10:05 MST