From: Ben Goertzel (ben@goertzel.org)
Date: Thu Mar 24 2005 - 11:58:47 MST
> People have only scratched the surface of the "selfish gene"
> idea. Looking only at the first level of implications. Does a "selfish
> gene" work only to copy those which are exact base pair for base
> pair copies of itself? Or does it work to copy those similar to itself?
What you're talking about here is not a "selfish gene" but rather a "selfish
base pair."
Based on what I understand about genetics (which is quite a lot, in fact), I
strongly suspect this is merely a minor correction to the selfish-gene
paradigm...
> For example the four color genes of the human eye (red,green,blue,night
> vision green), they are all very similar. A "selfish gene" might do well
> to support the copying of many variations of itself, as diversity of forms
> will tend to insure its survival, over a single form. Diversity of genetic
> variations is so important that we have sex, instead of cloning for all
> higher life forms, and many lower.
Yah, good point, but there's no way it's going to explain sacrificing
yourself for 10 distant genetic relatives, sorry...
> It comes out of Darwinism is you look at it closely. Survival of the
> fittest simply has to look beyond the individual organism, and the
> individual variant of a gene. Its a very very complex equation.
I don't believe you. Do some back of the envelope calculations, and you'll
see this idea is VERY far from working....
> > Altruism (in the sense I'm using it here) is a psychological
> attractor, and
> > the quasi-altruism that the selfish genome promotes has pushed
> some human
> > brains toward that attractor. Guiding AGI's into this psychological
> > attractor will be an important topic in AGI psychology...
>
> The psychological attractor is there because evolution built it in.
No -- it is there because it self-organized from other things that were
sculpted by evolution...
Not every trait in an evolved organism is there because it, specifically,
was adaptive (for the organism or the gene or any other level in the
selection hierarchy). This is one of the basic evolutionary-theory errors
that Gould was always (correctly) railing against.
> Altruism is selected for ultimately, and its not always 'nice', it also
> means altruistically fighting off parasites, for example people in games
> will tend to punish cheaters even if its not to there direct benefit to do
> so.
I'm aware of the evolutionary-theory explanations of altruism, and although
powerful it's apparent to me that they're incomplete.
I don't believe your hand-wavy workaround to my point at all ;-)
-- Ben G
This archive was generated by hypermail 2.1.5 : Tue Feb 21 2006 - 04:22:55 MST