Re: Evolved Preferences

Robin Hanson (hanson@hss.caltech.edu)
Tue, 29 Apr 1997 09:52:21 -0700 (PDT)


Steve Witham writes:
>>As we become better at exchanging information in ways other than via
>>sexual reproduction, it seems the longer time-horizon of asexual
>>reproduction should win out.
>
>If we skip details about *how* genes (that is, the information that
>is used to produce an offspring) get from here to there, the distinction
>between asexual and sexual reproduction boils down to *where* the
>information comes from. Does it all descend from the information that
>originally went into the parent, with no other input, *or does it come
>from more than one source*? If the latter, that's sex in my book.
>Robin, you are clearly using a different definition of "sexual
>reproduction" since you are talking about different ways of
>exchanging information (the essence of sex as I see it)--what is your
>distinction and why is it interesting?

Yes, your definition of "sexual" is vastly more liberal than mine.
My use of the term comes from precise mathematical models of evolution
of genes, while your use seems more metaphorical. The interest in
these precise models comes, as usual, from the fact that one can work
out concrete conclusions from them.

Let me suggest that it is useful to distinguish the evolution of
values from the evolution of "facts", or beliefs about facts. The
space of possible basic values is much smaller than the space of
possible fact beliefs; so I think it is plausible to think in terms of
an evolutionary equilibrium of values even when fact discovery is
still going strong. I started this discussion citing papers on basic
value evolution, and my conclusions are mainly regarding that: asexual
reproduction of values has strong advantages when investment matters.

Reproduction/dissemination/etc. of facts/beliefs is another matter.
Here I agree that organisms will not want to limit themselves to
beliefs inherited from a single parent, and while inheriting beliefs
from two parents is better, humans and their descendants can do much
better than this. Here I find it more valuable to look at genetic
selection of general policies for evaluating and sharing beliefs.

>>1) Until recently, organisms have had very little occasion to make
>>long-term investments other than children. So the investment benefit
>>of asexual reproduction has been moot. But now humans do have to
>>choose between "consuming" their own ability to have more children
>>now, and investing toward their children's ability to have more
>>children later.
>
>I think this just means, reproduce now (with current methods) vs.
>invest, then reproduce more wisely later? The sexual/asexual question
>seems independent here, even with your definition. For instance, I
>could become rich and marry a carefully selected mate who was
>capable of bearing large litters, and fund the raising of all these
>kids.

Yes you could, but the question is whether you would. Sexual vs.
asexual models have differing implications regarding selection of such
investment policies.

Robin D. Hanson hanson@hss.caltech.edu http://hss.caltech.edu/~hanson/