Re: META: One-week cool-off period

From: Keith Henson (hkhenson@rogers.com)
Date: Sat Jun 12 2004 - 23:04:11 MDT


At 04:58 AM 05/06/04 -0400, Eliezer wrote:
>Things have gotten way out of hand

snip

I would like to level shift this discussion, hopefully in the direction of
insight.

Unless there are some AIs on this list--we are all social primates whose
genes and resultant behavioral mechanisms were shaped by our ancestors
being reproductively successful during the millions of years when they
lived as hunter-gatherers. This is basic evolutionary psychology.

Because social status had such drastic effects on reproductive success
during the past, it is a major (if not the prime) driver of human motivation.

(Does anyone doubt that on average high social status men had more
wives/children? That the children of high social status women were more
likely to survive?)

Further, evolution may have favored those who are (at the conscious level)
blind or actively blind to their motivations. Why is not obvious. Perhaps
it was easier to enlist others in your tribe into cooperative activities if
you were not consciously aware of your deep motivations. Or perhaps such
awareness does not have a reproductive payoff for other reasons.

A consequence (of our blindness to our motivations) is a deep level bias
against discussing the evolutionary origin of our motivations--especially
if you apply it to yourself. I worked out this application of EP over a
period of years mostly on the alt.religion.scientology news group and took
some ugly flame for applying it to myself.

"I know you won't attempt to justify what you did, because you
acknowledge the only reason you did it was to increase your social
standing. I also know you'll continue to attempt putting a spin on
what I've done, because it's a threat to the social status you believe
you've gained.

"Evolutionary psychology is nothing more than a crock of shit, but
it certainly does go a long way in explaining why you do what
you do, Keith. Thanks!"

http://groups.google.ca/groups?selm=3b5b34d1.34198106%40news1.bway.net&output=gplain
and
http://groups.google.ca/groups?selm=3b4ff9f9.5884992%40news1.bway.net&output=gplain

I will be amusing to see how much flame I get here. I hope you can do
better than the self defeating argument of the above flamer.

Humans have higher order thinking skills, but they are far more used in
service of deep motivations (rationalizations) then they are for rational
thinking. Thus you have to take all arguments by social
primates--including this one--with a grain of salt.

Understanding our motivations and those of others at the EP level applies:

To the social context of people who are trying to understand the problem(s)
of creating an AI with social features we want;

To understanding what it is we want or should want and;

To the AI itself.

If there is interest I can go into further discussion on the first
point. Cooperation and consensus seeking on a list might not be as hard as
FAI, but it sure has not been solved.

Keith Henson

(Who is, of course, posting to raise his social status. :-) )



This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:47 MDT