Newsgroups: comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!europa.eng.gtefsd.com!emory!metro.atlanta.com!mhv.net!news.sprintlink.net!primenet!slip249.primenet.com!gmezero
From: gmezero@primenet.com (Game Zero)
Subject: Re: Reproduction as the defining characteristic of life
Message-ID: <gmezero.3.01A8AEE3@primenet.com>
Sender: news@primenet.com (News)
Organization: Primenet (602)395-1111
Date: Thu, 8 Sep 1994 08:51:15 GMT
References: <34dklr$kh9@gap.cco.caltech.edu> <34dnth$f97@scratchy.reed.edu> <hubey.778780195@pegasus.montclair.edu> <javhar.778859919@gwr> <34kb8t$nu0@canopus.cc.umanitoba.ca>
X-Newsreader: Trumpet for Windows [Version 1.0 Rev B final beta #1]
Lines: 100

In article <34kb8t$nu0@canopus.cc.umanitoba.ca> umholme0@cc.umanitoba.ca (Douglas Alan Holmes) writes:
>From: umholme0@cc.umanitoba.ca (Douglas Alan Holmes)
>Subject: Re: Reproduction as the defining characteristic of life
>Date: 7 Sep 1994 12:21:49 GMT

>In <javhar.778859919@gwr> javhar@Bausch.nl (Jack van Rijswijck) writes:

>> [...] But whenever
>>anyone comes up with one definition, someone else is bound to 
>>disagree with it "because in that case you could call X's alive" or
>>"because in that case Y's wouldn't be alive". That shows that we're
>>actually just doing the opposite thing. We *start out with* an idea
>>of what life is and then try to fit the definition to it. [ . . . ]

>Taking just the first part of your statement. I read that 
>1) we disagree on the definition of "life"
>2) we have an idea of what life is
>3) we are trying to fit a definition to life

>So can we build up a list of things we agree are "alive" and not "alive"
>and "in between" and build up a map of the domain. By looking at the
>examples which are contentious we might determine what the correct 
>properties are.

I think that perhaps we are getting too hung up on trying to determine what
DEFINES life, and instead should be trying to determine what DRIVES life.
We have to stop thinking like a human and start thinking like a cell. A cell
has no reasoning,  per se, however it still has drives and is life. If you look at 
sex chemically, and view it as an aspect of the basic desire of all molecular 
matter to reach entropy, it seems to make more sense. 

I have written a couple ai-video game shooters that learn through "genetic"
approximation, but the problem has been that they just don't learn fast 
enough to be useful. Reproducing as an end in and of itself just wasn't
enough to make it do what I wanted.

I haven't finished the coding, yet, but I have been working on a new theory
based around the idea of DESIRING sex instead of HAVING sex. When
you determine intellegence by reproduction, all an organism can do
is REACT to its environment. To be truly intellegent, an organism
needs to INTERACT and ANTICIPATE. But in order to make
decisions based on something that hasn't happened yet, an organism
has to have something it WANTS.

The organism I am coding has a set of logical fields that form a 
"pleasure-center". The more logicals an action flags to "yes", the 
greater the relative "pleasure" of the activity, and the greater the "desire" 
to do it again.

The INSTINCTIVE patterns of the organism represent "ideal stimulation",
which can be mathematically encoded.

The LEARNED patterns of the organism represent the percent which the
action approximates the "ideal"; which can also be mathematically
determined.

If you set the basic life logicals as:

Reproduce (primary drive)
*    must survive to reproduce
*    more than one ("like") must survive to reproduce

Consume (secondary drive that serves primary drive)
*     must consume that which contains energy
*     must not consume "self" to survive 
      (important to multi-cellular organisms)
*     must not consume "like" to reproduce

(I hope I have been clear enough on this.)

The logic obviously makes sense for a uni-cellular organism,
but it still readily applies for a human being.

If you look at nations and politics and racism, and instead of viewing
them as some sort of "higher reasoning", apply the concept that they are 
nothing more than mathematical resultant of the bulk catagorization of
food (you war with an opposing nation because they are less like the 
"self" and more like "food") it puts things in a different light. 

Many aspects of the human experience easily fit into this algorithim.
Even perturbed human reasoning fits into this notion; serial killers
are frequently known to either view themselves as .not. human or
view their victims as .not. human--therefore making them less like
the "self" and the experiance more "pleasurable". Normal humans
tend to date other people with similar interests--also more like
"self" and less like "food", making it inherently more pleasurable.

I theorize that if you create a series of nested relational databases and each 
level of the database controls more a more rudimentary level below it, you
can create alife and ai without having to prebuild very much
"knowledge" at all into the  system. If you give a system too much
knowledge, then you are biasing it on the human concept of what
is the more intellegent solution and we by  no means have all the
correct answers.

At any rate, reproduction does not define life, but it is 
very essential catalyst in motivating life. (IMHO)

Lisa Carter  
aka. The Red Salamander Zaruga
