Newsgroups: comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!fs7.ece.cmu.edu!hudson.lm.com!news.pop.psu.edu!news.cac.psu.edu!howland.reston.ans.net!gatech!bloom-beacon.mit.edu!eru.mt.luth.se!news.kth.se!sunic!news.chalmers.se!news.gu.se!gd-news!d6243
From: sa209@utb.shv.hb.se (Claes Andersson)
Subject: Re: "What is Life?"
Message-ID: <1995Feb5.124922.28250@gdunix.gd.chalmers.se>
Sender: usenet@gdunix.gd.chalmers.se (USENET News System)
Nntp-Posting-Host: d6243.shv.hb.se
Organization: Dept. of economy and computer science.
X-Newsreader: News Xpress Version 1.0 Beta #2.1
References: <3gjmlt$1ns@engnews2.Eng.Sun.COM> <3gk4cl$dpt@saturn.haverford.edu> <3gma5j$85l@scapa.cs.ualberta.ca> <3gmo28$l6j@engnews2.Eng.Sun.COM>
Date: Sun, 5 Feb 1995 19:07:57 GMT
Lines: 163

holtz@netcord.Eng.Sun.COM (Brian Holtz) wrote:
>In article <3gma5j$85l@scapa.cs.ualberta.ca>,
>Terry Van Belle <vanbelle@cs.ualberta.ca> wrote:
>
>>How about this:  "Life is the byproduct of evolution"
>
>Hmm, so there is no life present until the first mutation?  The word
>"byproduct" makes possible such chicken-and-egg quibbles...

Well, isn't it better to say: "Life is very unlikely to occur except as a result
of evolution"

>
>>"Life is the sum total of all strategies developed by replicators in
>> order to replicate more successfully.
>
>Depending on how you define "strategies", it might be hard to
>differentiate between, say, fire and viruses.  Also, "strategy" has
>strong connotations of intentionality.
>
>But I think you are on the right track.  Life is about reproduction
>(replication) and evolution (passing new traits, such as more
>successful reproductive strategies, across successive generations).

I think reproduction is a necessary way of obtaining life but still not
an absolutly required trait for life to exist. Even if it is highly unlikely,
a lifeform could just materialise. (I don't say that it will or something it's
just as thought experiment) Reproduction is required for evolution and
the only way life can come into existance is by evolution (of course, the
first replicator came around by chance but it wasn't more alive than a
virus). Therfor, my opinion is that: Reproduction is very likely to be a
trait of a lifeform.

 An analogy: A car would function perfectly even if noone have constructed
it but of course - it's very unlikely that a car would come into existance without
being manifactured. Still, we cannot say that construction is one of the definitions
of what is a car. We wouldn't dream about defining it like that. If the exceedingly
unlikely event that a car would materialize by itself occured, it would still would
work very well.

 Of course, I think it's impossible to find a lifeform that don't reproduce but
reproduction is an evolutionary mechanism, not a life mechanism. I consider
active entropy sinking to be the one and only life characteristic that could be
thought as a necessary trait of ALL possible lifeforms even if it's very likely that
all lifeforms reproduce as well.


>
>> Something is alive if it consists of these strategies."
>
>I'm not sure how an organism can contain, let alone "consist of",
>strategies.  What precisely does this mean?
>
>>One objection that came back was that under it mules aren't alive.
>
>It's helpful to separate the problems of defining life as a phenomenon
>and defining the criteria for whether an entity is alive instead of
>inanimate, unborn, or dead.  Grandfathering mules and vasectomy
>patients isn't too hard.
>
>>To run down the checklist of test cases:
>
>Hey, good idea.  I've modified my 'checklist' and appended it below.
>
>>- fire is not alive because its form is not the strategy of a
>>replicator.
>
>Why isn't it?  Is the problem that fire is procedural instead of
>substantial?  I should think there is not much of a difference to be
>found there.  Is the problem lack of strategizing on the part
>of fire?  Viruses do scarcely more strategizing than fires.
>Does the form have to contain the whole strategy?  The strategy
>enformed by a virus makes many cross-references to the biomolecular
>machinery of its host.  Couldn't we just say that the strategy of fire
>is even more efficiently enformed, with cross-references to the
>chemistry of the combustibles?
>
>>- Lt. Cmdr. Data is not alive (though I'd argue that he is sentient)
>>because his form is the byproduct of Dr. Soong's intelligence, not any
>>replicators.
>
>What if Data knew how to produce copies of himself, and what if new
>traits were passed on in the copying process?  Data and his progeny
>would surely be alive, even though bootstrapped with intelligent aid.
>Turning it around, would we all be suddenly considered non-living if
>extraterrestrials landed tomorrow and proved that they long ago
>kick-started terrestrial life with some custom-designed nucleic acids?
>
>>- replicators in an ALife experiment are alive.  In particular, the
>>genes in a genetic algorithm display a very primitive form of life.
>
>Which are alive, the genes themselves, or the gene sequences?  In most
>genetic algorithms, the space of possible change is so constricted
>that it's hard to say that any new traits can arise.
>
>>- biological viruses are alive, computer viruses aren't.  If computer
>>viruses ever became ubiquitous enough to compete for machines and
>>mutate to better competition strategies, then they too would become
>>alive.
>
>Are both mutation and competition required?  Is there no life in an
>Eden which suddenly gives rise to a single non-competitive species
>of perfect replicators?  I would answer yes, for the same reason that
>current computer viruses and Lotus 1-2-3 are not alive: they do not
>evolve.
>
>Checklist of life candidates
>(in no order I will admit to)
>-----------------------------
>tornados
>stars
>fires
>crystals
>tissue cultures
>
>sperm
>mules
>vasectomy patients
>
>a completely-sequenced wooly mammoth genome
>a completely-sequenced extinct virus's genome
>the last terrestrial virus, after earth dies
>prions
>the smallpox virus stored in Atlanta and Moscow
>viruses
>
>Lotus 1-2-3
>Conway Life gliders
>genetic algorithms
>computer viruses
>Tierrans
>Lt. Cmdr. Data
>the evolving robots from _The Blind Watchmaker_
>
>this checklist ;-)
>chain letters
>ideas
>myths
>words
>languages

 My usual objection: There is a fundamental differece between
a tornado and a tree. The tornado can be predicted to last
a certain amount of time, it doesn't have any mechanism to add
energy to its system. A tree just like any lifeform, comes in two
flavours: Alive and dead. There is an obvious difference between
a living tree and a dead tree when it comes to how long it will
last in its present state.
Compare a virus and a tornado. They have one thing in common:
They do not actively contribute to their low entropy in any way. But
the virus is still somewhat more alive, or perhaps has a lower
entropy: it is created in a way that allows it to stay in its low
entropy state for quite a time but that's nothing more really than
that stainless steel, an alloy that is higly unlikely to be found in
nature, remains as it is longer than iron. Stainless steel remains
as it is because it is designed by a low-entropy phenomenon (humans)
and virus because they are results of evolution.

 Those things you are talking about have a low entropy but they
aren't alive.


Claes Andersson. University of Bors. Sweden
