Are Meteors Selective?

Eugene.Leitl@lrz.uni-muenchen.de Eugene.Leitl@lrz.uni-muenchen.de
Thu, 24 May 2001 22:24:05 +0200


Lisa Dusseault wrote:
 
> That's really misunderstanding selection.  Evolution has nothing to do with
> this dirtball.  Selection works as long as entities selectively die or are

Oh yes it does, since it's one big compartment, coupled in with energy
and matter fluxes (and, come to think of it, not all that big a 
compartment at all) as contained within one gravitation well. <Insert a 180 
kByte treatise here>. Don't put all your eggs in one basket. The 
more eggs spread overe the more baskets, the less probability of 
all of them getting broken.

Right now we're running candidates for self-termination in this particular
basket. The time window so far is not severe, but I think it's there.
As soon as you're out, you've got several weakly, and high-latency 
at that coupled baskets.
If you'll model this, you'll notice that most eggs wind up whole, even
if you're only moderately smart about this, and even if your middle name 
happens to be Mallory.

> selectively born, whether or not they're on this planet, another planet, in
> a spaceship, or floating freely in space.  For a cheesy sci-fi example, take
> Cat on Red Dwarf -- a talking sentient being evolved on a ship from the
> original crew's pets.

Please understand that even mid-range impacts of current technologies 
(not to mention forthcoming instances of lunatic fringe technologies) are 
extremely counterintuitive. As much as I hate the spooks you might 
find you would want to consult some on some of the apparently
harmless experiments you're doing. The jackpots are sure rare, but some
of the jackpots might break one or several of the baskets (here's
another self-termination mechanism which is acting on a larger scale). 
It might or might not involve lots of special effects, but the results
might be rather dismal in any case. 
 
You can't make people not to play, so you have to start some serious
lookout for what is going on, because the geeks in question are all
too often happily oblivious of the potential impacts of the particular
experimental setups in question. No guilt assignment, but it doesn't
address the problem. No need to preventably stifle productivity by 
pointless (already massively bloated) administratory overhead, but 
it does need to get addressed. Oh yeah, fuck Billie Joy.

I think this requires some openness in experimental setup review
on both the community and the more deviously minded folks around.
As a corrollary, this means to reduce the pressure on grant 
allocation (yes, I do have an alibi), because it actively 
encourages information hiding. Experimental groups attempt to 
time-delay result release, or withheld part of information
to maintain their lead orelse do not persist too long on the 
grant-grubbing market. It's not industry, but they do feel the
heat.

> Even more than eternal life is required to stop the effects of evolution
> (among the species that achieved eternal life).  Also, you need some

You can't stop the evolution. The eternal life is just a theoretical
capability. It helps if you're a member of the pioneer shock
front, as sooner or later the local folks
do get nuked, either because of they've pulled a dud
on the local fitness lottery (sooner or later things turn out
Red Queen degenerate, with random spatiotemporal fitness
soup simmering), or because the particular pocket
of spacetime (think a very large basket) sees a rare but 
rather energetic event which totally wipes the local
biota, however hardass. It sure gets reseeded fast, but that's
no solace to parties nuked.

> deterministic, 100% sure way of starting new life.  To tie back to the awful

I think partly ditching Darwin in favour of Lamarck does need
looking into. Of course, you can never kick out the Darwin dude
sustainably, in worst case you wind up with a local metastable
pocket which runs danger of being overrun by a wave of folks
who chose to stay in the darwinian regime.

I think diversity should take care of some of that.

> abortion argument, take those 40% (quoting turpin I believe) zygotes that
> are unviable enough (or some other selection function) to be spontaneously
> aborted.  That's selection too.

I do really think wasting attention (a scarce resource) on wet context 
alone is dangerous. Your attention is limited, and wet context does 
increasingly seem out to become a candidate for an unsupported platform, 
to be phased out within nongeological time scale (if you're unlucky, 
you're going to see it in person).

Of course, you can still do run some funky exploits in the wet context,
so you do still have to allocate some of resources to address the issues.