[extropy-chat] What Human Minds Will Eventually Do

[Eugen Leitl]

On Sat, Jul 01, 2006 at 11:39:14AM -0700, Lee Corbin wrote:

> > It is almost always assumed that cruising aliens are super-advanced,
> > and super-intelligent. But just assuming a) relativistic flight
> > b) iterated selection over large distances it's pretty obvious
> > intelligence is not a trait selected for.
> 
> But what if---as seems probable to me---humans come to dominate
> every last cubic centimeter of the solar system? Now many will

It seems overwhelmingly improbable to me. Humans don't like
microgravity. Humans don't like radiation. Humans don't like
acceleration much beyond 1 g (and try accelerating a kT
of stuff at 1 g, versus a kg at 30 g for a month). Humans require 
lots of habitat and life support. Humans take 100 W to run, and 
need some ~100 kg of matter for what a few grams and <<10 W would do. 
Let's face it, canned monkeys don't travel at all well in space. 
*Really* short shelf half life, they have.

Even here selection effects already start. You don't see humans
leaving the solar system. Machines did and do. Dumb machines travel
lighter than super-smart machines. So when machines accelerate
at 30 g to cruise speeds of almost c, where does that leave humans
after a mere MLYr, or so? Data patterns take 30 ms to replicate,
humans 30 years.

> disagree (they being too sentimental in my opinion), but why
> preserver trillions of ants, for God's sake, when surely the
> emulation of a few hundred should satisfy anyone's conscience?

Huh? Which ants? 
 
> > The only traits selected for are short reproduction time and
> > expansiveness.
> 
> That is simply untrue. In R/K theory (biology), circumstances 
> and environments arise in which K predominates.

Biology doesn't deal with selection over very long distances.
Your closest analogy is organism sequence in volcanic island
colonization. Better, think of a 1 LYr worth of sterile rich medium, 
inoculated at one central point with vanilla E. coli. Which 
kind of critter will arrive after a (long) while on the other 
end? Sure as hell it won't be an E.coli, nor anything even vaguely
resembling E. coli. 
 
> > So while intelligence is necessary to produce expansive aliens, the
> > intelligence trait will be selected away over very large distances.
> 
> Our Von Neumann probes should overtake any escaping silly software.

Um, I am talking about selection effects in von Neumann probes
over very large distances. Except, there's no point calling them
von Neumann probes, or space chickens. They're postbiological life,
hopping from star to star, and from one galaxy to another, from
one cluster to another, from one supercluster to another.
They're our gift to the universe which keeps on giving.

> (And for the moralists, let me add: if one country's don't then
> another country's---or another planet's---will.)

How do you recall a chain letter? 

How does a 0.1 c probe catch a 0.9 c probe? 
How does a 0.9 c probe catch a 0.99 c probe? 
How does a 20-year replication probe catch a 2-month replication probe? 
How does a 5 LYr-hopping probe catch a 1 MLYr hopping probe?

Notice that none of it is about smarts. If this
spacetime doesn't allow you to punch traversible
portals intelligence is just baggage on the long
run. Just like in chromatography, the initial wavefront
has zero diversity. It's pure solvent, all the cargo
is trailing behind. The more complex, the slower.
There are successor waves trailing the pioneer wavefront.
Eventually, after many waves passing, you've got steady state.
Except, it's roiling, at a very high level of fitness. There
you've got viruses, and mice, and men, and deities. Maximum
diversity. Not pioneers. Pioneers are specialists, and only 
exist in wavefronts across pristine acres of congealed
star drek. If both such waves collide pioneers get
wiped out, because their niche is gone. There's much
less determinism in the following waves, and the omega
state does not conserve the information about the origin,
so it's degenerate. Regardless of the point of the
origin the omega state behind the trailing waves
looks all the same.
 
> > Er, not so altruistic towards the rest of them 
> > http://www.well.com/~davidu/sixthextinction.html
> 
> As I say, to paraphrase Ronald Reagan, once you've seen 10^5
> ants, you've seen 'em all. *Altruism* should be weighted by
> the complexity of who is on the receiving end. Is it altruistic

I am God, destroyer of stellar systems. Once you've seen 10^5
human primates, you've seen 'em all. Altruism is only relevant
if you rub shoulders with other powers. If you can prove me
wrong, I'll open a case of champagne, and get wasted in 
celebrating.

> for me to go through some trouble to kill a few houseflies so
> that my guest may be more comfortable?  Of course it is!

Exactly. Why should we concern outselves with nasty biofilm contaminating
the surface of our food? Yecch. Brzzt! See, the planet is all clean 
and ready to eat now. No more biofilm, all blasted into thin
plasma.
 
> > I think intelligence is very interesting, but I'm not fooling
> > myself for a moment that bulk of postbiomass will be 
> > intelligent. You can pack a lot of ops in a cubic micron, but
> > you can't put a lot of bits into it. 
> 
> I don't know what you mean. Is this your own idea?  I ask,

Most of these are my ideas. None of them are new, though
I keep rerunning them every few months, just to see whether
they stick. Most of them don't. But I understand others have put 
forward very similiar if not identical such here before, it must 
have been 1980s, or early 1990s. Some of it should be
even in the online archives.

> because I never heard of computronium being infested in the
> way that you suggest.

Computronium is just postbiology's brains, and habitats.
After a diversity bottleneck, you *always* radiate and
speciate. If you do that, you always got hosts and parasites.
Whether these are regular honest-to-god graphene mites
burrowing through god brains, or sneaky information patterns
preying on resources occupied by their hosts, it doesn't 
matter. 
 
> Just what keeps the intelligent parts from cannibalizing the
> less intelligent parts.  Looked at the growth curve of human

Just what keeps a species from radiating? You *can* strip
biology by hopping onto new substrate (this is what causes
the initial diversity bottleneck), but what keeps you coherent
ever after? Nothing, so far as I can see.

> protoplasm lately?  Then think of what an AI might do to the
> solar system.

There is never the one AI. There's is always a 1) population of 
2) imperfectly self-replicating systems 3) in a limited-resource 
context. There is never an isolated stellar system. There are huge 
islands of them. There are never isolated galaxies. They come in 
clusters.  And these come in superclusters. And when you're there, 
probably everything blows up. Or the opposite. Or, there's leverage for
life after all to make this never end, to lacking that, to be
reborn and amplified somewhere and/or somewhen else. 
 
> > You're a ruling intelligence. Do you control everything even
> > within your own body?
> 
> Oh come on now. Only my lack of adequate technology prevents me

If your technology is autonomous, who controls your technology?
Do you know what your Roomba is doing now? Are you sure your
computer is clean of malware? What about symbiotic programs,
which come and go? What about the terrestrial and space networks 
of 2036?

> from achieving TOTAL CONTROL over my property. And I will freely
> extend such control to the limits of the visible universe, staying
> within the limits of law, of course.

What about 10^9 of other Lees which are doing just the same? 
What about those who're warped, malicious, or insane? Or,
in neutral terms, are just sampling the behaviour space
randomly? If I personally want to start devolving into a 
billion of different species, most of them about the complexity of
bacteria and insects as a piece of performance art, will you stop 
me? More imprtantly, will you know, and will you be able to?
 
> > Do you control everything on your city block? Would you really
> > want to?
> 
> Why not?  If I own it, then I'm filling it up with people, or with
> computers. 

Would you want to micromanage those people or those computers?
If you *are* that city block, do you know what all your components
are doing, right at this moment? 
 
> > Why are you letting coyotes roam the New York Central Park?
> 
> Because I don't own the place. 

Okay, now you do. How do you keep them out now? How about the
cockroaches? The stuff living in the sewers? How do you sterilize
an entire city, and keep it sterile hereafter? And how do you
manage the rather radical means of sterilization keep from getting
frisky, growing feet and walking off somewhere else?

-- 
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
______________________________________________________________
ICBM: 48.07100, 11.36820            http://www.ativel.com
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 191 bytes
Desc: Digital signature
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20060701/daa578d8/attachment.bin>