From: Eugen Leitl (eugen@leitl.org)
Date: Sat Jun 08 2002 - 13:01:49 MDT
On Sat, 8 Jun 2002, Lee Corbin wrote:
> You appear to me to have succeeded in lumping our lifestyle
> into the same category as that of Earth's most primitive eras.
We *are* primitive, in relation to how far we still have to go. As long as
we're not a spacefaring species, we're both facing almost immediate
extinction (give or take a few Ma) and are completely irrelevant on the
cosmic scale.
> Yes, at the dawn of time Earth creatures began ingesting atoms,
> and then---thanks to the technological advent of jaws and tails
> ---proceeded to scoop up atoms with gusto and with real efficiency.
> This began to include other creatures, presently, bringing about
> the familiar forms of food chain.)
The analogy is quite apt: naturally emerged biological life is restricted
to a small subset of the entire periodical system of elements. Plus, it is
limited to a specific habitat, so scarcity becomes apparent very soon,
even through a negligible fraction of matter is being utilized.
> I described this as a struggle among algorithms. From Earth,
> or perhaps even a smaller domain (hopefully not Redmond), ever
> more advanced algorithms will be transmitted by EM radiation,
> and conquest by ingestion of atoms will be thus superseded by
> "new releases".
Running exploits at the hardware layer give too much edge to become an
extinct art (unless you've got a singleton seizing control of it and
blocking everybody's access to it -- without backdoors). Algorithms need
bits as a scarce resource, and bits are encoded in arrangements of atoms,
also a scarce resource (whatever bottleneck is narrower).
> High densities of computation per cubic meter will still take
> place in collections of atoms, don't you think? Perhaps versions
High, yes. Infinite, no.
> of me will endure beyond the Oort cloud, in tiny local collections
> of matter unimportant enough for physical conquest by advanced
> algorithms dominating nearby stars. N'est pas?
Atoms are atoms, and no neighbourhood is too shabby if you can make your
own power.
> But hasn't that always been true, about replication, I mean?
> Today, humans, for example, replicate by going to the supermarket
> and preying upon the resources of lesser creatures harvested from
> fields and oceans (and then with that energy, participating in
> conception and growth of more people).
Yes, and it's likely to remain that way. The part that becomes worse, is
that the bulk of atoms are then locked away in bodies/habitats of beings,
and pools of fuel.
> So you are pointing out that algorithms will prey on each other?
Keeping your patterns private looks like a good idea, so there is some
discouragement against a public exchange and recombination of pattern
parts without terminating the entire beastie (similiar to the current gene
(fragment) swap market, where microorganisms fish for fragments of useful
DNA in the biotic soup they're simmering in).
But all living algorithms need hardware to execute, or even to store.
Patterns are much quicker to propagate than it takes to grow habitats to
house them, so it doesn't take atom scarcity to arrive at a resource
scarcity.
> Now I'm lost completely. *Why* does some argument here limit the
> longevity of individuals? Why is it unlikely that some algorithms
> (maybe even me on a deep space snowball) might last a very, very long
> time? Why won't the universe be dominated by powerful algorithms
Because before very long, someone leaner and meaner will come along, and
rearrange those atoms, or even just rearrange the bits, thus erasing the
you-pattern (assuming, you can't prevent it from doing that to you, which
will even happen at the omega fitness scale, where fitness sees only
statistical fluctuations of small (but sufficient) amplitude).
> controlling the resources of massive objects like star systems, but
> rather immortal algorithms?
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 09:14:40 MST