[ExI] simulation as an improvement over reality
eugen at leitl.org
Sat Dec 25 10:56:05 UTC 2010
On Fri, Dec 24, 2010 at 04:17:28PM -0800, The Avantguardian wrote:
> > Presumably, the lifetimes for each species will follow a power law.
> So you readily admit that after a few minute or hours of real time, the upload
> would not be recognizable as human -- let alone Eugen Leitl? Then why bother?
You're no longer recognizable as a blastula stage embryo, yet
you still bothered enough to be still around.
'What is a human being, then?'
'An acorn that is unafraid to destroy itself in growing into a tree'
David Zindell, The Broken God
> > You're sharing the same physical layer. If it doesn't sound ominous,
> > it should.
> Yes. That is ominous. But those native to the physical layer would have an
> easier time manipulating the physical layer. Uploads would need an interface to
There is no fundamental difference between postbiology and you. All
cognition is embodied.
Only, postbiology is a lot fitter than you. Far more adept at manipulating
the physical layer. So we are the endangered species, for a change.
It would be an act of cosmic irony, if the perpetrators of the Holocene
extinction event themselves succumbed to habitat destruction and industrial
pollution at the hands of their offspring. Mere indifference would be enough.
> do so and interfaces can be manipulated.
What is your interface to reality? Why is it not being manipulated?
So what should be different?
> > Everyone seems to think everybody will be so very smart.
> > Diversity works in all directions.
> It's much worse than that. Imagine being locked in a (unix) box with John
Imagine suffering from locked-in syndrome.
> Malkovitch for a thousand subjective years. Could you guarantee that at the end
So don't do it, then.
> of that time, you yourself would not be John Malkovitch? Then consider that this
> might occur in mere minutes of real time. Then ask yourself what would become of
> Malkovitch-you after a million years of subjective time?
Imagine you being waterboarded for the rest of your life. It's going to
suck a lot, yes? How about solitary confinement, until your mind unravels
completely? There are a billion of possible alternatives which all suck
far more then the one you're in. Yet most people do not spend much time
thinking about how their lives could suck more, and then actively striving
to make them suck more.
> The uploads would be subject to internal selective pressures, from other uploads
You're also subject to selection pressures. Good old Darwin.
> or the virtual environment itself that would feedback on their configuration
The virtual environment does not differ from real environment in any
other way than it can be more easily changed. Just because it can
it doesn't mean it will, because you will need a consensus for continuing
interactions with others. Extreme solipsism is incompatible with
You are an embodied intelligence. You occupy space. You need Joules and
atoms, a metabolism to substitute those parts which are degraded. You
compete for resources with other embodied intelligences. You co-operate
with them, and you co-evolve with them.
In other way, the more things change, the more they stay the same.
> state at a much faster rate than would any changes in the physical layer. So
> after a while everybody would have adapted to John Malkovitch because he visits
How about Jesus? He seems to visit a lot of people often.
> once a week, but nobody knows how to deal with thunder shower in the real world
I don't think there will be thunder showers. Or atmosphere, for that matter.
Unless somebody really cares.
> because one of them only occurs every few million years of subjective time.
If you're maladapted, you die. Better adapted remain.
> > They're dead, Jim.
> So are all the "people" who uploaded. If similarity to the "original" person is
If it has a metabolism and if it can eat you it ain't dead.
> an issue. I do agree with you that not all the adaptation will be toward more
I'm not sure Lamarck would actually work. It would take a particular description,
and abandoment of a genome->phenotype mapping. If at all possible, the result
would be probably not compatible with Darwin operating in the same system. While
Lamarckian systems still being in competition with Darwinian systems. May the
best ones win? Something like that.
> sophistication and complexity. While some uploads might become M-Brains other
> uploads could revert to worms, busy beavers, or other stable algorithms.
Exactly. Though we're pretty close to the complexity floor,
so the reachable ceiling is ways off. But in terms of moles
of circuitry, the simpler postbiomass will probably
outweigh the gods by a fair fraction.
Eugen* Leitl <a href="http://leitl.org">leitl</a> http://leitl.org
ICBM: 48.07100, 11.36820 http://www.ativel.com http://postbiota.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A 7779 75B0 2443 8B29 F6BE
More information about the extropy-chat