Usenet discusssion - started April 15th. 2007 by me.
comments by Greg Egan and more.


THE SIMULATION ARGUMENT IN SCIENCE FICTION - AND BEYOND
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
 
From: SimonLaub.mail@gmail.com 
Date: 15 Apr 2007 10:10:56 -0700 
Local time: Sunday. 15 Apr. 2007 19:10 
NewsGroups: rec.arts.sf.written, rec.arts.sf.science, comp.society.futures 
Subject: The simulation argument in science fiction - and beyond 
-----------------------------------------------------------------------------------------

Once in a while comes along a book with breathtaking speculation. 
Certainly, Marcus Chowns "Dispatches from the frontline of science" 
fits the description. As Brian May (Queen guitarist) says: "Marcus 
Chown rocks". 

After you have read the Marcus chown book you certainly begin to 
wonder where the borderline between science and science fiction really 
is. I especially enjoy Chowns simulation speculations. 
Nick Bostroms simulation argument is covered over some chapters in the 
book 
(See http://home18.inet.tele.dk/silanian/Post/simulationarg.htm). To 
me this is pure Greg Egan stuff/speculations coming true -some 50 year 
before I had expected it. 
Sure, it is kind of difficult to tell whether we are living in a 
simulation or not - but these simulations seems to creep up on us from 
all sides in The Marcus Chown book - E.g. take the brilliant/ 
outrageous approach of the Frank Tipler resurrection scheme - at the 
end of time - on some infinite "end of time computer" - 
(see http://home18.inet.tele.dk/silanian/Post/neverendingdays.htm) 
Greg Egans simulation stuff has arrived and is now spreading - 
eventually we will all be convinced that we are really living in a 
simulation ..
 
-Simon 
http://www.fortunecity.com/skyscraper/lol/1165/index.html 

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
 
From: Erik Max Francis  
Date: Sun, 15 Apr 2007 17:31:39 -0700 
Local: Ma. 16 Apr. 2007 02:31 
NewsGroup: rec.arts.sf.written, rec.arts.sf.science, comp.society.futures 
Subject: Re: The simulation argument in science fiction - and beyond 

SimonLaub.m...@gmail.com wrote: 
> Once in a while comes along a book with breathtaking speculation. 
> Certainly, Marcus Chowns "Dispatches from the frontline of science" 
> fits the description. As Brian May (Queen guitarist) says: "Marcus 
> Chown rocks". 
> After you have read the Marcus chown book you certainly begin to 
> wonder where the borderline between science and science fiction really 
> is. I especially enjoy Chowns simulation speculations. 
> Nick Bostroms simulation argument is covered over some chapters in the 
> book 
> (See http://home18.inet.tele.dk/silanian/Post/simulationarg.htm). To 
> me this is pure Greg Egan stuff/speculations coming true -some 50 year 
> before I had expected it. 
> Sure, it is kind of difficult to tell whether we are living in a 
> simulation or not - but these simulations seems to creep up on us from 
> all sides in The Marcus Chown book - E.g. take the brilliant/ 
> outrageous approach of the Frank Tipler resurrection scheme - at the 
> end of time - on some infinite "end of time computer) - 
> (see http://home18.inet.tele.dk/silanian/Post/neverendingdays.htm) 
> Greg Egans simulation stuff has arrived and is now spreading - 
> eventually we will all be convinced that we are really living in a 
> simulation .. 

I've never found these types of conjectures very worthy of merit. If 
you're living in a properly-designed simulation, then you'd never know 
it. So what you're hoping to look for is flaws in the simulation, but 
you have no idea what those flaws might be, or how they might manifest 
themselves -- or, moreover, whether or not some peculiarity you might 
have discovered is indeed a flaw at all. Even ideas about where to look 
for flaws is suspect, since you've no guarantee that whatever universe 
the simulation is running in has laws even remotely similar to ours. 
It strikes me as completely unfalsifiable and about as worthy of real 
scientific study as speculation about whether God is real: Any "proof" 
you find is subject to interpretation and questionable at best. 
-- 
Erik Max Francis && m...@alcyone.com && http://www.alcyone.com/max/ 
San Jose, CA, USA && 37 20 N 121 53 W && AIM, Y!M erikmaxfrancis 
The slaying of multitudes should be mourned with sorrow. / A victory 
should be celebrated with the funeral rite. -- Laotse, ca. 6th C. BC

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

NewsGroups: rec.arts.sf.written, rec.arts.sf.science, comp.society.futures 
From: "Greg Egan"  
Date: 15 Apr 2007 19:29:20 -0700 
Local Time: Ma. 16 Apr. 2007 04:29 
Subject: Re: The simulation argument in science fiction - and beyond 


On Apr 16, 8:31 am, Erik Max Francis  wrote: 
[snip] 
> I've never found these types of conjectures very worthy of merit. If 
> you're living in a properly-designed simulation, then you'd never know 
> it. So what you're hoping to look for is flaws in the simulation, but 
> you have no idea what those flaws might be, or how they might manifest 
> themselves -- or, moreover, whether or not some peculiarity you might 
> have discovered is indeed a flaw at all. Even ideas about where to look 
> for flaws is suspect, since you've no guarantee that whatever universe 
> the simulation is running in has laws even remotely similar to ours. 
> It strikes me as completely unfalsifiable and about as worthy of real 
> scientific study as speculation about whether God is real: Any "proof" 
> you find is subject to interpretation and questionable at best. 

I agree with Erik on this 100%. I think Bostrom's arguments are 
logically flawed, but also (and I know this is subjective) immensely 
pessimistic. As I argued in another recent thread, evolution is not 
the kind of thing nice people inflict on anyone, and I don't believe 
"our" descendants would be so stupid and sadistic as to inflict what 
nature did to their ancestors upon a new set of beings, all over 
again. It becomes *slightly* more plausible if we assume that the 
simulators bear no relation to "us", so they're less likely to hold us 
in their affections, but it's still a repugnant prospect. At least 
Tipler wanted the simulation to be Heaven; Bostrom's vision implies 
that the number of deliberate simulations that include something like 
Auschwitz is vastly greater than the number of times this kind of 
thing has happened in nature

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

Fra: "Michael S. Schiffer"  
Dato: 16 Apr 2007 03:40:55 GMT 
Lokalt: Ma. 16 Apr. 2007 05:40 
Nyhedsgrupper: rec.arts.sf.written, rec.arts.sf.science, comp.society.futures 
Emne: Re: The simulation argument in science fiction - and beyond 

"Greg Egan"  wrote in 
news:1176690559.982598.27260@b75g2000hsg.googlegroups.com: 
> On Apr 16, 8:31 am, Erik Max Francis  wrote: 
> [snip] 
>> I've never found these types of conjectures very worthy of 
>> merit. If you're living in a properly-designed simulation, 
>> then you'd never know it. So what you're hoping to look for is 
>> flaws in the simulation, but you have no idea what those flaws 
>> might be, or how they might manifest themselves -- or, 
>> moreover, whether or not some peculiarity you might have 
>> discovered is indeed a flaw at all. Even ideas about where to 
>> look for flaws is suspect, since you've no guarantee that 
>> whatever universe the simulation is running in has laws even 
>> remotely similar to ours. 
>> It strikes me as completely unfalsifiable and about as worthy 
>> of real scientific study as speculation about whether God is 
>> real: Any "proof" you find is subject to interpretation and 
>> questionable at best. 
> I agree with Erik on this 100%. I think Bostrom's arguments are 
> logically flawed, but also (and I know this is subjective) 
> immensely pessimistic. As I argued in another recent thread, 
> evolution is not the kind of thing nice people inflict on 
> anyone, and I don't believe "our" descendants would be so stupid 
> and sadistic as to inflict what nature did to their ancestors 
> upon a new set of beings, all over again. 
>... 

I don't find the hypothesis convincing (and agree it seems 
nonfalsifiable), and am treating it as a thought experiment. 
However, even stipulating that all those views reach consensus, 
consensus isn't the same as universal. Slavery is outlawed in 
(AFAIK) every country by now, but there are still some slaves in 
the world. Murder is outlawed everywhere, but there are still 
murders. Depending on how resource-intensive such a simulation is 
by the standards of the civilization in question, and how many 
individuals it holds, it seems possible that some might slip 
through the cracks even if they're viewed the way the modern West 
views cannibalism. If one in a billion individuals racks up as 
many simulated Earths as a modern serial killer does bodies in his 
basement before getting caught... 
Mike

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

Greg Egan:
On Apr 16, 11:40 am, "Michael S. Schiffer" 
 wrote: 

[snip] 
> I don't find the hypothesis convincing (and agree it seems 
> nonfalsifiable), and am treating it as a thought experiment. 
> However, even stipulating that all those views reach consensus, 
> consensus isn't the same as universal. Slavery is outlawed in 
> (AFAIK) every country by now, but there are still some slaves in 
> the world. Murder is outlawed everywhere, but there are still 
> murders. Depending on how resource-intensive such a simulation is 
> by the standards of the civilization in question, and how many 
> individuals it holds, it seems possible that some might slip 
> through the cracks even if they're viewed the way the modern West 
> views cannibalism. If one in a billion individuals racks up as 
> many simulated Earths as a modern serial killer does bodies in his 
> basement before getting caught... 

I wouldn't dispute for a moment that there are *possible* scenarios 
where the numbers would stack up in favour of simulation. What bugs 
me is when people try to push the argument to the point of saying 
"Either you declare that humanity and all other advanced civilisations 
are short-lived, or you must accept that we're vastly more likely to 
be a simulation than the real thing". Bollocks to that. Rather like 
the Drake equation, there are enough free parameters in a serious 
treatment of this issue to make the numbers go any way you like. 
I also think there's a strong cultural bias creeping into the 
argument. Simulated worlds feature in current SF and futurology for 
obvious cultural reasons surrounding current technology. I don't 
doubt that our descendants will probably always be doing *some* things 
that come under the general banner of "realistic-ish VR", but I 
suspect most will be voluntary and as easy to step out of as switching 
off a TV. This particular current obsession might well end up as 
obscure a notion as having a human slave to be your chimney sweep. 
It's not just that it's evil, in the long run it's also rather silly. 
Unless you're either very dumb or very nasty, you can find other ways 
to get the job done

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

Damien Sullivan:
Greg Egan"  wrote: 
>where the numbers would stack up in favour of simulation. What bugs 
>me is when people try to push the argument to the point of saying 
>"Either you declare that humanity and all other advanced civilisations 
>are short-lived, or you must accept that we're vastly more likely to 
>be a simulation than the real thing". Bollocks to that. Rather like 

Especially given the computational resources needed to simulate us 
within our universe. Okay, so maybe a Jupiter-mass quantum computer 
could simulate the Earth and what we observe of the universe, but is 
someone really going to bother? 
Well, in the long term, maybe -- I could see wanting to simulate 
different planets. But that doesn't mean they'd be replaying our own 
history over and over again, or casually, or even having vastly many 
planetary simulators in general. 
-xx- Damien X-) 

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

Eivind Kjorstad:
Damien Sullivan wrote: 
> Especially given the computational resources needed to simulate us 
> within our universe. 

It's silly anyway, way into philosophy lala-land. 
IF you buy the argument that we're almost certainly simulated, then by 
the very same argument, you could argue that almost certianly, *you* are 
simulated, and all the rest doesn't exist, not even in the simulation. 
Surely, simulating a single human mind is much simpler than simulating 
an entire universe. 

Eivind Kjørstad

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

Michael Grosberg:
On Apr 16, 8:16 am, "Greg Egan"  wrote: 
> I also think there's a strong cultural bias creeping into the 
> argument. Simulated worlds feature in current SF and futurology for 
> obvious cultural reasons surrounding current technology. I don't 
> doubt that our descendants will probably always be doing *some* things 
> that come under the general banner of "realistic-ish VR", but I 
> suspect most will be voluntary and as easy to step out of as switching 
> off a TV. This particular current obsession might well end up as 
> obscure a notion as having a human slave to be your chimney sweep. 

That's a very good point. It reminds me of the fear of being burried 
alive, which was a surprisingly common concern in the 19th century. It 
spawned countless horror stories (Poe made quite a career out of it) 
and even some patents for devices that were meant to allow the not- 
quite-dead-yet to alert people to their situation from inside the 
coffin. I guess being trapped in a VR is our modern day equivalent.


XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

DJensen writes:
> "Greg Egan"  wrote: 
> >where the numbers would stack up in favour of simulation. What bugs 
> >me is when people try to push the argument to the point of saying 
> >"Either you declare that humanity and all other advanced civilisations 
> >are short-lived, or you must accept that we're vastly more likely to 
> >be a simulation than the real thing". Bollocks to that. Rather like 
> Especially given the computational resources needed to simulate us 
> within our universe. Okay, so maybe a Jupiter-mass quantum computer 
> could simulate the Earth and what we observe of the universe, but is 
> someone really going to bother? 
> Well, in the long term, maybe -- I could see wanting to simulate 
> different planets. But that doesn't mean they'd be replaying our own 
> history over and over again, or casually, or even having vastly many 
> planetary simulators in general. 

DJensen:
The simulated world/universe can be a fun thought experiment, but it's 
more 'useful' for coming up with unfalsifiable variations on the 
nature of the simulation. I don't think it's really worth 
investigating too deeply: I'm real and plugged in; I'm the only fully 
simulated person; humans (or a subset) are plugged in or fully 
simulated; only what we/I currently experience is being simulated; the 
whole world is fully simulated, but it ends one kilometre up and the 
rest is approximated... It's solipsism in high definition video with 
surround sound. And what's to be gained even if you could prove any of 
it? At some point, somewhere in the simulation's ground rules, there 
will be arbitrary values or a random number generator of some kind, 
throwing out any hope you might have for determining the nature of 
Real Reality or the Simulators. 
(In _The Algebraist_ [and perhaps other Culture books, I don't know] 
it's speculated that once a certain percentage of all intelligent 
organisms believe the universe is simulated, the simulation will end.) 
-- 
DJensen

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Erik Max Francis 
Organisation: Alcyone Systems
Date: 16. april 2007 11:22
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

"Erik Max Francis"  wrote in message news:...
> Eivind Kjorstad wrote:
> 
> > Surely, simulating a single human mind is much simpler than 
> > simulating
> > an entire universe.
> 

 There's Occam's razor dropping.  More like Occam's guillotine, really.
 
 --
Erik Max Francis && max@alcyone.com && http://www.alcyone.com/max/
San Jose, CA, USA && 37 20 N 121 53 W && AIM, Y!M erikmaxfrancis
Chastity the most unnatural of the sexual perversions.
 -- Aldous Huxley

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Michael Ash:

In rec.arts.sf.science Erik Max Francis  wrote:
> Eivind Kjorstad wrote:
> 
>> Surely, simulating a single human mind is much simpler than simulating 
>> an entire universe.
> 
> There's Occam's razor dropping. More like Occam's guillotine, really.

A single mind is simpler than the entire universe, and reality is simpler 
than simulation. Solipsism, here we come!

-- 
Michael Ash
Rogue Amoeba Software

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX


From: Sea Wasp

Damien Sullivan wrote:
> "Greg Egan"  wrote:
> 
> 
>>where the numbers would stack up in favour of simulation. What bugs
>>me is when people try to push the argument to the point of saying
>>"Either you declare that humanity and all other advanced civilisations
>>are short-lived, or you must accept that we're vastly more likely to
>>be a simulation than the real thing". Bollocks to that. Rather like
> 
> 
> Especially given the computational resources needed to simulate us
> within our universe. Okay, so maybe a Jupiter-mass quantum computer
> could simulate the Earth and what we observe of the universe, but is
> someone really going to bother?
> 
> Well, in the long term, maybe -- I could see wanting to simulate
> different planets. But that doesn't mean they'd be replaying our own
> history over and over again, or casually, or even having vastly many
> planetary simulators in general.

If the *actual* resources available to the civilization in question 
are a few dozen orders of magnitude greater than what *we* think is 
the universe, then your premise is flawed -- it's like, say, Ken 
Masters from Street Fighter arguing that there's no way he's part of a 
video game, because the entire universe he sees would require entire 
megabytes of data resources to simulate. If the equivalent of the 
simulator's PCs have the necessary resources, then it's not a problem.

ObSFVideoGame (Spoiler):



Spoiler ------------>


Ahoy!


In the video game "Star Ocean 3: To the End of Time", it turns out 
the main characters are ALL just AI characters in a video game run for 
entertainment by an exceedingly advanced civilization which is, 
nonetheless, basically run by human beings. The main characters are 
like their creators because their creators, obviously, would like to 
play games with characters that ARE like them. Just as I play human 
beings in my own RPGs, and assume human or humanlike characters 
throughout the world.

-- 
Sea Wasp
Live Journal: http://www.livejournal.com/users/seawasp/

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX


From James A. Donald 

On Mon, 16 Apr 2007 08:50:26 -0400, Sea Wasp wrote:
> If the *actual* resources available to the 
> civilization in question are a few dozen orders of 
> magnitude greater than what *we* think is the 
> universe, then your premise is flawed

Indeed, simulations are almost always run smaller than 
the original - a chinese garden simulates a wilderness, 
World of Warcraft simulates a universe in which you can 
walk from one end of the universe to the other in a few 
hours. Even if the main point of the simulation is to 
tell us something about the real world, you usually want 
to isolate a single point of interest concerning the 
real world.

If we were producing a simulation of the real world, 
occupied by AI NPCs, the main town would likely have 
half a dozen shops, and an inn in which you will likely 
meet every significant character sooner or later, and 
you could circumnavigate the planet in an afternoon 
stroll. If we wanted to simulate ancient greece, the 
three hundred Spartans would become the seven spartans, 
and they would constitute a good third of the entire 
Spartan army, and Sparta something like a third of the 
population of Greece, and Greece would have the great 
majority of all fully simulated or fully scripted NPCs, 
and a sizable fraction of all PCs. It would be a thirty 
minute run from Sparta to Athens. The battle of 
Thermopylae would pit seven spartans and two dozen Greek 
bit players against four or five hundred persian bit 
players, with one fully realized Xerxes, and a couple of 
fully realized minions of Xerxes. A single farm would 
be the size of a modern suburban house, and would be 
capable of supporting several townsmen.


--
----------------------
We have the right to defend ourselves and our property, because 
of the kind of animals that we are. True law derives from this 
right, not from the arbitrary power of the omnipotent state.

http://www.jim.com/ James A. Donald

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Mart Atwood:
Eivind Kjorstad  writes:
> 
> Surely, simulating a single human mind is much simpler than simulating
> an entire universe.

ObSF: "A Colder War"

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Gene Ward Smith
On Apr 15, 10:16 pm, "Greg Egan"  wrote:

> It's not just that it's evil, in the long run it's also rather silly.

It's also not known to be possible, a fact people keep forgetting
about such questions

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From : James Killian:

Eivind Kjorstad  wrote::
> Surely, simulating a single human mind is much simpler than simulating
> an entire universe.

Some of the stories in Lem's _Star Diaries_ use this idea. In one,
an inventor runs a set of simulated minds. In another (_Dr. Diagoras_),
there's a simulated mind (a clone of the guy from the first story, IIRC)
running in a talking clock.

-- 
mailto:jjk@acm.org As the air to a bird, or the sea to a fish,
http://www.bawue.de/~jjk/ so is contempt to the contemptible. [Blake]
http://del.icio.us/jjk

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: James A. Donald 
Date: 16. april 2007 22:27
NewsGroups:: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

"Greg Egan" wrote:
> I also think there's a strong cultural bias creeping
> into the argument. Simulated worlds feature in
> current SF and futurology for obvious cultural reasons
> surrounding current technology. I don't doubt that
> our descendants will probably always be doing *some*
> things that come under the general banner of
> "realistic-ish VR", but I suspect most will be
> voluntary and as easy to step out of as switching off
> a TV. This particular current obsession might well
> end up as obscure a notion as having a human slave to
> be your chimney sweep. It's not just that it's evil,
> in the long run it's also rather silly. Unless you're
> either very dumb or very nasty, you can find other
> ways to get the job done.

I am pretty sure our descendants will be doing simulated
worlds for entertainment and research. The simulation
is more interesting, to the extent that the NPCs
approximate consciousness.

But it probably is an excessive waste of resources to
simulate an entire civilization of several billion souls
- I would expect a few important NPCs to gett a full
personality that reflects their particular simulated
history, a rather larger number of NPCs to get a
scripted personality - they are one of several standard
personalities, and they are conscious only when the
story line impinges on them, and the rest to be
simulated by a large scale model that simulates economic
forces and the like, averages over large numbers of
people, rather than individual people, instantiated as
zombies and scripted personalities when a PC or a more
fully simulated NPC impinges on them. The world,
appears in detail when a significant character
approaches, and fades away into formula, stereotype, and
nothingness, when he departs. The more fully simulated
NPCs are programmed to overlook the more obviously
discrepancies, the discrepancies that the programmers
though of, and the script driven and zombie NPCs are
incapable of noticing any discrepancies.

And when our descendants do such simulations,
simulations simulating people having a really bad time
are likely to be more interesting - they will want
simulations of evil people where the evil people have a
real prospect of getting away with it and defeating the
heroes, because in real life, (the real life of those
doing the simulation) there is genuine difficulty in
detecting evil, and genuine danger from evil. A
simulation of Tipler's heaven is likely to be boring,
and therefore unlikely to be run.


--
----------------------
We have the right to defend ourselves and our property, because 
of the kind of animals that we are. True law derives from this 
right, not from the arbitrary power of the omnipotent state.

http://www.jim.com/ James A. Donald

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: James A. Donald
Date: 16. april 2007 22:37

Damien Sullivan wrote:
> Especially given the computational resources needed to
> simulate us within our universe. Okay, so maybe a
> Jupiter-mass quantum computer could simulate the Earth
> and what we observe of the universe, but is someone
> really going to bother?

It is likely our descendants will gain control of a
significant portion of the matter and energy in at least
the local group of galaxies. So every moderately
wealthy individual might have his own personal dyson
sphere, in which case running a full simulation of
humans on earth from prehistoric times to the
present probably costs him as much as a session of World
of Warcraft costs me.

--
----------------------
We have the right to defend ourselves and our property, because 
of the kind of animals that we are. True law derives from this 
right, not from the arbitrary power of the omnipotent state.

http://www.jim.com/ James A. Donald

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Gene War Smith

On Apr 16, 1:37 pm, James A. Donald  wrote:
> It is likely our descendants will gain control of a
> significant portion of the matter and energy in at least
> the local group of galaxies. 

Is there a line we can draw between Baysian probability
and random assertion of opinion here? I'm wondering what
evidence, if any, suggests this is "likely

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
From: James A. Donald

> > It is likely our descendants will gain control of a 
> > significant portion of the matter and energy in at
> > least the local group of galaxies.

"Gene Ward Smith" wrote:
> Is there a line we can draw between Baysian
> probability and random assertion of opinion here? I'm
> wondering what evidence, if any, suggests this is
> "likely".

It is physically possible, and there does not appear to
be anyone to stop us - unless of course, we are in a
simulation and the simulation stops a short distance
above our heads. Of course we will have to remake our
biological nature, but we are already getting started on
that. Because of the large distances between the
starts, we will have to have either immortality or FTL -
but immortality seems pretty much inevitable.

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: David DeLaney 
OraganizatioN: Formerly U. Tenn. Knoxville/Physics Dept.; presently extremely dis
Date: 17. april 2007 09:42
NewsGroup: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

James A. Donald  wrote:
>"Gene Ward Smith"
>> Is there a line we can draw between Baysian
>> probability and random assertion of opinion here? I'm
>> wondering what evidence, if any, suggests this is
>> "likely".
>
>It is physically possible, and there does not appear to
>be anyone to stop us

Disagree strongly. We are perfectly capable of stopping ourselves. (Alas.)

(The assumption that we don't stop ourselves turns it into 'will physics
stop us?', and since at present there don't seem to be invisible walls, if
we assume we don't screw it up on our own, we'll probably get Out There,
if slowly.)

Dave
-- 
\/David DeLaney posting from dbd@vic.com "It's not the pot that grows the flower
It's not the clock that slows the hour The definition's plain for anyone to see
Love is all it takes to make a family" - R&P. VISUALIZE HAPPYNET VRbeable > It is physically possible [to seize the nearby galaxies], and there does not appear to
> > be anyone to stop us

David DeLaney wrote:
> Disagree strongly. We are perfectly capable of stopping ourselves. (Alas.)

As long as we remain stuck on earth, we are quite likely to stop
ourselves. If, however, some of us can get off earth and make a
living out there, we are then pretty much unstoppable.
--
----------------------
We have the right to defend ourselves and our property, because 
of the kind of animals that we are. True law derives from this 
right, not from the arbitrary power of the omnipotent state.

http://www.jim.com/ James A. Donald

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: James A. Donald

On 16 Apr 2007 01:49:07 -0700, "DJensen"  wrote:
> The simulated world/universe can be a fun thought experiment, but it's
> more 'useful' for coming up with unfalsifiable variations on the
> nature of the simulation. I don't think it's really worth
> investigating too deeply: I'm real and plugged in; I'm the only fully
> simulated person; humans (or a subset) are plugged in or fully
> simulated; only what we/I currently experience is being simulated; the
> whole world is fully simulated, but it ends one kilometre up and the
> rest is approximated... It's solipsism in high definition video with
> surround sound. And what's to be gained even if you could prove any of
> it? At some point, somewhere in the simulation's ground rules, there
> will be arbitrary values or a random number generator of some kind,
> throwing out any hope you might have for determining the nature of
> Real Reality or the Simulators.

If you discover flaws in the simulation, you can do magic , though
this would probably result in you getting reset by a GM.

If you recognize the story line, you could get to talk to one of the
PCs, and he might break character and tell you about the real world,
or more likely you might drop in on an out of character conversation
between PCs.

--
----------------------
We have the right to defend ourselves and our property, because 
of the kind of animals that we are. True law derives from this 
right, not from the arbitrary power of the omnipotent state.

http://www.jim.com/ James A. Donald

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Damien Sullivan 
Organization: Murray's Mud Minions
Date: 17. april 2007 01:16
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond


Eivind Kjorstad  wrote:
>IF you buy the argument that we're almost certainly simulated, then by 
>the very same argument, you could argue that almost certianly, *you* are 
>simulated, and all the rest doesn't exist, not even in the simulation.
>
>Surely, simulating a single human mind is much simpler than simulating 
>an entire universe.

Except that you have to provide perceptions to the simulated mind. If
the simulated mind is to perceive other minds responding in appropriate
ways to its actions, simulating the other minds may well be the simplest
way of calculating the responses.

The argument might be extensible to the environment, though less
strongly so; at some point presumably heuristic filling in of detail
becomes cheaper than full emulation. But, going the other direction, at
some point the process you're trying to cheat on runs into Halting
Problem territory, and you need to emulate.

-xx- Damien X-)

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: dwight.thieme@gmail.com 
Organization: http://groups.google.com
Date: 17. april 2007 03:46
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

On Apr 16, 6:16 pm, phoe...@ofb.net (Damien Sullivan) wrote:
> Eivind Kjorstad  wrote:
> >IF you buy the argument that we're almost certainly simulated, then by
> >the very same argument, you could argue that almost certianly, *you* are
> >simulated, and all the rest doesn't exist, not even in the simulation.
>
> >Surely, simulating a single human mind is much simpler than simulating
> >an entire universe.
>
> Except that you have to provide perceptions to the simulated mind. If
> the simulated mind is to perceive other minds responding in appropriate
> ways to its actions, simulating the other minds may well be the simplest
> way of calculating the responses.
>
> The argument might be extensible to the environment, though less
> strongly so; at some point presumably heuristic filling in of detail
> becomes cheaper than full emulation. But, going the other direction, at
> some point the process you're trying to cheat on runs into Halting
> Problem territory, and you need to emulate.

Could be turtles all the way up both ways. An infinite universe is
home to a computer with infinite storage capacity running a a
simulation of an infinite universe in which there exists a
computer . . .

So, no problems with the detail work. If you don't like the ethics of
the situation, consider an alternate scenario: everyone is here
because they want to be, and they know full well what lies in store
for them, good, bad, and indifferent.

Gee, this is fun

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Eivind Kjorstad .
Oraganization: NetPower AS
Date: 17. april 2007 09:18
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

Damien Sullivan wrote:
> Eivind Kjorstad  wrote:
>> IF you buy the argument that we're almost certainly simulated, then by 
>> the very same argument, you could argue that almost certianly, *you* are 
>> simulated, and all the rest doesn't exist, not even in the simulation.
>>
>> Surely, simulating a single human mind is much simpler than simulating 
>> an entire universe.
> 
> Except that you have to provide perceptions to the simulated mind. If
> the simulated mind is to perceive other minds responding in appropriate
> ways to its actions, simulating the other minds may well be the simplest
> way of calculating the responses.

Perhaps. But still. Simulating those parts of the universe which I 
interact with in sufficient detail that *I* won't notice parts missing 
must be much simpler than simulating *all* parts of the universe in the 
same detail as my mind.

It does seem as if noone bothered simulating the Higgs, despite the fact 
that the standard model doesn't really make any sense without it. ;-)


Eivind Kjørstad

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Simon Laub
Date: 17. april 2007 18:52
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

"Greg Egan"  wrote:
>>Erik Max Francis wrote:
>> I've never found these types of conjectures very worthy of merit.  If
>> you're living in a properly-designed simulation, then you'd never know
>> it.
>> It strikes me as completely unfalsifiable and about as worthy of real
>> scientific study as speculation about whether God is real:  Any "proof"
>> you find is subject to interpretation and questionable at best.

>I agree with Erik on this 100%.  I think Bostrom's arguments are
>logically flawed, but also (and I know this is subjective) immensely
>pessimistic.  As I argued in another recent thread, evolution is not
>the kind of thing nice people inflict on anyone, and I don't believe
>"our" descendants would be so stupid and sadistic as to inflict what
>nature did to their ancestors upon a new set of beings, all over
>again.  It becomes *slightly* more plausible if we assume that the
>simulators bear no relation to "us", so they're less likely to hold us
>in their affections, but it's still a repugnant prospect.  At least
>Tipler wanted the simulation to be Heaven; Bostrom's vision implies
>that the number of deliberate simulations that include something like
>Auschwitz is vastly greater than the number of times this kind of
>thing has happened in nature

True, Evolution is potentially a bitch.
But, well - lets assume that we want to test a theory on the early
universe? And what is more natural than running a simulation.
Surely, we will run such a simulation on a quantum computer in the near 
future.

In quantum computations we are told you can't look at the processing while
it is going on - this will cause decoherence, and effectively
kill the computation. - So, we will have to wait for the end result ...

So, until you have your end result (whatever you are
looking for) - you won't know what is going on.
We are also told that the qubits in our computer are logical
equivalent to the particles they simulate. In the quantum world
there is total breakdown between simulated and real.

What then if some intermediate part of the simulation actually
has livings things - including murder and horrors?
We wouldn't know - because we can't look.  And
if we looked we might not understand that we were indeed
looking at such horrible things?
Our focus is that "end result of the quantum simulation" a billions years 
later.

I don't think you have to look very far for someone
wanting some sort of end result and being perfectly willing to let someone 
else pay the price.
Empathy for unknown stuff in your simulation ? Come on? :-)

Michael S. Schiffer wrote:
> I don't find the hypothesis convincing (and agree it seems
> nonfalsifiable), and am treating it as a thought experiment.

Surely, the universe is computing away. All the little
virtual particles pop up from the void and are given as input
to physical laws - whichs computes a new state. The universe
is one giant quantum computer.
So there is computation all over the place. To call
it a simulation is then a matter of words? You really only need a mapping 
from
the protons and electrons in our universe to what they
represent outside.  And then someone out there in
5 + dimensional brane world to understand the mapping?
Obviously the simulation will always look real to the ones in it,
unless you have signs popping up in midair saying "YOU ARE IN
A SIMULATION".

Greg Egan wrote:
>I wouldn't dispute for a moment that there are *possible* scenarios
>where the numbers would stack up in favour of simulation.  What bugs
>me is when people try to push the argument to the point of saying
>"Either you declare that humanity and all other advanced civilisations
>are short-lived, or you must accept that we're vastly more likely to
>be a simulation than the real thing".  Bollocks to that.  Rather like
>the Drake equation, there are enough free parameters in a serious
>treatment of this issue to make the numbers go any way you like.

Can we really imagine that a civilisation could make a thing
(a bomb, a simulation - whatever) - and then not let it go off, so to speak?
Saw one CERN guy recently explaining that their collider
was really safe - He couldn't see the dangers in that thing producing
mini black holes, or turning all Earth into some new and unknown phase
of matter. And he surely was pressing ahead with the experiment eager
to get that Higgs particle .. risks, small and insignificant...
Ok, maybe he was right, maybe the CERN risks are small -
but in general I think things imagined becomes real even when great
risks are involved.

>I also think there's a strong cultural bias creeping into the
>argument.  Simulated worlds feature in current SF and futurology for
>obvious cultural reasons surrounding current technology.  I don't
>doubt that our descendants will probably always be doing *some* things
>that come under the general banner of "realistic-ish VR", but I
>suspect most will be voluntary and as easy to step out of as switching
>off a TV.

:-) The better the simulations - the more connections between
you and the simulated world - the more difficult it is to step out of?
The harder it is for you to understand that you can actually step out of it?

>This particular current obsession might well end up as
>obscure a notion as having a human slave  to be your chimney sweep.
>It's not just that it's evil, in the long run it's also rather silly.
>Unless you're either very dumb or very nasty, you can find other ways
>to get the job done

I see it more as - it is only now that we are beginning to understand
simulations and what they really are. In the 1800hundreds
Boltzmann and his friends thought gases were exiciting.
We now have the S=Log W thing (or whatever it was) -
on to new stuff :-)

---

"DJensen"  wrote in message

> The simulated world/universe can be a fun thought experiment, but it's
> more 'useful' for coming up with unfalsifiable variations on the
> nature of the simulation. ... It's solipsism in high definition video with
> surround sound. And what's to be gained even if you could prove any of
> it?

Finding the stop button?. Remember the Clark story, where you just
have to mention all the possible names of some particular god ...
The the stars stop shining - the simulation stops .. and ....

Surely you want to knows the rules of the place you live in?


Eivind Kjørstad wrote:
>Surely, simulating a single human mind is much simpler than simulating
>an entire universe

Isn't that quantum physics for you? You don't know, before
you make a measurement. So yes, you have one conscious thing in there -
and whenever this conscious thing wants, sees, does something -
things have to "collapse" into a coherent world this conscious thing can be 
in.
Perhaps whats needs to be collapsed is such a tremendous task that you
could just as well simulate the whole thing in the first place ...

---------------------

Simon

Simon Laub
http://www.fortunecity.com/skyscraper/lol/1165/recentposts/mypostindex.html 

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Greg Egan 
Organization: http://groups.google.com
Date: 18. april 2007 01:03
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.future
Subject: Re: The simulation argument in science fiction - and beyond

On Apr 18, 12:52 am, "Simon Laub"  wrote:
[snip]
> True, Evolution is potentially a bitch.
> But, well - lets assume that we want to test a theory on the early
> universe? And what is more natural than running a simulation.

When we "simulate" the early universe, or other physical processes,
there is always a trade-off between detail, computing resources, and
the most import factor:  how smart we are at computing important
results without simulating boring and expensive details.  (Nobody
simulates an ideal gas by modelling every individual molecule
separately.)  When a science makes *real* progress, it is not about
buying a bigger computer and doing more brute-force computing, it is
about getting smarter at calculating more results by making use of
insights into the nature of the problem.

And even in messy fields involving mass human behaviour, simulating
the players down to the level where they'd be conscious would always
be vast overkill.  We can glean enough about human behaviour from
observations and historical records to abstract out useful higher-
level rules; this is a non-trivial project, but anyone who imagines
that simulating a billion people at *consciousness level* would be
easier is mistaken by many orders of magnitude.

> Surely, we will run such a simulation on a quantum computer in the near
> future.
>
> In quantum computations we are told you can't look at the processing while
> it is going on - this will cause decoherence, and effectively
> kill the computation. - So, we will have to wait for the end result ...
>
> So, until you have your end result (whatever you are
> looking for) - you won't know what is going on.
> We are also told that the qubits in our computer are logical
> equivalent to the particles they simulate. In the quantum world
> there is total breakdown between simulated and real.
>
> What then if some intermediate part of the simulation actually
> has livings things - including murder and horrors?
> We wouldn't know - because we can't look.  And
> if we looked we might not understand that we were indeed
> looking at such horrible things?
> Our focus is that "end result of the quantum simulation" a billions years
> later.

Quantum computers are likely to give significant speedups over
classical computers in only a narrow range of specialised problems,
like factoring.  (It has been rigorously proved that for a completely
general, "unstructured" computing task, they will *not* give an
exponential speedup.)

The advantages of monitoring intermediate results and diverting
resources to a different project if a particular program is going
nowhere interesting would far outweigh any advantage of letting a
quantum computer remain in a superposition.

Quantum computers might turn out to be useful for simulating the exact
behaviour of medium-scale quantum systems, but remember, if you want
to simulate the quantum-level behaviour of a portion of the universe,
your quantum computer is going to have to be bigger (with realistic
technology, many, many orders of magnitude bigger) than the portion of
the universe you're simulating.

> Surely, the universe is computing away. All the little
> virtual particles pop up from the void and are given as input
> to physical laws - whichs computes a new state. The universe
> is one giant quantum computer.
> So there is computation all over the place. To call
> it a simulation is then a matter of words?

You can describe anything you like as being a computation.  What this
thread is arguing about is whether there is a conscious being who has
deliberately arranged for this computation to take place.

It might well be that future civilisations gain the ability to create
pocket universes of various kinds, and to choose the laws of physics
in them to some degree (e.g. Gregory Benford's _Cosm_).  That is a
different scenario, though, to having computer-level control over what
happens in a simulation.

> Can we really imagine that a civilisation could make a thing
> (a bomb, a simulation - whatever) - and then not let it go off, so to speak?

Numerous commentators believed that a nuclear holocaust was inevitable
before the end of the 20th century.  Not every foolish thing that is
physically possible has to happen.

> >This particular current obsession might well end up as
> >obscure a notion as having a human slave  to be your chimney sweep.
> >It's not just that it's evil, in the long run it's also rather silly.
> >Unless you're either very dumb or very nasty, you can find other ways
> >to get the job done
>
> I see it more as - it is only now that we are beginning to understand
> simulations and what they really are. In the 1800hundreds
> Boltzmann and his friends thought gases were exiciting.
> We now have the S=Log W thing (or whatever it was) -
> on to new stuff :-)

Yes, and we will figure out smarter ways to get interesting answers
for most of the things we presently waste time simulating, or
contemplate simulating, and then "on to new stuff".  Everyone's
excited about this Christmas's toys (both real and imagined), but
there will be something else to obsess about long before we have the
technology to do the kind of dumb, boring things that people are
currently declaring to be "inevitable".

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX


From: Mark Atwood 
Organization: EasyNews, UseNet made Easy
Date: 18. april 2007 02:55
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

Greg Egan  writes:
> results without simulating boring and expensive details.  (Nobody
> simulates an ideal gas by modelling every individual molecule
> separately.)

I have.

Mainly as an exercise in writing a physics model, but still...

-- 
Mark Atwood                 When you do things right, people won't be sure
me@mark.atwood.name         you've done anything at all.
http://mark.atwood.name/   http://fallenpegasus.livejournal.com/

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Simon Laub
Date: 18. april 2007 14:21
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

Greg Egan"  wrote in message 
news:1176851031.262750.178900@o5g2000hsb.googlegroups.com...

>When a science makes *real* progress, it is not about
> buying a bigger computer and doing more brute-force computing, it is
> about getting smarter at calculating more results by making use of
> insights into the nature of the problem.

To me it seems that Science used to be - going back and forth between
theory and experiment (real world) - but for various reasons
simulations are coming in as ersatz for the real world.
I.e. a significant part of  the computing power
of the best computers in the world goes into simulating nuclear
bombs - as it is political incorrect to actual explode these things
(now one can make do with one real explosion every 10 years
to make sure the damn things ignites .. but the real weapons
development is through simulations) - and you would want to
simulate as close to the actual thing as possible ?

> And even in messy fields involving mass human behaviour, simulating
> the players down to the level where they'd be conscious would always
> be vast overkill.  We can glean enough about human behaviour from
> observations and historical records to abstract out useful higher-
> level rules; this is a non-trivial project, but anyone who imagines
> that simulating a billion people at *consciousness level* would be
> easier is mistaken by many orders of magnitude.

Oh, this is way into Master Asimovs Psychohistory. And his prime radiant .. 
:-)
Of course we should use Occams razor as much as possible. If stuff
is of no consequence for what we want to do - away it goes, we don't want to 
simulate that.
But sometimes the models are to simple for what we want to achieve  - take 
communism or
other political systems that had certain (to) simple models of what
humans are - then they fail.
What is too simple? Are parts of the human brain not really necessary
for our simulation purposes? -
If the simulation is about how society works,
perhaps the actual thing is as algebraic compessed as it can be -
you really need to simulate everything. Even though game theory -and what 
have
you- might provide a lot of input.. If evolution is a bitch - then certainly 
complexity
is also one (writing in 2007).

And btw. how do we really imagine to make the first supercomputers? Wouldn't
they be build as pretty extensive models of what we have and how people 
think?
The super coming from the possibility of speeding the thought processes
up (making it faster, not smarter) - and then connecting them to a society.
Where you would interface the society. Then we are already on the road to
running simulations of a lot of brains.

> Quantum computers might turn out to be useful for simulating the exact
> behaviour of medium-scale quantum systems, but remember, if you want
> to simulate the quantum-level behaviour of a portion of the universe,
> your quantum computer is going to have to be bigger (with realistic
> technology, many, many orders of magnitude bigger) than the portion of
> the universe you're simulating.

Which seems quite challenging for now ... But if we are serious about a 
Jupitor sized
quantum computer .... I am game for almost anything then.

> You can describe anything you like as being a computation.  What this
> thread is arguing about is whether there is a conscious being who has
> deliberately arranged for this computation to take place.

I am afraid I think there is intelligence implied in computation?
Somehow even the most mindless computations can end up being
complex and intelligent. If we can't answer how these
mindless computations got started - maybe, we might be able to answer
where they will end - Giving us a why and how, Tipler style. But ok thats 
normally way
outside Science, that doesn't like to talk about purpose and
direction (for good reasons).
However, you run a million simulations of this thing and they all
end up in the same corner ... maybe this leads you to conclude something?
Maybe you can't draw any conclusions - the simulations end up all
over the place - thats a sort of conclusion as well.
Maybe these vastly overkill simulations are necessary to get the insights
you need to process and predict in a smarter way. To get to that
new theory that would make the very extensive and complex simluations
unnesessary.

-Simon

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

From: Gerry Quinn 
Organization: Bindweed Entertainment Software
Date: 18. april 2007 15:31
NewsGroups: rec.arts.sf.written,rec.arts.sf.science,comp.society.futures
Subject: Re: The simulation argument in science fiction - and beyond

In article <1176851031.262750.178900@o5g2000hsb.googlegroups.com>, 
gregegan@netspace.net.au says...
> On Apr 18, 12:52 am, "Simon Laub"  wrote:
> [snip]
> > True, Evolution is potentially a bitch.
> > But, well - lets assume that we want to test a theory on the early
> > universe? And what is more natural than running a simulation.
> 
> When we "simulate" the early universe, or other physical processes,
> there is always a trade-off between detail, computing resources, and
> the most import factor:  how smart we are at computing important
> results without simulating boring and expensive details.  (Nobody
> simulates an ideal gas by modelling every individual molecule
> separately.)  When a science makes *real* progress, it is not about
> buying a bigger computer and doing more brute-force computing, it is
> about getting smarter at calculating more results by making use of
> insights into the nature of the problem.

Indeed - look at the current denate on global warming, which is largely 
due to the fact that simulating the climate is extremely difficult and 
unreliable.

- Gerry Quinn

    Source: geocities.com/sheepxpower/simulation

               ( geocities.com/sheepxpower)