![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
elfsImagine the following: You're watching ten television screens. Each one of them shows the exact same scene from the exact same show with the exact same actors, but each one is slightly different in some way: different camera angle, or different pitch and timing when lines are read, or so on. It might be ten different takes from the same taping session, but it isn't. Only one of those shows has actually been taped live; the other nine are variants in which the original scene was digitized and revised algorithmically.
Not exactly the stuff of science fiction. This is the art of machinima and while it's hardly automated it's a lot easier to automate parts of it than it was a decade ago, and the decade before that it didn't even exist.
So, let's take it into the realm of speculative fiction, you know, the kind that the laws of physics say can happen, we just haven't figured out how to do it yet. It might be possible someday to do with streams of consciousness what the above scenario does with light and action: be convincing to the viewer. This is the basis of much "upload" SF, from the absolutely terrible Lawnmower Man to the popular Battlestar Galactica and Matrix series, along with the more deeply considerate Diaspora.
So here's the basic argument of simulationism: given that nothing in the laws of physics as we know them says we can't create simulations of consciousness in a medium other than the grey meat between our ears, it is entirely likely that, at some point in the future, there will be more streams of consciousness experiencing customized simulations of our reality than there will be experiencing the reality you and I are currently sharing.
Therefore, if it is likely that you and I are living in a reality that has this capacity (a very high likelihood), it is equally likely that this has already happened and you and I are not living in the original reality.
The alternative is actually harder for me to believe: not only are we living in the Basement Universe, but we are the first species to do so. Ever. Since the beginning of eternity.
Wikipedia calls this the Simulation Argument and puts forth a number of criticisms. Some are valid, but I disagree that what the author puts forth as the "most likely" contrapremise doesn't fly: that no civilization, ever, develops the technology necessary to do a simulation. The technology exists: it works with that quick-and-dirty hack called the brain, dependent upon chemistry to operate correctly. The trick is figuring out how to do it in a more efficient medium.
Not exactly the stuff of science fiction. This is the art of machinima and while it's hardly automated it's a lot easier to automate parts of it than it was a decade ago, and the decade before that it didn't even exist.
So, let's take it into the realm of speculative fiction, you know, the kind that the laws of physics say can happen, we just haven't figured out how to do it yet. It might be possible someday to do with streams of consciousness what the above scenario does with light and action: be convincing to the viewer. This is the basis of much "upload" SF, from the absolutely terrible Lawnmower Man to the popular Battlestar Galactica and Matrix series, along with the more deeply considerate Diaspora.
So here's the basic argument of simulationism: given that nothing in the laws of physics as we know them says we can't create simulations of consciousness in a medium other than the grey meat between our ears, it is entirely likely that, at some point in the future, there will be more streams of consciousness experiencing customized simulations of our reality than there will be experiencing the reality you and I are currently sharing.
Therefore, if it is likely that you and I are living in a reality that has this capacity (a very high likelihood), it is equally likely that this has already happened and you and I are not living in the original reality.
The alternative is actually harder for me to believe: not only are we living in the Basement Universe, but we are the first species to do so. Ever. Since the beginning of eternity.
Wikipedia calls this the Simulation Argument and puts forth a number of criticisms. Some are valid, but I disagree that what the author puts forth as the "most likely" contrapremise doesn't fly: that no civilization, ever, develops the technology necessary to do a simulation. The technology exists: it works with that quick-and-dirty hack called the brain, dependent upon chemistry to operate correctly. The trick is figuring out how to do it in a more efficient medium.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2006-09-26 01:12 am (UTC)What's wrong with the concept that any species capable of simulating reality will do so for a time, and then eventually discard it as they move to other newer distractions with different dynamics? Or is there an unwritten assumption that, if simulationism were possible, it would universally not be desirable or possible to do anything else?
no subject
Date: 2006-09-26 02:17 am (UTC)no subject
Date: 2006-09-26 03:02 pm (UTC)Obviously, the universe is "running": atoms and their constituents continue to interact to the beat of Planck-scale time quanta, and the consequences of that interaction are visible at our scale. Borstrom's point is, I believe, straightforward: if it's "running," it's running on something, and given the fundamental point that there is no reason to believe we could not someday soon create similar running environments, there's also no reason not to suppose we aren't already in a created environment.
Who'd bother simulating MY life?
Date: 2006-09-26 03:37 am (UTC)First, this kind of simulation would require computers with fifteen to twenty orders of magnitude higher performance than today's machines. It is not clear that this kind of performance is possible given the interaction of speed-of-light delays and minimum feature-size requirements.
A current system is at least nine orders of magnitude too slow to simulate the visual field of a single human eye using the best known methods (which, I must admit, suggests we may see this milestone reached, since 3D graphics chips have been getting about 1,000 times faster every five years since about 1993).
Simulating the other human senses-- of which there are at least ten in total-- is probably easier but doing it all with imperceptible synchronization errors plus imperceptible chained reactive latencies would require orders of magnitude more total performance. Cosimulation of multiple virtual persons for ordinary social interactions such as football games and orgies would obviously multiply the processing effort. (Just keeping you awake. :-)
Second, we'd have to assume that our simulation already comprehends every scientific experiment we might perform so that it doesn't give apparently impossible results. That last requirement puts a practical limit on the level of scientific advancement that can be simulated, and that leads us to a sort of contradiction-- a society with a level of technology otherwise sufficient for simulating reality couldn't simulate itself.
Ultimately this whole idea is just silly. It's just as hard to believe that anyone would create a simulation in which I'd be living this life as it is to believe that a divine Creator would create a world in which I'd be living this life. And I really would rather not think that anyone would create a simulation in which billions of (real or simulated) entities are forced to live and die in pain. It could happen, but I prefer to live my life as if nobody's intention is controlling it.
. png
Re: Who'd bother simulating MY life?
Date: 2006-09-26 03:05 pm (UTC)There's this running hubris throughout many of the comments that the simulation is for our benefit, or at the very least is "about us." Unfortunately, there's no reason to believe this. The consistency of the simulation may simply be a side effect of the rule set.
The real question is...
Date: 2006-09-26 06:52 am (UTC)