3 ms·

What I would like to discuss is what purpose would simulating a universe have for the advanced civilisation? As the civilisation becomes closer to being capabl

by polshaw 14y ago

What I would like to discuss is what purpose would simulating a universe have for the advanced civilisation? As the civilisation becomes closer to being capable of creating a perfectly simulated universe, they get less benefit from having one. If we assume the universe we are in is at least similar to 'the first', then the existence of entropy / thermodynamic laws would surely create a significant cost to universe simulation. So I can't see it being done without good reason (eg as a toy), and as the master would have to understand all the factors involved in the universe already to create the simulation, so what would they gain? Is it right to say that it would be impossible to simulate the universe at full speed? (ie. the computer would need to be powerful enough to simulate itself and more (ie. the rest of the universe)). If so, this would not only make it less useful, but put significant (increasing) limits on the local age of the descendent universes; assuming (fairly, i think) that the master universe does not have an infinite existence with infinite time between shocks. Also, how would it be possible to store all the information for a universe within itself? Or perhaps we are talking about imperfect simulations, which would make it much more feasible? but then the infinite exponential chain would not be possible as each child would become more crude. Or perhaps only a (diminishing) proportion of space(-time) is simulated.. or maybe the speed of light decreases as simulation levels get deeper? In writing this I think I have come to the conclusion that it is impossible (in so far as anything is given our imperfect knowledge), at least a complete universe simulation or an endless chain, but I would love for someone to show otherwise.

jonnathanson 14y ago
"What I would like to discuss is what purpose would simulating a universe have for the advanced civilisation?" The Douglas Adams-style answer would be that we're some kid's physics homework. Perhaps, in a sufficiently advanced civilization, building a working computer model of an advanced universe is child's play. An alternative is that the simulators inhabit a very different universe from our own, and that our simulation is something entirely new. It's not a replica of their universe; it's a completely different creation. They could be studying various scenarios, of which ours is just one. We need to bear in mind that we're approaching this problem from our own, subjective understanding of our universe. It's not a given that the simulators' universe is anything like ours.
- polshaw 14y ago
  If we are to say it is a perfect or near perfect simulation ('copy') then I don't see how, as stated in my first post. It's interesting to consider if we were a different simulation. But it would still have the same limits (that the sum of the activity ('reality') in each simulation must be less large/complex than it's host), prohibiting the 'infinite simulations' concept. It surely isn't necessary that we are similar to a host universe, but I wouldn't you say it was a necessary part of this theory?; because it all starts with the vastness of our own space(-time) as the justification for the likelihood of super-advanced civilisations. You could argue that the size of our universe implies simulations must happen, and that if our universe were a simulation the host must be larger, i suppose. If the simulations are not similar to their host, though, the foundation suggestion that a simulation could be our universe seems a little more shaky (as there would be no direct theory ('evidence') that a universe like ours could be simulated). There is perhaps a semantic issue too.. if somewhere in this universe we are simulating something that works on different rules to our universe, is it in fact a universe, or what makes it so?
  - jonnathanson 14y ago
    "It surely isn't necessary that we are similar to a host universe, but I wouldn't you say it was a necessary part of this theory?" I don't think it's a necessary part of the theory, but it's a necessary part of the test we would need to conduct. And given that such a test is simply testing for inconsistencies, artifacts and bugs inherent to our own universe/simulation, it's possible to detect whether we're in a simulation without having to test for the precise nature of it (or the nature of the simulators). But our tests are inherently being conducted from our own frame of reference, with our own understanding of what should and shouldn't be found. Ergo, we can prove a simulation by finding artifacts of one as recognizable to us, but we can't necessarily disprove a simulation by not finding artifacts. That being said, the proper way to think about the test wouldn't be "Physicists are testing to determine if we're in a simulation." It would be "Physicists are testing to find evidence that we're in a simulation."
ep103 14y ago
You started this post asking one question, and ended it asking a different (IMO much more interesting) question. To the first one though, so long as politics and war exist, there would always be reason to attempt to run simulations, regardless of the degree of advancement of a civilization. By the By, Iain Banks plays around / has a recurring thread of simulations and high powered computers in his culture series, if you enjoy sci fi.
- polshaw 14y ago
  I absolutely agree, it got much more interesting as I started fleshing out my thought process. As for the motives, for a simulation to be helpful, it would have to run at greater than real time, no? Which.. is surely impossible, at least for a complete 'perfect' simulation. Right? Imperfections (and the limit of a simulation having to be simpler than its host) would infer a finite chain of simulations, which would vastly reduce the likelihood of us existing in a simulation. Also if these wars (etc) exist, the simulation is going to be knocked offline at some point, destroying all descendant universes, shrinking the pool of possible existent universes much further still. So (AFAICT) we have gone from it being almost a certainty we are part of a simulation (only one true universe from infinite similar simulations), to being very unlikely. I may check out the culture series, but i'm not one for reading much fiction, sadly.. impatient internet world that has formed me and all.
  - ep103 14y ago
    > it would have to run at greater than real time, no? Not for the people in the simulation : ) We'd just be self-conscious byte code. Which actually is the question Iain Banks brings up. They want to run simulations on how a society will respond to a political action, but that would imply creating self-conscious byte-code in the simulation. And wouldn't that be genocide when they turn off the sim? So maybe the only ethical thing is to never turn off the sim...

by polshaw 14y ago

jonnathanson 14y ago

"What I would like to discuss is what purpose would simulating a universe have for the advanced civilisation?" The Douglas Adams-style answer would be that we're some kid's physics homework. Perhaps, in a sufficiently advanced civilization, building a working computer model of an advanced universe is child's play. An alternative is that the simulators inhabit a very different universe from our own, and that our simulation is something entirely new. It's not a replica of their universe; it's a completely different creation. They could be studying various scenarios, of which ours is just one. We need to bear in mind that we're approaching this problem from our own, subjective understanding of our universe. It's not a given that the simulators' universe is anything like ours.

polshaw 14y ago
If we are to say it is a perfect or near perfect simulation ('copy') then I don't see how, as stated in my first post. It's interesting to consider if we were a different simulation. But it would still have the same limits (that the sum of the activity ('reality') in each simulation must be less large/complex than it's host), prohibiting the 'infinite simulations' concept. It surely isn't necessary that we are similar to a host universe, but I wouldn't you say it was a necessary part of this theory?; because it all starts with the vastness of our own space(-time) as the justification for the likelihood of super-advanced civilisations. You could argue that the size of our universe implies simulations must happen, and that if our universe were a simulation the host must be larger, i suppose. If the simulations are not similar to their host, though, the foundation suggestion that a simulation could be our universe seems a little more shaky (as there would be no direct theory ('evidence') that a universe like ours could be simulated). There is perhaps a semantic issue too.. if somewhere in this universe we are simulating something that works on different rules to our universe, is it in fact a universe, or what makes it so?
- jonnathanson 14y ago
  "It surely isn't necessary that we are similar to a host universe, but I wouldn't you say it was a necessary part of this theory?" I don't think it's a necessary part of the theory, but it's a necessary part of the test we would need to conduct. And given that such a test is simply testing for inconsistencies, artifacts and bugs inherent to our own universe/simulation, it's possible to detect whether we're in a simulation without having to test for the precise nature of it (or the nature of the simulators). But our tests are inherently being conducted from our own frame of reference, with our own understanding of what should and shouldn't be found. Ergo, we can prove a simulation by finding artifacts of one as recognizable to us, but we can't necessarily disprove a simulation by not finding artifacts. That being said, the proper way to think about the test wouldn't be "Physicists are testing to determine if we're in a simulation." It would be "Physicists are testing to find evidence that we're in a simulation."

ep103 14y ago

You started this post asking one question, and ended it asking a different (IMO much more interesting) question. To the first one though, so long as politics and war exist, there would always be reason to attempt to run simulations, regardless of the degree of advancement of a civilization. By the By, Iain Banks plays around / has a recurring thread of simulations and high powered computers in his culture series, if you enjoy sci fi.

polshaw 14y ago
I absolutely agree, it got much more interesting as I started fleshing out my thought process. As for the motives, for a simulation to be helpful, it would have to run at greater than real time, no? Which.. is surely impossible, at least for a complete 'perfect' simulation. Right? Imperfections (and the limit of a simulation having to be simpler than its host) would infer a finite chain of simulations, which would vastly reduce the likelihood of us existing in a simulation. Also if these wars (etc) exist, the simulation is going to be knocked offline at some point, destroying all descendant universes, shrinking the pool of possible existent universes much further still. So (AFAICT) we have gone from it being almost a certainty we are part of a simulation (only one true universe from infinite similar simulations), to being very unlikely. I may check out the culture series, but i'm not one for reading much fiction, sadly.. impatient internet world that has formed me and all.
- ep103 14y ago
  > it would have to run at greater than real time, no? Not for the people in the simulation : ) We'd just be self-conscious byte code. Which actually is the question Iain Banks brings up. They want to run simulations on how a society will respond to a political action, but that would imply creating self-conscious byte-code in the simulation. And wouldn't that be genocide when they turn off the sim? So maybe the only ethical thing is to never turn off the sim...