Conditional on this universe being a simulation, the universe doing the stimulating has laws vastly different from our own. For example, it might contain more than 3 extended-spacial dimensions, or bear a similar relation to our universe as our universe does to second life. 99.999%
I was originally going to include an additional 9, but decided I should compensate for overconfidence bias. :)
But, seriously, I don’t understand why people are so reluctant to quote large probabilities. For some statements, e.g., “the sun will rise tomorrow”, 99.999% seems way underconfident.
I wouldn’t have said the number of nines indicated overconfidence if you were talking about the sun rising. I do not believe you have enough evidence to reach that level of certainty on this subject. I would include multiple nines in my declaration of confidence in that claim.
Upvoted for disagreement. The most detailed simulations our current technology is used to create (namely, large networks of computers operating in parallel) are created for research purposes, to understand our own universe better. Galaxy/star formation, protein folding, etc. are fields where we understand enough to make a simulation but not enough that such a simulation is without value. A lot of our video games have three spatial dimensions, one temporal one, and roughly Newtonian physics. Even Second Life (which you named in your post) is designed to resemble our universe in certain aspects.
Basically, I fail to see why anyone would create such a detailed simulation if it bore absolutely no resemblance to reality. Some small differences, yes (I bet quantum mechanics works differently), but I would give a ~50% chance that, conditional on our universe being a simulation, the parent universe has 3 spatial dimensions, one temporal dimension, matter and antimatter, and something that approximates to General Relativity.
This is much less than obvious—if the parent universe has sufficient resources, it’s entirely plausible that it would include detailed simulations for fun—art or gaming or some costly motivation that we don’t have.
Considering how much stuff like convays game of life which bears no resemblance to our universe is played I’d put the probability much lower.
Whenever you run anything which simulates anything turing compatible (Ok. Finite state machine is actually enough due to finite amount of information storage even in our universe) there is a chance for practically anything to happen.
Basically, I fail to see why anyone would create such a detailed simulation if it bore absolutely no resemblance to reality.
I have seen simulators of Conway’s Game of Life (or similar) that contain very complex things, including an actual Turing machine.
I could see someone creating a simulator for CGL that simulates a Turing machine that simulates a universe like ours, at least as a proof of concept. With ridiculous amounts of computation available I’m quite sure they’d run the inner universe for a few billion of years.
If by accident a civilization arises in the bottom universe and they found some way of “looking above” they’d find a CGL universe before finding the one similar to theirs.
I’m supposed to downvote if I think the probability of that is >= 99.999% and upvote otherwise? I’m upvoting, but I still the probability of that is > 90%.
Well, someone might agree with wedrifid (that second-order comments are to be voted on normally) but still disapprove of his comment for reasons other than disagreement (for example, think it clarifies what would otherwise have been a valuable point of confusion), and downvote (normally) on that basis.
Okay, given the confusion, I’ve retracted my downvote. I’ve also seen a comment get about 27 karma on this thread replying to another post, and that comment was certainly not massively irrational, so I assume we vote normally if it’s not a first-order comment.
I disagree with this one more than any other comment by far. Have you looked into Tegmark level 4 cosmology? It’s really important to take into account concepts like measure and the utility functions of likely simulating agents when reasoning about this kind of thing. Upvoted.
My reasoning is that it would take more then a universe’s worth of computronium to completely stimulate a comprable universe.
One could argue that they’re taking shortcuts with, e.g., the statistics of bulk matter, but I think we’d notice the edge cases caused by something like that.
Conditional on this universe being a simulation, the universe doing the stimulating has laws vastly different from our own. For example, it might contain more than 3 extended-spacial dimensions, or bear a similar relation to our universe as our universe does to second life. 99.999%
Upvoted for excessive use of nines. :)
(ie. Gross overcondidence.)
I was originally going to include an additional 9, but decided I should compensate for overconfidence bias. :)
But, seriously, I don’t understand why people are so reluctant to quote large probabilities. For some statements, e.g., “the sun will rise tomorrow”, 99.999% seems way underconfident.
I wouldn’t have said the number of nines indicated overconfidence if you were talking about the sun rising. I do not believe you have enough evidence to reach that level of certainty on this subject. I would include multiple nines in my declaration of confidence in that claim.
You think there’s a 999,999⁄100,000 chance the sun will rise tomorrow? I think you may be overconfident here...
Upvoted for disagreement. The most detailed simulations our current technology is used to create (namely, large networks of computers operating in parallel) are created for research purposes, to understand our own universe better. Galaxy/star formation, protein folding, etc. are fields where we understand enough to make a simulation but not enough that such a simulation is without value. A lot of our video games have three spatial dimensions, one temporal one, and roughly Newtonian physics. Even Second Life (which you named in your post) is designed to resemble our universe in certain aspects.
Basically, I fail to see why anyone would create such a detailed simulation if it bore absolutely no resemblance to reality. Some small differences, yes (I bet quantum mechanics works differently), but I would give a ~50% chance that, conditional on our universe being a simulation, the parent universe has 3 spatial dimensions, one temporal dimension, matter and antimatter, and something that approximates to General Relativity.
This is much less than obvious—if the parent universe has sufficient resources, it’s entirely plausible that it would include detailed simulations for fun—art or gaming or some costly motivation that we don’t have.
True. I would estimate that our universe resembles the parent universe with probability ~50%.
Considering how much stuff like convays game of life which bears no resemblance to our universe is played I’d put the probability much lower.
Whenever you run anything which simulates anything turing compatible (Ok. Finite state machine is actually enough due to finite amount of information storage even in our universe) there is a chance for practically anything to happen.
I have seen simulators of Conway’s Game of Life (or similar) that contain very complex things, including an actual Turing machine.
I could see someone creating a simulator for CGL that simulates a Turing machine that simulates a universe like ours, at least as a proof of concept. With ridiculous amounts of computation available I’m quite sure they’d run the inner universe for a few billion of years.
If by accident a civilization arises in the bottom universe and they found some way of “looking above” they’d find a CGL universe before finding the one similar to theirs.
I’m supposed to downvote if I think the probability of that is >= 99.999% and upvote otherwise? I’m upvoting, but I still the probability of that is > 90%.
Army1987: Not sure what the rules are for comments replying to the original, but hell. Voted down for agreement.
(I think we just vote normally in these replies. I agree with army too.)
Why in the world would the parent be downvoted? I’m having difficulty unraveling the paradox.
Well, someone might agree with wedrifid (that second-order comments are to be voted on normally) but still disapprove of his comment for reasons other than disagreement (for example, think it clarifies what would otherwise have been a valuable point of confusion), and downvote (normally) on that basis.
Okay, given the confusion, I’ve retracted my downvote. I’ve also seen a comment get about 27 karma on this thread replying to another post, and that comment was certainly not massively irrational, so I assume we vote normally if it’s not a first-order comment.
I’m not actually sure why there was ever confusion. From the OP: “comment voting works normally for comment replies to other comments.”
I’d be with you with that much confidence if the proposition were “the top layer of reality has laws vastly different from our own.”
One level up, there’s surely at least an 0.1% chance that Snowyowl is right.
I disagree with this one more than any other comment by far. Have you looked into Tegmark level 4 cosmology? It’s really important to take into account concepts like measure and the utility functions of likely simulating agents when reasoning about this kind of thing. Upvoted.
My reasoning is that it would take more then a universe’s worth of computronium to completely stimulate a comprable universe.
One could argue that they’re taking shortcuts with, e.g., the statistics of bulk matter, but I think we’d notice the edge cases caused by something like that.
In realtime, maybe, but what if we’re running at one simulated planck time per many time units of calculation?