Nick Bostrom once argued that you must choose between three options re the possibility that you are now actually living in and experiencing a simulation created by future folks to explore their past: (A) its true, you are most likely a sim person living in a sim, either of this sort or another, (B) future folk will never be able to do this, because it just isn’t possible, they die first, or they never get rich and able enough, or (C) future folk can do this, but they do not choose to do it much, so that most people experiencing a world like yours are real humans now, not future sim people.
Ancient Rome, Ancient Egypt, and Sparta/Ancient Greece are about equal in interest to the Renaissance American civil war and WW1 judging by google trends. Middle ages and ww2 are 3 an 10 times more popular. Napoleonic wars and dark ages and battle of agincourt are a sixth as popular. Seems more like there's a bunch of interest placed in a bunch of times. And we aren't sure how popular recent events will be in a century or 2. Don't get me wrong little interest pre writing , and this is comparing eras wars and battles, but it seems random.
Would honestly be interesting to get an interest per year metric. However the simulation argument works even if you're the only concious being, or we're in a multiplayer game stocked with people who just wanted to forget their lives for a day. Then it's just % time spent playing video games or number of characters made if the probability is per independent memory, rather then observer moments. And yeah people would prefer more interesting stuff to do, but we don't know what people will find interesting in the future.If we use how people use video games, movies , books or live their lives as an analogy; yeah some people are only going to want to be millionaires or billionaires, some people are going to want to be a general in a war,or soldiers in a war if they're thrill seekers/masochists, people might only want to be a king or some sort of leader. However there's also people who idealise a simple life, some people like a challenge or want to earn becoming wealthy.
Re: "Why isn't the answer (D) the cost of simulations is greater than the the cost of the base reality, and so any simulation en masse will always be at a lower resolution?"
Maybe look into "historical reenactment". High res sims do happen. However, if we are in a sim, it is probably a relatively low-resolution sim. Nobody claimed that the original world was a fuzzy as this world is.
Re: "we spend very little time today doing any sort of simulation of typical farming or forager-era folks, for example." - it isn't clear that this is true. Jane Eyre is currently being pretty widely simulated, for example. Resources we spend on historical sims are significant. If our civilization scales up, it is not a big stretch to think that sim histories could exceed actual ones. Expecially so for those who live in interesting times.
I don't understand your answer. You're an interesting person who has produced a lot of material that could be used to fit a reasonably accurate model of your mind. No one in the whole future of the universe will ever decide to make a Robin Hanson simulation? There only needs to be 1 for you to have a 50% chance of being simulated. So, it seems like you should be pretty confident that you are being simulated A random nobody maybe a bit less. A really consequential figure like Donald Trump or Elon Musk should be absolutely certain that they are in a simulation.
Why isn't the answer (D) the cost of simulations is greater than the the cost of the base reality, and so any simulation en masse will always be at a lower resolution?
We already know what being in a simulation feels like. We do it in our dreams, and a lot of us experience up to a third of our lives in it (not even counting day dreams).
But we can usually tell dreams from reality (even if not at that very moment) because they're always at a lower resolution. I don't see the probability of high resolution simulations being that much greater than the probability of base reality.
Even to support our dreaming, we have to eat, which means we have to work, which means we need activity time in base reality (or at least a less costly simulation). The lowest cost simulation is base reality. The amount of experienced reality * resolution will always be greatest at the lowest level.
Ah, well, here you find a weak joint in the argument: There is not yet any scale model recreation of human society. That seems beyond present technology, so the argument that predictions of the "realness" of our universe should take scale model options into count faces the counter-argument, "What scale models?!"
That conceded, computer sims are hardly ready to recreate the universe, either. Also: the examples of wind tunnels and wave tanks show that there can be very large incentives to recreate even rather smallish (if critical) parts of the universe at reduced scale.
Agreed that a future society--one with scale universe technology--would likely be most interested in modeling itself, rather than us. However, the very nature of the metaphor suggests that it is likely they will have to recreate us to get to them. A wind tunnel has to rev up; a wave tank takes time to start sloshing. So too might we be living now in a scale model created by (possible) future humans waiting for us to mature into them.
It could be made clear that these are being lumped together.
I really think the survey's next to useless if people don't understand the distinction. I suspect quite a lot of people are filing it under D - Bostrom is wrong.
Also, from a skim of Bostrom's abstract, he doesn't seem to have seen the distinction or issue himself.
Is this something you think he's addressed/included?
Thanks. I now guess that in your initial response you were guessing that there are levels of behavioral detail which our progeny will frequently want to simulate and whose simulation can most easily be done by (or perhaps even requires) also simulating consciousness. In which case my response would be D, because a guess isn't an argument.
Hmm I don't see how this answers my question, which I will try rephrasing: Why do you assume that as capabilities expand it won't be cheaper to simulate human behavior, than human behavior + consciousness?
I gather based on your answer above that you think that the only way to simulate B is to simulate B + C , then change the config setting from B + C to B, and that such a config setting wnt ever be implementable, as we don't know how to do that now..
But why do you assume that's the only way to simulate B (*and* that even if it were, that future capabilities would be limited by our current knowledge)?
Ancient Rome, Ancient Egypt, and Sparta/Ancient Greece are about equal in interest to the Renaissance American civil war and WW1 judging by google trends. Middle ages and ww2 are 3 an 10 times more popular. Napoleonic wars and dark ages and battle of agincourt are a sixth as popular. Seems more like there's a bunch of interest placed in a bunch of times. And we aren't sure how popular recent events will be in a century or 2. Don't get me wrong little interest pre writing , and this is comparing eras wars and battles, but it seems random.
Would honestly be interesting to get an interest per year metric. However the simulation argument works even if you're the only concious being, or we're in a multiplayer game stocked with people who just wanted to forget their lives for a day. Then it's just % time spent playing video games or number of characters made if the probability is per independent memory, rather then observer moments. And yeah people would prefer more interesting stuff to do, but we don't know what people will find interesting in the future.If we use how people use video games, movies , books or live their lives as an analogy; yeah some people are only going to want to be millionaires or billionaires, some people are going to want to be a general in a war,or soldiers in a war if they're thrill seekers/masochists, people might only want to be a king or some sort of leader. However there's also people who idealise a simple life, some people like a challenge or want to earn becoming wealthy.
Re: "Why isn't the answer (D) the cost of simulations is greater than the the cost of the base reality, and so any simulation en masse will always be at a lower resolution?"
Maybe look into "historical reenactment". High res sims do happen. However, if we are in a sim, it is probably a relatively low-resolution sim. Nobody claimed that the original world was a fuzzy as this world is.
Re: "we spend very little time today doing any sort of simulation of typical farming or forager-era folks, for example." - it isn't clear that this is true. Jane Eyre is currently being pretty widely simulated, for example. Resources we spend on historical sims are significant. If our civilization scales up, it is not a big stretch to think that sim histories could exceed actual ones. Expecially so for those who live in interesting times.
I don't understand your answer. You're an interesting person who has produced a lot of material that could be used to fit a reasonably accurate model of your mind. No one in the whole future of the universe will ever decide to make a Robin Hanson simulation? There only needs to be 1 for you to have a 50% chance of being simulated. So, it seems like you should be pretty confident that you are being simulated A random nobody maybe a bit less. A really consequential figure like Donald Trump or Elon Musk should be absolutely certain that they are in a simulation.
Why isn't the answer (D) the cost of simulations is greater than the the cost of the base reality, and so any simulation en masse will always be at a lower resolution?
We already know what being in a simulation feels like. We do it in our dreams, and a lot of us experience up to a third of our lives in it (not even counting day dreams).
But we can usually tell dreams from reality (even if not at that very moment) because they're always at a lower resolution. I don't see the probability of high resolution simulations being that much greater than the probability of base reality.
Even to support our dreaming, we have to eat, which means we have to work, which means we need activity time in base reality (or at least a less costly simulation). The lowest cost simulation is base reality. The amount of experienced reality * resolution will always be greatest at the lowest level.
This was the punchline of Charles Stross' excellent SF story "Missile Gap" iirc.
Ah, well, here you find a weak joint in the argument: There is not yet any scale model recreation of human society. That seems beyond present technology, so the argument that predictions of the "realness" of our universe should take scale model options into count faces the counter-argument, "What scale models?!"
That conceded, computer sims are hardly ready to recreate the universe, either. Also: the examples of wind tunnels and wave tanks show that there can be very large incentives to recreate even rather smallish (if critical) parts of the universe at reduced scale.
Agreed that a future society--one with scale universe technology--would likely be most interested in modeling itself, rather than us. However, the very nature of the metaphor suggests that it is likely they will have to recreate us to get to them. A wind tunnel has to rev up; a wave tank takes time to start sloshing. So too might we be living now in a scale model created by (possible) future humans waiting for us to mature into them.
This doesn't seem like a reply to what Dave wrote!
And very much in the theme of avoiding reasonable questions/points etc. :)
It could be made clear that these are being lumped together.
I really think the survey's next to useless if people don't understand the distinction. I suspect quite a lot of people are filing it under D - Bostrom is wrong.
Also, from a skim of Bostrom's abstract, he doesn't seem to have seen the distinction or issue himself.
Is this something you think he's addressed/included?
Thanks. I now guess that in your initial response you were guessing that there are levels of behavioral detail which our progeny will frequently want to simulate and whose simulation can most easily be done by (or perhaps even requires) also simulating consciousness. In which case my response would be D, because a guess isn't an argument.
I've never said or implied that high detail sims are the only possible or useful sims;
With only 4 options in Twitter polls, there's no way not to lump options together.
Hmm I don't see how this answers my question, which I will try rephrasing: Why do you assume that as capabilities expand it won't be cheaper to simulate human behavior, than human behavior + consciousness?
I gather based on your answer above that you think that the only way to simulate B is to simulate B + C , then change the config setting from B + C to B, and that such a config setting wnt ever be implementable, as we don't know how to do that now..
But why do you assume that's the only way to simulate B (*and* that even if it were, that future capabilities would be limited by our current knowledge)?
That doesn't mean they're all wrong.
We can't reject all simply because they're mostly wrong. We have to evaluate each claim on its merits.
But everyone has always found it too easy to see they & associates as special.