and yet there is no trace of anyone 'programming' or 'guiding' us anywhere. — jasonm

I'm unsure this is an argument. It may also be empirically wrong, and we misdescribe, or mislabel the evidence of such.

Similarly, why don't we sometimes notice violations of the laws of physics? — jasonm

A huge number of people claim this is so. I can't possibly vet every claim. So, it's an open question.

Are we talking about computers that are bigger than the universe itself? Is this possible even in principle? — jasonm

I'm not sure where the intimation comes from. In principle, there is no reason why we couldn't produce simulations either larger, or more complex than our own. Seems implausible, though.

Nevertheless, I think the best answer comes from Occam's Razor: "Explanations that posit fewer entities, or fewer kinds of entities, are to be preferred to explanations that posit more." — jasonm

I am not entirely convinced this is the best Razor to use when it comes to speculative cosmologies. It seems fairly clear that if anything other than materialism via random distribution at the big bang is true, it must necessarily be a theory which Occam would reject, prima facie.

Also, I think many do not realize that the “God hypothesis” has come back in a stealthy sort a way. Instead of the watch needing a designer, the simulation needs a simulator. — Richard B

Good point.
Come to think of it, the argument could be seen as a version of Berkeley's "proof" of God.

If we are a simulation and there is a world outside ours, how would we know what is possible? Since we know nothing of the world outside the simulation, we don't even know if it is done via computers. Would it not be a mistake to assume that what applies in our world applies outside it? This seems an odd position to take. — Tom Storm

It rather depends on what your project is. If the project is to make a space for fantasies, then the fact that we don't know is an opportunity, not a problem. The point where knowledge runs out is always an opportunity for myth to fill the gap. This is a myth.

"We are living in a 'simulation' and such a virtual world is the same as the 'real world' in every respect, except that it is simulated and therefore 'not real." I have a few arguments against this notion: — jasonm

From the viewpoint of fish in an aquarium, is their existence a simulated life, in that the aquarium simulates the ocean, or is it a real life, in that their environment is all they know. As with every life-form, their lives are both simulations and real. Takeaway pizzas simulate real food, social media simulates real life, sports events simulate medieval battles, surveillance cameras simulate the nosy neighbour in a small village, modern government tries to simulate the parents in a traditional family (whether father of the nation or mamala to the people) and corporate employers simulate the process of having to work for essential food and shelter.

As regards argument 1), the designers do pop out at times, but the life-form is not aware because of their necessarily limited intelligence.
As regards 2), similarly, the life-form doesn't notice violations in the laws of physics also because of their necessarily limited intelligence
As regards 3), the same amount of computing power would be required to house a "simulated" world as a "real" world.
As regards 4), the simplest explanation for fish in an aquarium is that they exist in a simulated world

You may respond that humans are the supreme intelligence in the universe and are all knowing, yet humans have only been around for 2.8 million years whilst fish have managed to successfully survive in a hostile world for more than 500 million years, regardless of whether this world happens to be "simulated" or "real".

If we are a simulation and there is a world outside ours, how would we know what is possible? Since we know nothing of the world outside the simulation, we don't even know if it is done via computers. Would it not be a mistake to assume that what applies in our world applies outside it? This seems an odd position to take. — Tom Storm

It rather depends on what your project is. If the project is to make a space for fantasies, then the fact that we don't know is an opportunity, not a problem. Myths have always been stories told where knowledge was not possible. This is a myth.

From the viewpoint of fish in an aquarium, is their existence a simulated life, in that the aquarium simulates the ocean, or is it a real life, in that their environment is all they know. — RussellA

Both make some sort of sense. It would also be possible to say that the aquarium is a little bit of the ocean. It would also be possible to say that, since the water they swim in and the food they eat are both real that their world, though small, is not simulated.

Takeaway pizzas simulate real food, social media simulates real life, sports events simulate medieval battles, surveillance cameras simulate the nosy neighbour in a small village, modern government tries to simulate the parents in a traditional family (whether father of the nation or mamala to the people) and corporate employers simulate the process of having to work for essential food and shelter. — RussellA

You are confusing "simulate" with "is like".

As regards argument 1), the designers do pop out at times, but the life-form is not aware because of their necessarily limited intelligence.
As regards 2), similarly, the life-form doesn't notice violations in the laws of physics also because of their necessarily limited intelligence
As regards 3), the same amount of computing power would be required to house a "simulated" world as a "real" world.
As regards 4), the simplest explanation for fish in an aquarium is that they exist in a simulated world — RussellA

1) Do you have any evidence for that?
2) Do you have any evidence for that?
3) The real world does not need to be housed and therefore does not require any computing power.
4) What is the question the fish are asking? If there's no question, there's no need for an explanation.

You may respond that humans are the supreme intelligence in the universe and are all knowing, yet humans have only been around for 2.8 million years whilst fish have managed to successfully survive in a hostile world for more than 500 million years, regardless of whether this world happens to be "simulated" or "real". — RussellA

What does that have to do with anything?

Bostrom's Simulation Argument is that one of these is almost certainly true:

1. The fraction of human-level civilizations that reach a posthuman stage (that is, one capable of running high-fidelity ancestor simulations) is very close to zero, or
2. The fraction of posthuman civilizations that are interested in running simulations of their evolutionary history, or variations thereof, is very close to zero, or
3. The fraction of all people with our kind of experiences that are living in a simulation is very close to one.

He then argues that if (3) is true then we are almost certainly living in a simulation.

Surely the problem is the one frequently pointed out, with the word "simulate" being ambiguous between "describe or theoretically model" and "physically replicate or approximate". — bongo fury

You know, map vs replicated territory.

This

high-fidelity ancestor simulations — Michael

being a good example. Amazingly detailed descriptions/theoretical models of ancestors; or physical replicas/approximations of them?

Or something else?

Artificial consciousnesses programmatically fed phenomenal experience, e.g. man-made brains-in-a-vat.

The first is that the whole of our world could not be simulated, because the hardware would have to be bigger than the whole (real) world. — Ludwig V

Yes, the world would have to be bounded, probably more than once. Bostrom for instance suggests the detailed simulation be bounded at human brains (all of them). A less detailed simulation of bodies, animals (all animals will apparently be NPCs), purposeful devices and such. Probably at least 5 levels of this, ending with 'everything else' which simulates the stars in the sky and such, more in detail only when purposefully being paid attention to.
This makes it hard to see physics be different here from there since the physics of a thing changes whenever you try to investigate it.

The second is that exact simulation of even a small part of the real world, down to sub-atomic and near-light-speed events could not be constructed, for the same reason.

It has to be done at that level if someone is paying attention to it. But you choose an easy interpretation like Copenhagen, and it's usually only one particle (like the electron being sent through the double slits) that has to be simulated.
Consider smoke detectors, which very much depend on quantum indeterminism to work. Those I suppose can be classically simulated when nobody is observing them closely.

So it would not be possible to simulate the progress of research in physics over the last 100 years or so?

That isn't an isulated system. One could put together an approximation of the state of Earth in 1924 and simulate it from there. That (the setting up of a plausible world) would require for instance a full understanding of physical consciousness and how memories work so that each person is created will a full memory of his past and has no idea that he just came into existence. The people there pushing the view of 'Last Tuesdayism' would be correct without knowing it.

Then you run the simulation for a simulated century and you get some world that differs completely from ours, but if they haven't killed themselves, the physics would likely have developed more or less at pace with our own history. All the names of the main contributors that hadn't been born by 1925 would be different, and the contributions of those that did exist would change.

I think you'll have to say that the hardware of this simulation we live in must be much, much more powerful than anything we can conceive of and that QM and GR are false. No?

Bostrom makes some outlandish suggestions that say otherwise, like for instance that Moore's law will continue indefinitely.
Don't know what you mean by QM and GR being wrong. They're not, but they're not necessarily the physics of whatever is simulating us.

The paradox of the situation is that believers in it have to put more faith in their fancies than in their experience — Ludwig V

You got it. I also see no motivation for our simulators to run this simulation. Bostrom suggests the 'ancestor history' thing, but it wouldn't be our history being simulated, just 'a' history, and a very different one. The only purpose of that might be to see how things might otherwise have turned out. How lucky are we to have survived to the point of being able to put together these simulations?

So I agree: the far simpler model is to presume our experiences are valid evidence of how things are, since the alternative is making up conclusions from zero evidence.

I don't beleive we are in a simulation, but this is my reaction to your points. — Tom Storm

As I've pointed out already, you're speaking to air. jasonm doesn't contribute to his own topics.

If we are a simulation and there is a world outside ours, how would we know what is possible? Since we know nothing of the world outside the simulation, we don't even know if it is done via computers. — Tom Storm

Exactly. Everybody online that pushes something like this presumes unreasonably that the world simulating us has similar physics.

Bostrom's Simulation Argument is that one of these is almost certainly true:

1. The fraction of human-level civilizations that reach a posthuman stage (that is, one capable of running high-fidelity ancestor simulations) is very close to zero, or
2. The fraction of posthuman civilizations that are interested in running simulations of their evolutionary history, or variations thereof, is very close to zero, or — Michael

I find both these to be highly unlikely, for the reason stated in this topic and mine. Bostrom of course has motivation to rationalize a higher probability for both of these, but rationalizing is not being rational.

:up:

I find both these to be highly unlikely, for the reason stated in this topic and mine. Bostrom of course has motivation to rationalize a higher probability for both of these, but rationalizing is not being rational. — noAxioms

I'm confused by what you're saying.

Bostrom is saying that one of these is almost certainly true:

1. Almost every intelligent civilisation is incapable of creating simulations
2. Almost every intelligent civilisation doesn't want to create simulations
3. Almost every conscious person is living in a simulation

Because if lots of civilisations are capable of and willing to make simulations then they will, and so simulated persons will greatly outnumber non-simulated persons.

Therefore, if simulated persons do not greatly outnumber non-simulated persons then most civilisations are either incapable of or unwilling to make simulations.

He doesn't say which of the three is most likely to be true.

the physics would likely have developed more or less at pace with our own history. — noAxioms

I don't see why you say that. I think you are assuming at least a soft determinism? But given that the starting-point of the history in the simulation is not more than roughly the same, I don't think you have any real basis for that assumption. I grow even more sceptical when I remember the argument that small differences, over time, can result in big differences. Remember, there were times during the Cold War when nuclear holocaust hung by a thread.
You say that you wouldn't necessarily run detailed simulations of everything at the same time, but switch to closer simulations when necessary to maintain the illusion. That's all very well, though it imposes an extra burden on the machinery because it will have to be aware of what people are attending to at all times.
But what I'm thinking about is what would happen if the people inside the simulation decided to do some physics, including, of course, experiments. That would require interaction between the person and the machinery even when they are carrying out experiments designed to reveal the physics at work. If the progress of science is anything to go by, it wouldn't be easy to fool them all the time.

Don't know what you mean by QM and GR being wrong. They're not, but they're not necessarily the physics of whatever is simulating us. — noAxioms

Didn't you say something to the effect that quantum mechanics and general relativity couldn't be simulated? Perhaps I misunderstood.
There are two physics involved. One is the physics of the simulated world, which would need to be quite like ours. The other is the physics of the world in which the simulators exist.

How lucky are we to have survived to the point of being able to put together these simulations? — noAxioms

Even luckier to have the resources to waste on such a project.

Bostrom is saying that one of these is almost certainly true: — Michael

Yes he does, and, as you say, he doesn't say which of them he thinks most likely - though many people seem to have decided that 3) is the best bet. I've no idea why.
But I couldn't see why Bostrom thought that one of those three must be true.

But I couldn't see why Bostrom thought that one of those three must be true. — Ludwig V

If lots of civilisations are capable of and willing to make simulations then they will, and so simulated persons will greatly outnumber non-simulated persons.

Therefore, if simulated persons do not greatly outnumber non-simulated persons then most civilisations are either incapable of or unwilling to make simulations.

man-made brains-in-a-vat. — Michael

I see.

You are confusing "simulate" with "is like". — Ludwig V

According to Merriam Webster, "like" means the same or nearly the same, whilst "simulate" means to give the same or nearly the same appearance often with the intent to deceive. Therefore, simulation involves some form of deception.

The world as experienced by the human mind can only ever be a pale representation of any real world that may or may not exist outside the mind, yet we deceive ourselves that we can directly know such a world. It is inevitable that any world in our mind can only ever be a simulation of any world that may or may not exist outside our minds.

Bostrom is saying that one of these is almost certainly true:

1. Almost every intelligent civilisation is incapable of creating simulations — Michael

Bostrom does not say this. We create simulations today. He calls the state 'posthuman', and it apparently means a device capable of simulating all of human civilization to a level sufficient for the full consciousness of the humans, and also a full simulation of more complex things like the simulation hardware itself.

2. Almost every intelligent civilisation doesn't want to create simulations

He doesn't say that either. He says that nobody will run 'ancestor simulations', which is defined as simulations (however long or brief) of our own evolutionary history. But such a simulation is impossible since no intiial state they give it would evolve anything like our actual history. They can run a sim of an arbitrary alternate outcome from the initial state, but that won't be our ancestry history, it will just be a simulation of fiction. Depending on where they put the initial state, there might not ever be humans at all.

So two premises, each of which has odds of being almost exactly 1.

3. Almost every conscious person is living in a simulation

That is a valid suggestion if the odds of the above two are small.

Point is, you are misstating Bostrom's premises. Item 3 doesn't follow from the premises as you word them.

He doesn't say which of the three is most likely to be true.

He does. Most of the paper focuses on rationalizing low probabilities for the first two premises to the point of 3 being likely.

Therefore, if simulated persons do not greatly outnumber non-simulated persons then most civilisations are either incapable of or unwilling to make simulations. — Michael

Incapable or unwilling to simulate a lot of them. I see purpose in simulating one person, or a very small group in a closed environment. There's value to that. But not to simulating that group that has decided to have its own simulating machine and running the same simulation.

I don't see why you say that. I think you are assuming at least a soft determinism? — Ludwig V

Scientific discover is sort of inevitable. Einstein stated somewhere that relativity theory was totally ripe after M&M experiment showed the apparent frame invariance of light speed. Minkowski would have come up with SR, but not GR. Others would have had to finish it.

The progress of physics is yes, a sort of fatalistic thing, much like Asimov's foundation series: It will happen inevitably, presuming there is the means to make progress. Much of progress hinges on the political state of things, which cannot be fully predicted.

Remember, there were times during the Cold War when nuclear holocaust hung by a thread.

Oh yes. That's what I mean above by 'presuming there is the means to make progress'. Plenty of viable outcomes have us all nuked away, or a pandemic or something. Asteroid is not likely since that isn't a chaotic function over times as short as centuries.

You say that you wouldn't necessarily run detailed simulations of everything at the same time, but switch to closer simulations when necessary to maintain the illusion.

Bostrom suggests that, yes. It's a necessary thing for an open system. Most simulations we run today are not open. Not always the case. I used to run computer chip simulations which has to be an open system since (most) chips need external input to drive them. We needed to see how the chip would function before going to the great expense of actually manufacturing a batch.

That's all very well, though it imposes an extra burden on the machinery because it will have to be aware of what people are attending to at all times.

You got it. Also what their devices are attending to, even when the people are not around.

What if the human decides to dig at location X? That location was trivially simulated up until now, but suddenly the machine has to decide if there should be a dinosaur there or something, even when the digging is not being done for purposes of looking for them.

it wouldn't be easy to fool them all the time.

Nope. It would be dang difficult, which is a decent reason why nobody would attempt such simulations, simulations good enough to fool its occupants, even the very smart but skeptical ones.

Didn't you say something to the effect that quantum mechanics and general relativity couldn't be simulated?

QM can't easily be simulated, but it can be done. My example of the cc of water was an example beyond some limits, but it depends on the interpretation being simulated.
I don't see much difficulty with relativity theory being simulated. They do that all the time in astronomical simulations like what it looks like to fall into a black hole, or a sim of our collision with Andromeda.

The difficultly in simulating QM is not in any way evidence that it is wrong. It's just evidence that it isn't classical, and most simulations as we know them are classical simulations.

There are two physics involved. One is the physics of the simulated world, which would need to be quite like ours.

If we are simulated, then the physics of the simulated word IS our physics, by definition. They can't be wrong. They might be only an approximation of what the runners of the simulation actually wanted.

For instance, any simulation run by us is discreet. Humans and machines only have access to a countable set of numbers, leaving most of the real numbers inaccessible. For instance, a typical floating point number is but 64 bits in a computer, more if you want more precision. There are only so many values that a finite number of bits can represent. The rest are off limits.
Real physics seems to work with real numbers, not these discreet numbers. But we can't prove that.

Point is, you are misstating Bostrom's premises. Item 3 doesn't follow from the premises as you word them. — noAxioms

They are not premises. (3) isn't intended to follow from (1) and (2).

He does. Most of the paper focuses on rationalizing low probabilities for the first two premises to the point of 3 being likely. — noAxioms

What paper are you reading? From the conclusion to Are You Living in a Computer Simulation?:

A technologically mature “posthuman” civilization would have enormous computing power. Based on this empirical fact, the simulation argument shows that at least one of the following propositions is true: (1) The fraction of human‐level civilizations that reach a posthuman stage is very close to zero; (2) The fraction of posthuman civilizations that are interested in running ancestor‐simulations is very close to zero; (3) The fraction of all people with our kind of experiences that are living in a simulation is very close to one.

If (1) is true, then we will almost certainly go extinct before reaching posthumanity. If (2) is true, then there must be a strong convergence among the courses of advanced civilizations so that virtually none contains any relatively wealthy individuals who desire to run ancestor‐simulations and are free to do so. If (3) is true, then we almost certainly live in a simulation. In the dark forest of our current ignorance, it seems sensible to apportion one’s credence roughly evenly between (1), (2), and (3).

They are not premises. (3) isn't intended to follow from (1) and (2). — Michael

Right you are.
They are three options, and the premise (the one and only and false one) is that one of the three options is very likely to be true. In fact, all three as you stated them are unlikely and a 4th option is the true one: There is neither capability nor desire to run sims of more conscious humans than there are real humans.

As for 1, his assignment of low probabiltiy to that is due to becoming extinct before it happens, not to it not being plausibly possible. Discussion of item 2 seems to suggest that it is for some reason, something that a wealthy person would want to do. I have no idea why. I guess it implies that despite this arbitrarily advanced technology, it's still costly to use it.

Yes, I agree that nowhere in the paper (except an impication in the title) does he conclude that the third option is the most likely. He's done talk shows and such, and there's very much an argument for its likelihood, but maybe such assertions are necessary to get him on the paying talk show.

That it’s just reheated Descartes and silly.

Nothing to do with Descartes except when equating simulation hypothesis with virtual reality hypothesis and analogising the latter with a brain-a-vat argument, which still has little to do with Descartes.
Otherwise I invite you to reference Descartes' writings.

That this reality may be a simulation and “in every way the same” as the “real world” is simply the deus deceptor. Which is all the Matrix was; which is all simulation “hypothesis” nonsense is: reheated, reworked Descartes.

In the Matrix, the humans' minds and consciousness are still coming from their brains. Simulation theory goes way beyond that.

In the Matrix, the humans' minds and consciousness are still coming from their brains. Simulation theory goes way beyond that. — RogueAI

I know— says it’s from a computer. The same basic concept. Replace the computer with a dream of a malicious demon, and it’s basically the same thing.

All of it is cooked up by the human mind, including the meaning of reality, simulation, computer, etc. Just a silly waste of time.

I am sure that all of you have heard it before: "We are living in a 'simulation' and such a virtual world — jasonm

Of course, and too often. And it always raises this very logical question in my mind: "If we are, how would we know it?"
Likewise regarding parallel universes, etc.

For me, really wondering about such things and taking them seriously, even as hypotheses, is wasting "gray matter" and time. Such things are good only for having fun and creating sci-fi stories.

First, if the world is simulated, why don't its 'designers' simply 'pop out' at times and leave us with some trace of their existence? — jasonm

How do you know if they do? And if you have such an experience, how would you distinguish it from illusion, delusion or hallucination?

If it's just a simulation, does it matter if the laws of physics are perfectly consistent? — jasonm

You couldn't know what laws of physics would apply to other universes ...

All these questions have the same basic answer: "How would you know"?

(At least, your position/conclusion on the subject is correct. Well, for me at least. :smile:)

That this reality may be a simulation and “in every way the same” as the “real world” is simply the deus deceptor — Mikie

The simulation assumes a "physical" world, the evil genius hypothesis doubts the physical. The virtual reality hypothesis (called simulation here) is pretty much the same to the brain-in-a-vat argument, which finds its closest parallel in Descartes in the always-dreaming hypothesis — which are still different, as in BIAV the sensations are caused strictly by an outside mechanism (machine) while in the AD the sensations are caused by one's own mind and more extensively not caused by a real world (denial of extended AD implies denial of BIAV but not conversely, strict AD and BIAV are independent claims).
It reminds of Descartes, but it is not strictly the same.

Why can't the term 'simulation' refer to things as it normally would by dictionary definition - why must we assume some solitary metaphysical significance?

Light reflecting off of objects and producing color and form in mind is a kind of simulation. My idea about the sun being a frame that literally frames space and locks those in its locale into a physical reality, whilst other stars, who's gravitational influence we are not part of, have no physical reality to us, is another example of a possible simulation(but it's just a random theory I had).

Why is the original post an argument against a more normalized use of the term simulation? Why can't the universe work reversed from a off-center of the final product? Why must the big bang had to of happened rather than it's just the anti-thesis to what was actually produced by something more miniscule?

A technologically mature “posthuman” civilization would have enormous computing power. Based on this empirical fact, the simulation argument shows that at least one of the following propositions is true: (1) The fraction of human‐level civilizations that reach a posthuman stage is very close to zero; (2) The fraction of posthuman civilizations that are interested in running ancestor‐simulations is very close to zero; (3) The fraction of all people with our kind of experiences that are living in a simulation is very close to one.

If (1) is true, then we will almost certainly go extinct before reaching posthumanity. If (2) is true, then there must be a strong convergence among the courses of advanced civilizations so that virtually none contains any relatively wealthy individuals who desire to run ancestor‐simulations and are free to do so. If (3) is true, then we almost certainly live in a simulation. In the dark forest of our current ignorance, it seems sensible to apportion one’s credence roughly evenly between (1), (2), and (3).

I suppose it is not unreasonable to apply a judicious, but not radical, scepticism to this argument or at least some of it. Long posts risk not being read, so I shall focus:-

(3) The fraction of all people with our kind of experiences that are living in a simulation is very close to one.

I charitably assume that "all people" means all people past, present and future, including artificial people developed as part of a holistic simulation - if there be any such. What is the evidence that there are any people with our kind of experiences living in a simulation - apart from NASA experiments, archaeological research and trials for deep-sea mining? None that I know of.

But the Doomsday Clock is at 90 seconds. Extinction without any evolutionary descendants is a real possibility and should have been included as a fourth possibility - amongst many others.

If (3) is true, then we almost certainly live in a simulation.

That seems to follow. However, let us note that "we almost certainly live in a simulation" assigns a probability of, say, 0.99 to "we live in a simulation".

In the dark forest of our current ignorance, it seems sensible to apportion one’s credence roughly evenly between (1), (2), and (3).

In probability theory, it is indeed regarded as sensible to do that if we have listed all the possibilities. But the rule only really applies in mathematical probability, which this exercise is certainly not. The Bayesian notion of credence is based on an admittedly subjective evaluation of the evidence for each outcome. But there is no real evidence for anything here, so the assignment is arbitrary, rather than sensible.

In any case, what we actually have is a probability of 0.33 that the probability we are living in a simulation is 0.99. That seems a good deal less than almost certain and almost certainly less that 0.5.

Such things are good only for having fun and creating sci-fi stories. — Alkis Piskas

Yes. But some people have peculiar ideas of fun. Other people get annoyed and engage in the forlorn hope of persuading them to stop being so silly.

It reminds of Descartes, but it is not strictly the same. — Lionino

True. The point of the comparison is to introduce some perspective and suggest that these thought-experiments are subject to similar criticisms.

Light reflecting off of objects and producing color and form in mind is a kind of simulation. — Barkon

Really? What is it a simulation of?