There isn't really much of a connection between this and Bostrom's simulation argument. The simulation argument is about calculating the probability that we live in a simulated universe based on certain assumptions about human behavior and technological development. Bostrom's argument doesn't make any metaphysical claims other than assuming that consciousness is substrate independent.
I don't buy into metaphysical theories that claim to deduce the existence of worlds outside our own based on armchair reasoning. We know that the physical universe exists and we can explain everything that we experience in terms of quantum field theory and general relativity. Any theory that wants to challenge this view of the world needs to modify those existing theories, or design an experiment that shows why they aren't adequate to explain reality.
Modal Realism was the inspiration for the somewhat infamous, slightly tongue-in-cheek "Possible Girls" paper where Neil Sinhababu argues that people across different modal realities can fall in love with each other (and that explains imaginary relationships)
As I understand it, the major candidates for dark matter are new elementary particles, so they would still fall under quantum field theory.
Of course, there is always the possibility that another revolution in physics happens, but even then our current theories will still be valid in most domains, in the same way that Newtonian mechanics is still valid in most domains.
> we can explain everything that we experience in terms of quantum field theory and general relativity
Everything we experience, except experience itself. Conscious/qualia/whatever is still… well, none but God knows what it is, and I have no evidence for the existence of any god let alone that one.
That's because consciousness is an emergent phenomenon, not part of fundamental physics. Every atom in your brain behaves according to the laws of particle physics, and somehow consciousness emerges out of that. Our theories of neuroscience aren't developed enough to explain it yet, but there isn't any reason to believe that there is something magical or non-physical going on.
How do we know consciousness is real, and we're not just the same as robots or a tv series? Couldn't every moment throughout the universe have been encoded in every atom before the Big Bang?
https://en.wikipedia.org/wiki/Qualia[1] - is what you are now describing mixed with a little bit of Maxwell's Demon[2] - something is intercepting between you and a true experience.
Real, in the way I believe you are trying to express, is a complicated topic.
Because we can change brain states (ie, manipulating a material object) to go from conscious to unconscious, which is strong evidence for the former. What evidence is there for the latter?
This is evidence that self-conscious beings can stop being self-conscious or that the content of what they are conscious of can change when the stuff that enables self-consciousness to occur is manipulated. It does not prove that consciousness is an emergent phenomena. Suggestions that consciousness is prior to physicalism (and in general prior to the whole division into subjective and objective realities made by a rational self-conscious agent) are not something new and are worth exploring.
> consciousness is an emergent phenomenon
That's a theory that has yet to be tested, not a certainty.
<digression>
> Every atom in your brain behaves according to the laws of particle physics...
These are assumptions. They're very plausible and useful ones. They may turn out to be absolutely correct, but it's still worth noting that they are theories.
"Every atom behaves..." so far, so good, yes it does.
"...according to the laws of particle physics"-- presumably yes, with this caveat: The existence of a complete set of inviolable "laws" that explain _all_ physical behavior is an article of faith. It's a plausible potential outcome of the extremely practical process of collective scientific inquiry that so far seems to hold up, but still-- the concept of a set of laws as programs and constraints for absolutely all physical phenomena is fundamentally a philosophy. It has proven to be extremely useful, but it's not a proven or even probable fact. It's a powerful axiom.
</digression>
> ...and somehow consciousness emerges out of that.
Back to the assumption that consciousness is a phenomenon that emerges from the physical activity of a brain:
Is a (functional) brain both necessary and sufficient for consciousness to "emerge"? I don't know, I'm just a person interested in this stuff, but I think it's an important question without a definite answer.
We tend to assume that having a brain is a necessary requirement for consciousness, but testing for consciousness (not to be confused with mere rationality) is difficult, as far as I know. I can't prove that a stone, plant, or region of spacetime has no consciousness.
Even if we came up with a test for "consciousness", and that this test _did_ prove that a brain (or a similarly complex network) is necessary for consciousness to appear, there is still the issue of whether the brain is sufficient. How would we know?
Whatever it is that we call consciousness may emerge from the complex physical phenomena of brain activity, as you said. Or the brain may be a catalyst for the phenomena we label consciousness. Or a brain (or human body) may only be a prerequisite for us, with our current capacities, to be able to observe the experience we have come to label "consciousness".
"Light" used to mean only the visible spectrum, until we understood there was more to it: Light went beyond the limits of what we had up til then used to recognize its existence, and indeed it occurred even in what we had called "darkness", and blue things didn't contain blueness in and of themselves, nor were blue things producing blue light.
I'm not advocating for any particular alternate theory, or proposing that consciousness operates outside of known physical constraints, I'm just a stickler for reminding ourselves what concepts are articles of faith, even in science itself
I don't think "article of faith" is the right phrase. For every possible hypothesis, we attach a prior probability and then we update those probabilities when new evidence comes in. If accepting a given hypothesis would require changing a lot of other hypotheses that we have high confidence in, then we should attach a low prior probability to it. I don't have to take it on faith that a low probability hypothesis is false; the default position is to assume that it is false until compelling evidence to the contrary appears.
For example, I can hypothesize that a fifth fundamental force is needed to explain consciousness, but then I need to modify the standard model of particle physics to account for new interactions. I have a high confidence that the standard model is correct in the domains relevant to the evolution of life on Earth, namely chemistry and biology. So my default assumption is that we don't need to introduce new laws of physics to explain consciousness.
>Is a (functional) brain both necessary and sufficient for consciousness to "emerge"?
I think you misread my post as implying that a brain is required for consciousness. Our baseline for talking about consciousness is consciousness in humans, and all of the medical evidence suggests that human consciousness is associated with brain activity. But I see no reason to believe that something made out of silicon or other non-organic materials couldn't be conscious if it implemented the same kind of processes that we find in the brain.
This argument seems to mix up "existence" and "construction".
The number states do not magically appear in the physical universe merely by thinking up the construction. The numbers could be configured as (temporary) patterns in physical objects, such as brains, books, or in ink molecules on paper. But the states are not physical objects themselves.
Also, if our universe happens to be universal, in the sense that it encompasses all of existence, then how could a calculation device exist outside of it? I'm not saying this is necessarily the case, but it's an option that many simulation-believers overlook. The calculation device might be part of the existence, but it seems rather unlikely that it can then predict reality faster than it unfolds.
Much ink has been spilled by many a philosopher on the topic of whether or not numbers "magically exist." Plato was the obvious example of a philosopher who believed numbers "exist" independent of our universe. Though no one is saying they exist "in the physical universe", but it's not a given that they cna't possibly exist if not "within our universe."
Think of it this way. Graham's number is an absolutely enormous number, right? Let's assume for the sake of argument that nobody has ever computed the Graham's-number-th digit of pi. We know for certain that there is a Graham's-number-th digit of pi. And we know that if two people calculated it independently, they'd get the same digit. But (at least in this hypothetical) nobody has actually ever done the calculation to see what the Graham's-number-th digit of pi is. Given all I've said so far, the act of finding out the Graham's-number-th digit of pi seems more like an act of discovery of something that already existed than an act of invention of something that didn't already exist. So, it seems quite reasonable to many to conclude that numbers "exist."
Also, Iah's view does imply that our universe does not encompass all of existence. It also implies that no calculation device need exist anywhere.
In response to your second point, at a high level I believe the calculation device would exist inside _one_ universe but calculate another one... the idea that you could calculate your own universe and use that to predict future events does seem covered in paradoxes. For one, the universe you're predicting would (recursively) have to include the computer you're using to predict it.
... but that's different to what I've argued here. I'm not claiming the states are physical objects, but just the existing of the pattern, even if temporary or intangible, would feel real to the humans/actors inside it.
My premise was that the universe was universal, i.e. there is no other universe. Again, I'm not saying that this is the case, or that there is any reason to believe that it is so, but I don't see a good reason why there would be more than one universe. (Note that all of this depends a bit on the definition of a universe -- for the sake of argument, I'm assuming our universe to be a system closed under physical interactions. Happy to argue about other universes, but perhaps it's best to save that for another time.)
If you are trying to prove the existence of this universe by requiring the existence of another universe, then it's turtles all the way down.
How do you define "existing" of a pattern? Does it exist inside a physical thing? If so, then how does that physical container come into existence? And if it exists only conceptually, then how is it possible for concepts to exist? In the universe that I know, concepts only exist in the minds of human beings, and perhaps in some other animals. To me, it seems rather unlikely (and a bit anthropocentric or egocentric) that concepts are something truly universal.
For me, it helped to meditate a lot on what it'd be like to be a rock. The rock does not have memory, no sensory input, and therefore most likely no concept of time, space, logic, nor mathematics. It makes you wonder whether the rock exists at all. In any case, it probably doesn't care as much about it as we humans do. There might be a hint there.
> The calculation device would exist inside one universe but calculate another one...
Hmm... calculation device? I thought the premise here is that actually doing the calculation is not necessary - a tenet you invoke in order to say say elsewhere [1] that the computational complexity does not matter, as the notionally simulated universe exists anyway.
But if that were so, would it not do away with the distinction between simulating and simulated universes, creating the situation where every possible universe exists in every extant universe?
One one hand it's absurd (it means that everything that can be imagined and many more things exist).
On another hand the opposite (requiring a mapping from that computation to real-world objects) is absurd too, because for any sequence of numbers you can always find a mapping to physical objects (notice that you can make the mapping arbitrarily complex). So why require the extra steps?
My opinion is that it follows that asking about existence without specifying the domain in which sth exists is meaningless.
You can say that the number 42 exists in the domain of integers. You cannot say whether the number 42 exists in general. It wouldn't mean anything.
Similarly you can say that Harrison Ford exists in our universe but Han Solo doesn't. But you cannot say whether one or the other exist in general.
Should have said "physical" not "real" probably. I didn't meant to imply it's more "real" than the integers or any other domain, my point was that you have to specify the domain and that there are many options.
It seems to take a highly reductionist pathway: reality/experience can be simulated -> simulation is computation -> computation is mathematics -> mathematical objects exist regardless of whether anybody has discovered them.
This implies that all conceivable universes (including the ones where a lot of really bad things happen on an eternal loop) are possessed of the exact same reality as ours.
Am I missing something here? I can suppose a universe where the premise is false (This is a conceivable universe, I’d argue). Doesn’t that mean that this premise really is false?
Someone who holds this view would probably have to make the definitions more precise, so that if you live in a deterministic reality, you actually cannot conceive of a universe that isn't deterministic. You can throw around words like "indeterministic", but you cannot precisely simulate something indeterministic using only deterministic ingredients, and hence, for some precise definition of "conceive", cannot conceive such a universe.
The author's tenets seem to insist upon the reality of universes expressed in inconsistent formal systems, so I don't think this is fatal to the argument - though it suppose it might render pointless any attempt to treat the premise analytically.
I believe so, yes. I think it's somewhat likely that mathematically minded people would come up with that idea independently -- it's essentially Platonism taken to its logical conclusion.
Had this thought in the bath ~25 years ago. Not an idea you can actually do anything with, though, other than go 'huh, interesting' and add more hot water.
You might also like to check out Greg Egan's Permutation City[0] which (in the form of entertaining scifi) presents some related arguments about the need for a computational substrate, or lack thereof.
Take a smaller example: weather on Earth. There are a LOT of particles, but still classically simulatable. Chaos ensures that we still cannot know all future states of the weather. This is a remarkable truth, and one should give it time to sink in.
Note that quantum modelling those effects go as O(a^N). If you want to hand-wave away exponential computational cost, then I cry foul: the details matter, and I posit that you cannot build a computer that is more powerful than the universe itself.
Sure, but this argument is for simulation theory... not what is described here.
What I'm suggesting is that the calculation never needs to be done, which means the complexity of it does not matter. Whether it's O(1) or O(a^N) they're both far smaller than the infinite number of potential states.
Well, that's your preference then. But personally I want to see the Mets play the Yankees because it's profoundly unsatisfying to believe that all possible outcomes are computable and therefore on an equal basis. I daresay that idea is so bad that if you took it seriously, it would die out with your genes/memes. (Not that that matters since you could travel to and impregnate every woman on Earth).
If Iah's view is correct, after the game, there will be at some versions of you who witnessed the Mets winning, some that witnessed the Yankees winning, and some that saw the game end in a tie. (As well as some for which the game was interrupted by aliens landing on the field and some where all the players spontaneously disappeared etc.)
So, subjectively, it'd still feel as if you'd gone to the game without knowing what was going to happen and then exactly one outcome (and not all outcomes) happened. But from a more objective perspective, all outcomes are indeed on an even footing.
I think the author's premise would lead to the conclusions that you really are hungry and you really eat the pizza, but the former is not causal in the latter (they are just theorems of the formalism) and the latter is not causal in anything else.
I thought of the similar thing before, including all of the details they mention, and I agree. (Of course you still might not konw what it is unless a simulation can be made, but it can exist (mathematically) whether or not a simulation is made.)
Furthermore, I think that anything existence (including mathematics, physics, etc) are by relation to everything else that is existence; there is no such thing as an "existence" that is truly independently from everything else, including mathematics (and, the various possible mathematical structures, and physical universes), etc. (If it is truly independent, how can numbers be prime and composite, and how can objects collide with each other, and how can planets moving around the Sun by gravity, and how can people have ideas about the gods, etc?)
(However, there is then the consideration if such a simulation is being made which is then being interacted with by outside (e.g. as in Star Trek); in that case, it is I/O, so you can make "Mathematical Universe Hypothesis with I/O". This could be called as a kind of nondeterminism, I suppose; although, a constraint may be produced by their combination.)
[0] Boltzmann brains are cognitively unstable, because the logic leading to them says you should expect to be one, but also that they have exponentially increasingly false "memories" the further back they "remember", including the beliefs leading to the conclusion that you probably are one. If all maths is real, this is worse to at least the degree that Aleph one is bigger than Aleph null, possibly more.
The Boltzmann brain theory is not something I have heard before but have to say I love the idea.
The thing is if the universe is infinite and you believe that, then by default many other things also have to be true.
I do hope the universe is infinite which means I exist many times over in many different states of being, hopefully one such state where the Canadian housing market isn’t so crappy and I get to afford a house one day ;)
Infinite universe doesn’t necessarily imply what you’re suggesting. Removing all even numbers from a set gives you another infinite set - but it is still an infinite set that will never contain a multiple of 2 in it.
Similarly, our universe may be infinite but isn’t at all guaranteed to produce multiple Earths with identical evolutionary history such that another or even many “versions” of you exist somewhere/when.
I’m not sure I agree on your take. There are of course different definitions of infinity which would have different rules based on said definition. But if our universe is infinite and physics as we know it remains true then we would have to run into all the variations of every molecule combination possible again and again.
Saying we won’t see any even numbers in an infinite set of odd numbers is obvious as the rules of that infinite set have been defined. There are no such rules in the universe.
The universe appears to be expanding close to e^t. If it continues to do so, each fundamental matter particle may end up isolated in a Hubble volume where the only photons from the rest of the universe have been red shifted to wavelengths larger than that Hubble volume.
"The argument assumes that given one state of human existence, it is possible to calculate a distinct next state. And then a state after that. That is, if we had a powerful enough computer to store the state of every single atom, and calculate their interactions, we could play through time in that existence."
Why are you assuming this to be true? Due to the Heisenberg Uncertainty Principle, it's impossible to perfectly know "one state of human existence". So the rest of the thought experiment seems moot.
It's also naive to think that everything can be calculated by a computer, at some point the calculations become so enormous that the energy requirement approaches infinity.
> at some point the calculations become so enormous that the energy requirement approaches infinity.
No, they really don't. You might say that the energy requirement approaches Graham's Number, or TREE(3), and maybe that's somewhat plausible, but Graham's Number and TREE(3) are both minisculely insignificant compared to infinity. If you ever counted that high, you'd be no closer to reaching infinity than you were when you started.
But who says that Graham's Number or TREE(3) are large numbers in the universe doing the simulating? Why would the simulating universe be as incomprehensibly tiny and simple as ours is?
Oh sure, our universe looks big to us. But who are we to judge its size?
Is it moot though? The uncertainty principle makes it impossible for us to _measure_ both speed and position, but if you're simulating it that's not necessary.
Basically it's a question of whether math exists. If it exists then the states of the simulation exist and look like reality "from the inside".
I really like Iain M. Banks moral argument against universe being simulated. No intelligence capable of simulating it would be immoral enough to create something so horrible. Unless they are a total bastard. So it's not a 100%.
Humanity gets more benign as history progresses. This general trend is momentarily reversed from time to time by recent technological developments but still stands.
In science fiction it's common to have species that are so far advanced that humans are like ants to them. For all we know, our reality could be a petri dish sitting under the couch of an alien dorm room.
The actual performance of calculating in the OPs example is analogous to the passage of time, as expressed by entropic processes.
Op’s conjecture can be thought of as the same conjecture as proposing that things in the past and future still / already exist, a conjecture that I find compelling based on dimensional reductionist thought experimentation.
> That is, if we had a powerful enough computer to store the state of every single atom, and calculate their interactions, we could play through time in that existence. It would feel completely real to everyone in that simulation.
Telling persuasive stories can achieve essentially the same outcome, is much easier, and is already an accepted convention.
It's like saying buildings existed before it's been built because there's already drawings on it, so why build anything.
The missing component in the thesis is that it argues that consciousness is just an abstract, immaterial object that consists of states. That's not true because to be conscious you need matters.
Author should also look into complexity science, and its consequences that even if everything is determined, most information about next state quickly evaporates. If it's impossible to know most things about next state, does it even matter that it is deterministic/simulated?
One useful way to think of the halting problem is that the future is not computable. There is no shortcut to finding out what will happen other than actually running the computer. I think all computational theories sweep this issue under the rug. Even if the universe is a computer (quantum or otherwise) it makes no difference to how people should approach their decisions. There are no royal roads so regardless of how much a situation is analyzed by symbolic simulations at the end if the day you still have to act of your own volition to actually find out what will happen.
That coincides with my belief. If the next state exists mathmatically, it exists.
The "simulation not required" hypothesis eliminates the inconvenience of infinite regress: what simulator runs the simulator, and what runs that one ...
I believe that you are more real than Donald Duck if and only if the future is not yet determined.
In philosophy, theories of time are categorized in A- and B- theories [1].
In B-theoretic time, the difference between past, present and future is only subjective. Objectively, all points in time exist and are equally real. I view this to mean that there isn't any particular difference in the degree of reality between our reality and any mathematical model or imagined reality. Only with A-theoretic time are you objectively more real than Donald Duck.
One of the similarities between philosophy and art is that in both fields, non experts seem to think they are qualified to give their own opinions. Unfortunately this means the public discourse becomes more about what people enjoy than what is actual.
We're building digital twins of much of the universe around ourselves at different scales. We're even building digital twins of ourselves.
We're improving the technology that enables that twinning quite rapidly.
We're improving the technology that allows for emulating a virtual environment.
And in fact, there's a remarkable overlap in how we are running those virtual environments and the fundamental building blocks we've experimentally validated in our own universe.
The part that I think people get caught up in is the assumption that we'd need to ourselves calculate a 1:1 copy for us to be in a calculated copy.
But at macro scales the universe behaves as if things are continuous, not discrete/quantized. And the mechanics of quantization remains incompatible with the mechanics of continuous theories like GR.
The nuances of that quantization map to how we fudge fidelity in our own virtual representations.
So we need not create a 1:1 copy for us to be in a copy any more than one would need to create Minecraft at a 1:1 scale within Minecraft for Minecraft to exist.
Additionally, looking at mechanics isn't the only way to investigate whether we are in a virtual copy. For example, in many copies we make, there's some acknowledgement of that state woven into the world lore.
Indeed, in our own world in antiquity was a set of beliefs attributed to one of the most well known figures in history that espoused that we were in a copy of an original world in which an original humanity came to exist spontaneously and then brought forth the creator of this copy within light with us in their archetype before they ultimately perished. The full text at the heart of these beliefs was lost for centuries before being found again the same month as ENIAC ran its first computer program.
This text claimed the proof for what it said about a creator of light would be found in motion and rest (the domain now called Physics), and the group following it claimed that the ability to find an indivisible point within bodies would only be possible in the non-physical.
That's quite the coincidence in an age when we are moving towards putting the AI emulating our digital twins literally into light with optoelectronics and are very focused on the discovery of indivisible points within the domain of study of motion and rest.
(The name of the text is literally translated as "the good news of the twin" and its main point is that if one understands its content they will not fear death or worry about the soul's dependence on a physical body.)
https://en.wikipedia.org/wiki/Modal_realism
There isn't really much of a connection between this and Bostrom's simulation argument. The simulation argument is about calculating the probability that we live in a simulated universe based on certain assumptions about human behavior and technological development. Bostrom's argument doesn't make any metaphysical claims other than assuming that consciousness is substrate independent.
I don't buy into metaphysical theories that claim to deduce the existence of worlds outside our own based on armchair reasoning. We know that the physical universe exists and we can explain everything that we experience in terms of quantum field theory and general relativity. Any theory that wants to challenge this view of the world needs to modify those existing theories, or design an experiment that shows why they aren't adequate to explain reality.