Philosophically speaking...

[Fung Koo]

So Elon says we're in a simulation. But philosophers argue otherwise.

https://thearcmag.com/were-all-living-in-a-video-game-64fca860d688#.om5vovaqi

I've often heard complaints that SF has failed to keep up with the cutting edge of science and technology. But if this is what the cutting edge looks like, then is the next step for SF.... philosophy?

 

[Dawnstorm]

So Elon says we're in a simulation. But philosophers argue otherwise.

https://thearcmag.com/were-all-living-in-a-video-game-64fca860d688#.om5vovaqi

I've often heard complaints that SF has failed to keep up with the cutting edge of science and technology. But if this is what the cutting edge looks like, then is the next step for SF.... philosophy?

That's less the cutting edge and more the smothering sponge, and it's probably neither science nor philosophy. I didn't see any argument in there at all. Of course, I would say that, as I'm just an NPC programmed to hide the truth. I mean, look at my forum title.

 

[Stephen Palmer]

You would expect a world-denying "explanation" (by which i mean total fantasy) from somebody like this. The world-denying religions were doing this sort of thing 2,000 years ago, and doing it with at least an attempt at some ethics.

 

[Fung Koo]

I find it curious that concepts like the Many Worlds interpretation or the Holographic Universe, for example, are being arrived at via the science, though.

SF-wise, I'm thinking of stories like Iain M. Banks' Transition. Or, TV-wise, something like Fringe. The latter, like X-Files, was often more fantasy-ish than sciencey, and the former mostly just avoided a lot of the more technical nitty-gritty's of the science fictional elements.

Philosophically, I think Buddhism has a lot to say in this regard. Jainism, too.

It just seems like the far end of this brand of theoretical physics collapses any possible differentiation between genres.

In the background, I have a story world I've been working on for a little while now and it's a challenge to keep it SFnal. So I'm just looking for an out loud puzzling-out of where the science begins and ends.

 

[RX-79G]

Old epistemological arguments don't become more relevant or real when you can point at a computer. The world could be "fake" because my hypnotist is good at his job or our brains explain time to us all wrong. These are all just ways of thinking about the problem of "knowing".

The multiverse thing is interesting, in much the same way that the fact that every solid object is actually 99.99999% empty space. The facts of the underlying nature of reality aren't the same as those facts being able to interact with us. I can no more expect to travel to another timeline than I should expect to be able to walk through that mostly empty wall.

I love Iain Banks and enjoyed Transition - but it really isn't a lot different than urban fantasy. William Gibson's The Peripheral is a more science based and thoughtful look at the way two universes could interact, without promising violations of our laws. Please read both if you haven't.

Elon Musk's belief system is no more "science" than any other statistical prediction. It is worth exploring, and has been by, many, many authors. It is essentially the same as the Fermi Paradox or the Singularity - great thought experiments to make a story out of, or explain away.

 

[N. E. White]

I can no more expect to travel to another timeline than I should expect to be able to walk through that mostly empty wall.

Good one.

I like to think you all (the universe and everything) are just figments of my imagination. And damn you all for not bending to my will.

(And I think Musk's argument is basically that - just an idea. Interesting, but not likely.)

 

[Dawnstorm]

I find it curious that concepts like the Many Worlds interpretation or the Holographic Universe, for example, are being arrived at via the science, though.

Well, science - to some extent - defamiliarises the world. It has to, because we need something to pay attention to that we can all treat more or less in the same way. But at some point, we end up with something so unfamiliar that it gets hard to remember why we're actually doing this. And thus we re-familiarise this stuff with things like the Many World Interpretation of quantum theory. Whether these metaphors end up being fruitful for thinking up stuff you can again de-familiarise and look at scientifically is another question.

Note that some stuff, even in maths, looks quite eerie. Take for example this:

1 + 2 + 3 + 4 + 5 + ... = - 1/12

I didn't make that up. This is not the maths we know; it's based on some additional assumptions. But it's something that mathematicians accept. And that's a number we apparently use in string theory.

However strange all this is, we do end up with the world we live in. So when we write SF we basically fake empirical data for non-existant science (some of which is closer to what we have now).

So when we go to Elon Musk's quote in the article, the metaphor has become oddly specific, to the point that it ceases to function as a metaphor. But at the same time it's not specific enough: Whenever I play a game, whether it's Pong or a full 3-D game, I carry with me memories from the real world, and without those memories the game wouldn't make sense. If we are in a man(?)-created simulation, why don't we know about it?

It's questions like this that make for good SF, but it's rather unlikely that we stumble on "the truth" that way.

It just seems like the far end of this brand of theoretical physics collapses any possible differentiation between genres.

To me, too. Which is why a clear cut demarcation never made sense to me. Also, when you describe "what happens", you don't have to explain that either way. You can, basically, be interested in the experiential aspect of the speculative events, without taking sides in a magic/science debate (which you can just import into that world to laugh at both).

 

[RX-79G]

The other problem I see with the simulation thing is the assumption that a programmed simulation is more likely than a natural occurring simulation. We have long accepted the idea that self-organizing life came about accidentally from complex pre-conditions. Why shouldn't complex algorithms come about from complex computing conditions.

Our universe might be nothing more than math, but that math could just be running as part of an anti-virus program somewhere, or a completely accidental logic matrix embedded in a much less complex universe than our own.

And who's to say our simulated universe isn't greatly preferable in its subtle complexities to the one whose matrix ours is running on?

 

[Justin Swanton]

Taking the simulation outlook seriously means systematically rejecting the fundamental and consistent datum that all our sensory imput gives us, namely that the external world and everything in it is real, and our existence in it is as our sensory input tell us - a biological self-aware entity with very definite abilities and limitations. There's nothing - ever - in all the vast stream of information we get second-by-second during our entire lives that contradicts this. Thus we have no founded reason to doubt external reality.

To seriously doubt it puts one into a state of irrationality. Not only does the individual have to disregard the information of his five senses for no good reason, he also has no good reason to conform his life to the demands and constraints of that illusory external world. He can run into a wall, shoot a fellow passenger in a bus in a fit of boredom, or cut his throat out of idle curiosity to see what happens. If one rejects the external reality of the universe one must in consequence reject any imagined constraints imposed by that universe. If my existence is not determined or conditioned by the phenomena around me then I must shed the illusory conformity imposed by those phenomena. That is logical. Elon Musk is not.

Antoine Roquentin tries act on his detachment from external reality in Sartre's La Nausée by stabbing his friend for no good reason, but he can't bring himself to do it. When the chips are down, one takes the external world seriously enough.

 

[Fung Koo]

Thus we have no founded reason to doubt external reality.

The point is not based on doubting external reality. It's actually based on not doubting it, then looking at our own abilities within real reality.

We have two possibilities. In this real reality, either it is possible for us to create a true consciousness through technology, or it is not. (There is a third hybrid option, but for argument's sake lets avoid that one for a moment).

If we assume that it is not possible to create a truly conscious AI, then there will be all sorts of philosophical issues arising. Chiefly, this would strongly argue that our own consciousness is not a product of material reality. Per your point about our senses, whilst our senses would give us the ability to perceive the universe, our reasoning faculty (cogito, ergo sum) -- or, consciousness -- would not be a by-product of a complex interaction of physical features and forces of this universe, but would be dependent on something-else.

In other words, to posit that AI/simulated true-consciousness is impossible becomes proof of some sort of ordinal/subordinal relationship between our universe and another reality that ours exists within. This would be a proof of some sort of god. We will come back to this.

The other assertion, that it is possible for us to create a simulated true consciousness, will have other philosophical issues attached. First, it will prove that material reality is the only necessary condition for consciousness, which will, on the surface, suggest an atheistic reality. But there is an obvious point here -- if our consciousness arises not through some magical property delivered to us by some higher-order consciousness (a god, per the example above) but through the factual forces of the universe itself, then in order to create a simulated true consciousness there are two possible ways of doing it:

Option 1 - we create an AI in this world based on the same parameters that give us consciousness, which is, effectively, us. The only currently known forms of consciousness in the world are those which we see here on earth, so the AI we make will take shape and function in some way like life as we see it here. Another way of thinking of this would be a sort of accelerated evolution.

Option 2 - we create a simulation of a universe that is, for all intents and purposes, factual and "real." While it may only be a simulation to us, it has all the necessary conditions, per our own universe, to give rise to true consciousness within it.

The consciousness that emerges within that simulation would perceive the universe it lives within as being the real reality. And to ensure there is no serious issue of information gain that destabilizes the simulated universe's balance of forces, it would seem likely that the consciousness within the simulated realty would only be aware of the reality it is within, and not of the reality that its reality is within. That position of awareness would already be occupied by us.

If we assume that consciousness itself, per your sensory input argument, coupled with DesCartes' reasoning of the importance of reason in conjunction with the senses, is a form of verification of reality itself, then for the consciousness within the simulated universe the simulation will be a real reality.

Consciousness, as a function of existence, is either a divine property or it is a physical property of the universe itself. If the latter, that means that universe gives itself reality by having the ability to perceive itself.

Therefore, this universe remains real regardless of whether or not it is a simulation in a subordinal position to another universe.

So we aren't talking about a Matrix-style "you exist elsewhere but your mind is in a simulation" assertion. We are talking about the wholesale manufacture of real realities.

Assuming the logic above is sound to even a small extent, then it is seems not only plausible but likely that consciousnesses within a simulated real reality would have no need to know, or even ability to know, that their real reality was a simulated reality in some other reality.

Given enough time and assuming "God" is not a prerequisite for consciousness, then the likeliest explanation of any universe is the existence of an ordinal universe that created it.

----

Coming back to the matter of AI/true sim consciousness not being possible, the trouble here should be evident. If we can't create real consciousness, proving it not to be a property of this world but of another, then all we've actually done is prove the same thing -- that this universe/existence is subordinate to another and our consciousness is dependent on some other existing thing.

That would effectively demonstrate that this reality is, for all intents and purpose, a simulation.

No less real for being a simulation, but a simulation nonetheless.

 

[Fung Koo]

And who's to say our simulated universe isn't greatly preferable in its subtle complexities to the one whose matrix ours is running on?

Preferable?

Strange word choice. Preference would presume transference of consciousnesses between realities/simulations. Naturally arising (per the Many Worlds Interpretation) versus Created (per the possibility of AI) seems, then, a moot point. If such universes can be real independently of the process, then they are all equally real regardless of their origin.

What does "preference" mean, then, in this sense?

 

[RX-79G]

Preferable?

Strange word choice. Preference would presume transference of consciousnesses between realities/simulations. Naturally arising (per the Many Worlds Interpretation) versus Created (per the possibility of AI) seems, then, a moot point. If such universes can be real independently of the process, then they are all equally real regardless of their origin.

What does "preference" mean, then, in this sense?

There is a certain "monkeys in the zoo" cloud that hangs over any discussion of living in a created/simulated universe, and talk about the programmers as "gods". But what if each programmed iteration gave rise to much greater levels of complexity and intelligence than what exists in the computing matrix universe?

If that were the case, we may well be the "gods", rather than the specimens. Our creators would be unable to understand or truly appreciate what they've created, let alone be able to interfere or influence our world.

 

[Justin Swanton]

Coming back to the matter of AI/true sim consciousness not being possible, the trouble here should be evident. If we can't create real consciousness, proving it not to be a property of this world but of another, then all we've actually done is prove the same thing -- that this universe/existence is subordinate to another and our consciousness is dependent on some other existing thing.

Or it proves something else: that consciousness is an aspect of human nature that is not reductible to purely biological or physiological causes. It certainly won't ever be replicated by a computer. Computers are an array of microscopic transistors that act as gateway switches to electric current, with two settings: on and off. This enables them to produce strings of interrupted current that replicate a binary number system. Creating and manipulating a base 2 number system enables computers to imitate mathematical processes. Notice that computers imitate mathematical processes - they don't actually understand the relations between numbers as humans do. The imitative mathematics enables them to imitate other processes of thinking without ever actually thinking themselves.

What is thinking? Thinking is the ability to mentally grasp the nature of something, in a sense to become that nature. The Latin word for knowledge is cognitio from which comes the French word connaissance, to be 'born with' something. One can develop the sophistication of computers indefinitely on the horizontal level, enabling them to execute more and more complex mathematical operations, but they do not by that rise one millimeter vertically towards true cognitive thought and hence towards consciousness.

The existence of human consciousness doesn't prove that the human mind is in any way separate from the physical world, just that reality is not all reductible to atomic structure and the laws of physics. Matter, too, has an immaterial component to it - something that cannot be visualised but only intellectually grasped. I gave an example somewhere else: a black hole. A black hole has no volume at all. It occupies a mathematical point with zero width, breadth and height. It has no atomic or subatomic structure, in fact it has no structure at all. It has only mass and gravity. It's impossible to visualise it. It can only be grasped intellectually.

 

[RX-79G]

Justin,
You seem to be arguing that a human mind has something in it that allows consciousness, but that thing is completely hidden, inaccessible and undefinable. If it is something in our universe, what is it and why could an analogue not be constructed?

I don't agree with your assessment of computers. Base 2 expression of math is no different than any other expression of math - any math can be expressed in an base system. Human beings can manipulate mathematical concepts, and they can do arithmetic, but we have trouble actually picturing and keeping track of much more than three objects at a time. We can only do math because we turn objects into concepts, then perform logic transformations on those concepts.

A black hole has no volume at all. It occupies a mathematical point with zero width, breadth and height.

This isn't true. Where did you hear this?

 

[Igor]

The -1/12 thingie is nonsense. I even wrote about it:
http://www.dedoimedo.com/physics/sum-1-2-3-4-bollocks.html

But as to whether we are in a simulation or not, it makes no difference really.

Igor

 

[Dawnstorm]

Given enough time and assuming "God" is not a prerequisite for consciousness, then the likeliest explanation of any universe is the existence of an ordinal universe that created it.

So that's what this is about?

I think what's been confusing me is that Musk was talking about indistinguishability. I don't see how that matters. A consciousness that runs on a machine doesn't have to be anything like the world we simulate. If simple consciousnesses are possible, then we can have conscious Pong bats. Why not? A consciousness operates on what it knows. What do photo-realistic 3D environments have to do with it?

Also, is it universes all the way down?

ETA: @ - 1/12: Here. I've seen something more complicated and abstract, too. Basically, it's cheating that works, but can also create alternative mathematics theory, or something (at some point I had to bow out).

 

[Justin Swanton]

You seem to be arguing that a human mind has something in it that allows consciousness, but that thing is completely hidden, inaccessible and undefinable. If it is something in our universe, what is it and why could an analogue not be constructed?

Not indefinable, just not reductible to any physical process.

I don't agree with your assessment of computers. Base 2 expression of math is no different than any other expression of math - any math can be expressed in an base system. Human beings can manipulate mathematical concepts, and they can do arithmetic, but we have trouble actually picturing and keeping track of much more than three objects at a time. We can only do math because we turn objects into concepts, then perform logic transformations on those concepts.

The base doesn't matter. My point is that a CPU can mechanically produce the processes of mathematical calculation, but it has no conception of mathematics as we do. It's just an automatic execution of programmed instructions. In the same way a garage door doesn't 'recognise' that you've arrived home when you hit the remote button. The remote just triggers a series of automated actions that open the door.

A black hole has no volume at all. It occupies a mathematical point with zero width, breadth and height.

This isn't true. Where did you hear this?

Sorry, I made an assumption. A little digging turns up the fact that there is no certain knowledge of what is going on past the event horizon. There might be a singularity with zero volume and infinite density as predicted by General Relativity, but this may not be the case. Nature doesn't like infinities. So scratch that example.

 

[Fung Koo]

Not indefinable, just not reductible to any physical process.

Per the Anthropic Principle, though, the physical conditions of the universe must physically allow for sapience. Otherwise, you're putting consciousness into the realm of the supernatural.

I gather, then, you're arguing from a theistic perspective.

In the same way a garage door doesn't 'recognise' that you've arrived home when you hit the remote button. The remote just triggers a series of automated actions that open the door.

Right. The article I posted addresses this.

At the purely functional level, our bodies are only slightly more complex machines than the garage door opener, really. We could, in theory, simulate an "apparently intelligent" human that is, in effect, nothing more advanced than a really fine-tuned garage door opener. It could look, function, and act like us. Given enough inputs and enough variables, there is every reason to expect that Siri will become even better than the "Computer Lady" in Star Trek, and if we give Siri a human body, there is certainly a line at which she will become indistinguishable from a human, which is what Dick gets at in DADOES.

This is also the whole idea at play in Asimov's Robot Novels. If we can create a human-analogue that appears sentient/conscious/sapient, then how do we really know that it's not?

If the robot asserts that it is thinking (I think, therefore I am), then who are we disbelieve it?

That's only one kind of sentience we might make. That one we would know we made, and it would live in our world.

But if we go to the Anthropic Principle and consider our own sentience as naturally occurring, then if we can simulate a human with a robot that is indistinguishable form a human, then we can certainly eventually simulate an entire reality that is indistinguishable from reality. And more than that, we could simulate unreal realities -- realities unlike our own. And within those conditions, if our reality can give rise to consciousness, then there's no reason to expect a simulation could not.

Sorry, I made an assumption. A little digging turns up the fact that there is no certain knowledge of what is going on past the event horizon. There might be a singularity with zero volume and infinite density as predicted by General Relativity, but this may not be the case. Nature doesn't like infinities. So scratch that example.

And remember that zero, in some senses, is another way of expressing infinity.

 

[Fung Koo]

I think what's been confusing me is that Musk was talking about indistinguishability. I don't see how that matters. A consciousness that runs on a machine doesn't have to be anything like the world we simulate. If simple consciousnesses are possible, then we can have conscious Pong bats. Why not? A consciousness operates on what it knows. What do photo-realistic 3D environments have to do with it?

I think "indistinguishability" here means kind of what you do -- the consciousness within a given reality can't discern that its reality is simulated relative to some other reality.

Also, is it universes all the way down?

Taken to the logical conclusion of the argument: yes. Also all the way up.

It's basically the "grain of sand" theory. But if we add sentience to our consideration of physics, then it ceases to be only a natural phenomenon. There might be a universe in every grain of sand, but there are perhaps also collections of grains of sand that have been arranged/designed by sentient forms to work in concert that also give rise to universes with sentient forms in them.

So maybe it's not just all the way up and down, but also all the way in and out.

 

[RX-79G]

The base doesn't matter. My point is that a CPU can mechanically produce the processes of mathematical calculation, but it has no conception of mathematics as we do.

A CPU has no conception of anything. But it isn't that way because it does its "thinking" in binary. An organic brain is still the most complex device in the universe. Until a comparably complex CPU is built there shouldn't be any assumption about what a computer can or can't do.