I challenge you: your best alternate to materialism

[Tassman]

That's a fun one. The problem with electronic consciousness is it leads to an absurd conclusion.

If a machine can become conscious, then it's possible for a system of electronic switches, transformers, rheostats, etc. to become conscious (if it's sufficiently complex and everything is arranged a certain way). If THAT'S possible, then it's possible to replace the moving electrons with moving water molecules so that a system of vales, pumps, pressure regulators, etc. is the functional equivalent of the system that is producing machine consciousness. If it's functionally equivalent (i.e., does the same thing that our brains do (or the conscious machine brain does), but with a different substrate) there's no reason it shouldn't be conscious.

Which LEADS to the inevitable conclusion that if you move rocks around in the right way, you can produce a simulation of the entire universe, conscious beings and all.
https://xkcd.com/505/

No more absurd than consciousness emerging from brains made of meat, surely?

Great cartoon BTW.

 

[MRC_Hans]

That's a fun one. The problem with electronic consciousness is it leads to an absurd conclusion.

If a machine can become conscious, then it's possible for a system of electronic switches, transformers, rheostats, etc. to become conscious (if it's sufficiently complex and everything is arranged a certain way). If THAT'S possible, then it's possible to replace the moving electrons with moving water molecules so that a system of vales, pumps, pressure regulators, etc. is the functional equivalent of the system that is producing machine consciousness. If it's functionally equivalent (i.e., does the same thing that our brains do (or the conscious machine brain does), but with a different substrate) there's no reason it shouldn't be conscious.

Correct. What is absurd about that?

Which LEADS to the inevitable conclusion that if you move rocks around in the right way, you can produce a simulation of the entire universe, conscious beings and all.
https://xkcd.com/505/

Well, the universe does indeed consist of "rocks" that are moved around in the right way. So, what's the problem?

Hans

 

[Fudbucker]

No more absurd than consciousness emerging from brains made of meat, surely?

Great cartoon BTW.

That is very true, which is why consciousness is often referred to as the "hard problem". Explaining how a bunch of neurons can give rise to subjective experiences has made some pretty smart people say some pretty far-out things.

 

[Fudbucker]

Correct. What is absurd about that?



Well, the universe does indeed consist of "rocks" that are moved around in the right way. So, what's the problem?

Hans

The problem is that if the cartoon is true (that moving rocks in some particular manner can produce consciousness), then consciousness can result from anything being moved around: abacus beads, electrons, meteor swarms, sand particles, etc.

It's basically panpsychism, and it's the biggest problem for materialism at the moment. No one has any idea why arranging a bunch of switches in a particular way gives rise to the experience of pain, pleasure, self-awareness, etc.

That's why you have people like Christof Koch speculating that consciousness (like mass) is a fundamental property of the universe, and any simulated consciousness will be "black inside", and not able to experience anything.

 

[Myriad]

That's why you have people like Christof Koch speculating that consciousness (like mass) is a fundamental property of the universe, and any simulated consciousness will be "black inside", and not able to experience anything.

If consciousness (like mass) is a fundamental property of the universe, the simulator, existing within the universe, would be conscious anyhow. So if the simulation wasn't, it wouldn't matter.

From the same premises, it's your brain that's conscious, not your mind. Is that a distinction that makes any difference? If anything it seems more materialist than otherwise.

 

[MRC_Hans]

The problem is that if the cartoon is true (that moving rocks in some particular manner can produce consciousness), then consciousness can result from anything being moved around: abacus beads, electrons, meteor swarms, sand particles, etc.

Well, maybe it can. Although it is difficult to imagine how you could achieve the necessary complexity, with such simple components. But what is the problem?

It's basically panpsychism, and it's the biggest problem for materialism at the moment. No one has any idea why arranging a bunch of switches in a particular way gives rise to the experience of pain, pleasure, self-awareness, etc.

I'm sorry, but I fail to see the problem. If a sufficiently complex calculating mechanism interacts with its environment to give those results, why should we doubt, let alone see it as a problem, that such a system is conscious?

That's why you have people like Christof Koch speculating that consciousness (like mass) is a fundamental property of the universe, and any simulated consciousness will be "black inside", and not able to experience anything.

That is merely assertions, and unfounded.

Hans

 

[Fudbucker]

If consciousness (like mass) is a fundamental property of the universe, the simulator, existing within the universe, would be conscious anyhow. So if the simulation wasn't, it wouldn't matter.

How could consciousness be a fundamental property of the universe? Koch compares it to mass. There would have to be some type of field (like the Higgs field) responsible for consciousness. How would that work? Should we start looking for the Higgs Boson of consciousness?

From the same premises, it's your brain that's conscious, not your mind. Is that a distinction that makes any difference? If anything it seems more materialist than otherwise.

Well, some materialists would claim that the brain (or parts of it, at least) are conscious. There are other materialists who don't think mental states are the same as brain states. The brain causes mental states but a brain-state is not the equivalent of a mental-state. I don't think those materialists would be comfortable saying the brain is conscious. More like the brain causes consciousness. I think it's called "property dualism".

It's the "how" that is the problem. How do neurons sending and receiving electrical impulses produce the feeling of being conscious? And if neurons can do it, then anything functionally equivalent to a neuron should be conscious, if you arrange them in a way that is functionally equivalent to a working brain. There are all sorts of odd substrates that can be the functional equivalent of a working brain: ropes and pulleys, water and valves, lightswitches, etc.

 

[Fudbucker]

Well, maybe it can. Although it is difficult to imagine how you could achieve the necessary complexity, with such simple components. But what is the problem?



I'm sorry, but I fail to see the problem. If a sufficiently complex calculating mechanism interacts with its environment to give those results, why should we doubt, let alone see it as a problem, that such a system is conscious?

It's obviously not a problem if you think shifting sand dunes or cloudbursts can produce consciousness. Or that if rocks are moved around in a certain way, you can simulate a whole universe of conscious beings (I've always called that idea "rock-consciousness").

But I think rock-consciousness, and the possibility that there may be "flashes" of consciousness going on all around us as particles randomly move around is ad-hoc. The only reason some materialists take it seriously is that "rock-consciousness" is the logical end-step in a sequence that begins with whatever is functionally equivalent to a working brain is conscious.

The possibility of conscious sand dunes is like Bigfoot- yeah it's possible, but why should I even entertain the notion? Where's the proof? The possibility of a conscious landslide seems very woo-ish, and not something we should consider plausible unless there is very good evidence for it.

That is merely assertions, and unfounded.

Hans

Well, they're assertions, but you'll have to argue for why they're unfounded. Koch is one of the top scientists in the world when it comes to consciousness.
There is no problem qualifying him as a preeminent expert on consciousness (both computational and biological).

 

[Myriad]

The problem is that if the cartoon is true (that moving rocks in some particular manner can produce consciousness), then consciousness can result from anything being moved around: abacus beads, electrons, meteor swarms, sand particles, etc.

It's basically panpsychism, and it's the biggest problem for materialism at the moment. No one has any idea why arranging a bunch of switches in a particular way gives rise to the experience of pain, pleasure, self-awareness, etc.

Koch aside, you make quite an unjustified leap from the idea that a conscious entity could be composed of abacus beads or electrons, to the idea that any old collection of abacus beads or electrons must be conscious. That's like saying if a tuna salad is only a particular kind of arrangement of ordinary atoms (without any requirement of ineffable non-material tuna salad nature), all collections of matter are tuna salads.

Only certain very specific arrangements of rocks or other things have been shown to be computationally universal, which as far as we know is a requirement for consciousness (and a very minimal one at that, like saying that a riot must at least have one person present). Those systems have certain features in common, such as evolution based on nonlinear threshold-based state changes, sufficient numbers of simultaneous independent elements to represent a set of symbols, and ability to both sustain stable configurations and alter them. That leaves out your meteor swarms and piles of sand particles.

What these mere-lifeless-rocks kind of arguments leave out is the scale of computational complexity involved. It's one thing to show that a very large arrangements of track and railroad switches (and some additional mechanisms that can flip a given switch as a result of the train passing a given point), with a single train running along it, can be computationally universal, and therefore conscious in theory. Because of scale, that doesn't mean that there's any possibility for your local railroad yard to become (or turn out to be) conscious. To actually make a universal railroad machine emulate the simplest processors we use in present day computers, you'd need an array of track the size of a continent. To scale that up to the computational capability of something like IBM's Watson, the entire earth's surface would be far from sufficient for all the tracks; the surface of a sphere the size of earth's orbit around the sun might do. But actual consciousness probably requires more computation than that, so Saturn's orbit might do instead. (That would be way more railroad switches than neurons in the brain, but that's because each neuron does the work of many thousand switches.)

This system would also be extremely slow. To see it do anything interesting, you'd have to speed up the train (or slow down your perception of time) so that the train zips from switch to switch and from one side of the solar system to another far too fast to see; you'd instead see switches changing everywhere nearly all at once, like neurons firing in a brain. Then, maybe, the consciousness could occur at a suitable time scale for you to interact with it, especially if you also shrink the sphere down to maybe the size of a basketball.

That's more or less what you'd have to do to make a railroad system conscious. But it doesn't much resemble a train with a few tracks and switches any more, does it? What you'd actually have would be a sphere covered with complex sub-microscopic patterns of incomprehensibly vast complexity, where you can barely perceive (using sensitive instruments) hundreds of trillions of switches opening and closing in constant waves of back-and-forth chatter. Much like we see when we study a brain.

Brains are more efficient, though, and can therefore operate at the time scale we're familiar with without the radical slowing down/speeding up. (No one said railroad tracks and switches were a good engineering choice for building a conscious entity.)

 

[MRC_Hans]

It's obviously not a problem if you think shifting sand dunes or cloudbursts can produce consciousness. Or that if rocks are moved around in a certain way, you can simulate a whole universe of conscious beings (I've always called that idea "rock-consciousness").

Two points:

1) I'm not talking about sand dunes or cloudbursts. Those are algrithmically quite simple and mainly random events. I'm talking about a sufficiently complicated computation device. That is something entirely different.

2) Rocks are a subset of the whole universe. It is not likely that a subset of the universe can mimick the whole. That is why I put rocks in quotation marks when I said the universe consists of moving "rocks".

But I think rock-consciousness, and the possibility that there may be "flashes" of consciousness going on all around us as particles randomly move around is ad-hoc. The only reason some materialists take it seriously is that "rock-consciousness" is the logical end-step in a sequence that begins with whatever is functionally equivalent to a working brain is conscious.

No, it is not a logical end-step. It is an extreme example, which is very unlikely to be real: The conscious brain is extremely complex and systematic, and randomly moving rocks are no more likely to produce consciousness than a random letter generator is likely to produce the works of Shakespeare.

The possibility of conscious sand dunes is like Bigfoot- yeah it's possible, but why should I even entertain the notion? Where's the proof? The possibility of a conscious landslide seems very woo-ish, and not something we should consider plausible unless there is very good evidence for it.

It IS entirely implausible. Just not impossible.

Hans

 

[Fudbucker]

Koch aside, you make quite an unjustified leap from the idea that a conscious entity could be composed of abacus beads or electrons, to the idea that any old collection of abacus beads or electrons must be conscious. That's like saying if a tuna salad is only a particular kind of arrangement of ordinary atoms (without any requirement of ineffable non-material tuna salad nature), all collections of matter are tuna salads.

I don't think you can brush aside Koch's comments so easily. Koch is an expert that has spent decades studying consciousness. For someone like Koch to speculate there might a conscious field in the universe, similar to a Higgs field, has to be taken seriously. It's like one of the top climatologists in the world claiming the Earth's warming is due to dark energy fluctuations.

But putting Koch aside, yeah, whatever is being moved around (rocks, electrons, water molecules) has to be part of a sufficiently complex system that is arranged in the right way. But that begs an obvious question: why does one arrangement of neurons (and their corresponding activity) correspond to the pain of stubbing your toe, while another arrangement corresponds to the pleasure of eating a a nice tuna salad.

In the comic I linked about moving rocks around and simulating a universe of conscious beings, the rocks are moved a certain way, but why should one particular pattern of rock movement be more significant than another? Why does rock movement pattern 1100101010010101010101010101 produce a moment of consciousness while pattern 0010100101001010010100010101 doesn't produce anything?

Also, it's worth noting that in the comic, it's a person moving the rocks around with some pattern in mind, but of course a person isn't required, because whatever pattern a person comes up with can be duplicated by a random rock placing machine, given an infinite amount of time (or rock movers). What that means is that there is the possibility that a shifting sand dune (for example), just by random chance, could reproduce a pattern of sand movement that corresponds exactly to a specific pattern of rock movement that is supposed to produce some instant of consciousness.

Now, what are the odds that a specific sand dune (or rock slide or cloudburst or bunch of dust specks) could be conscious? Unknown, but the real possibility that flashes of consciousness are occurring all around us is a startling one, and one you wouldn't expect on materialist model. The materialistic account of consciousness is that it is something brains do, and you can't get much further away from a brain than a bunch of moving sand particles.

Consciousness is something brains do, and moving rocks around isn't remotely similar to what brains do, so why should we entertain the notion that consciousness can be produced by moving a bunch of rocks around "in the right way"? Where's the evidence for that claim? What's the causal mechanism whereby a bunch of rocks being moved around produces a simulation of a universe with conscious beings?

Only certain very specific arrangements of rocks or other things have been shown to be computationally universal, which as far as we know is a requirement for consciousness (and a very minimal one at that, like saying that a riot must at least have one person present). Those systems have certain features in common, such as evolution based on nonlinear threshold-based state changes, sufficient numbers of simultaneous independent elements to represent a set of symbols, and ability to both sustain stable configurations and alter them. That leaves out your meteor swarms and piles of sand particles.

How so? Let's take an instance of rock movement that corresponds to a moment in the simulated universe (I'll use binary, but it really doesn't matter). According to the comic, if I move the rocks around in a certain way, I can simulate a universe of conscious beings. So let's suppose that rock pattern 1001010010100010101001010011100...1 (put however many rocks you want in the ellipsis) produces a nanosecond of simulated universe time.

Why can't that be duplicated by a random process with another substrate? If that pattern of rock movement produces a nanosecond of a simulated universe, why can't we do the same with sand particles? Or water molecules? Is there something special about rocks? To avoid special pleading, any unique movement of particles should produce a moment of a simulated universe, as long as the particles match the movement of the rocks "in the right way" ("the right way" is whatever particular arrangement of rocks produces some instance of a simulated universe) .

And if that's not convincing, then forget the comic. Do you agree that we can make a mechanical brain that is the functional equivalent of a working organic brain? Would't the functionally equivalent mechanical brain be conscious, no matter what it's made out of? Transistors? Valves? Light switches? What is so special about organic substrates?

What these mere-lifeless-rocks kind of arguments leave out is the scale of computational complexity involved. It's one thing to show that a very large arrangements of track and railroad switches (and some additional mechanisms that can flip a given switch as a result of the train passing a given point), with a single train running along it, can be computationally universal, and therefore conscious in theory. Because of scale, that doesn't mean that there's any possibility for your local railroad yard to become (or turn out to be) conscious.

Right, but in the comic, there really are no mechanisms, other than a guy moving rocks around. Are you saying that's not enough to simulate a universe with conscious beings? Because if it is enough, then you don't need trains and switches. You can scale the whole thing down exponentially, so that the quadrillions of drops in a rain shower could randomly move the right way to produce an instant of consciousness.

But to take your example of a railroad, the only thing stopping it from being conscious is scale, correct? So if you had a large enough railroad, and arranged it the right way, it should become conscious. Doesn't that seem a little strange? Again, I go back to the only thing we know can produce consciousness: an organic brain. The same objections for railroad-consciousness apply as rock-consciousness: trains moving on railroads aren't anything like what an organic brain does, so why should we even entertain the idea that a railroad (even if sufficiently complex and arranged the right way) can become conscious? And if it become conscious, what would it feel? It has no nerves, so how could it feel pain or pleasure? Would it be murder to dismantle it?

To actually make a universal railroad machine emulate the simplest processors we use in present day computers, you'd need an array of track the size of a continent. To scale that up to the computational capability of something like IBM's Watson, the entire earth's surface would be far from sufficient for all the tracks; the surface of a sphere the size of earth's orbit around the sun might do. But actual consciousness probably requires more computation than that, so Saturn's orbit might do instead. (That would be way more railroad switches than neurons in the brain, but that's because each neuron does the work of many thousand switches.)

It's not the scale that's the issue. It's the assertion that a railroad, no matter how big and complex, could become conscious. At what point would it become conscious? How many miles of track would be required? How many trains? These are the same questions we can ask about brains (e.g., how many neurons are required to produce a conscious experience), but at least we know collections of neurons can produce consciousness. Where is the evidence that collections of trains and tracks can do the same?

This system would also be extremely slow. To see it do anything interesting, you'd have to speed up the train (or slow down your perception of time) so that the train zips from switch to switch and from one side of the solar system to another far too fast to see; you'd instead see switches changing everywhere nearly all at once, like neurons firing in a brain. Then, maybe, the consciousness could occur at a suitable time scale for you to interact with it, especially if you also shrink the sphere down to maybe the size of a basketball.

Yeah, it wouldn't be a quick-thinker, but it's the possibility of a railroad system being a thinker at all that's a problem for materialism.

That's more or less what you'd have to do to make a railroad system conscious. But it doesn't much resemble a train with a few tracks and switches any more, does it? What you'd actually have would be a sphere covered with complex sub-microscopic patterns of incomprehensibly vast complexity, where you can barely perceive (using sensitive instruments) hundreds of trillions of switches opening and closing in constant waves of back-and-forth chatter. Much like we see when we study a brain. Brains are more efficient, though, and can therefore operate at the time scale we're familiar with without the radical slowing down/speeding up. (No one said railroad tracks and switches were a good engineering choice for building a conscious entity.)

Right, the railroad brain would have to be functionally equivalent to a working brain, so it would probably start to resemble one before too long. But does that mean you object to the comic I linked? In the case of the comic, it doesn't resemble a brain in any way. It's just a guy moving rocks around on an endless plain.

 

[Fudbucker]

Two points:

1) I'm not talking about sand dunes or cloudbursts. Those are algrithmically quite simple and mainly random events. I'm talking about a sufficiently complicated computation device. That is something entirely different.

2) Rocks are a subset of the whole universe. It is not likely that a subset of the universe can mimick the whole. That is why I put rocks in quotation marks when I said the universe consists of moving "rocks".




No, it is not a logical end-step. It is an extreme example, which is very unlikely to be real: The conscious brain is extremely complex and systematic, and randomly moving rocks are no more likely to produce consciousness than a random letter generator is likely to produce the works of Shakespeare.




It IS entirely implausible. Just not impossible.

Hans

1. If it's not impossible, then it's possible. If it's possible, then given enough sand dunes, some will be conscious (if only for a brief instant).
Or, to put it in modal terms, there is some possible world where sand dunes are conscious. Do you agree with that?

2. What makes it "entirely implausible. Just not impossible"? What are the odds of a sand dune becoming conscious, even for a moment? Show your work, please.

 

[MRC_Hans]

1. If it's not impossible, then it's possible. If it's possible, then given enough sand dunes, some will be conscious (if only for a brief instant).
Or, to put it in modal terms, there is some possible world where sand dunes are conscious. Do you agree with that?

2. What makes it "entirely implausible. Just not impossible"? What are the odds of a sand dune becoming conscious, even for a moment? Show your work, please.

Don't be silly, how should I be able to calculate the odds? However, there is a different factor: Time. Consciousness is not just a product of complexity, it is also a product of experience, and that requires time.

So a sufficiently complex and suitable configured system will not be conscious the moment it exists, it will only have the potential to be. It's concsiousness will evolve as it gathers and coordinates experiences. Even a human infant is not self-aware according to current tests till it is a couple of years old.

I'm sorry if you perceive this as moving the goalposts, but I actuallt expected you already knew this.

- Obviously, a random collection of stuff, while hypothetically in the proper configuration, is not likely to become conscious. It must also have the opportunity to gather and process information.

The materialist position is that consciousness is not dependent on hardware type, but it is certainly dependent on configuration and experience.

As for Koch, just because he is an important authority, he does not have to be right in everything he says. Or, because he forwards a hypothesis, it doesn't even mean he himself assumes it is the truth.

Hans

 

[punshhh]

why?

Because the cell is the unit of life, each life form consists of one cell, or a colony of cells.

 

[punshhh]

No, not chemical. It s called an algorithm.

Is an algorithm virtual, or physical?

Why don't you think it can be defined as an algorithm? If it can't, how will you define it at all?

In each person "Russell's Teapot" is known by a unique algorithm, or combination of neural activity. Perhaps theoretically the concept could be represented in an algorithm, it would however be necessary for the algorithm to include the entire life experience of the person in order to analytically identify the algorithm resulting in what is known by that person.

In other words, the concept "Russell's Teapot", is known (read) by a fully developed person and by no other route. It can be represented by a person with a symbol or two, or communicated between persons. But any representation outside the sphere of persons/humanity, or any external analysis of a physical brain, (in the absence of consultation with the person) would be exceedingly complicated and probably could not be done without a replication of a whole person, the intellectual learning of their culture and including conscious entities.

 

[punshhh]

(About the robot). The ant is acting out for the benefit of the antheap.

Yes and it has life experience like myself.

Define experience, and I'll tell you.

Experience is what occurs to a living entity.

 

[MRC_Hans]

Is an algorithm virtual, or physical?

It is data.

In each person "Russell's Teapot" is known by a unique algorithm, or combination of neural activity. Perhaps theoretically the concept could be represented in an algorithm, it would however be necessary for the algorithm to include the entire life experience of the person in order to analytically identify the algorithm resulting in what is known by that person.

Since every person has his/her individual algorithm/concept, obviously an artificial entity could have its own, equally valid, individual algorithm/concept for it.

Hans

 

[MRC_Hans]

Yes and it has life experience like myself.

Hardly. I should hope that your life experience is vastly different from that of an ant.

Experience is what occurs to a living entity.

Circular argumentation.

Hans

 

[MuDPhuD]

That's why you have people like Christof Koch speculating that consciousness (like mass) is a fundamental property of the universe, and any simulated consciousness will be "black inside", and not able to experience anything.

I think you may be misunderstanding his position. Yes, he does say "simulations" will not be conscious, but in the same interview he also says,

"If you were to build a computer that has the same circuitry as the brain, this computer would also have consciousness associated with it. It would feel like something to be this computer."

It is a matter of computation involving integrated information in a manner similar to what the brain does that will produce consciousness. If the binary computation involving moving rocks carried out the same computational functions as the brain it would generate a consciousness. Personally I believe this to be obviously practically impossible, as I think we all recognize, but the point is that any computation which replicates the necessary parts of brain activity, regardless of substrate, will produce consciousness.
I think there may be some confusion about what "simulation" means.

 

[LarryS]

I'm curious what if any motivations there would be for artificial consciousness, that is, what intentionality would there possibly be for a conscious device w/o curiosity, hunger or thirst, sex drive, etc?