6 ms·

I like how quickly this line of thinking leads to panpsychism. Alright, so your dog has some level of cognition – sensations, emotion, joy, pain. Perhaps less

by skosch 8y ago

I like how quickly this line of thinking leads to panpsychism. Alright, so your dog has some level of cognition – sensations, emotion, joy, pain. Perhaps less complex than a human, but nevertheless nonzero. Because whatever computations happens in its brain feel like something from the inside. How about a frog? Less, but still nonzero. How about a worm? Less, but still nonzero. But hey: we can already model a worm's nervous system [0]. And if consciousness is just what computation feels like from the inside, then torturing the model worm creates real pain (or the worm-equivalent of pain, whatever that emotion is). Something interesting to think about. [0] http://openworm.org http://openworm.org

pygy_ 8y ago
You can get that down to single cells, actually. Transcription factors (genes whose product activate or inhibit other genes) form regulation network that are equivalent in function to neural nets. There's also evidence of memory in single cell organisms.
- skosch 8y ago
  Absolutely, and why stop there? Neural nets just add up numbers – other physical systems do that, too.
  - pygy_ 8y ago
    Indeed!
hjek 8y ago
> But hey: we can already model a worm's nervous system. It surely is interesting to think about, at least as long as this Chinese Room way of thinking isn't used to dismiss or justify pain or suffering inflicted on other living beings.
- chr1 8y ago
  There is one important difference between other living beings and the model, in the case of other living beings you do not own the computer, you are not the one who started the computation, and you do not have a power to undo your actions. But in case of the model nothing would exist if you didn't build the computer, and you can take snapshots to restore any state you want. It seems like people in some way always understood this difference, because they were ok with god sending arbitrary tortures to them, and then granting them eternal life undoing the effect of the torture.
Retra 8y ago
You're just going down the road which leads to the failures of moral absolutism. The conclusion is not that pain can't exist in a machine, or that it is wrong to torture a simulated being. The conclusion is that torture is not wrong simply because it causes pain, but because the pain it causes leads to otherwise undesirable consequences. Eventually you're going to have to decide what is intrinsically undesirable, which means understanding what your goals are before you ascribe moral value to some process. You can't say pain is good or bad without first understanding how it gets in the way of your goals.
koboll 8y ago
We can, without lapsing into full panpsychism, state some ethical paradoxes about the way we can observe animal emotion, versus the way we actually treat animals. For example, it's widely accepted that pigs are extremely intelligent animals, capable of feeling pain and suffering to a degree more sophisticated even than young children. And yet treating young children with the same level of disregard-for-suffering that we routinely, banally treat pigs with is considered a capital crime. At a certain point, we'll have to resolve this societal cognitive dissonance. I strongly suspect it will end up being resolved with animals being granted a far greater degree of legal protection from pain and suffering than they have now.
- pbhjpbhj 8y ago
  >At a certain point, we'll have to resolve this societal cognitive dissonance. // It appears that the majority of the World, setting aside the question of the the truth of the notion for now, believes that this is entirely solved as the child has a soul and the pig does not? So, for most there is no dissonance?
  - slfnflctd 8y ago
    Most of the world probably does believe this, which is something I find profoundly disturbing. We have a long way to go as a species in developing a sane ethics system. However, considering the progress we've already made, I have hope that we can make more.
    - chr1 8y ago
      Perhaps ethics is not something absolute, and is a game thorethic optimal strategy that changes as the rules of the game change. If hunting is the only thing you can do sane ethics system would allow killing the whale no matter how intelligent creature it is. And letting members of another tribe die instead of trying to share and risking members of your own tribe to die is a sane thing too. But things change if you a member of a civilization that can produce as much food as it wants without killing anyone.
chr1 8y ago
This is indeed very interesting line of thinking, and if one agrees that there is no soul, and there is no quantum process essential to life that can not be modeled on a classical computer, then this is the only remaining way of thinking. But this is different from dictionary definition of panpsychism, because this doesn't need to assign consciousness to simpler computations (rocks, buildings, simple cells). And this indeed opens really interesting questions. For instance if you have a simulation of many people, is it ok to continue the simulation when they are killing each other? Is it ok to ever stop the simulation? Is it ok to create a model of a person you do not like and torture him? What if your brain is augmented with a powerful computer, and torturing happens literally in your imagination?
- skosch 8y ago
  > this doesn't need to assign consciousness to simpler computations (rocks, buildings, simple cells). That necessarily follows, I think. The quantum processes in a rock are also a form of computation and, therefore, inhabited by some rudimentary form of experience we might call, dunno, "rockness". I don't see an easy way around it. Perhaps at absolute zero all experience stops? I have no answers, but your questions are precisely why I find this fascinating to think about!
  - chr1 8y ago
    I would expect there to be very marked jump between the rockness and the computations with the simplest consciousness, that will allow to assign them to different classes instead of having a continuous spectrum going to zero with temperature. Kind of similar to the way how behavior of cellular automata changes between the predictable patterns and chaotic ones. (But i wouldn't bet all my money on it, only the half:)
    - skosch 8y ago
      > Kind of similar to the way how behavior of cellular automata changes between the predictable patterns and chaotic ones. Hm, reminds me of Permutation City. That's an interesting parallel I hadn't considered at all.
incadenza 8y ago
I don’t think it necessitates pansychism. You can still think of consciousness as a by product of a specific configuration, or just sufficient quantity of neurons or whatever. However it emerges remains a mystery, but it’s possible that some configuration of them does not produce the consciousness affect. For example, a million neurons is not enough to turn the lights on, but X amount is.
- skosch 8y ago
  Self-aware consciousness is a pretty high bar, but how about the ability to experience suffering? The ability to experience reward? That question very quickly turns into "what is experience", or rather, "what isn't experience". I don't find any of this intuitive, nor do I have answers to the ethical questions that follow (besides not eating animals, which I already don't). But I find it revealing how most people will fight tooth and nail to maintain that there is some magical threshold separating dead matter from precious, conscious life, without having given the topic much thought at all.
  - incadenza 8y ago
    Yeah, it is pretty remarkable. I consider it non controversial at this point that many of humans favorite meals are certainly capable of a lot of suffering.
    - pas 8y ago
      I think that was never a real question. Lobsters might or might not feel pain, okay, but bovines and other mammals certainly do. And birds are (can be) pretty intelligent (eg parrots). So that's why they are supposed to be kept, raised and slaughtered in relatively humane conditions. (Eg. free range, and painless death.)
  - pas 8y ago
    The really nasty question is, do animals have a(n intuitive or other) concept of the future, and/or of causality? Do they feel dread? And when they experience suffering, can they know what does that mean to them? Elephants might ( https://en.wikipedia.org/wiki/Elephant_cognition#Death_ritual https://en.wikipedia.org/wiki/Elephant_cognition#Death_ritua... ). Of course, life in the wild is not pretty. Not many of the wild animals are apex predators in their respective territory, and thus their deaths are usually not peaceful, and being hunted, having close calls with death almost daily is a very integral part of their life, so they probably have an evolved way to deal with the psychological burden of this. And that's probably what domestication changed.
c_shu 8y ago
when an error occurs in a machine and the machine tries to recover, maybe that's also a form of pain. A machine is not necessarily less sophisticated than any living creature.
gurumeditations 8y ago
That is one of the most interesting things to me about this. Consciousness seems to arise from the communication and computation in and between sensory inputs, so do computers feel anything? I can understand the logic of input 1 causes output 2 and output 5 to do something, providing the objective indicators of consciousness, but the internal experience puzzles me. Why do we feel? Why do our pain or pleasure receptors result in agonizing pain or pleasurous feeling instead of just throwing messages saying “you are experiencing 57% pleasure” or “pain at 25%”? Why do we see what is going through our visual system? Why are we conscious at all? Why aren’t we just autonomous? Systems of inanimate material seem to have this bizarre innate property where they create subjective ephemeral forces from objective inanimate atoms. Which raises the question: do computers feel?
muzani 8y ago
What about something like an anthill? It changes states, it follows patterns similar to other anthills, and probably wishes to preserve itself in some way. It's probably extremely low, but is it nonzero?

by skosch 8y ago

pygy_ 8y ago

You can get that down to single cells, actually. Transcription factors (genes whose product activate or inhibit other genes) form regulation network that are equivalent in function to neural nets. There's also evidence of memory in single cell organisms.

skosch 8y ago
Absolutely, and why stop there? Neural nets just add up numbers – other physical systems do that, too.
- pygy_ 8y ago
  Indeed!

hjek 8y ago

> But hey: we can already model a worm's nervous system. It surely is interesting to think about, at least as long as this Chinese Room way of thinking isn't used to dismiss or justify pain or suffering inflicted on other living beings.

chr1 8y ago
There is one important difference between other living beings and the model, in the case of other living beings you do not own the computer, you are not the one who started the computation, and you do not have a power to undo your actions. But in case of the model nothing would exist if you didn't build the computer, and you can take snapshots to restore any state you want. It seems like people in some way always understood this difference, because they were ok with god sending arbitrary tortures to them, and then granting them eternal life undoing the effect of the torture.

Retra 8y ago

You're just going down the road which leads to the failures of moral absolutism. The conclusion is not that pain can't exist in a machine, or that it is wrong to torture a simulated being. The conclusion is that torture is not wrong simply because it causes pain, but because the pain it causes leads to otherwise undesirable consequences. Eventually you're going to have to decide what is intrinsically undesirable, which means understanding what your goals are before you ascribe moral value to some process. You can't say pain is good or bad without first understanding how it gets in the way of your goals.

koboll 8y ago

We can, without lapsing into full panpsychism, state some ethical paradoxes about the way we can observe animal emotion, versus the way we actually treat animals. For example, it's widely accepted that pigs are extremely intelligent animals, capable of feeling pain and suffering to a degree more sophisticated even than young children. And yet treating young children with the same level of disregard-for-suffering that we routinely, banally treat pigs with is considered a capital crime. At a certain point, we'll have to resolve this societal cognitive dissonance. I strongly suspect it will end up being resolved with animals being granted a far greater degree of legal protection from pain and suffering than they have now.

pbhjpbhj 8y ago
>At a certain point, we'll have to resolve this societal cognitive dissonance. // It appears that the majority of the World, setting aside the question of the the truth of the notion for now, believes that this is entirely solved as the child has a soul and the pig does not? So, for most there is no dissonance?
- slfnflctd 8y ago
  Most of the world probably does believe this, which is something I find profoundly disturbing. We have a long way to go as a species in developing a sane ethics system. However, considering the progress we've already made, I have hope that we can make more.
  - chr1 8y ago
    Perhaps ethics is not something absolute, and is a game thorethic optimal strategy that changes as the rules of the game change. If hunting is the only thing you can do sane ethics system would allow killing the whale no matter how intelligent creature it is. And letting members of another tribe die instead of trying to share and risking members of your own tribe to die is a sane thing too. But things change if you a member of a civilization that can produce as much food as it wants without killing anyone.

chr1 8y ago

This is indeed very interesting line of thinking, and if one agrees that there is no soul, and there is no quantum process essential to life that can not be modeled on a classical computer, then this is the only remaining way of thinking. But this is different from dictionary definition of panpsychism, because this doesn't need to assign consciousness to simpler computations (rocks, buildings, simple cells). And this indeed opens really interesting questions. For instance if you have a simulation of many people, is it ok to continue the simulation when they are killing each other? Is it ok to ever stop the simulation? Is it ok to create a model of a person you do not like and torture him? What if your brain is augmented with a powerful computer, and torturing happens literally in your imagination?

skosch 8y ago
> this doesn't need to assign consciousness to simpler computations (rocks, buildings, simple cells). That necessarily follows, I think. The quantum processes in a rock are also a form of computation and, therefore, inhabited by some rudimentary form of experience we might call, dunno, "rockness". I don't see an easy way around it. Perhaps at absolute zero all experience stops? I have no answers, but your questions are precisely why I find this fascinating to think about!
- chr1 8y ago
  I would expect there to be very marked jump between the rockness and the computations with the simplest consciousness, that will allow to assign them to different classes instead of having a continuous spectrum going to zero with temperature. Kind of similar to the way how behavior of cellular automata changes between the predictable patterns and chaotic ones. (But i wouldn't bet all my money on it, only the half:)
  - skosch 8y ago
    > Kind of similar to the way how behavior of cellular automata changes between the predictable patterns and chaotic ones. Hm, reminds me of Permutation City. That's an interesting parallel I hadn't considered at all.

incadenza 8y ago

I don’t think it necessitates pansychism. You can still think of consciousness as a by product of a specific configuration, or just sufficient quantity of neurons or whatever. However it emerges remains a mystery, but it’s possible that some configuration of them does not produce the consciousness affect. For example, a million neurons is not enough to turn the lights on, but X amount is.

skosch 8y ago
Self-aware consciousness is a pretty high bar, but how about the ability to experience suffering? The ability to experience reward? That question very quickly turns into "what is experience", or rather, "what isn't experience". I don't find any of this intuitive, nor do I have answers to the ethical questions that follow (besides not eating animals, which I already don't). But I find it revealing how most people will fight tooth and nail to maintain that there is some magical threshold separating dead matter from precious, conscious life, without having given the topic much thought at all.
- incadenza 8y ago
  Yeah, it is pretty remarkable. I consider it non controversial at this point that many of humans favorite meals are certainly capable of a lot of suffering.
  - pas 8y ago
    I think that was never a real question. Lobsters might or might not feel pain, okay, but bovines and other mammals certainly do. And birds are (can be) pretty intelligent (eg parrots). So that's why they are supposed to be kept, raised and slaughtered in relatively humane conditions. (Eg. free range, and painless death.)
- pas 8y ago
  The really nasty question is, do animals have a(n intuitive or other) concept of the future, and/or of causality? Do they feel dread? And when they experience suffering, can they know what does that mean to them? Elephants might ( https://en.wikipedia.org/wiki/Elephant_cognition#Death_ritual https://en.wikipedia.org/wiki/Elephant_cognition#Death_ritua... ). Of course, life in the wild is not pretty. Not many of the wild animals are apex predators in their respective territory, and thus their deaths are usually not peaceful, and being hunted, having close calls with death almost daily is a very integral part of their life, so they probably have an evolved way to deal with the psychological burden of this. And that's probably what domestication changed.

c_shu 8y ago

when an error occurs in a machine and the machine tries to recover, maybe that's also a form of pain. A machine is not necessarily less sophisticated than any living creature.

gurumeditations 8y ago

That is one of the most interesting things to me about this. Consciousness seems to arise from the communication and computation in and between sensory inputs, so do computers feel anything? I can understand the logic of input 1 causes output 2 and output 5 to do something, providing the objective indicators of consciousness, but the internal experience puzzles me. Why do we feel? Why do our pain or pleasure receptors result in agonizing pain or pleasurous feeling instead of just throwing messages saying “you are experiencing 57% pleasure” or “pain at 25%”? Why do we see what is going through our visual system? Why are we conscious at all? Why aren’t we just autonomous? Systems of inanimate material seem to have this bizarre innate property where they create subjective ephemeral forces from objective inanimate atoms. Which raises the question: do computers feel?

muzani 8y ago

What about something like an anthill? It changes states, it follows patterns similar to other anthills, and probably wishes to preserve itself in some way. It's probably extremely low, but is it nonzero?