4 ms·

In theory, and this is a pretty airtight theory, it is computation. Because if it isn't, it's magic. In practice, it is only implemented in one type of animal,

by Zigurd 5y ago

In theory, and this is a pretty airtight theory, it is computation. Because if it isn't, it's magic. In practice, it is only implemented in one type of animal, or maybe a few types, depending on how it is defined. And we do not have a practical answer for how and why. Partial at best. And not in a way that would enable enough understanding about what kind of computation it is to implement similar-enough computation in a machine. It feels like the scale of computation needed is now, or soon will be, available. The 100 billion neurons with 100 trillion connections seem daunting until you look at the connection and neuron (and sensory) data rates. It was clearly never going to happen in a Symbolics 3600. It is not clear it could not happen in a modern data center.

brandonmenc 5y ago
> it is computation OK, but what is "computation"? If you perform the computation with a lot of pencils and a lot of paper, is it conscious?
- bryan0 5y ago
  This is basically the Chinese Room argument and the answer is: yes, the system as a whole is conscious.
  - sramsay 5y ago
    That's one answer. Another is that you would be within your rights to draw the inference that the system you are observing is conscious (which is slightly different, and deeply unsatisfying to many).
    - bryan0 5y ago
      Can you explain the difference?
      - sramsay 5y ago
        For Turing, the criteria for intelligence was a matter of observers (if it walks like a duck . . .); what is "actually happening" inside is irrelevant. Searle's objection was that what is happening "inside" is surely not irrelevant. So while some answered that in the Chinese Room argument the "whole system is conscious," others argued that Turing's point still stands. Someone observing the system would have to conclude that the system was intelligent, regardless of what was happening inside. The two answers are roughly on the same side of the argument, but subtly different in important ways.
        bryan0 5y ago
        Ok thanks. Yeah I think that Turing came up with the best solution for detecting intelligence. If it passes that test then either the system is truly intelligent or a “zombie” which only appears intelligent. As mentioned in some other replies on this post, I think “zombies” are something created by philosophers but do not exist in real life. So i think the observation of intelligence is equivalent to actual intelligence.
  - Zigurd 5y ago
    The Chinese Room argument is tendentious because scaling a Chinese Room metaphor to a human brain's neurons and connections is hard to imagine. It is still hard-ish to imagine because, currently, it would take a large data center and 4-6 orders of magnitude the power consumption. The other side of the coin is that we are asked to believe that what happens in an animal brain turns into magic when the animal is human. That is even harder to imagine.
    - Jensson 5y ago
      > The other side of the coin is that we are asked to believe that what happens in an animal brain turns into magic when the animal is human. That is even harder to imagine. Nobody is saying that animals aren't conscious. Or maybe a few are, and maybe human brains have some unique properties, we can't tell since we don't know what we are looking for. But animals display many of the same things as humans, and we know for a fact that humans are conscious so it isn't a stretch to imagine animals being as well. > magic Yes, we have no formula to describe consciousness, so as far as we know today it is magic. But if we manage to properly map that out in the future then consciousness will be science.
sramsay 5y ago
I think you only regard that theory as "airtight" because you accept the premises of ontological naturalism. What you're basically saying is that it must be this way, because there is no other way to for it to be. But it remains the case that there is exactly no evidence at all for the belief that consciousness is an emergent property of computation. Not some evidence -- no evidence.
- ncmncm 5y ago
  We know factually that there is computation, because we can observe it happening using wires inserted at the inputs and outputs. The burden is on people who insist there is something else besides computation. Thus far, no evidence means, exactly, no phenomenon, and nothing to be explained.
  - sramsay 5y ago
    You may be right about where the burden lies, but your explanation renders the thermostat conscious. Edit: I should put that differently. Your sense of the criteria by which the problem may be considered solved does not preclude thermostat consciousness. And if that is allowed, then yes: the problem is solved. But I sense, again, the assumption of physicalism (or ontological naturalism), and I don't know why that premise admits of no dissent or argument.
    - ncmncm 5y ago
      When a term is undefined, any application of it is undecidable. Are you conscious? Is the thermostat? Y/Y, Y/N, N/Y, N/N are all equally valid conclusions.
      - sramsay 5y ago
        Would you accept the statement "If physicalism is true, then there is nothing here to explain." I'm not setting a trap. I'm just trying to clarify your position.

by Zigurd 5y ago

brandonmenc 5y ago

> it is computation OK, but what is "computation"? If you perform the computation with a lot of pencils and a lot of paper, is it conscious?

bryan0 5y ago
This is basically the Chinese Room argument and the answer is: yes, the system as a whole is conscious.
- sramsay 5y ago
  That's one answer. Another is that you would be within your rights to draw the inference that the system you are observing is conscious (which is slightly different, and deeply unsatisfying to many).
  - bryan0 5y ago
    Can you explain the difference?
    - sramsay 5y ago
      For Turing, the criteria for intelligence was a matter of observers (if it walks like a duck . . .); what is "actually happening" inside is irrelevant. Searle's objection was that what is happening "inside" is surely not irrelevant. So while some answered that in the Chinese Room argument the "whole system is conscious," others argued that Turing's point still stands. Someone observing the system would have to conclude that the system was intelligent, regardless of what was happening inside. The two answers are roughly on the same side of the argument, but subtly different in important ways.
      - bryan0 5y ago
        Ok thanks. Yeah I think that Turing came up with the best solution for detecting intelligence. If it passes that test then either the system is truly intelligent or a “zombie” which only appears intelligent. As mentioned in some other replies on this post, I think “zombies” are something created by philosophers but do not exist in real life. So i think the observation of intelligence is equivalent to actual intelligence.
- Zigurd 5y ago
  The Chinese Room argument is tendentious because scaling a Chinese Room metaphor to a human brain's neurons and connections is hard to imagine. It is still hard-ish to imagine because, currently, it would take a large data center and 4-6 orders of magnitude the power consumption. The other side of the coin is that we are asked to believe that what happens in an animal brain turns into magic when the animal is human. That is even harder to imagine.
  - Jensson 5y ago
    > The other side of the coin is that we are asked to believe that what happens in an animal brain turns into magic when the animal is human. That is even harder to imagine. Nobody is saying that animals aren't conscious. Or maybe a few are, and maybe human brains have some unique properties, we can't tell since we don't know what we are looking for. But animals display many of the same things as humans, and we know for a fact that humans are conscious so it isn't a stretch to imagine animals being as well. > magic Yes, we have no formula to describe consciousness, so as far as we know today it is magic. But if we manage to properly map that out in the future then consciousness will be science.

sramsay 5y ago

I think you only regard that theory as "airtight" because you accept the premises of ontological naturalism. What you're basically saying is that it must be this way, because there is no other way to for it to be. But it remains the case that there is exactly no evidence at all for the belief that consciousness is an emergent property of computation. Not some evidence -- no evidence.

ncmncm 5y ago
We know factually that there is computation, because we can observe it happening using wires inserted at the inputs and outputs. The burden is on people who insist there is something else besides computation. Thus far, no evidence means, exactly, no phenomenon, and nothing to be explained.
- sramsay 5y ago
  You may be right about where the burden lies, but your explanation renders the thermostat conscious. Edit: I should put that differently. Your sense of the criteria by which the problem may be considered solved does not preclude thermostat consciousness. And if that is allowed, then yes: the problem is solved. But I sense, again, the assumption of physicalism (or ontological naturalism), and I don't know why that premise admits of no dissent or argument.
  - ncmncm 5y ago
    When a term is undefined, any application of it is undecidable. Are you conscious? Is the thermostat? Y/Y, Y/N, N/Y, N/N are all equally valid conclusions.
    - sramsay 5y ago
      Would you accept the statement "If physicalism is true, then there is nothing here to explain." I'm not setting a trap. I'm just trying to clarify your position.