7 ms·

I've often been frustrated by the seeming polarization in the debate, with dualism on one end, and substrate independent information processing on the other. I

by psyc 9y ago

I've often been frustrated by the seeming polarization in the debate, with dualism on one end, and substrate independent information processing on the other. I wish I knew how to spark a discussion about hypotheses that are physical, but don't reduce to information processing. Consider that incandescent light is not substrate independent - it requires tungsten or some other specific chemistry that works. Couldn't consciousness be simultaneously fundamental, like a physical field, but also have specific physical or chemical preconditions to manifest in a location? Gravity coalesces around mass concentrations. Could consciousness gather around local "concentrations" of hyper-connected networks with a lot of feedback circuits in them? To be clear, I am spit-balling vague hypotheses here to make a point, so please don't ask me for citations. What I'm looking for are hypotheses that are physical, maybe substrate dependent, and non-reducible to computation in the abstract (e.g. conscious water- or domino-based computers, a la Searle).

naasking 9y ago
> I wish I knew how to spark a discussion about hypotheses that are physical, but don't reduce to information processing You can't. Information is physical. Matter irreducibly contains information: https://www.scottaaronson.com/blog/?p=3327 https://www.scottaaronson.com/blog/?p=3327 What you need to do is get over the prejudice that information processing is undesirable as an explanation for consciousness.
- psyc 9y ago
  I simply don't have a strong reason at this point to believe it is sufficient. That everything is information is, IMO a tautology, and citing it as an explanation is like citing matter and/or energy as an explanation. But is it inherently conscious, or is there a minimum necessary condition in terms of arrangement, data flow, or other physical evolution of a localized system?
  - naasking 9y ago
    All good questions, but do you really have strong reasons for thinking consciousness is not physical? There seems little need to posit consciousness as some irreducible property of matter, except for those who are mysteriously uncomfortable with the idea of being purely physical beings. Most arguments against physicalism are smoke and mirrors. As for specifically what physical process produces consciousness, it's an open question that's being studied. We're currently like Ada Lovelace studying a web browser and marvelling at what it does, having little idea how it actually works, yet we're mostly convinced it's not magic. I think you'd be more forgiving if Miss Lovelace took some time trying to reverse engineer the operation of a computer and it's program. Personally, I think [1] is a great starting point on this journey. The problem of apparent subjectivity was a huge stumbling block, and I think that paper is a great stab at it. [1] http://journal.frontiersin.org/article/10.3389/fpsyg.2015.00500/full http://journal.frontiersin.org/article/10.3389/fpsyg.2015.00...
    - monktastic1 9y ago
      Nice to see you again, naasking :) It doesn't require any "mysterious discomfort." Believing in the existence of a physical reality may seem like a totally reasonable extrapolation to make from our experiences, but it's not the only one.
      - naasking 9y ago
        Not the only one, simply the most plausible one! I can't answer the ultimate skeptic except to say that not only do skeptical arguments against knowledge depend on that very same knowledge, and so undermine themselves, but even if that were not so, abstract musings will never be more plausible than the concrete reality staring you in the face.
        monktastic1 9y ago
        That's fair, but the skeptical argument isn't my main concern. The point is that subjective phenomena are the most basic facts of my existence, so it's not even meaningful for me to explain (or worse, deny) them with respect to (in favor of) things I extrapolated from them. I think last time you didn't agree that subjective phenomena were the most immediate facts of my experience. If that's still the case, perhaps I can try to be more precise.
- goatlover 9y ago
  But what reason is there to suppose that information processing is any better at explaining consciousness than normal physics or neuroscience, for that matter? The hard problem of concrete subjectivity remains no matter which objective, abstract explanation you prefer.
  - naasking 9y ago
    Because the language of physics and neuroscience is mathematics aka information processing.
    - goatlover 9y ago
      That might be so. It's intriguing. But that still doesn't close the gap on explaining consciousness.
      - naasking 9y ago
        There are many gaps in our scientific knowledge, yet for nearly every such gap except consciousness, we are quite certain that science will close those gaps. Why is that?
        goatlover 9y ago
        The argument is that consciousness is subjectively experienced, while science is an objective explanation. There's a very fundamental split regarding objectivity and subjectivity. And the nature of what it means to explain something. As for closing gaps, do we really expect science to be able to explain everything in terms of some fundamental theory, from quarks to societies? Or maybe in terms of emerging complexity going from chemistry all the way up to social organization?
        naasking 9y ago
        > The argument is that consciousness is subjectively experienced, while science is an objective explanation. Certainly consciousness appears to be subjective experience. That doesn't make it so. See: http://journal.frontiersin.org/article/10.3389/fpsyg.2015.00500/full http://journal.frontiersin.org/article/10.3389/fpsyg.2015.00... > As for closing gaps, do we really expect science to be able to explain everything in terms of some fundamental theory, from quarks to societies? That certainly seems to be the trend. Every step forward in some field sheds light on other fields, and cross-disciplinary work is becoming ever more relevant.
AgentME 9y ago
>Consider that incandescent light is not substrate independent - it requires tungsten or some other specific chemistry that works. Light is a physical thing. Physical light is interesting/useful because it interacts with physical objects. Intelligence (or "intelligent behavior"; going to just talk about that and not exclusively-human-level for this comparison before touching "consciousness" here) is usually characterized by the way people/things choose to act. By the information (such as the information encoded in a series of nerve impulses by a brain, or the purchase and sell orders by a trading bot) produced in response to the inputs. A simulation of intelligence that produces the same information in reaction to the same inputs as a different physical implementation shares certain essential qualities that make the simulation of intelligence as interesting as other implementations, as long as you can translate the I/O into physical/worldly interactions. As compared to the much less usefulness (or restricted usefulness) of a simulation of a light bulb (unless you can translate its output into physical form, but the main way we know to do that is with a physical lightbulb or something even more expensive). Now it's possible that maybe consciousness doesn't share this same substrate-independence as intelligence; I just think there's a likely connection between intelligence and consciousness and that extends to qualities like this. >Could consciousness gather around local "concentrations" of hyper-connected networks with a lot of feedback circuits in them? Does it influence then influence the particles to act physically (or virtually) different then? The rules it uses to identify particles to manipulate and the ways it manipulates them would then just be considered a part of physics. (It could be that such a "consciousness field" could enable higher classes of computation that Turing machines are capable of. The natural thing would be to then try to figure out if a computer could be built that takes advantage of the consciousness field. But then... would an algorithm that uses the consciousness field to calculate a specific value and then complete be considered conscious or to have the moral-standing like a human just by virtue of interacting with the consciousness field? What about a computer that produces intelligent behavior and interacts with the consciousness field? I feel like people uncomfortable with this would then propose the existence of an undetected consciousness² field that gives humans something that makes them unique that the computer doesn't interact with, because it seems like that's the main reason to propose a regular consciousness field to begin with.) If the consciousness field doesn't manipulate the particles but just passively exists near them... Then it can't have anything to do with why we talk about consciousness. Some chain of thought occurs in our brains that results in nerve impulses that makes us open our mouths to talk about "consciousness", and a non-interacting field doesn't have to do with that process by definition. Maybe it could exist separately, but it doesn't provide any answers in conversations we have.

by psyc 9y ago

naasking 9y ago

> I wish I knew how to spark a discussion about hypotheses that are physical, but don't reduce to information processing You can't. Information is physical. Matter irreducibly contains information: https://www.scottaaronson.com/blog/?p=3327 https://www.scottaaronson.com/blog/?p=3327 What you need to do is get over the prejudice that information processing is undesirable as an explanation for consciousness.

psyc 9y ago
I simply don't have a strong reason at this point to believe it is sufficient. That everything is information is, IMO a tautology, and citing it as an explanation is like citing matter and/or energy as an explanation. But is it inherently conscious, or is there a minimum necessary condition in terms of arrangement, data flow, or other physical evolution of a localized system?
- naasking 9y ago
  All good questions, but do you really have strong reasons for thinking consciousness is not physical? There seems little need to posit consciousness as some irreducible property of matter, except for those who are mysteriously uncomfortable with the idea of being purely physical beings. Most arguments against physicalism are smoke and mirrors. As for specifically what physical process produces consciousness, it's an open question that's being studied. We're currently like Ada Lovelace studying a web browser and marvelling at what it does, having little idea how it actually works, yet we're mostly convinced it's not magic. I think you'd be more forgiving if Miss Lovelace took some time trying to reverse engineer the operation of a computer and it's program. Personally, I think [1] is a great starting point on this journey. The problem of apparent subjectivity was a huge stumbling block, and I think that paper is a great stab at it. [1] http://journal.frontiersin.org/article/10.3389/fpsyg.2015.00500/full http://journal.frontiersin.org/article/10.3389/fpsyg.2015.00...
  - monktastic1 9y ago
    Nice to see you again, naasking :) It doesn't require any "mysterious discomfort." Believing in the existence of a physical reality may seem like a totally reasonable extrapolation to make from our experiences, but it's not the only one.
    - naasking 9y ago
      Not the only one, simply the most plausible one! I can't answer the ultimate skeptic except to say that not only do skeptical arguments against knowledge depend on that very same knowledge, and so undermine themselves, but even if that were not so, abstract musings will never be more plausible than the concrete reality staring you in the face.
      - monktastic1 9y ago
        That's fair, but the skeptical argument isn't my main concern. The point is that subjective phenomena are the most basic facts of my existence, so it's not even meaningful for me to explain (or worse, deny) them with respect to (in favor of) things I extrapolated from them. I think last time you didn't agree that subjective phenomena were the most immediate facts of my experience. If that's still the case, perhaps I can try to be more precise.
goatlover 9y ago
But what reason is there to suppose that information processing is any better at explaining consciousness than normal physics or neuroscience, for that matter? The hard problem of concrete subjectivity remains no matter which objective, abstract explanation you prefer.
- naasking 9y ago
  Because the language of physics and neuroscience is mathematics aka information processing.
  - goatlover 9y ago
    That might be so. It's intriguing. But that still doesn't close the gap on explaining consciousness.
    - naasking 9y ago
      There are many gaps in our scientific knowledge, yet for nearly every such gap except consciousness, we are quite certain that science will close those gaps. Why is that?
      - goatlover 9y ago
        The argument is that consciousness is subjectively experienced, while science is an objective explanation. There's a very fundamental split regarding objectivity and subjectivity. And the nature of what it means to explain something. As for closing gaps, do we really expect science to be able to explain everything in terms of some fundamental theory, from quarks to societies? Or maybe in terms of emerging complexity going from chemistry all the way up to social organization?
        naasking 9y ago
        > The argument is that consciousness is subjectively experienced, while science is an objective explanation. Certainly consciousness appears to be subjective experience. That doesn't make it so. See: http://journal.frontiersin.org/article/10.3389/fpsyg.2015.00500/full http://journal.frontiersin.org/article/10.3389/fpsyg.2015.00... > As for closing gaps, do we really expect science to be able to explain everything in terms of some fundamental theory, from quarks to societies? That certainly seems to be the trend. Every step forward in some field sheds light on other fields, and cross-disciplinary work is becoming ever more relevant.

AgentME 9y ago

>Consider that incandescent light is not substrate independent - it requires tungsten or some other specific chemistry that works. Light is a physical thing. Physical light is interesting/useful because it interacts with physical objects. Intelligence (or "intelligent behavior"; going to just talk about that and not exclusively-human-level for this comparison before touching "consciousness" here) is usually characterized by the way people/things choose to act. By the information (such as the information encoded in a series of nerve impulses by a brain, or the purchase and sell orders by a trading bot) produced in response to the inputs. A simulation of intelligence that produces the same information in reaction to the same inputs as a different physical implementation shares certain essential qualities that make the simulation of intelligence as interesting as other implementations, as long as you can translate the I/O into physical/worldly interactions. As compared to the much less usefulness (or restricted usefulness) of a simulation of a light bulb (unless you can translate its output into physical form, but the main way we know to do that is with a physical lightbulb or something even more expensive). Now it's possible that maybe consciousness doesn't share this same substrate-independence as intelligence; I just think there's a likely connection between intelligence and consciousness and that extends to qualities like this. >Could consciousness gather around local "concentrations" of hyper-connected networks with a lot of feedback circuits in them? Does it influence then influence the particles to act physically (or virtually) different then? The rules it uses to identify particles to manipulate and the ways it manipulates them would then just be considered a part of physics. (It could be that such a "consciousness field" could enable higher classes of computation that Turing machines are capable of. The natural thing would be to then try to figure out if a computer could be built that takes advantage of the consciousness field. But then... would an algorithm that uses the consciousness field to calculate a specific value and then complete be considered conscious or to have the moral-standing like a human just by virtue of interacting with the consciousness field? What about a computer that produces intelligent behavior and interacts with the consciousness field? I feel like people uncomfortable with this would then propose the existence of an undetected consciousness² field that gives humans something that makes them unique that the computer doesn't interact with, because it seems like that's the main reason to propose a regular consciousness field to begin with.) If the consciousness field doesn't manipulate the particles but just passively exists near them... Then it can't have anything to do with why we talk about consciousness. Some chain of thought occurs in our brains that results in nerve impulses that makes us open our mouths to talk about "consciousness", and a non-interacting field doesn't have to do with that process by definition. Maybe it could exist separately, but it doesn't provide any answers in conversations we have.