Re: Consciousness is information?
Kelly wrote: On Apr 22, 12:24 pm, Bruno Marchal marc...@ulb.ac.be wrote: So for that to be a plausible scenario we have to say that a person at a particular instant in time can be fully described by some set of data. Not fully. I agree with Brent that you need an interpreter to make that person manifest herself in front of you. A bit like a CD, you will need a player to get the music. It seems to me that consciousness is the self-interpretion of information. David Chalmers has a good line: Experience is information from the inside; physics is information from the outside. I still don't see what an interpreter adds, except to satisfy the intuition that something is happening that produces consciousness. Which I think is an attempt to reintroduce time. But I don't see any advantage of this view over the idea that conscious states just exist as a type of platonic form (as Brent mentioned earlier). At any given instant that I'm awake, I'm conscious of SOMETHING. And I'm conscious of it by virtue of my mental state at that instant. In the materialist view, my mental state is just the state of the particles of my brain at that instant. I think we need some definition of state. Supposing your brain were a Newtonian system the state would be the position and velocity of all the particles. Physically this leads to the next state by the Newtonian dynamics. But those dynamics operate in a continuum. If we discretize your brain, say slice it into Planck units of time as Jason suggested, now we need to have something to connect one state to another. The states are no longer part of a continuum. In a computer running your brain this is provided by the hardware of the computer. In Bruno's theory it is provided by a relation in Platonia, i.e. a computational rule. In idealism, the content of a state consciousness (a Planck slice, not of a brain, but of a stream of consciousness) seems to me to be very small and it doesn't so far as I can see have anything analogous to dynamical equations to connect it to another state. You say it is connected by the correlation of information content, but is that unique? Is there a best or most probable next state or what? Brent But I say that what this really means is that my mental state is just the information represented by the particles of my brain at that instant. And that if you transfer that information to a computer and run a simulation that updates that information appropriately, my consciousness will continue in that computer simulation, regardless of the hardware (digital computer, mechanical computer, massively parallel or single processor, etc) or algorithmic details of that computer simulation. But, what is information? I think it has nothing to do with physical storage or instantiation. I think it has an existence seperate from that. A platonic existence. And since the information that represents my brain exists platonically, then the information for every possible brain (including variations of my brain) should also exist platonically. Conscious experience is with the information. Conscious experience is more the content, or the interpretation of that information, made by a person or by a universal machine. If the doctor makes a copy of your brain, and then codes it into a bit string, and then put the bit string in the fridge, in our probable history, well in that case you will not survive, in our local probable history. Given the platonic nature of information, this isn't really a concern. In Platonia, you always have a next moment. In fact, you experience all possible next moments. The no cul-de-sac rule applies I think. If you say yes to a doctor for a digital brain, you will ask for a brain which functions relatively to our probable computational history. No? I won't worry about it too much, as there is no doctor, only my perceptions of a doctor. Every possible outcome of the brain replacement operation that I can perceive, I will perceive. Including outcomes that don't make any sense. Additionally, every possible outcome of the operation that the doctor can percieve, he will perceive. Including outcomes that don't make any sense. So it seems to me that a lot of your effort goes into explaining why we don't see strange white rabbit universes. Thus the talk of probabilities and measures. I'm willing to just say that all universes are experienced. Strange ones, normal ones, good ones, bad ones, ones with unbreakable physical laws, ones with no obvious physical laws at all. It's all a matter of perception, not a matter of physical realization. Yes there is a world in which you computer will transform itself into a green flying pig. The scientific, but really everyday life question, is, what is the probability this will happen to me here and now. I'm not sure what it means to ask,
Re: Consciousness is information?
On 24 Apr 2009, at 02:37, Kelly wrote: On Apr 22, 2:02 pm, Brent Meeker meeke...@dslextreme.com wrote: I was with you up to that last sentence. Forward or backward, we just experience increasing entropy as increasing time, but that doesn't warrant the conclusion that no process is required and an instant within itself has an arrow of time. It seems to me that each instant DOES contain within itself an arrow of time, in the form of memories. Later instances are related to earlier instances by virtue of having memory-information about those earlier instances. That's what ties the various states together. The nature of the computations that might transition you from instant to instant are not relevant. What matters is where you end up, not how you got there. What matters is the first person probability you find yourself ending up there. And this will depend on all computations going through your current state (or below) and going through the state up there. If a transition causes you to assume a state that contains no information about earlier events (i.e., no memory of these events), then you have lost a crucial part of what makes you who you are. If you save your brain state at time A and then you save state again at a subsequent time B, there is a relationship and an objectively measureable degree of correlation between the information contained in the two saved data sets. It is, I think, the degree of correlation between states that provides the illusion of a flow of consciousness. This has nothing to do with the type of computation that could be used to transition between the two data sets. Again, it seems to me that the arithmetic logic that Bruno refers to just serves to describe the relations between datasets. It doesn't produce consciousness. If there are many algorithms that could be used to transition from state A to state B, it seems to me that all of them would produce the same conscious experience. If you end up at state B, then it doesn't matter how you go there...your memory of the experience will be identical regardless of what path you took. And since all states (not just A and B) exist platonically, then every possible process can be inferred to connect them in every possible way. But I don't think this means that the processes are the source of consciousness. They are just descriptions of the ways that states could be connected. http://iridia.ulb.ac.be/~marchal/ --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Everything List group. To post to this group, send email to everything-list@googlegroups.com To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/everything-list?hl=en -~--~~~~--~~--~--~---
Re: Consciousness is information?
Jason Resch wrote: Kelly, Your arguments are compelling and logical, you have put a lot of doubt in my mind about computationalism. I have actually been in somewhat of a state of confusion since Bruno's movie graph argument coupled with a paper by Max Tegmark. In Tegmark's paper, he was explaining that there is an appeal to many people of associating the time dimension with the computational clock, but argued there is no reason to do so, time is just another dimension after all, and everything being an atemporal platonic/mathematical object any perception of change is illusory. That's a common model but it's certainly not a settled question in physics. Just recently Sean Carroll wrote a paper titled What if Time Really Exists? http://arxiv.org/abs/0811.3772. And even in a block universe model the time dimension is still different from the space dimensions. Later, when Bruno explained his movie graph argument, it came to the point where we were asked: Is a recording of Alice's brain activity itself conscious? I first thought obviously no, but then realized the contradiction with space-time. Could the block-time view of the universe not be considered a recording? Perhaps the difference between a recording (like Tape or CD) and the universe (or a computer program/simulation) is that with a physical recording it is possible to alter a state at one point in time without affecting future/past states. This implicitly assumes that you can dispense with the continuum and treat the process as a succession of discrete states. I question that. It is how we think and how we write and describe computer programs and we know that if we make the time step small enough in the simulation we can accurately reproduce processes. But I think we are fooling ourselves by taking the description in terms of discrete states to be sufficient - actually we are relying on the physics of the computer to join one state to the next. Bruno proposes to abstract this whole process up to Platonia where the role of the computer in interpreting the program is taken over by abstract computations. But then to avoid any choice he must allow all possible (countably infinite) computations between any two states. ISTM this implies a strange topology of states and I'm not clear on how it models consciousness. Or maybe consciousness is only created from platonic objects / information or relationships that exist within them. The appeal of computationalism for me is that it creates a self-interpreting structure, the information or state has meaning only because it is part a state machine. We, being creatures who can only experience through time might be fooled into thinking change over time is necessary for consciousness, but what if we could make a computer that computed over the X-dimension instead of T, what would such a computer look like and how would it be logically different from a recording (which is static over T), and how is it logically different from a computer that computes accross the T dimension? I don't think it is *logically* different. Before computers, a computation was something written out on sheets of paper (I know because my first summer job in college was calculating coordinates and depths for a geological research company and my official job title was Computer.) :-) Brent I very much look forward to reading your and others' opinions on this. Jason On Thu, Apr 23, 2009 at 11:14 PM, Kelly harmon...@gmail.com wrote: On Apr 22, 12:24 pm, Bruno Marchal marc...@ulb.ac.be wrote: So for that to be a plausible scenario we have to say that a person at a particular instant in time can be fully described by some set of data. Not fully. I agree with Brent that you need an interpreter to make that person manifest herself in front of you. A bit like a CD, you will need a player to get the music. It seems to me that consciousness is the self-interpretion of information. David Chalmers has a good line: Experience is information from the inside; physics is information from the outside. I still don't see what an interpreter adds, except to satisfy the intuition that something is happening that produces consciousness. Which I think is an attempt to reintroduce time. But I don't see any advantage of this view over the idea that conscious states just exist as a type of platonic form (as Brent mentioned earlier). At any given instant that I'm awake, I'm conscious of SOMETHING. And I'm conscious of it by virtue of my mental state at that instant. In the materialist view, my mental state is just the state of the particles of my brain at that instant. But I say that what this really means is that my mental state is just the information represented by the particles of my brain at that instant. And that if you transfer that information to a computer and run a simulation that updates that information
Re: Consciousness is information?
2009/4/24 Brent Meeker meeke...@dslextreme.com: Boltzmann brains are improbable, but the example of the punchcards is not. The operator could have two punchcards in his pocket, have a conversation with someone on the way from M1 to M2 and end up forgetting or almost forgetting which is the right one. That is, his certainty of picking the right card could vary between 0.5 and 1. Would you say that only if his certainty is 1 would the conscious process be implemented, and not if it is, say, 0.9? I said it would be implementing *the same* consciousness if he was following the rule. If not he might be implementing a different consciousness by using a different rule. Of course if it were different in only one moment that wouldn't really be much of a difference. I don't think it depends on his certainty. Even more difficult we might ask what it means for him to follow the rule - must he do it *consciously*; in which case do we have to know whether his brain is functioning according to the same rule? You're asking a lot of questions, Stathis. :-) What do you think? I don't think the rule matters, only the result, which could consist of a series of disconnected states. The utility of a process is that it reliably brings about the relevant states; but if they arose randomly or by a different process that would be just as good. If not, then you could have an apparently functionally identical machine which has a different consciousness. One half of your brain might function by a different process that gives the same neuronal outputs, and you would have a feeling that something had radically changed, but your mouth would seemingly of its own accord continue to declare that everything is just the same. So, I agree with Kelly that the consciousness consists in the information. -- Stathis Papaioannou --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Everything List group. To post to this group, send email to everything-list@googlegroups.com To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/everything-list?hl=en -~--~~~~--~~--~--~---
Re: Consciousness is information?
2009/4/25 Brent Meeker meeke...@dslextreme.com: This implicitly assumes that you can dispense with the continuum and treat the process as a succession of discrete states. I question that. So are you saying that, because we are conscious, that is evidence that reality is at bottom continuous rather than discrete? It is how we think and how we write and describe computer programs and we know that if we make the time step small enough in the simulation we can accurately reproduce processes. But I think we are fooling ourselves by taking the description in terms of discrete states to be sufficient - actually we are relying on the physics of the computer to join one state to the next. Bruno proposes to abstract this whole process up to Platonia where the role of the computer in interpreting the program is taken over by abstract computations. But then to avoid any choice he must allow all possible (countably infinite) computations between any two states. ISTM this implies a strange topology of states and I'm not clear on how it models consciousness. Or maybe consciousness is only created from platonic objects / information or relationships that exist within them. The appeal of computationalism for me is that it creates a self-interpreting structure, the information or state has meaning only because it is part a state machine. We, being creatures who can only experience through time might be fooled into thinking change over time is necessary for consciousness, but what if we could make a computer that computed over the X-dimension instead of T, what would such a computer look like and how would it be logically different from a recording (which is static over T), and how is it logically different from a computer that computes accross the T dimension? I don't think it is *logically* different. Before computers, a computation was something written out on sheets of paper (I know because my first summer job in college was calculating coordinates and depths for a geological research company and my official job title was Computer.) :-) Do you think a computation would feel different from the inside depending on whether it was done with pencil and paper, transistors or vacuum tubes? -- Stathis Papaioannou --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Everything List group. To post to this group, send email to everything-list@googlegroups.com To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/everything-list?hl=en -~--~~~~--~~--~--~---
Re: Consciousness is information?
Stathis Papaioannou wrote: 2009/4/24 Brent Meeker meeke...@dslextreme.com: Boltzmann brains are improbable, but the example of the punchcards is not. The operator could have two punchcards in his pocket, have a conversation with someone on the way from M1 to M2 and end up forgetting or almost forgetting which is the right one. That is, his certainty of picking the right card could vary between 0.5 and 1. Would you say that only if his certainty is 1 would the conscious process be implemented, and not if it is, say, 0.9? I said it would be implementing *the same* consciousness if he was following the rule. If not he might be implementing a different consciousness by using a different rule. Of course if it were different in only one moment that wouldn't really be much of a difference. I don't think it depends on his certainty. Even more difficult we might ask what it means for him to follow the rule - must he do it *consciously*; in which case do we have to know whether his brain is functioning according to the same rule? You're asking a lot of questions, Stathis. :-) What do you think? I don't think the rule matters, only the result, which could consist of a series of disconnected states. The utility of a process is that it reliably brings about the relevant states; but if they arose randomly or by a different process that would be just as good. If two processes always produce the same sequence they are the same process in the abstract sense. If not, then you could have an apparently functionally identical machine which has a different consciousness. One half of your brain might function by a different process that gives the same neuronal outputs, and you would have a feeling that something had radically changed, but your mouth would seemingly of its own accord continue to declare that everything is just the same. So, I agree with Kelly that the consciousness consists in the information. But is it the information in consciousness and is it discrete? If you include the information that is in the brain, but not in consciousness, I can buy the concept of relating states by similarity of content. Or if you suppose a continuum of states that would provide a sequence. It is only when you postulate discrete states containing only the contents of instants of conscious thought, that I find difficulty. Brent Brent --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Everything List group. To post to this group, send email to everything-list@googlegroups.com To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/everything-list?hl=en -~--~~~~--~~--~--~---
Re: Consciousness is information?
Stathis Papaioannou wrote: 2009/4/25 Brent Meeker meeke...@dslextreme.com: This implicitly assumes that you can dispense with the continuum and treat the process as a succession of discrete states. I question that. So are you saying that, because we are conscious, that is evidence that reality is at bottom continuous rather than discrete? It is how we think and how we write and describe computer programs and we know that if we make the time step small enough in the simulation we can accurately reproduce processes. But I think we are fooling ourselves by taking the description in terms of discrete states to be sufficient - actually we are relying on the physics of the computer to join one state to the next. Bruno proposes to abstract this whole process up to Platonia where the role of the computer in interpreting the program is taken over by abstract computations. But then to avoid any choice he must allow all possible (countably infinite) computations between any two states. ISTM this implies a strange topology of states and I'm not clear on how it models consciousness. Or maybe consciousness is only created from platonic objects / information or relationships that exist within them. The appeal of computationalism for me is that it creates a self-interpreting structure, the information or state has meaning only because it is part a state machine. We, being creatures who can only experience through time might be fooled into thinking change over time is necessary for consciousness, but what if we could make a computer that computed over the X-dimension instead of T, what would such a computer look like and how would it be logically different from a recording (which is static over T), and how is it logically different from a computer that computes accross the T dimension? I don't think it is *logically* different. Before computers, a computation was something written out on sheets of paper (I know because my first summer job in college was calculating coordinates and depths for a geological research company and my official job title was Computer.) :-) Do you think a computation would feel different from the inside depending on whether it was done with pencil and paper, transistors or vacuum tubes? No, I don't think the medium makes a difference. But interpretation makes a difference. Most computations we do, on pencil and paper or transistors or neurons, have an interpretation in terms of our world. Kelly is supposing there is a self-interpreting structure I'm not sure what he means by this, but I imagine something like an elaborate simulation in which some parts of the computation simulate entities with values or purposes - on some mapping. But what about other mappings? Brent --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Everything List group. To post to this group, send email to everything-list@googlegroups.com To unsubscribe from this group, send email to everything-list+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/everything-list?hl=en -~--~~~~--~~--~--~---