Re: Maudlin's Machine and the UDist
On Fri, Aug 12, 2005 at 05:21:50PM +0200, Bruno Marchal wrote: This is a personal copy and I would ask you not to redistribute it. I will try to get some authorization. It will be hard for me not putting that paper in my webpage. Did you just scanned it. I would acknowledge the fact. You can give me suggestion for preventing your sending into jail ;-) My copy I obtained through JStor (http://www.jstor.org), but I'm not allowed to redistribute it under the terms of the license. I detest this balkanisation of scientific knowledge, but aside from ensuring my own work is available through open channels (eg arXiv), there is not too much one can do about it. At least my institution can still afford JStor! Cheers -- *PS: A number of people ask me about the attachment to my email, which is of type application/pgp-signature. Don't worry, it is not a virus. It is an electronic signature, that may be used to verify this email came from me if you have PGP or GPG installed. Otherwise, you may safely ignore this attachment. A/Prof Russell Standish Phone 8308 3119 (mobile) Mathematics0425 253119 () UNSW SYDNEY 2052 [EMAIL PROTECTED] Australiahttp://parallel.hpc.unsw.edu.au/rks International prefix +612, Interstate prefix 02 pgpRLlLaEvvcq.pgp Description: PGP signature
Re: Maudlin's Machine and the UDist
Le 07-août-05, à 22:20, Hal Finney a écrit : Rutgers philosopher Tim Maudlin has a paper intended to challenge certain views about consciousness and computation, which we have discussed occasionally on this list. Indeed. Maudlin's paper is without doubt one of the most important paper in the philosophy of mind literature. Note that Barnes' answer, in the same journal, is worth reading too. It is called Computation and Consciousness, Journal of Philosophy v86, pp. 407-432. I have temporarily put a copy online at http://www.finney.org/~hal/maudlin.pdf . Many thanks for this. I urge people to download it at once! Then we will have all the time to discuss it. It is equivalent (logically) with my movie-graph argument. An analysis of Maudlin's in term of movie graph is done (in french) in my Brussel's thesis and a shorter one in my french Lille thesis. Maudlin is responsible for relating this issue with the counterfactual issue, which can be related quasi-directly to quantum logic, thanks to a very cute and readable paper by Hardegree: Hardegree, G. M. (1976). The Conditional in Quantum Logic. In Suppes, P., editor, Logic and Probability in Quantum Mechanics, volume 78 of Synthese Library, pages 55-72. D. Reidel Publishing Company, Dordrecht-Holland. I even tend to think that a refinement of Maudlin's paper (using Hardegree) could lead to an elimination of the need of the Universal Dovetailer Argument (but that is still a little bit speculative) to derive the quantum from comp. Actually Russell did provide an hint in that direction in his answer to Hal's post. This is a personal copy and I would ask you not to redistribute it. I will try to get some authorization. It will be hard for me not putting that paper in my webpage. Did you just scanned it. I would acknowledge the fact. You can give me suggestion for preventing your sending into jail ;-) I will comment your post asap. Meanwhile, just note that what Maudlin calls supervenience, I prefer to call it physical supervenience, so that I keep a notion of comp-supervenience. In a nutshell, Maudlin, like me, proves the incompatibility of digital mechanism and materialism. Maudlin presupposes materialism, so he concludes there is a problem with comp. I presuppose mechanism and conclude there is a problem with materialism. Have a nice week-end, Hal and all, Bruno http://iridia.ulb.ac.be/~marchal/
Re: Maudlin's Machine and the UDist
Russell Standish writes: The take home message I get from Maudlin's experiment is that a computationalist consciousness is supervenient on a physical process _spread_ over the multiverse, ie the counterfactuals must really exist as alternate branches of the Multiverse. So what does that tell you about Olympia? Is she conscious or not, by this criterion? I guess that you would say that if the unused counterfactual machinery would actually work if tested, then she is conscious; but if the counterfactual machines were broken or blocked such that they wouldn't work (even though they are not used) then she is unconscious. And perhaps you can say that the machines are in fact tested in other branches of the multiverse, so the criterion is more than merely a hypothetical difference between unused working machines and unused broken machines. I see some difficulties with this position but I better first hear whether this is what you have in mind before trying to extrapolate further. A far as your UDist argument goes, the fact that a conscious HLUT, or a conscious clock has very low measure simply means it is very unlikely for us to be one of these things. They would still be conscious. However accepting the Multiverse would eliminate these objects from being conscious at all, because tof the lack of counterfactuals. From my perspective it doesn't make sense to ask whether a system is conscious, per se. Consciousness exists platonically in the multiverse. Any given consciousness exists, whether a particular system implements it or not. What we want to know is whether running a certain program or process will add to the measure of a given consciousness. Running a clock will not add any noticeable measure to any consciousness. Running a neural simulation or some AI program may well add significant measure. We can deduce these facts without considering counterfactuals. It is only necessary to see how short a program can compute a representation of the abstract conscious calculation by starting from the program or process that we initiate. My understanding is that the main argument for requiring counterfactuals in the definition of implementation is to escape the argument that a clock implements every finite state machine. I believe that other responses are better, such as the one by Jacques Mallah. Unfortunately Mallah's works seems to have largely disappeared from the web, as has Mallah himself, but I found an early copy of one of them on archive.org and have put it here, http://www.finney.org/~hal/mallah1.html. This version does not lay out the argument as clearly as the later ones, and merely hints at the role Kolmogorov complexity can play, but the basic ideas are present. Hal Finney
Re: Maudlin's Machine and the UDist
On Sun, Aug 07, 2005 at 11:35:42PM -0700, Hal Finney wrote: Russell Standish writes: The take home message I get from Maudlin's experiment is that a computationalist consciousness is supervenient on a physical process _spread_ over the multiverse, ie the counterfactuals must really exist as alternate branches of the Multiverse. So what does that tell you about Olympia? Is she conscious or not, by this criterion? Not Olympia, nor Olympia + Klara with Klara blocked an immobile. Only Olympia and Klara without the blocks is conscious. I guess that you would say that if the unused counterfactual machinery would actually work if tested, then she is conscious; but if the counterfactual machines were broken or blocked such that they wouldn't work (even though they are not used) then she is unconscious. And perhaps you can say that the machines are in fact tested in other branches of the multiverse, so the criterion is more than merely a hypothetical difference between unused working machines and unused broken machines. I see some difficulties with this position but I better first hear whether this is what you have in mind before trying to extrapolate further. That is indeed my meaning. What difficulties do you see? Cheers -- *PS: A number of people ask me about the attachment to my email, which is of type application/pgp-signature. Don't worry, it is not a virus. It is an electronic signature, that may be used to verify this email came from me if you have PGP or GPG installed. Otherwise, you may safely ignore this attachment. A/Prof Russell Standish Phone 8308 3119 (mobile) Mathematics0425 253119 () UNSW SYDNEY 2052 [EMAIL PROTECTED] Australiahttp://parallel.hpc.unsw.edu.au/rks International prefix +612, Interstate prefix 02 pgpUGNthj7XWU.pgp Description: PGP signature
Re: Maudlin's Machine and the UDist
I speculated: I guess that you would say that if the unused counterfactual machinery would actually work if tested, then she is conscious; but if the counterfactual machines were broken or blocked such that they wouldn't work (even though they are not used) then she is unconscious. And perhaps you can say that the machines are in fact tested in other branches of the multiverse, so the criterion is more than merely a hypothetical difference between unused working machines and unused broken machines. I see some difficulties with this position but I better first hear whether this is what you have in mind before trying to extrapolate further. Russell replied: That is indeed my meaning. What difficulties do you see? I see a few problems. The first is the concept that the multiverse will contain copies of the machine that execute the counterfactuals. While this could happen, it would normally be to a very, very tiny degree, and it seems highly implausible that this could cause any actual effects. In the case of Maudlin's thought experiment, the intention of the operator is to run the machine on the specific inputs that it is pre-programmed to handle. In effect, it is doing a replay of a previously computed calculation. The counterfactual machinery is not intended to be activated. All of the effort and skill of the operator will be devoted to pre-setting the machine into the precise state necessary for an exact replay, as Maudlin describes in detail. Your interpretation depends on there being worlds in the multiverse where these circumstances do not hold. But the only way such a world could exist is if the operator somehow fails to achieve his goal. The measure of such worlds, if the operator is skillful and effective, will be extremely tiny. Only in worlds where the machine breaks or the operator makes a mistake (or perhaps the operator's brain malfunctions so he changes his mind about what to do) will the counterfactuals be explored. My problem is that such worlds are so few and so remote from the great majority of worlds where the machine is operating as designed, i.e. without triggering the counterfactual mechanisms, that it is hard to see how the presence of consciousness could depend on activity in those worlds. We might also note that those worlds are causally disconnected from the one in question due to decoherence, and again it seems questionable how a real phenomenon like consciousness can be affected by what is happening in another universe. Now, we would do an experiment which was designed to more effectively test and exercise the machine in a mode where it would seem more plausible to say that the multiverse was testing the counterfactuals. We could in effect treat Olympia as a quantum computer, somehow initializing it in a superposition of states, both the replay state and the many states which would trigger counterfactuals, and then let it run like that. In that case it would make more sense to say that the multiverse was testing the counterfactuals and perhaps to say that the presence or absence of consciousness was dependent on what happens across the many worlds in this superposition. But that is not the usual case of exercising Olympia. Instead, it is put, as certainly as we are able, into a single initial state. If there do appear variations in the multiverse it is unintentional, and such variations will be remote and of small measure. A related problem is that if we do consider the entirety of the multiverse, undoubtedly there will be corners where virtually anything happens. In a normal run of the Olympia machine augmented with the counterfactuals, indeed if we look hard enough we can find corners of the multiverse where the counterfactuals are explored, as just discussed. Perhaps on this basis you want to say that the machine is conscious. But what if we have blocked the counterfactual mechanism as Maudlin described? You probably want to say that the machine is unconscious now. Yet there are corners of the multiverse where the blockage fails and the counterfactual mechanism continues to operate. Doesn't that mean that in fact the machine should be conscious, after all? In general, there will be all kinds of variations happening throughout the multiverse. There will be worlds where the counterfactuals are tested and they work, but also worlds where the counterfactuals are tested and they fail (due to blockage or just plain broken machinery). In yet other worlds the machine may fail or misbehave in any number of bizarre ways. It could even implement a different calculation, perhaps even a different consciousness, elsewhere in the multiverse. My basic problem is that the multiverse is so big, and so much variation is possible, that it will not work to say that consciousness depends on activity that is spread across the multiverse. If you try to pin it down so that it depends on what happens merely on a particular world and its near neighbors,
Re: Maudlin's Machine and the UDist
On Mon, Aug 08, 2005 at 09:42:06AM -0700, Hal Finney wrote: Russell replied: That is indeed my meaning. What difficulties do you see? I see a few problems. The first is the concept that the multiverse will contain copies of the machine that execute the counterfactuals. While this could happen, it would normally be to a very, very tiny degree, and it seems highly implausible that this could cause any actual effects. In the case of Maudlin's thought experiment, the intention of the operator is to run the machine on the specific inputs that it is pre-programmed to handle. In effect, it is doing a replay of a previously computed calculation. The counterfactual machinery is not intended to be activated. All of the effort and skill of the operator will be devoted to pre-setting the machine into the precise state necessary for an exact replay, as Maudlin describes in detail. This is actually impossible to achieve in the Multiverse. For every branch where a given input that is 1, there are as many branches where the input is 0. In order to achieve the situation you're talking about, you would need to narrow down the Multiverse to just those branches in which the environment was fixed to a given sequence. If the environment is deterministic, and the conscious entity is deterministic, the universe is deterministic - in other words there is there is only one branch in your sub-Multiverse. My point is that Maudlin's argument shows that (assuming computationalism) the projection of the conscious entity onto that single branch is not conscious, but does not invalidate computationalism itself. Look, nobody expects a single neuron to be conscious. So is it not such an extreme position to require consciousness to be spread across multiple Multiverse branches. My basic problem is that the multiverse is so big, and so much variation is possible, that it will not work to say that consciousness depends on activity that is spread across the multiverse. If you try to pin it down so that it depends on what happens merely on a particular world and its near neighbors, then that will not in general test the counterfactuals. You do need to subset the Multiverse to those systems that contain a stable Olympia + unblocked Klara. I don't see this as posing a problem though. All counterfactuals not fatal to Olympia + Klara will be tested. If you expand it to include a wide range of possibilities, then there is too much going on, too many variations and bizarre outcomes, so that the criteria you are trying to use for consiciousness are met in some worlds and contradicted in others. All in all I don't think this approach will work as a general method for making consciousness supervene on physicality. Hal Finney If you ask my personal opinion, it is that certain amount of indeterminism is required within the machinery for consciousness to supervene on it (ie computationalism is strictly false). Whilst I've strongly argued that this is essential for free will, unfortunately I haven't seen an argument demonstrating its necessity for consciousness. Maudlin's argument does have this escape clause (computationalism supervenient on multiple branches of a Multiverse), weird though that may be. Cheers -- *PS: A number of people ask me about the attachment to my email, which is of type application/pgp-signature. Don't worry, it is not a virus. It is an electronic signature, that may be used to verify this email came from me if you have PGP or GPG installed. Otherwise, you may safely ignore this attachment. A/Prof Russell Standish Phone 8308 3119 (mobile) Mathematics0425 253119 () UNSW SYDNEY 2052 [EMAIL PROTECTED] Australiahttp://parallel.hpc.unsw.edu.au/rks International prefix +612, Interstate prefix 02 pgp06vCfFLvBX.pgp Description: PGP signature
RE: Maudlin's Machine and the UDist
Stephen has a number of fine questions about Hal's paper (way *too* many, really) and while I am still working on what one or two questions I may pose, there is one of Stephen's questions that perhaps I can answer: I am still worried about how a measure can exist over a set, collection, class, or whatever of computations! Does not the notion of a measure require the existence of a space where each point is an object of the class and the measure itself defines the similarity/difference between one object, here a computation, and some given other? Here is what I would guess to be the UDist answer: The notion of a measure *does* require the existence of a space, and a space, by usual mathematics conventions does consist of points, as you imply. But no, I would *not* say that a measure defined on this space has anything to do with the similarity/difference between points of the space. (You refer to these as objects which seems like a land mine of confusion, at least to me :-) I think that what is meant is very similar to the ordinary mathematical definition of measure which you know. Each point is mapped to a real number that sort of indicates its weight. (In math, a measure is defined on some *subsets* of the space, but here only a single point at a time needs to have a weight.) So for two computations (in the UDist view), each may be thought of as a point in the space metaphor, if you want. Suppose that computation X has greater measure than computation Y. X might be, for example, a calculation that proves 1,000,000,000,061 is prime. And Y might be Lee Corbin. As it takes a much, much shorter program to give rise to X than to Y (I say not with just a trace of pride), the measure of X is greater than that of Y (unfortunately for me). What ontological status does Computation Space have? The space of all computations is a subset of the space of all math patterns, and we math Platonists regard them as exceedingly real. Lee P.S. Platonists != UDist-ers != computationalists != COMP
RE: Maudlin's Machine and the UDist
Russell said (Hal's paraphrase) I guess that you would say that if the unused counterfactual machinery would actually work if tested, then she is conscious; but if the counterfactual machines were broken or blocked such that they wouldn't work (even though they are not used) then she is unconscious. And perhaps you can say that the machines are in fact tested in other branches of the multiverse, so the criterion is more than merely a hypothetical difference between unused working machines and unused broken machines. I see some difficulties with this position but I better first hear whether this is what you have in mind before trying to extrapolate further. Russell replied: That is indeed my meaning. What difficulties do you see? There is an almost practical objection. After we have robots who appear to be very conscious in every way, it would be very peculiar to inquire from one hey, just how *deterministic* are you?. If the robot says, Oh, don't worry, I have a quantum random number generator built in that guarantees that what I say, do, and think is slightly but firmly unpredictable in principle. So you see, I am really conscious. The other robot says, I used to believe that. But then one day I discovered that the chip containing my randomizer had recently developed problems, and during odd hours I'm fully deterministic, but not the during even hours. But I *never* notice any difference! I swear that I'm still conscious [looks inside his head] uh, right now. Yes, from one philosophical position the second robot's narrative must be dismissed, but I just don't believe that position. Lee
RE: Maudlin's Machine and the UDist
I wrote P.S. Platonists != UDist-ers != computationalists != COMP and meant != to have the programming meaning of not equal. For example, I am a (math) Platonist and also a computationalist, but don't know enough about (Bruno's) COMP to say anything, and am skeptical of UDist. Surely all UDist-ers are Platonists, and philosophical computationalists are a superset of the followers of COMP. I am wary of venturing any other inclusions. Lee
Re: Maudlin's Machine and the UDist
On Sun, Aug 07, 2005 at 01:20:22PM -0700, Hal Finney wrote: ... In a nutshell, Maudlin argues that these two common views on the matter are actually in contradiction. But frankly, although Maudlin's argument is complicated and involves all kinds of thought experiments and elaborate, imaginary machines, I think it is actually quite an obvious point. Supervenience means that implementation depends on what P does, while support for counterfactuals means that implementation depends on what P doesn't do. Q.E.D.! Maudlin merely takes great time and care to illustrate the contradiction in detail. ... Now to bring in the multiverse perspective, in the flavor I have been pursuing. From the viewpoint that says that all information objects exist and are distributed according to the Universal Distribution (UDist), what can we say about these questions? The take home message I get from Maudlin's experiment is that a computationalist consciousness is supervenient on a physical process _spread_ over the multiverse, ie the counterfactuals must really exist as alternate branches of the Multiverse. A far as your UDist argument goes, the fact that a conscious HLUT, or a conscious clock has very low measure simply means it is very unlikely for us to be one of these things. They would still be conscious. However accepting the Multiverse would eliminate these objects from being conscious at all, because tof the lack of counterfactuals. -- *PS: A number of people ask me about the attachment to my email, which is of type application/pgp-signature. Don't worry, it is not a virus. It is an electronic signature, that may be used to verify this email came from me if you have PGP or GPG installed. Otherwise, you may safely ignore this attachment. A/Prof Russell Standish Phone 8308 3119 (mobile) Mathematics0425 253119 () UNSW SYDNEY 2052 [EMAIL PROTECTED] Australiahttp://parallel.hpc.unsw.edu.au/rks International prefix +612, Interstate prefix 02 pgp0AEGwue60E.pgp Description: PGP signature
Re: Maudlin's Machine and the UDist
Hi Hal, Thank you very much for you work in writing this review and commentary of the Maulding paper. I have not read it yet, but would like to ask some questions and interject some comments, even if I end up looking like a fool. ;-) Interleaving - Original Message - From: Hal Finney [EMAIL PROTECTED] To: everything-list@eskimo.com Sent: Sunday, August 07, 2005 4:20 PM Subject: Maudlin's Machine and the UDist Rutgers philosopher Tim Maudlin has a paper intended to challenge certain views about consciousness and computation, which we have discussed occasionally on this list. It is called Computation and Consciousness, Journal of Philosophy v86, pp. 407-432. I have temporarily put a copy online at http://www.finney.org/~hal/maudlin.pdf . This is a personal copy and I would ask you not to redistribute it. [SPK] It is sad that copyrights are what they have become such that the free flow and accessibility of papers is becoming more and more like the guilds of yore. :_( The background question is when a given physical system can be said to implement a given computation (especially, a conscious computation). We imagine that the computation is specified in abstract terms, perhaps as a program in a conventional computer language, or perhaps in a lower-level form as a program for a Turing Machine or some other model of computation. But it is abstract. We then ask, given a certain physical system P, is it implementing computation C? [SPK] This seems to touch on the question that Chalmers asked here: http://consc.net/papers/rock.html In practice, it seems that this is an easy question to answer. Our computers implement the programs which are fed into them. No one denies that. But philosophers have argued that there is a way of viewing the activity of a physical system that can make it appear to be implementing *any* computation C. We can think of C as passing through a series of states: C1, C2, C3, and so on. And we can think of the physical system P as passing through a series of states: P1, P2, P3. So if we map P1 to C1, P2 to C2, and so on, we can argue that P is implementing C, for any C and for pretty much any P. [SPK] It seems to me that there should be nothing special about the ordering of the P_i for the COMP assumptions to hold, OTOH, there seems to be some requirement that some aspect of P_n be relatable to P_n-1 in a way that is independent of how the particular P_i are extent, no? The philosophers' response to this is that it is not enough to be able to set up such a mapping. What is also necessary, to claim that P implements (or instantiates) C, is that the *counterfactuals* would have to exist in correspondence as well. That is, not only the states C1, C2, C3 but also other states of C that would have been taken if the inputs had been different, have to have correspondences in P. It is claimed that the kind of arbitrary mapping described above will not work once counterfactuals are taken into account. (I'm not sure I fully understand or agree with this rebuttal but it is well accepted in the field.) [SPK] The contrafactuals have been shown, at least in QM experiements, to be just as *real* when it comes to notion of causation as the factuals. So if we are to take empirical evidence as a guide, it seems that there is some reason to expect that contrafactuals can not be dismissed out of hand. http://www.nonlocal.com/quantum-d/v2/kbowden_03-15-97.html The principle that whether P implements C depends on these counterfactuals is one of the issues that Maudlin addresses. When referring to conscious computations, this principle is generally considered part of the computationalist hypothesis, that the instantiation of an appropriate computation C in a physical system P actually produces a corresponding sensation of consciousness. Implicit in this hypothesis is that to be said to be instantiating C, P must have had enough structure to also produce counterfactuals, if they had occured. Well, that's a lot of background! And there's more. The other thesis that Maudlin considers is called supervenience, a philosophical word for what is fortunately a pretty straightforward concept. It is that whether a physical system implements a given computation depends solely on the activity of that physical system. No mystical or non-physical elements or attributes need to be considered in judging whether P implements C. All that matters is P's physical activity. [SPK] We can hope that no obscurum per occultum is involved! ;-) In a nutshell, Maudlin argues that these two common views on the matter are actually in contradiction. But frankly, although Maudlin's argument is complicated and involves all kinds of thought experiments and elaborate, imaginary machines, I think it is actually quite an obvious point. Supervenience means that implementation depends on what P does, while support for counterfactuals means that
