Bruno, Thanks for the references.
--Abram ps- it is final exam crunch time, so I haven't been checking email so much as usual... I may get around to more detailed replies et cetera this weekend or next week. On Sun, Dec 7, 2008 at 1:12 PM, Bruno Marchal <[EMAIL PROTECTED]> wrote: > > On 07 Dec 2008, at 06:19, Abram Demski wrote: > > Bruno, > > Yes, I think there is a big difference between making an argument more > detailed and making it more understandable. They can go together or be > opposed. So a version of the argument targeted at my complaint might > not be good at all pedagogically... > > I would be pleased if you can give me a version of MAT or MEC to which > > the argument does not apply. For example, the argument applies to most > > transfinite variant of MEC. It does not apply when some "magic" is > > introduced in MAT, and MAT is hard to define in a way to exclude that > > magic. If you can help, I thank you in advance. > > My particular brand of "magic" appears to be a requirement of > counterfactual/causal structure that reflects the > counterfactual/causal structure of (abstract) computation. > > Sometimes I think I should first explain what a "computation" is. I take it > in the sense of theoretical computer science, a computation is always define > relatively to a universal computation from outside, and an infinity of > universal computations from inside. This asks for a bit of computer science. > But there is not really "abstract computation", there are always relative > computation (both with comp and Everett QM). They are always concrete > relatively to the universal machine which execute them. The starting point > in no important (for our fundamental concerns), you can take number with > addition and multiplication, or lambda terms with abstraction and > application. > > > > Stathis has > pointed out some possible ways to show such ideas incoherent (which I > am not completely skeptical of, despite my arguments). > > I appreciate. > > > Since this type > of theory is the type that matches my personal intuition, MGA will > feel empty to me until such alternatives are explicitly dealt a > killing blow (after which the rest is obvious, since I intuitively > feel the contradiction in versions of COMP+MAT that don't require > counterfactuals). > > Understanding UD(1...7) could perhaps help you to figure out what happens > when we abandon the physical supervenience thesis, and embrace what remains, > if keeping comp, that is the comp supervenience. It will explain how the > physical laws have to emerge and why we believe (quasi-correctly) in brains. > > > > > > Of course, as you say, you'd be in a hard spot if you were required to > deal with every various intuition that anybody had... but, for what > it's worth, that is mine. > > > I respect your intuition and appreciate the kind attitude. My feeling is > that if front of very hard problems we have to be open to the fact that we > could be surprised and that truth could be counterintuitive. The > incompleteness phenomena, from Godel and Lob, are surprising and > counterintuitive, and in the empirical world the SWE, whatever > interpretation we find more plausible, is always rather counterintuitive > too. > I interpret the "self-referentially correct scientist M" by the logic of > Godel's provability predicates beweisbar_M. But the intuitive knower, the > first person, is modelled (or defined) by the Theatetus trick: the machine M > knows p in case "beweisbar_M('p') and p". Although extensionally equivalent, > their are intensionally different. They prove the same arithmetical > propositions, but they obey different logics. This is enough for showing > that the first person associated with the self-referentially correct > scientist will already disbelieve the comp hypothesis or find it very > doubtful. We are near a paradox: the correct machine cannot know or believe > their are machine. No doubt comp will appear counterintuitive for them. I > know it is a sort of trap/ the solution consists in admitting that comp > needs a strong act of faith, and I try to put light on the consequences for > a machine, when she makes the bet. > > The best reference on the self-reference logics are > Boolos, G. (1979). The unprovability of consistency. Cambridge University > Press, London.Boolos, G. (1993). The Logic of Provability. Cambridge > University Press, Cambridge.Smoryński, P. (1985). Self-Reference and Modal > Logic. Springer Verlag, New York.Smullyan, R. (1987). Forever Undecided. > Knopf, New York. > > The last one is a recreative book, not so simple, and rather quick in the > "heart of the matter" chapter. Smullyan wrote many lovely books, recreative > and technical on that theme. > The bible, imo, is Martin Davis book "The undecidable" which contains some > of the original papers by Gödel, Church, Kleene, Post and indeed the most > key starting points of the parts of theoretical computer science we are > confonted to. It has been reedited by Dover. > Bruno > Other references here: > http://iridia.ulb.ac.be/~marchal/lillethesis/these/node79.html#SECTION001300000000000000000 > > --Abram > > On Sat, Dec 6, 2008 at 9:32 AM, Bruno Marchal <[EMAIL PROTECTED]> wrote: > > > Le 05-déc.-08, à 22:11, Abram Demski a écrit : > > > Bruno, > > Perhaps all I am saying is that you need to state more explicitly the > > assumptions about the connection between 1st and 3rd person, in both > > MEC and MAT. Simply taking them to be the general ideas that you take > > them to be does not obviously justify the argument. > > > I don't see why nor how. The first person notions are defined in the > > three first steps of the UDA. Wait I come back on this in the > > discussion with Kim perhaps. In AUDA I define the first person by the > > "knower", and I use the classical definition proposed by Theaetetus in > > the Theaetetus of Plato. Keep in mind that you arrived when I was > > explaining the real last step of an already long argument. > > Of course you may be right, and I would really appreciate any > > improvements. But making things more precise could also be a red > > herring sometimes, or be very confusing pedagogically, like with the > > easy 1004 fallacy which can obviously crop here. > > When I defended the thesis in France, it was already a work resulting > > from 30 years of discussions with open minded physicists, engineers, > > philosophers and mathematicians, and I have learned that what seems > > obvious for one of them is not for the others. > > I don't think there is anything controversial in my work. I got > > academical problems in Brussels for not having find an original result > > (but then I think they did not read the work). Pedagogical difficulties > > stem from the intrinsical difficulty of the mind body problem, and from > > the technical abyss between logicians and physicists to cite only them. > > It is more easy to collide two protons at the speed of light (minus > > epsilon) than to arrange an appointment between mathematical logicians > > and mathematical physicists (except perhaps nowadays on quantum > > computing issues thankfully). > > > > Furthermore, stating the assumptions more clearly will make it more > > clear where the contradiction is coming from, and thus which versions > > of MEC and MAT the argument applies to. > > I would be pleased if you can give me a version of MAT or MEC to which > > the argument does not apply. For example, the argument applies to most > > transfinite variant of MEC. It does not apply when some "magic" is > > introduced in MAT, and MAT is hard to define in a way to exclude that > > magic. If you can help, I thank you in advance. > > Bruno > > > > --Abram > > On Fri, Dec 5, 2008 at 4:36 AM, Bruno Marchal <[EMAIL PROTECTED]> > > wrote: > > > On 04 Dec 2008, at 15:58, Abram Demski wrote: > > > PS Abram. I think I will have to meditate a bit longer on your > > (difficult) post. You may have a point (hopefully only pedagogical > > :) > > A little bit more commentary may be in order then... I think my point > > may be halfway between pedagogical and serious... > > What I am saying is that people will come to the argument with some > > vague idea of which computations (or which physical entities) they > > pick out as "conscious". They will compare this to the various > > hypotheses that come along during the argument-- MAT, MEC, MAT + MEC, > > "Lucky Alice is conscious", "Lucky Alice is not conscious", et > > cetera... These notions are necessarily 3rd-person in nature. It > > seems > > like there is a problem there. Your argument is designed to talk > > about > > 1st-person phenomena. > > The whole problem consists, assuming hypotheses, in relating 1-views > > with 3-views. > > In UDA, the 1-views are approximated by 1-discourses (personal diary > > notes, memories in the brain, ...). But I do rely on the minimal > > intuition needed to give sense to the willingness of saying "yes" to a > > digitalist surgeon, and the believe in a comp survival, or a belief in > > the unchanged feeling of "my" consciousness in such annihilation- > > (re)creation experiences. > > > > > If a 1st-person-perspective is a sort of structure (computational > > and/or physical), what type of structure is it? > > The surprise will be: there are none. The 1-views of a machine will > > appears to be already not expressible by the machine. The first and > > third God have no name. Think about Tarski theorem in the comp > > context. A sound machine cannot define the whole notion of "truth > > about me". > > > If we define it in > > terms of behavior only, then a recording is fine. > > We certainly avoid the trap of behaviorism. You can see this as a > > weakness, or as the full strong originality of comp, as I define it. > > We give some sense, albeit undefined, to the word "consciousness" > > apart from any behavior. But to reason we have to assume some relation > > between consciousness and possible discourses (by machines). > > > If we define it in > > terms of inner workings, then a recording is probably not fine, but > > we > > introduce "magical" dependence on things that shouldn't matter to > > us... ie, we should not care if we are interacting with a perfectly > > orchestrated recording, so long as to us the result is the same. > > It seems like this is independent of the differences between > > pure-comp / comp+mat. > > > > This is not yet quite clear for me. Perhaps, if you are patient > > enough, you will be able to clarify this along the UDA reasoning which > > I will do slowly with Kim. The key point will be the understanding of > > the ultimate conclusion: exactly like Everett can be said to justify > > correctly the phenomenal collapse of the wave, if comp is assumed, we > > have to justify in a similar way the wave itself. Assuming comp, we > > put ourself in a position where we have to explain why numbers > > develops stable and coherent belief in both mind and matter. We can > > presuppose neither matter, nor mind eventually, except our own > > consciousness, although even consciousness will eventually be reduced > > into our "believe in numbers". > > Bruno > > > http://iridia.ulb.ac.be/~marchal/ > > > > > > > > > > http://iridia.ulb.ac.be/~marchal/ > > > > > > > > > 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 [EMAIL PROTECTED] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/everything-list?hl=en -~----------~----~----~----~------~----~------~--~---