On Sun, Oct 7, 2012 at 2:39 PM, Stephen P. King <stephe...@charter.net>wrote:

> On 10/7/2012 4:26 AM, Bruno Marchal wrote: > > > On 06 Oct 2012, at 21:27, Stephen P. King wrote: > > On 10/6/2012 2:51 PM, Bruno Marchal wrote: > > > On 06 Oct 2012, at 17:40, Stephen P. King wrote: > > On 10/6/2012 4:25 AM, Bruno Marchal wrote: > > > On 06 Oct 2012, at 09:35, Stephen P. King wrote: > > Hi Bruno, > > You wrote: > > As the cow-boy guessed right this is assuming too much, both for the > formalism used (which is OK), and the ontology, so it uses implicitly > non-comp hypothesis, which is less OK, as comp is also assumed implicitly. > IT is not uninteresting for possible progress, but it is unaware that > matter as to be explained by statistics on computations "seen from inside". > The role of "Russell operator" is played by the Kleene second recursion > theorem, which encapsulates the "non foundation" well enough. > > > I disagree. His operators are "looking from the outside" at A (the > physical universe). > > > > What do you mean here by "physical universe"? > > > What do you think it means? The common subject of observation by a > collection of observers. > > > What are observers? Where do they come from? > > Hi Bruno, > > It depends on what feature you wish to find an explanatory model > of. My point is that "what is an observer" depends on the features that one > wishes to explain. > > > > The context is the search of a TOE, which does not avoid the problem of > explaining consciousness and matter, and the relation between, or put in > another way the relation between first person views and third person views. > You mentioned the "physical universe", but this is something that we > cannot take for granted in such a context. You defined it by using the > notion of observer, that's OK ... if you define the notion of observer > without mentioning "physical universe" (or if you do it, you have to solve > the recursion, with the second recursion theorem, or, if you want, with Set > Theory + non foundation, à la Barwise, but this must be eliminable in comp, > or put in the machine's epistemology). > > > In my thinking, a physical world = a reality = that which is > incontrovertible (free of contradictions = Boolean Satisfiable) for some > finite collection of observers, where observers are defined as "bundles of > computations". Physical worlds are not actual in the absence of observers. > I also stipulate that there are an infinite (uncountable) number of > physical worlds. This demands that there exists an uncountable infinity of > observers <=> an infinite number of bundles of computations. Please recall > how I define "exist"; it is *necessary possibility*. > > > > You like the modal logical explanatory model, > > > This is not correct. I just model "belief" by the "instantional" manner, > à-la Dennet. > > > One day I might be able to recall your terminology exactly. ;-) > > > A machine believes p if the machine assert p, which makes sense as I > limit myself to machine talking first order language, ideally > arithmetically sound, and being able to believe the logical consequences of > its beliefs in arithmetic. Then modal logic just happens to describe > completely, at the propositional level, the logic of provability of such > machine, thanks to the work of Gödel, Löb and Solovay (and others). > I have never choose to use modal logic, I use only machine self-reference, > where a very special modality imposes itself (G). > > > Sure, you use that is necessitated by non-contradiction principle. ;-) > > > > so there we might think of observers as bundles of computations. > > > OK. That's nice, but what is a computation? > > > A computation is **any transformation of information**. Information = > any difference between two that makes a difference to a third. It is > interesting that this definition demand that there exists at least three > entities or processes or whatever for information and thus computation to > exist. I have not considered the further consequences of this idea so far. > It might be completely fallacious. > > > > Your preceding post were using a notion of physical computation, which > would not cut the regress. > > > I disagree. I am arguing that only if we retain a connection between > computation as a platonic abstraction and the requirement of physical > resources we will have a viable cut off for the regress. A computer that > can only process a finite number of recursions or iterations of > self-modeling will not have an infinite regress for obvious reasons. What I > am trying to do to make this a more formal statement is to tie together the > Kolmogorov > complexity of a description a > system<http://en.wikipedia.org/wiki/Kolmogorov_complexity>, > abstract or physical, with the physical degrees of freedom of a physical > system (for example the dimensions of its Hilbert space or Hamiltonian). In > this way we have a way to define a physical system as a bounded bundle of > computations. This would be a lower bound on the necessary physical > resources required to implement an arbitrary computations. > Following this idea we can see that it implies that physical systems > that require infinite computations to be exactly simulated only can exist > is very special circumstances! > > > But as I answered you can take the original definition of computation > (by Post, Turing & Co.), in which case you can assume only arithmetic, and > the regression is cut, by defining the "bundle of computations" with the > first person indeterminacy. Then you are back to sane04, and you describe > the comp theory. > > > This is where I agree with comp. I only disagree with your step 8: > http://iridia.ulb.ac.be/~marchal/publications/SANE2004MARCHAL.pdf > > "what if we don’t grant a concrete robust physical universe? Up to this > stage, > we can still escape the conclusion of the seven preceding reasoning steps, > by > postulating that a ‘‘physical universe’’ really ‘‘exists’’ and is too > little in the sense of not being > able to generate the entire UD*, nor any reasonable portions of it, so > that our usual physical > predictions would be safe from any interference with its UD-generated > ‘‘little’’ computational > histories." > > Here you seem to only consider a "physical universe" to be an entity > that "exists" with no observers associated with it and reason that such is > an irrelevant concept that can and should be dismissed from consideration. > I am OK with that! I define a "physical universe" differently! (see above) > I am only asking for a weaker version of a "physical universe" to exist; > one that is not ontologically primitive, but to demand this is to also > demand that arithmetics and numbers are not ontologically primitive as > well. Both abstractions, such as numbers and their truths, and physical > worlds must emerge together from a primitive ground which is neutral in > that it has no innate properties at all other that necessary possibility. > It merely exists. > > from your paper: > > "It will follow that a much weaker and usual form of Ockham’s razor > can be used to conclude that not only physics has > been epistemologically reduced to machine psychology, but that > ‘‘matter’’ has been > ontologically reduced to ‘‘mind’’ where mind is defined as the object > study of fundamental > machine psychology." > > But "physics" =/= "physical world". Here is where we diverge in our > agreement! Physics is behavior not "being". Matter =/= "physical world". > Matter is the spatial aspect of sets, once we remove all of the > contingencies of physics explanations. You are claiming to "reduce" matter > (Aristotelian "substance") to "mind" (that which fundamental machine > psychology presupposes) and thus eliminate "matter". But this seems to be > possible only if fundamental machine psychology allows the existence of a > single machine and thus a singular mind. I have tried to point out that > this is exactly consistent with Tennenbaum's > Theorem<http://web.mat.bham.ac.uk/R.W.Kaye/papers/tennenbaum/tennrosser>and > tried to argue that this is a huge problem as it makes comp (or UDA) > into a description of the psychology of a consistent solipsistic machine. > One possibility requires that we "relativize" countable models of PA > using the non-standand arithmetics but in a way that perpetually "hides" > the non-standard component, the constant, of the model from the machine > itself. The inability for a computable model of consciousness to "know" > which computations "it is", is an expression of this property, IMHO. > Stephen, I'm not sure that there is any real disagreement between your view and Bruno's. It seems more to be a language thing, if anything. When Bruno refers to a physics, he means the appearance of a physical world from the perspective of observers. But observers never see a single "physical world", they exist in and see an infinite number of interfering (physical worlds / dreams of numbers). I sympathize with your view that many (most?) observers exist in larger mathematical structures that contain multiple observers, but I don't see that this invalidates step 8. You say "one that is not ontologically primitive, but to demand this is to also demand that arithmetics and numbers are not ontologically primitive as well." Do you agree that at least something has to be ontologically primitive? If so, what do you think that something is? You referred to something that has no innate properties, but I don't find much explanatory power in that. Jason -- 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.