On Mar 24, 4:32 am, Bruno Marchal <marc...@ulb.ac.be> wrote: > On 23 Mar 2012, at 00:06, Craig Weinberg wrote:
> >>> How does a digital artificial intelligence make sense of it's world > >>> without converting or sampling every truth about that world > >>> available > >>> to it into digital? > > >> First, the fact that the digital machine converts "truth about the > >> world" into digital, does not make that world digital. > > > Which world, the one being converted or the one that has been > > converted? > > The one being converted. Then yes, of course. I am not claiming that the entire universe becomes digital because I turn a computer on. > > > > >> UDA explains > >> why the contrary occurs, through the first person indeterminacy > >> bearing on a very huge and complex arithmetical reality. > > > Why does hugeness, complexity, first person, or indeterminacy affect > > whether something is digital or not? > > Because there is a continuum of computational histories (computations) > going through your state in arithmetic, or in the UD. There are histories. OK. Why does that make them digital or not? >The rest follows > from the 1-indeterminacy and its invariance for the huge delays in the > UD virtual reconstitutions. Ask more if this is unclear, but you are > supposed to have study the UDA. Yes, I don't really get where 'delays' come from. Does the UDA exist in 'time'? Is time an inevitable epi- of +, *, and n? Still not seeing a connection with whether something is digital or not. > > > > >> Second, the first person impression of the machine might be (and is > >> necessarily, once you accept Theaetetus' insight) a non digitalizable > >> truth, from the machine point of view. > > > Which of Theaetetus' insight do you mean? > > The definition of knowledge by "true belief". Kp = Bp & p. I think I know what that is, but since Google shows nothing at all for it, please spell it out for me one more time. > > > > > To me speculating that a machine has a first person impression is just > > a way to plug the problem. > > With the Theaetetus definition, modeling the first person by the > knower, and modeling belief by provability (as the incompleteness can > justify), the machine has a first person experience. To me all that says is that since knowledge, 'belief', and provability have a relation, then abracadabra: first person experience. >It is a theorem, > not a speculation. If it is impossible to test the theorem, then what is the difference? > Non comp makes the machine into a mechanical > variant of a zombie. Non-comp is the atheism of computationalism. it does not make a machine anything, it lets machines be the inanimate puppets that they have always been. It is comp which introduces the expectation of sentience in machines and then balks at the idea of their absence. > > > Since subjectivity doesn't make sense > > mechanically or digitally, > > This is what we ask an argument for. You beg the question. I understand, I'm just offering insight into the psychology behind the formulation of the idea of first person machine states. > > > we'll say it's non digital and hang a > > Mission Accomplished sign. It takes advantage of the privacy and > > ineffability of subjectivity to misrepresent its absence as possible > > presence, even though our experience with machines thus far has not > > supported any presence at all. > > >> That is actually the case, > >> because the impression of the machine is a conjunct of both a > >> digitalizable belief and (some) truth, with the greek's suggested > >> notion of knowledge. > > > beliefs, truths, knowledge...these are abstractions to me. Tertiary > > level commentaries on experience which barely exist. Important, sure, > > but lacking in any power to initiate direct action. Which is maybe why > > a machine has no such power. > > >>>>>> If you negate this, it means that > >>>>>> you assume the level to be "infinitely low", > > >>>>> No, it means I understand that your assumption that description > >>>>> can be > >>>>> quantified is simplistic and inaccurate. > > >>>> Description of my (generalized) brain. With your theory we have > >>>> zombie. > > >>> Never zombie, only puppet. Zombie is like calling water 'wet fire'. > > >> I don't think so. With the Heisenberg Matrix of the Milky way, no-one > >> would be able to distinguish me or you from the behavior of the > >> entity > >> simulated. If comp is false, those entity who behaves "exactly" like > >> you or me, are not conscious, and so they are by definition p-zombie. > > > The word zombie privileges the expectation of consciousness, while the > > word puppet does not. > > So you beg the question here too. By definiion a zombie has no > consciousness, but behaves like a conscious being. A puppet does the same thing. Why not use puppet instead? There must be some reason why you resist this change. > > > The fact that the Heisenberg Matrix of the Milky > > way fails completely in detecting consciousness > > I said that we fail to do that. Not the (virtual) Milky Way. It's our virtual Milky Way, so that's the same thing. > > > does not mean that the > > simulations it takes for genuine are missing something that they > > should have. > > > That they are simulations in the first place means that we intend them > > to resemble something that they are not. > > We don't know that. Why not? I make a simulation of the Milky Way but don't know that I intend the resemblance to it? > > > I make a witch's head out of > > a shrunken apple - that is a simulation of a witch; a puppet of a > > witch. It isn't an actual witch that has been rendered a zombie. It > > freaks me out that I keep having to explain this, it seems stunningly > > obvious to me. I draw a dog on a paper bag, is it now a dog zombie? > > > The deeper insight here of course is that simulation relates only to > > the intended audience. > > Nobody has been supposed to look at the simulation. Then why create it? > > > It looks like a witch or a dog only to us. It's > > a text. > > Not at all. It is a computation. It concerns true relations which > cannot be considered in a text. What true relations are you talking about? That it truly looks like a witch to us means that it truly looks like a witch to a house fly? http://deuxfilleslemercredisoir.files.wordpress.com/2011/09/ceci-nest-pas-une-pipe.jpg > > > It's symbols are not grounded in the firmament of the cosmos > > like matter or biology, they are only suggestions that help 'us' > > fabricate a fictional association for ourselves. > > Not at all. Why not? > > > > >>>> Study the sane04 paper. Or search in the archive. It is a > >>>> consequence > >>>> of comp that physics emerge from the way numbers can bet on > >>>> arithmetical relations. It is not entirely obvious. > > >>> I don't have a problem with physics emerging from comp, I have a > >>> problem with consciousness emerging from either one. > > >> I understand. > >> This is related with the fact that we can explain why consciousness > >> is > >> not entirely explainable by the machine. > >> The machine also find hard to believe that consciousness "emerge" > >> from > >> arithmetical truth. But she cannot accept this because she can > >> understand the necessary transcendent aspect of any notion of truth > >> encompassing the truth about herself. > > > It sounds like you mean that machines can tell what is true about > > themselves so they can't accept that there is an external source of > > all truth. That makes a little sense, but I have no problem accepting > > that there could or could not be an external source of all truth, I > > just don't see why it would be arithmetic in either case. > > This mean you have not study the UDA. I would if I could, but it really doesn't work for me when I try to do that. My understanding of UDA goes something like this: Instead of the big bang, the universe may be a single program running each and every possible program (which means making each possible program within each program) such that we are a function of the particular sheath of nested programs which have given rise to a machine capable of sustaining machines like us. Is that not more or less the case? > > > > >> What might help you is the idea that with comp, contrary to the > >> (contradictory) belief that consciousness is a natural product of > >> some > >> physical activity, consciousness is an atemporal global feature of > >> the > >> arithmetical truth. Machine does not produce consciousness, they > >> borrow it to the truth. With comp, there is a sense to say that only > >> God is conscious, but to manifest itself relatively to machines, he > >> has to be amnesic to who he is. Nature invented the brain, not to > >> make > >> machines conscious, but to make them amnesic to their true identity. > >> If not, the prey would not mind escaping the predator, and life could > >> not evolve. > > > I don't have a problem with any of that, but why would it be only > > arithmetic and not multisensory? > > Because we assume comp. The rest is logic. Ok, so you admit it is purely arbitrary from the beginning. You don't need to consider assuming comp instead because you choose to comp and that's that. Doesn't matter what the universe actually is, only what theory you are invested in. That's ok, but I'm only interested in the theory that explains why multisense can or can't be true. > > > > >>>> They don't need to dream. But they can (trivially with comp that I > >>>> assume all the time). Physics appears because deep linear dreams > >>>> are > >>>> shared by relatively persistent universal numbers. > > >>> Couldn't physics still theoretically appear without anything > >>> dreaming > >>> anything? Isn't that the point of physics, to explain the world as a > >>> structured relation which makes sense entirely in physical terms? > > >> That's is very natural to believe. Somehow we are programmed to > >> believe this. > > > My view explains the 'Somehow'. Symmetry makes us see the universe on > > one level as that which we are not. > > >> But it is false in the comp theory, as the UDA > >> explains, and as most mystics seems to grasp from introspection. > > > Yes, it is false in one sense, but it is also true in another - hence > > the 'Somehow'. > > >> Note that this makes also possible to explain the physical in a non > >> circular way, that is without assuming the physical. > > > Yes, but it only explains it in a deeper but still just as circular > > way. > > Why. If physics is explain from numbers, the explanation is not > circular. Because it then fails to explain numbers. What's the difference whether you explain numbers as physics or physics as numbers? Neither of them need to feel like anything, and that is the only thing that would make the universe matter. What makes numbers matter? > > > > >> This is counter-intuitive, and has to be counter-intuitive. We are > >> not > >> programmed to believe this, quite the contrary, as suggested by the > >> above remark. > > >>>>>>> and if we have dreams we don't need numbers. > > >>>>>> ? > > >>>>> If our lives are being dreamed by Platonic universal principles, > >>>>> why > >>>>> do we want to turn them into dust by seeking out the dreamers? > > >>>> Because we are curious. > > >>> In Comp that can make sense that we would be curious about what can > >>> only reveal our own lives meaningless data. In a sense realism, we > >>> are > >>> only likely to be curious because on some level we already know that > >>> comp is just another new dimension of meaning to explore and create > >>> in. We know, on some level, that we don't really have to worry about > >>> computers coming to life or developing feelings. > > >> There is nothing to worry indeed. About machines coming to life. They > >> will treat us as we treat them. > > > Not worry about them being hostile, I mean worry about it ever coming > > to pass that any machine ever comes to life at all - ie develops > > intention and self interest. I'm saying that deep inside, we know that > > is not possible in reality. > > We know nothing of that sort. In science, we never know. We believe > temporarily. I'm not talking about what we think we know, I'm talking about out deep intuitive connections that are true. > > > > >> We can worry only about ourselves. > >> Brains and universal numbers are relative dynamical mirror. If we > >> fear > >> them, they will fear us, and that can lead to a diabolical vicious > >> circle. This is true also in between humans. > > > Is there a mathematical proof that addresses fear of universal > > numbers? > > Many. Google on emotion, machine, AI. One of my student has make a > thesis on machine emotion (Giovona Colombetti), a long time ago. I'll take your word for it, since I wouldn't be able to understand what it said anyhow. > > > > >> Not really, because the "real feeling of certainty" comes from the > >> first person experience. > > > But isn't the content of that feeling all about transcending or > > sharing beyond first person experience? > > No. First person singular feeling are not sharable. How do you know? Love? Hate? These things are isolated parameters in the dark? > > > > >> But the publicly communicable certainty is > >> only in the sharable deduction, because the first person experience > >> is > >> just not publicly sharable. > >> To communicate the 1-truth would be like trying to communicate that > >> "in reality you are the one having been reconstituted in Washington, > >> which is either a trivial 3p statement for your interlocutor, or a > >> sort of delusion as the doppelganger in Moscow can understand > >> (because > >> *he* lives the delusion of being "in reality" the guy in Moscow). > >> This > >> can explain the frustration we can have about communicating that we > >> are what we are. That's a 3-tautology, lived as an incredible non > >> tautological 1-experience. But then this can give rise to deep poem > >> and music. > > > You don't have to communicate the 1-truth, you communicate a 3-symbol > > and the audience shares your reflected 1-truth as their own locally > > isomorphic 1-truth. > > ? You don't have to transplant your feelings into another person, you give them a signal with a shared association and they experience their version of your feeling (or their reaction to your feeling, or some combination thereof). > > > > >>>>> I'm assuming them though. As you assume addition and > >>>>> multiplication, I > >>>>> could assume odor and color and go from there. > > >>>> Except that addition is far simpler than a notion of odor and > >>>> color. > > >>> Tell that to a two year old. > > >> It is 3-simpler. A simple abacus can illustrate the point. The odor > >> is > >> only 1-simpler, for the owner of a sophisticate universal machine- > >> brain. Here I was using "simple" in the usual 3p sense. "1+5 = 6" is > >> simpler than the universal machinery to get the smell of something. > > > That's not comparing apples to apples though. The universal machinery > > to get the smell may be incredibly primitive, while for something to > > actually understand that 1+5 = 6 (rather than to simply be embody it) > > may require some kind of a brain. The abacus doesn't know that 1+5 = > > 6. It doesn't even know + or =. It only knows xxoooxxxx. Same as a > > computer. Odor could go all the way down to quarks for all we know. > > You confuse "1+5=6" with "I understand 1+5=6". You confuse "bacon smells good" with "I smell bacon". > > > > >>>> We need a big human brain to get easily the human color. > > >>> How do you know that the color we see isn't what the molecules in a > >>> cone cell see? > > >> I think this is not plausible from what we can believe about brain. > >> But even if true, a molecule is by itself a very complex object/ > >> notion, far more complex than natural numbers. > > > We don't know that natural numbers exist independently of the totality > > of arithmetic or consciousness. That is far more complex than a > > molecule. > > You confuse "1+5=6" with the whole of arithmetic. No need for that in > my argument above. How does 1+5=6 without 6-5=1, 1+2+3=5+1, etc? > > > > >> I am OK, up to the vagueness present. > >> Comp leads in that direction too. > > > It shouldn't be vague. I think any vagueness comes from comp being > > insensitive to the universally obvious. > > Only for you, apparently. I think presentation is a fairly clear and universal term. > > > > >>>>> I could say arithmetic > >>>>> is part of the dream of universal qualia. > > >>>> Everybody agree and grasp arithmetic. for "qualia" I regularly met > >>>> people who don't grasp the notion. > > >>> That's what makes qualia the counter-intuitive universal truth. > > >> This contradicts the intuitive nature of qualia, that both you and I > >> agree already on. > > > The referent of qualia is intuitive, but the notion of it is, as you > > agreed, not grasped by most (counter-intuitive). > > >> We just interpret it differently, which is normal given that we work > >> in different theories. > > >>> Of > >>> course we don't readily grasp ourselves in the world since the world > >>> is the reflection of exactly what we are not, not what we are. > > >>>>> All digital patterns could > >>>>> be reduced to stereo distinctions of simpler whole qualia into > >>>>> which > >>>>> complexity is injected as an afterthought. > > >>>> Please do it. > > >>> 1 = feeling of flow > >>> 0 = feeling of holding back > > >>> done > > >> Hmm... > > > 1 = actively assert > > 0 = passively allow > > >>>>> I don't assume non-comp, I assumed comp for many years until I > >>>>> eventually saw reason to stop assuming it. > > >>>> All the reason you gave are more related to the reductionist > >>>> conception of nulbers and machine, which can no more be sustained > >>>> after Gödel's and Turing's discoveries. > > >>> I don't pretend to understand Gödel's and Turing's discoveries as > >>> you > >>> do or as even a casual math enthusiast does, but I don't think I > >>> misunderstand them. I don't share your interpretation that > >>> incompleteness proves the boundless fertility of arithmetic truth, I > >>> think it proves the reverse, that all arithmetic logics are limited > >>> and incapable of justifying their own axioms, therefore impossible > >>> to > >>> ground fundamental universal principles. > > >>> Is there any relevant point of Gödel's and Turing's work that you > >>> can > >>> tell me about in layman's terms that you think would change my > >>> position if I understood it? I think that you think I underestimate > >>> the capacity of Universal Machines, but I think that is not the case > >>> at all. I am fully prepared to see Turing based logic emulate > >>> everything from Jello pudding to sex slaves eventually, but I don't > >>> think any of it will be really real, even to us. They will always be > >>> CGI churn because they aren't grounded in the overall narrative of > >>> the > >>> cosmos. They have no unified matter-energy-sense-motive-time-space > >>> address, just assemblies of objects that are forever strangers to > >>> each > >>> other. > > >> You are right that Gödel's incompleteness shows the limitations of > >> the > >> finite entities (logic, machine) with respect to truth. But machines > >> can prove their own incompleteness theorem and learn to overcome > >> their > >> limitations, and accept the price of it, giving them the possibility > >> for some courage and grandeur. > > > Are you saying that a particular enactment of a machine can be more > > courageous than another identical machine? Can a program running on my > > computer do something that the same program running on a different but > > identical computer lacks the courage to do? > > Yes. Comp is compatibilist. In theory. Have we seen chronically timid iterations of identical machines? > > > > >> You don't convince me that unified matter-energy-sense-motive-time- > >> space even exists, nor even that it makes sense, and still less that > >> something like that would be necessary for consciousness. > > > See if this new metaphor I posted today helps: > > > Lights - Camera - Action! > > > Corresponding to the Who and Why, "Lights" direct attention as they > > enable access to the content of attention. Light is sense and motive. > > > "Camera" is the What and How of motion picture production. It is the > > objective vehicle or body with an entirely non-directorial > > cinematographic mechanism. The camera doesn’t care what the film looks > > like. Camera is matter and energy; sense and motive seen from the > > outside looking back in. > > > "Action!" is the When and Where. It is the coming together of > > cinematic subjects and objects in a syncretic tableau that is unique > > and unrepeatable. It is the LIVE ACTION Mise-en-scène which is > > recorded for later use as a vicarious sense experience to the camera’s > > audience. > > ? > > > > >> It looks to much like a prejudice against them, or a feeling of > >> superiority from your part. > > > Not at all. That's the funny part. I'm much more a fan of computers > > than human beings. I have no emotional attachment one way or the other > > to comp, but I do have an emotional attachment to my understanding of > > why comp cannot be true. > > But you fail to convey that understanding. Because it isn't important. > > > > >> I don't think either that we can ground any fundamental universal > >> principle in a manner such that we feel them as truth. They can only > >> be a sort of hope or bet, and we might find better one by listening > >> to > >> others and looking into the unknown. The rest seems to me like > >> wishful > >> thinking or pseudo-religion. It is marmalade to hide the pill. It > >> seems to me. > > > That's exactly the way it has to look from the comp-favoring (or > > "Camera"-favoring, if you prefer) perspective. Comp seems like a > > wishful thinking or pseudo-religion to me too. How could it not? > > Because we don't pretend it is true. We reason from it. I don't pretend multisense is true, I understand from it. Craig -- 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.
