Re: The Super-Intelligence (SI) speaks: An imaginary dialogue
Hello again Jesse, I am going to assume that by trashing computationalism that Marc Geddes has enough ammo to vitiate Eleizer's various predilections so... to that end... Your various comments (see below) have a common thread of the form I see no reason why you can't ..X.. So let's focus on reasons why you can't ...X. These are numerous and visible in real - empirically verifiable physics...let's look at it from a dynamics point of view. In saying 'you can see no reason' would mean that if you chose a computationalist abstraction level (you mentioned atoms) that you would claim the resultant agent able to demonstrate scientific behaviour indistinguishable from a human. I would claim that to be categorically false and testably so. OK. Firstly call the computationalist artificial scientist, COMP_S. Call the human scientist HUMAN_S. Call computationalism COMP. This saves a lot of writing! The test regime: HUMAN_S constructs laws of nature tn using the human faculty for observation (call it P) delivered by real atoms in the brain of HUMAN_S. If COMP_S and HUMAN_S are to be indistinguishable then the state dynamics (state vector space) of COMP_S must be as sophisticated, accessible as HUMAN_S and ALSO /convergent on the same outcomes as those of HUMAN_S/. Our test is that they both converge on a law of nature tn, say. Note: tn is a abstracted statement of an underlying generalisation in respect of the distal external natural world (such as tn = ta, a model of an atom). Yes? That is what we do... the portability of laws of nature tn proves that we have rendered such abstractions invariant to the belief dynamics of any particular scientist.. Yes? HUMAN_S constructs a model of atoms a 'law of nature' = ta. Using that model ta we then implement a sophisticated computational version of HUMAN_S at the level of the model: atoms. We assemble an atomic-level model replica of HUMAN_S. We run the computation on a host COMP substrate. This becomes our COMP_S. We expect the two to be identical to the extent of delivering indistinguishable scientific behaviour. We embody COMP_S with IO as sophisticated as a human and wire it upIf the computationalist position holds, by definition, the dynamics of COMP_S must be (a) complex enough and (b) have access to sufficient disambiguated information to construct tn indistinguishably from HUMAN_S. If computationalism is true then given the same circumstance of original knowledge paucity (which can be tested), A demand for a scientific outcome should result in state-vector dynamics adaptation resulting in the delivery of the same tn (also testable), which we demand shall be radically novel If they are really equivalent this should happen. This is the basic position (I don't want to write it out again!) I would claim the state trajectory of COMP_S to be fatally impoverished by the model ta. (abstracted atoms). That is, the state-trajectory of COMP_S would fail to consistently converge on a new law of nature tn and would demonstrate instability (chaotic behaviour). Just like ungrounded power supplly voltage drift about, a symbolically ungrounded COMP_S will epistemically drift about. Indeed I would hold that would be the case no matter what the abstraction level: sub-atomic, sub-sub atomic , sub sub sub atomic .. etc ... the result would be identical. Remember: there's no such 'thing' as atoms...these are an abstraction - of a particular level of the organisational hierarchy of nature. also note ... so-called' ab-initio quantum mechanics of the entire HUMAN_S would also fail because QM is likewise just an abstraction of reality, not reality. COMP would claim that the laws of nature describing atoms behave identically to atoms. The model ensemble of ta atoms should be capable of expressing all the emergent properties of an ensemble of real atoms. This already makes COMP a self-referential question-begging outcome. HUMAN_S is our observer, made of real atoms. COMP assumes that P is delivered by computing ta when there is no such 'thing' as atoms! Atoms are an abstraction of a thing, not a thing. Furthermore, all the orighinal atoms of HUMAN_S have been replaced with the atoms of the COMP_S substrate. What is NOT in law of nature ta is the relationship between the abstraction ta and all the other atoms in the distal world outside COMP_S. (beyond the IO boundary). Assume you supplied all the data about all the atoms in the environment of the original human HUMAN_S used to construct and initialise COMP_S. You know all these relationships at the moment you measured all the atoms in HUMAN_S to get you model established. However, after initialisation, when you run the COMP_S, all relationships of the model with the distal world (those intrinsic to the atoms which the model replaced) are GONE the instant the abstraction happens, from that
Re: The Super-Intelligence (SI) speaks: An imaginary dialogue
Colin Hales wrote: Hello again Jesse, I am going to assume that by trashing computationalism that Marc Geddes has enough ammo to vitiate Eleizer's various predilections so... to that end... Your various comments (see below) have a common thread of the form I see no reason why you can't ..X.. So let's focus on reasons why you can't ...X. These are numerous and visible in real - empirically verifiable physics...let's look at it from a dynamics point of view. In saying 'you can see no reason' would mean that if you chose a computationalist abstraction level (you mentioned atoms) that you would claim the resultant agent able to demonstrate scientific behaviour indistinguishable from a human. I would claim that to be categorically false and testably so. OK. Firstly call the computationalist artificial scientist, COMP_S. Call the human scientist HUMAN_S. Call computationalism COMP. This saves a lot of writing! The test regime: HUMAN_S constructs laws of nature tn using the human faculty for observation (call it P) delivered by real atoms in the brain of HUMAN_S. If COMP_S and HUMAN_S are to be indistinguishable then the state dynamics (state vector space) of COMP_S must be as sophisticated, accessible as HUMAN_S and ALSO /convergent on the same outcomes as those of HUMAN_S/. Our test is that they both converge on a law of nature tn, say. Note: tn is a abstracted statement of an underlying generalisation in respect of the distal external natural world (such as tn = ta, a model of an atom). Yes? That is what we do... the portability of laws of nature tn proves that we have rendered such abstractions invariant to the belief dynamics of any particular scientist.. Yes? HUMAN_S constructs a model of atoms a 'law of nature' = ta. Using that model ta we then implement a sophisticated computational version of HUMAN_S at the level of the model: atoms. We assemble an atomic-level model replica of HUMAN_S. We run the computation on a host COMP substrate. This becomes our COMP_S. We expect the two to be identical to the extent of delivering indistinguishable scientific behaviour. We embody COMP_S with IO as sophisticated as a human and wire it upIf the computationalist position holds, by definition, the dynamics of COMP_S must be (a) complex enough and (b) have access to sufficient disambiguated information to construct tn indistinguishably from HUMAN_S. If computationalism is true then given the same circumstance of original knowledge paucity (which can be tested), A demand for a scientific outcome should result in state-vector dynamics adaptation resulting in the delivery of the same tn (also testable), which we demand shall be radically novel If they are really equivalent this should happen. This is the basic position (I don't want to write it out again!) I would claim the state trajectory of COMP_S to be fatally impoverished by the model ta. (abstracted atoms). That is, the state-trajectory of COMP_S would fail to consistently converge on a new law of nature tn and would demonstrate instability (chaotic behaviour). Just like ungrounded power supplly voltage drift about, a symbolically ungrounded COMP_S will epistemically drift about. Indeed I would hold that would be the case no matter what the abstraction level: sub-atomic, sub-sub atomic , sub sub sub atomic .. etc ... the result would be identical. Remember: there's no such 'thing' as atoms...these are an abstraction - of a particular level of the organisational hierarchy of nature. also note ... so-called' ab-initio quantum mechanics of the entire HUMAN_S would also fail because QM is likewise just an abstraction of reality, not reality. COMP would claim that the laws of nature describing atoms behave identically to atoms. The model ensemble of ta atoms should be capable of expressing all the emergent properties of an ensemble of real atoms. This already makes COMP a self-referential question-begging outcome. HUMAN_S is our observer, made of real atoms. COMP assumes that P is delivered by computing ta when there is no such 'thing' as atoms! Atoms are an abstraction of a thing, not a thing. Furthermore, all the orighinal atoms of HUMAN_S have been replaced with the atoms of the COMP_S substrate. What is NOT in law of nature ta is the relationship between the abstraction ta and all the other atoms in the distal world outside COMP_S. (beyond the IO boundary). Assume you supplied all the data about all the atoms in the environment of the original human HUMAN_S used to construct and initialise COMP_S. You know all these relationships at the moment you measured all the atoms in HUMAN_S to get you model established. However, after initialisation, when you run the COMP_S, all relationships of the model with the distal world (those intrinsic to
Re: The Super-Intelligence (SI) speaks: An imaginary dialogue
On Sep 2, 6:27 pm, Colin Hales [EMAIL PROTECTED] wrote: Hello again Jesse, I am going to assume that by trashing computationalism that Marc Geddes has enough ammo to vitiate Eleizer's various predilections so... to that end... To make it clear, I'm not trashing computaionalism. I maintain that comp is true (See what Bruno said). It's Bayesianism I'm trashing. And yes, I now have enough 'intellectual ammo' to crush Yudkowsky. Cheers --~--~-~--~~~---~--~~ 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 -~--~~~~--~~--~--~---
Re: The Super-Intelligence (SI) speaks: An imaginary dialogue
Hi Marc, */Eliezer/*'s hubris about a Bayesian approach to intelligence is nothing more than the usual 'metabelief' about a mathematics... or about computation... meant in the sense that cognition is computation, where computation is done BY the universe (with the material of the universe used to manipulate abstract symbols) /search?hl=ensa=Xoi=spellresnum=0ct=resultcd=1q=Eliezer+Yudkowskyspell=1. *You don't have to work so hard to walk away from that approach...* Computationalism is FALSE in the sense that it cannot be used to construct a scientist. A scientist deals with the UNKNOWN. If you could compute a scientist you would already know everything! Science would be impossible. So you *can* 'compute/simulate' a scientist, but if you could the science must already have been done... hence you wouldn't want to. Computationalism is FALSE in the sense of 'not useful', not false in the sense of 'wrong'. You cannot model a modeller of the intrinsically unknown. As a computationalist manipluator of abstract symbols you are required to deliver a model of how to learn - in which you must specify how all novelty shall be handled! In other words you can;t deal with the REAL unknown - where you have no such model! ie. a computationalist scientist is an oxymoron: a logical contradiction. If you say you can then you are question begging computationalism whilst failing to predict an a-priori unsupervised observer (a scientist). The Bayesian 'given' (the conditional) assumes knowledge of a given which is a-priori not available. It assumes observation of the kind we have.. otherwise how would you know any options to choose as givens?. furthermore it assumes that if somehow we were to experiment to resolve a choice of 'givens' (Bayesian conditionals) as being the 'truth' - then there are potentially an enormous collection of 'givens', all of which can be inserted in the same bayesian predictor... resulting in degenerate knowledge you know NOTHING because you fail to resolve anything useful about the world outside. You don't even know there's an 'outside'. The bayesian (all computationalist) approach fails to predict observation (in the sense of ANY observation/an observer, not a particular observation) and fails to predict the science that might result from an observer. This is the achilles heel of the computationalist argument. The computationalist delusion (dressed up in Bayesian or any other abstract symbol-manipulator's clothes) has to stop right here, right now and for good. BTW This does not mean that 'cognition is not computation' I hold that cognition is NATURAL symbol manipulation, not ABSTRACT symbol manipulation. But that's a whole other story... The natural symbols are the key. Please feel free to deliver the above to Eliezer. He'll remember me! Tell him the AGI he is so fearful of are a DOORSTOP and will be pathetically vulnerable to human intervention. The whole AGI fear-mongering realm needs to get over themselves and start being scientific about what they do. It's all based on assumptions which are false. cheers, colin --~--~-~--~~~---~--~~ 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 -~--~~~~--~~--~--~---
Re: The Super-Intelligence (SI) speaks: An imaginary dialogue
Hi! Assumptions assumption assumptionstake a look: You said: Why would you say that? Computer simulations can certainly produce results you didn't already know about, just look at genetic algorithms. OK. here's the rub... /You didn't already know about.../. Just exactly 'who' (the 'you') is 'knowing' in this statement? You automatically put an external observer outside my statement. *My observer is the knower.* *There is no other knower:* The scientist who gets to know is the person I am talking about! There's nobody else around who gets to decide what is known... you put that into my story where there is none. My story is of /unsupervised/ learning. Nobody else gets to choose Bayesian priors/givens. And nobody else is around to pass judgement... the result IS the knowledge. Tricky eh? A genetic algorithm (that is, a specific kind of computationalist manipulation of abstract symbols) cannot be a scientist. Even the 'no free lunch' theorem, proves that without me adding anything but just to seal the lid on itI would defy any computationalist artefact based on abstract symbol manipulation to come up with a law of nature ... ... by law of nature I mean an ABSTRACTION about the distal natural world derived from a set of experiences of the distal natural world (NOT merely IO signals... these are NOT experienced). The IO is degenerately related to the distal natural world by the laws of physics... a computationalist IO system is fundamentally degenerately related to the distal natural world...so it doesn't even know what is 'out there' at all, let alone that there's a generalisation operating BEHIND it. A law of nature, to a genetic algorithm or any other abstract/computationalist beast... would merely predict IO behaviour at its sensory boundary. It may be brilliantly accurate! But that *IS NOT SCIENCE* because there's no verifiable deliverable to pass on...and it has nothing else to work with. An artefact based on this may survive in a habitat... but that is NOT science. Sothere's no scientist here. (BTW IO = input/output). cheers, colin Jesse Mazer wrote: Colin Hales wrote: Computationalism is FALSE in the sense that it cannot be used to construct a scientist. A scientist deals with the UNKNOWN. If you could compute a scientist you would already know everything! Science would be impossible. So you can 'compute/simulate' a scientist, but if you could the science must already have been done... Why would you say that? Computer simulations can certainly produce results you didn't already know about, just look at genetic algorithms. --~--~-~--~~~---~--~~ 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 -~--~~~~--~~--~--~---
Re: The Super-Intelligence (SI) speaks: An imaginary dialogue
On Sep 2, 1:56 pm, Colin Hales [EMAIL PROTECTED] wrote: Hi Marc, */Eliezer/*'s hubris about a Bayesian approach to intelligence is nothing more than the usual 'metabelief' about a mathematics... or about computation... meant in the sense that cognition is computation, where computation is done BY the universe (with the material of the universe used to manipulate abstract symbols) /search?hl=ensa=Xoi=spellresnum=0ct=resultcd=1q=Eliezer+Yudkowskyspell=1. *You don't have to work so hard to walk away from that approach...* Hi Colin, The chess computer 'Deep Blue' was computational, and could play chess better than the (then) chess world champion, Gary Kasparov. But that didn't mean that the programmers understood all the chess, or all the chess had already been played. So I don't think your argument is a good one. You can't rebut Yudkowsy's approach as easily as that ;) But I kind of understand your sentiment, and agree that science can't (and shouldn't be) reduced to mere Bayesian probability shuffling. There are aesthetic judgements involved in science, and I don't think any precise mathematical definition of these aesthetic notions is possible, as Bruno has already opined.Yudkowsky's excessive faith in Bayesian Induction is definitely his weakness. But that doesn't mean we can't make a computational super-intelligence. Cheers --~--~-~--~~~---~--~~ 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 -~--~~~~--~~--~--~---