Re: [agi] Instead of an AGI Textbook
Hmm.. well, but at least, using words related to robotics gives a flavour of embodiment :-). Anyhow, I still prefer sharing terminology with robotics, as opposed to narrow AI. Narrow AI and AGI are perhaps closer, so the risk of confusion is bigger. /R 2008/3/29, Ben Goertzel [EMAIL PROTECTED]: 4. In fact. I would suggest that AGI researchers start to distinguish themselves from narrow AGI by replacing the over ambiguous concepts from AI, one by one. For example: knowledge representation = world model. learning = world model creation reasoning = world model simulation goal = life goal (to indicate that we have the ambition of building something really alive) If we say something like world model creation, it seems pretty obvious that we do not mean anything like just tweaking a few bits in some function. Yet, those terms are used for quite shallow things in many Good Old Fashioned robotics architectures ;-) ben --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=98558129-0bdb63 Powered by Listbox: http://www.listbox.com
Re: [agi] Instead of an AGI Textbook
A few things come to my mind: 1. To what extent is learning and reasoning a sub topic of cognitive architectures? Is learning and reasoning a plugin to a cognitive architecture, or is in fact the whole cognitive architecture about learning and reasoning. 2. I would like a special topic on AGI goal representation. More specifically, a topic that discusses how a goal specified by any human designer, can be related to the world model and actions that an AGI system creates? For example, how can the human specified goal, be related to a knowledge representation that is constantly developed by the AGI system? 3. Why do AI/AGI researchers always talk about *knowledge representation.*It gives such a strong bias towards static or useless knowledge bases. Why not talk more about *World modelling*. Because of the more active meaning of the word modelling as opposed to representation, it implies that things such as inference etc. need to be considered. Since the word modelling is also used to denote the process of creating a model, it also implies that we need mechanisms for learning. I really think we should consider if not knowledge representation is a concept straightly borrowed from dumb-narrow AI, or if it really is a key concept for AGI. Sure enough, there will always be knowledge representation, but the question is whether it is an important/relevant/sufficient/misleading concept for AGI. 4. In fact. I would suggest that AGI researchers start to distinguish themselves from narrow AGI by replacing the over ambiguous concepts from AI, one by one. For example: knowledge representation = world model. learning = world model creation reasoning = world model simulation goal = life goal (to indicate that we have the ambition of building something really alive) If we say something like world model creation, it seems pretty obvious that we do not mean anything like just tweaking a few bits in some function. 2. I am thinking of whether it would be a good idea with a topic like methods for quelling combinatorial explosions in AGI world model processes. That topic could outline basic principles like meta-adaptation and parallelisation of adaptation (meaning that the AGI system needs to separate objects in reality that can be studied separatley). Like someone mentioned, such principles might be overly simple to many already in the field, and thereby not worth mentioning, but if we aim at writing documents for beginners, we really need to get the basics right. Simple/basic principles are still interesting, as long as they are not narrow. Maybe Ben Goertzel could add some more difficoult material under such a topic also. Hope any of these ideas could could be helpful. Thanks. /R 2008/3/26, Ben Goertzel [EMAIL PROTECTED]: BTW I improved the hierarchical organization of the TOC a bit, to remove the impression that it's just a random grab-bag of topics... http://www.agiri.org/wiki/Instead_of_an_AGI_Textbook ben --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=98558129-0bdb63 Powered by Listbox: http://www.listbox.com
Re: [agi] reasoning knowledge
d) you keep repeating the illusion that evolution did NOT achieve the airplane and other machines - oh yes, it did - your central illusion here is that machines are independent species. They're not. They are EXTENSIONS of human beings, and don't work without human beings attached. Manifestly evolution has taken several stages to perfect tool/machine-using species - of whom we are only the latest version - I refer you to my good colleague, the tool-using-and-creating Caledonian crow. Yes, somehow, we are going to create the first independent machine species - but there's a big unanswered set of questions as to how . It can be said that the emergence of human intelligence and human cultures set of another kind of technological evolution on top of the biological one. That these two forms of evolution can be seen as separate, can be explained as follows: Biological evolution works through DNA sequences, genes. The survivability of genes, depend on whether they are a part of successful biological lifeforms. Technological evolution works through sets of ideas, or memes that grow in our culture and in the minds of human beings. The survivability of memes depend on whether they are appealing to human minds. Whether a meme is appealing or not, could depend on a number of factors, such as whether the meme could help humans to achieve some of their goals, whether they are self-contradicting, or whether we can understand them etc. Memes can even survive outside the brain of humans, stored in books etc. The reason why technological innovations works with such great strides, is first because memes are produced at an incredible rate compared to genes; they are software based instead of hardware based. But more importantly, because memes can be selected based on logical deduction and the consideration of a predicted future. Thus, the survivability of memes depend on how well we believe them to help us in the future. I think it would be more accurate to say that technological meme evolution *was caused by *the biological evolution, rather than being *the extension of it*, since they are in fact two quite different evolutionary systems, with different kinds of populations/survival conditions. I would say that in some sense, there is already a machine species, even if not independent. This machine species just have not yet found a way of staying alive and breed outside human minds. Is this a helpful perspective? :-)... One key issue here, is whether we want to consider hardware and software evolutionary systems, or just hardware based evolutionary systems. Also, I admit that maybe I am not using the concept of species in any stringent way. --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=95818715-a78a9b Powered by Listbox: http://www.listbox.com
[agi] AGI Metaphysics and spatial biases.
out from a particular object. Since these metaphysics are based on 3D space, they can be easily modified to be suitable for 2D space. Maybe it could be possible to build AGI prototypes using 2D space biases to lessen the demand for hardware, and then when we have gained more experience, it could be possible to shift to full 3D space metaphysics based on the experience from 2D AGI. So, my questions now are: Has anyone else had similar ideas about what biases/metaphysics should be encoded into an AGI system. What could be good with them, bad with them? Does anyone agree with the fact that object isolation could be an important principle for achieving AGI learning? Also, some specific questions for Ben Goertzel: I understand Novamente is based on patternist metaphysics. In what ways is patternist metaphysics different/similar to the metaphysics sketched at here? As I understand it, the patternist metaphysics is based on events. Would it be possible/easy to model data-flow dependencies between objects using the Novamente metaphysics? Also, I remember once seeing a Novamente demonstration where the AGI system was learning the concept of object persistancy. The fact that objects hidden, remain in hiding until next time it is shown again (I hope this give a correct description of what was shown). But in that case I guess that there must have been some initial concepts encoded already into the AGI system, for example the concepts of how dense objects can move through space objects. Using informal words, how would you describe the metaphysics or biases currently encoded into the Novamente system? /Robert Wensman --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=95818715-a78a9b Powered by Listbox: http://www.listbox.com
[agi] Primates using tools.
This could perhaps be relevant to understanding human level intelligence. One interpretation here is that the brain of primates considers tools as part of their body, which makes them good at using them: http://sciencenow.sciencemag.org/cgi/content/full/2008/128/2 This of course, still leaves the question of how a generally intelligent system uses its body in the first place, and what special hardware there is to deal with this problem. :-). Personally I believe that a general intelligence, such as the human mind, still have some specialized processors to deal with very common situations. Another thing that I guess could use some special hardware, is the ability to feel empathy and understand other human beings or animals. To understand other intelligent beings is so important for humans, yet if done in a general way it seems so incredibly expensive and difficult. Also, a human is in many ways very similar to the intelligent beings it tries to simulate, so it is my firm belief that a human uses parts of its own cognitive process to simulate other intelligent beings. I think that a social AGI system needs to be able to instantiate its own cognitive process in a kind of role-play. Assume that I know this, that I want this, and that I am in this kind of situation, what would I do. And then use this role playing to assess others actions. The fact that empathy seems to be more strongly connected to biological heritage, rather than by social influence could indicate that the ability to feel empathy needs special hardware in our brain. I think I heard of a study that showed a very strong correlation between the empathic ability of identical twins, which should indicate that their social upbringing has less influence on this particular ability. However, I don´t remember the source of that that information. /Robert Wensman - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=91461624-5f7744
Re: [agi] Study hints that fruit flies have free will
1. Brembs and his colleagues reasoned that if fruit flies (Drosophila melanogaster) *were simply reactive robots entirely determined by their environment*, in completely featureless rooms they should move completely randomly. Yes, but no one has ever argued that a flier is a stateless machine. It seems like their argument ignores the concept of internal state. If they went through all this trouble just to prove that the brain of the flies has an internal state, it seems they wasted a lot of time on something trivial. I cannot see how the concept of free will has got anything to do with this. /R - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=89402809-0e047a
Re: [agi] Study hints that fruit flies have free will
I don't think anyone with knowledge of insect nervous systems would argue that they're stateless machines. Even simple invertebrates such as slugs can exhibit classical condition effects which means that at least some minimal state is retained. To me the idea of free will suggests that a number of possible behaviors can be triggered at any moment in time and that the system in some way chooses between those possibilities. The system can only move easily from one state to another if its dynamics are perched on an edge between pure randomness and determinism. If any one behavior is too strong an attractor then the system overall may become dysfunctional. I don't understand what you mean by randomness. To people who believe in determinism, there is no true randomness. What you might mean randomness is smooth distribution or lack of complexity. I am curious whether the same scientific method also would conclude that the following fractal tree, also has a consciousness? http://commons.wikimedia.org/wiki/Image:Fractal_tree_%28Plate_b_-_3%29.jpghttp://upload.wikimedia.org/wikipedia/commons/4/41/Fractal_tree_(Plate_b_-_2).jpg Do they mean that a system complex enough obtains free will? /R - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=89412989-a5f2a7
Re: [agi] A Simple Mathematical Test of Cog Sci.
Mike Tinter, If you really do not think that digital computers can be creative by definition, I do not understand why you would like to join a mailing list with AGI researchers? Computers operate by using software, thus, they need to be programmed. It just seems to me that you do not understand what the word program means. Even if you use use a computer that do not need to be loaded with a program, guess what, such a computer could be considered to have an initial program. The very determinism of the universe implicates that everything runs according to a program, including your ramblings here about creativity. I have to ask you a question, do you think the universe and everything in it runs according to deterministic laws of nature? Do you accept that you are a part of this deterministic reality? Well, in that case Ive got news for you, you are a program also! As evidence I would present your DNA, a program encoded and stored in molecular structures. Have you ever heard of computational equivalence? Do you know what it means? Also, I feel annoyed that you compare the Novamente architecture with something that just takes instructions, like do this, do that, then do this etc. It seems you need to spend greater effort in studying this architecture, for example by reading The Hidden Pattern. I feel you are in great need of widening your mind to understand chaotic or fractal processes. Take a forest for example, even in all its complexity and diversity, it is still governed by very simple and basic laws namely the laws of nature. By mimicking some of these laws at an appropriate level, such as shape level, programmers can create forests that to a very large extent looks like real forests: http://www.speedtree.com/. A generator such as speedtree could generate entire forests of miles and miles of trees, with no single two trees looking the same. Even though the lines of code producing the trees are pretty simple, the outcome in creativity and originality is vast. The same thing applies to a human mind. Even though the output of a human mind is amazingly diverse and creative, its program is still goverened by the basic laws of nature, and the DNA program. What AGI designers tries to do is to is to mimic this process. The concepts of program and determinism are pretty well established within the scientific community, please do not try to redefine them like you do. It just creates a lot of confusion. I think what you really want to use is the concept of adaptability, or maybe you could say you want an AGI system that is *programmed in an indirect way* (meaning that the program instructions are very far away from what the system actually does). But please do not say things like we should write AGI systems that are not programmed. It hurts my ears/eyes. /Robert Wensman 2008/1/7, Mike Tintner [EMAIL PROTECTED]: Well we (Penrose co) are all headed in roughly the same direction, but we're taking different routes. If you really want the discussion to continue, I think you have to put out something of your own approach here to spontaneous creativity (your terms) as requested. Yes, I still see the mind as following instructions a la briefing, but only odd ones, not a whole rigid set of them., a la programs. And the instructions are open-ended and non-deterministically open to interpretation, just as my briefing/instruction to you - Ben go and get me something nice for supper - is. Oh, and the instructions that drive us, i.e. emotions, are always conflicting, e.g [Ben:] I might like to.. but do I really want to get that bastard anything for supper? Or have the time to, when I am on the very verge of creating my stupendous AGI? Listen, I can go on and on - the big initial deal is the claim that the mind isn't - no successful AGI can be - driven by a program, or thoroughgoing SERIES/SET of instructions - if it is to solve even minimal general adaptive, let alone hard creative problems. No structured approach will work for an ill-structured problem. You must give some indication of how you think a program CAN be generally adaptive/ creative - or, I would argue, squares (programs are so square, man) can be circled :). Mike, The short answer is that I don't believe that computer *programs* can be creative in the hard sense, because they presuppose a line of enquiry, a predetermined approach to a problem - ... But I see no reason why computers couldn't be briefed rather than programmed, and freely associate across domains rather than working along predetermined lines. But the computer that is being briefed is still running some software program, hence is still programmed -- and its responses are still determined by that program (in conjunction w/ the environment, which however it perceives only thru a digital bit stream) I don't however believe that purely *digital* computers are capable of all the literally imaginative powers (as already
Re: [agi] A Simple Mathematical Test of Cog Sci.
Mike, Let me clarify further. What me and other computer scientists mean by program, is probably something like *A formal and non-ambigous description of a deterministic system that operates over time*. Thus, if you can describe something in nature with enough detail, your description is a program. As another example, if you write a book that describes the human mind formally in enough detail, that book in itself would become a program. So when you say that we cannot write a program that is creative on the same level as humans, you basically state that it would be impossible to describe the human mind in a detailed enough way. This is certainly bogus, as this could be done theoretically by simply scanning and recording the state and connections of every neuron in a human mind. Another way to put it, is that your suggestion implies that we could never *understand *the human mind on a fine enough level, which is pretty upsetting and certainly not revolutionary. What computers have or have not done up until this point is completely besides the question, if we discuss the definition of program. Yes, enough powerful AGI would be revolutionary, but they would still be programs. What you is suggesting is equivalent to asking a painter to paint a revolutionary painting, without using paint. What should he do, stare intensely at the canvas until what happens? He could try to cheat, using dirt, or mud to paint. But most people would then just say he invented another kind of paint, namely the dirt paint, or the mud paint. It is just impossible to paint a painting without paint (unless your painting is intended to look the same as the empty canvas). Painting paintings without paint is not a radical idea, it is just plain futile or incorrect, depending on perspective. Why this topic is frustrating, is because you are roughly right in one aspect. Yes, computers and AI systems up until this point has been programmed in a much too direct way, where the connection between programmers lines of code, and the systems actions has been too close. E.g. there is a line of code saying *if(handIsHot()) moveHand(),* and where the robot system moves its hand when it becomes hot. But this is what we here call narrow AI and what we all here try to distance ourselves from. From what I can tell, Novamente is for example miles and miles and miles away from this kind of programming. In contrast, systems like Novamente studies input and builds and relates concepts to abstract goals and later form actions using different kinds of subtle methods. But a system like that is * complex* and you cannot expect Ben Goertzel to blurt out all this complexity in an email on this mailing list. You have to study the design in detail if you are interested in it. But the bottom line is, it is still programming in any way you choose to look at it (unless you want to use the word programming in some way that no other person on earth is using it, but in that case, be prepared to feel alone). You should focus on HOW we could make programs creative, rather loosing yourself in a strange quest to redefine well established terminology. It is completley besides the point. /Robert Wensman 2008/1/7, Mike Tintner [EMAIL PROTECTED]: Robert, Look, the basic reality is that computers have NOT yet been creative in any significant way, and have NOT yet achieved AGI - general intelligence, - or indeed any significant rulebreaking adaptivity; (If you disagree, please provide examples. Ben keeps claiming/implying he's solved them or made significant advances, but when pressed never provides any indication of how). These are completely unsolved problems. Major creative problems. And I would suggest you have to be prepared for the solutions to be revolutionary and groundshaking. If you are truly serious about solving these problems, I suggest, you should prepared to be hurt - you should be ready to consider truly radical ideas - for the ground on which you stand to be questioned - and be seriously shaken up. You should WELCOME any and all of your assumptions being questioned. Even if, let's say, what I or someone else suggests is in the end nutty, drastic ideas are good for you to contemplate at least for a while. Having said all this, I accept that what I have been saying offends this community - I wasn't trying originally to push it, I got dragged into some of that last discussion.by Ben. And I also accept that most of you are not interested in going for the revolutionary, from whatever source. And I shall try to restrict my comments unless someone wishes to engage with me - although BTW I am ever more confident of my broad philosophical/ psychological position - the mind really doesn't work that way. I may possibly make one last related post in the not too distant future about the nature of problems, and which are/aren't suitable for programs - but just ignore it. Mike Tinter, If you really do not think that digital
Re: [agi] A Simple Mathematical Test of Cog Sci.
Mike, To put my question in another way. Would you like to understand intelligence? Understand it to such a degree, that you can give a detailed and non-ambiguous description of how an intelligent system operates over time? Well, if you do want that, then you want -using standard terminology- to create an intelligent program. Why we get upset is because we feel you basically say I don't want to understand intelligence alternatively intelligence can never be clearly understood. You have to understand how computer scientists use the word program to understand how we perceive your statements. From our perspective, your position is not revolutionary, just depressing. /Robert Wensman 2008/1/7, Mike Tintner [EMAIL PROTECTED]: Robert, Look, the basic reality is that computers have NOT yet been creative in any significant way, and have NOT yet achieved AGI - general intelligence, - or indeed any significant rulebreaking adaptivity; (If you disagree, please provide examples. Ben keeps claiming/implying he's solved them or made significant advances, but when pressed never provides any indication of how). These are completely unsolved problems. Major creative problems. And I would suggest you have to be prepared for the solutions to be revolutionary and groundshaking. If you are truly serious about solving these problems, I suggest, you should prepared to be hurt - you should be ready to consider truly radical ideas - for the ground on which you stand to be questioned - and be seriously shaken up. You should WELCOME any and all of your assumptions being questioned. Even if, let's say, what I or someone else suggests is in the end nutty, drastic ideas are good for you to contemplate at least for a while. Having said all this, I accept that what I have been saying offends this community - I wasn't trying originally to push it, I got dragged into some of that last discussion.by Ben. And I also accept that most of you are not interested in going for the revolutionary, from whatever source. And I shall try to restrict my comments unless someone wishes to engage with me - although BTW I am ever more confident of my broad philosophical/ psychological position - the mind really doesn't work that way. I may possibly make one last related post in the not too distant future about the nature of problems, and which are/aren't suitable for programs - but just ignore it. Mike Tinter, If you really do not think that digital computers can be creative by definition, I do not understand why you would like to join a mailing list with AGI researchers? Computers operate by using software, thus, they need to be programmed. It just seems to me that you do not understand what the word program means. Even if you use use a computer that do not need to be loaded with a program, guess what, such a computer could be considered to have an initial program. The very determinism of the universe implicates that everything runs according to a program, including your ramblings here about creativity. I have to ask you a question, do you think the universe and everything in it runs according to deterministic laws of nature? Do you accept that you are a part of this deterministic reality? Well, in that case Ive got news for you, you are a program also! As evidence I would present your DNA, a program encoded and stored in molecular structures. Have you ever heard of computational equivalence? Do you know what it means? Also, I feel annoyed that you compare the Novamente architecture with something that just takes instructions, like do this, do that, then do this etc. It seems you need to spend greater effort in studying this architecture, for example by reading The Hidden Pattern. I feel you are in great need of widening your mind to understand chaotic or fractal processes. Take a forest for example, even in all its complexity and diversity, it is still governed by very simple and basic laws namely the laws of nature. By mimicking some of these laws at an appropriate level, such as shape level, programmers can create forests that to a very large extent looks like real forests: http://www.speedtree.com/. A generator such as speedtree could generate entire forests of miles and miles of trees, with no single two trees looking the same. Even though the lines of code producing the trees are pretty simple, the outcome in creativity and originality is vast. The same thing applies to a human mind. Even though the output of a human mind is amazingly diverse and creative, its program is still goverened by the basic laws of nature, and the DNA program. What AGI designers tries to do is to is to mimic this process. The concepts of program and determinism are pretty well established within the scientific community, please do not try to redefine them like you do. It just creates a lot of confusion. I think what you really want to use is the concept of adaptability, or maybe
Re: [agi] A Simple Mathematical Test of Cog Sci.
2008/1/7, David Butler [EMAIL PROTECTED]: How would an AGI choose which things to learn first if given enough data so that it would have to make a choice? This is a simple question that demands a complex answer. It is like asking How can a commercial airliner fly across the Atlantic?. Well, in that case you would have to study aerodynamics, mechanics, physics, thermodynamics, computer science, electronics, metallurgy and chemistry for several years, and in the end you would discover that one single person cannot understand such a complex machine in its entire detail. True enough, one person could understand all basic principles for such a system, but explaining them would hardly suffice as evidence that it would actually work in practice. If you lived in the medieval times, and someone asked you how is it possible to cross the Atlantic in a flying machine carrying several hundred passengers?, what would you answer? Even if you had the expertise knowledge it would be very hard to explain thoroughly, just because the machine is so complex and you would have to explain every technology from the beginning. Where would you start? Maybe some person with less insight would interrupt you after a few sentences and say well, clearly you cannot present evidence that it will ever work and make fun of the idea, but how does insufficient time/space to explain a complex system prove that something is not possible? The same goes for AGI, for example when someone asks how can we create a program that is creative and can choose what to learn?. In response to this it is possible to present a lot of different principles, such as adaptability, genetic programming, quelling of combinatorial explosions etc. But will the principles work in practice when put together? Well, at this stage we simply cannot tell. *So every person just has to make a choice in whether to believe it is possible, or whether to believe it is not possible. *But just because no AGI researcher can answer that question in a few words. how can we create a programs that is creative and can choose what to learn, it doesn't mean it is not possible when all these principles come together. We just have to wait and see. To those who do not believe: Please just go away from this mailing list and do not interfere with the work here. Don't demand proof that it would work, because when we have such proof, i.e. a finished AGI system, we wont need to defend our hypothesises anyway. If two AGI's (again-same hardware, learning programs and controlled environment) were given the same data would they make different choices? Is a deterministic system deterministic? I do not understand what you are getting at. Why this question? I think Benjamin answered this question pretty thoroughly already. /Robert Wensman - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=82686766-4e2400
Re: [agi] An AGI Test/Prize
Regarding testing grounds for AGI. Personally I feel that ordinary computer games could provide an excellent proving ground for the early stages of AGI, or maybe even better if they are especially constructed. Computer games are usually especially designed to encourage the player towards creativity and exploration. Take a simple platform game for example, at every new stage new graphics and monsters are introduced, and in large, the player undergoes a continuous self training that last throughout the whole game. Game developers carefully distribute rewards and challenges to make this learning process as smooth as possible. But also I would like to say that given any proving ground for the first stages of AGI could be misused if AGI designers bring specialized code into their system. So if there is to be a competition for first generation AGI, there would have to be some referee that evaluates how much domain specific knowledge has been encoded to any given system. For the late development stages of AGI, where we basically have virtual human minds, then we could use so hard problems that specialized code could not help the AGI system anymore. But I guess that at that time we have basically already solved the problem of AGI, and competitions where AGI systems compete in writing essays on some subject, could only be used to polish some already outlined solution to AGI. I am a fan of Novamente, but for example when I watched the movie where they trained an AGI dog, I was left with the question about what parts of its cognition was specialization. For example, the human teacher used natural language to talk to the dog. Did the dog understand any of it, and in that case, was there any special language module involved? Also, training a dog is quite open ended, and it is difficult to assess what is progress. This shows just how difficult it is to demonstrate AGI. Any demonstration of AGI would have to support a list of what cognitive aspects are coded, and which are learnt. Only then you can understand whether it is impressive or not. Also, because we need to have firm rules about what can be pre-programmed, and what needs to be learnt, it is easier if we used some world with pretty simple mechanics. What I basically would like to see is an AGI learning to play a certain computer game, starting by learning the fundamentals, and then playing it to the end. Take an old videogame classic like The Legend of Zelda. http://www.zelda.com/universe/game/zelda/. I know a lot of you would say that this is a far to simplistic world for training an AGI, but not if you prohibit ANY pre-programmed knowledge. You only allow the AGI system to start with proto-knowledge representation, and basically hard-wire the in-game rewards and punishemnts to the goal of the AGI. The AGI system would then have to learn basic concepts such as: objects moving around on the screen which graphics correspond to yourself walls where you can go keys that opens doors the concept of coming to a new screen when walking of the edge of one how screens relate to each other teleportation (the flute for anyone who remembers) If the AGI system then can learn to play the game to the end and slay Ganon based on only proto-knowledge, then maybe we have some interesting going on. Such an AGI could maybe be compared to a rodent running in a maze, even if the motoric and vision system are more complicated. Then we are ready to increase the complexity of the computer game, adding communication with other characters, more complex concepts and puzzles, more dimensions, more motorics etc.. Basically, I would like to se Novamente and similar AGI systems play some goal oriented computer game, since AGI in itself needs to be goal oriented. /R 2007/10/20, Benjamin Goertzel [EMAIL PROTECTED]: I largely agree. It's worth pointing out that Carnot published Reflections on the Motive Power of Fire and established the science of thermodynamics more than a century after the first working steam engines were built. That said, I opine that an intuitive grasp of some of the important elements in what will ultimately become the science of intelligence is likely to be very useful to those inventing AGI. Yeah, most certainly However, an intuitive grasp -- and even a well-fleshed-out qualitative theory supplemented by heuristic back-of-the-envelope calculations and prototype results -- is very different from a defensible, rigorous theory that can stand up to the assaults of intelligent detractors I didn't start seriously trying to design implement AGI until I felt I had a solid intuitive grasp of all related issues. But I did make a conscious choice to devote more effort to utilizing my intuitive grasp to try to design and create AGI, rather than to creating better general AI theories Both are worthy pursuits, and both are difficult. I actually enjoy theory better. But my sense is that the heyday of AGI
Re: Economic libertarianism [was Re: The first-to-market effect [WAS Re: [agi] Religion-free technical content]
scientific facts that we can verify are correct. These scientific facts can then be used for our production but never connect the AGI system to the production facilities directly. If the production facility needs intelligence, we choose a separate more dumb AGI system that is just suited to its task of running the factory. There are a number of safety measures like this that could greatly improve the safety of AGI usage. I believe we could make it quite difficult for an AGI system to obtain power by using the age old idea of divide and conquer. Also, the history shows that intelligence is no guarantee for power. The Russian revolution and the genocide in Cambodia illustrates effectively how intelligent people were slaughtered by apparently less intelligent people, and later how they were controlled to the extreme for decades. Most communist dictatorships end because instability caused by poverty, not because the control structure itself failed. This just reveals something raw and basic about existence on earth that I think many AGI enthusiasts and futurists wants to deny: What good are wits when you are looking down the barrel of a gun? /Robert Wensman - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=51821901-7abc6f
Re: Economic libertarianism [was Re: The first-to-market effect [WAS Re: [agi] Religion-free technical content]
(off topic, but there are something relevant for AGI) My fears about economical libertarianism could be illustrated with a fish pond analogy. If there is a small pond with a large number of small fish of some predatory species, after an amount of time they will cannibalize and eat each other until at the end there will just remain one very very fat fish. The instability occurs because a fish that already has managed to eat a peer, becomes slightly larger than the rest of the fish, and therefore has a better position in continuing to eat more fish, thus its progress can accelerate. Maybe if the pond is big enough, a handful of very big fish would remain. This is of course just an illustration and by no means a proof that the same thing would occur in a laissez-faire/libertarianism economy. Libertarians commonly put blame for monopolies on government involvement, and I guess some would object that I unfairly compares fish that eat each other with a non-violent economy. But lets just say I do not share their relaxed attitude towards the potential threat of monopoly, and a bigger fish eating a smaller fish do have some similarity to a bigger company acquiring a smaller one. First of all, the consequence of monopoly is so serious that even if the chance is very slight, there is a strong incentive to try to prevent it from ever happening. But there are also a lot of details to suggest that a laissez-faire economy would collapse into monopoly/oligopoly. Effects of synergy and mass production benefits would be one strong reason why a completely free market would benefit those companies that are already large, which could make them grow larger. *Especially when considering AGI and intelligence enhancement I believe a libertarian market could be even more unstable. In such a setting, the rich could literally invest in more intelligence, that would make them even more rich, creating a positive economic feedback loop. A dangerous accelerating scenario where the intelligence explosion could co-occur with the rise of world monopoly. We could call it an AGI induced monopoly explosion. Unless democracy could challenge such a libertarian market, only a few oligarchs might have the position to decide the fate of mankind, if they could control their AGI that is. Although it is just one possible scenario.* A documentary I saw claimed that Russia was converted to something very close to a laissez-faire market in the years after the Soviet Union collapse. However I don't have any specific details about it, such as exactly how free the market of that period was. But apparently it caused chaos and gave rise to a brutal economy with oligarchs controlling the society. [ http://en.wikipedia.org/wiki/The_Trap_(television_documentary_series)]. Studying what happened in Russia after the fall of communism could give some insight on the topic. /R 2007/10/8, Bob Mottram [EMAIL PROTECTED]: Economic libertarianism would be nice if it were to occur. However, in practice companies and governments put in place all sorts of anti-competitive structures to lock people into certain modes of economic activity. I think economic activity in general is heavily influenced by cognitive biases of various kinds. On 06/10/2007, BillK [EMAIL PROTECTED] wrote: On 10/6/07, a wrote: A free market is just a nice intellectual theory that is of no use in the real world. - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=51384923-1d1de1
[agi] A problem with computer science?
, and dares to let in a little bit of the psychological vagueness in their paper writing jargon. By that I do not mean to encourage any kind of Freud-like incoherent crackpot theories, but just the kind of vagueness that is associated with any kind of complex engineering, like this system seems to be better than that system, or it seems this design could benefit a certain capability etc. Maybe an increased focus on AGI would encourage such a development. /Robert Wensman These are not clearly separable things. One of the reasons many people do the system synthesis and balanced approximations so badly is because they tend to use minor variations of the same function representations they would use when playing with those functions in isolation. The assumption that a particular set of functions are only expressible as a particular narrow form can frequently make it impossible to synthesize a useful system because the selected form imposes limits and tradeoffs specific to its form in practice that are not required to achieve equivalent function. A lot of computer science tends to be like this in practice (e.g. the ever ubiquitous balanced tree). Cheers, J. Andrew Rogers - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=47989577-57c6b2
Re: [agi] a2i2 news update
ownership on all the land that we need to live on, on all the food we need to eat, and on all the air we need to breathe. Then it could just kill us in self-defence because we trespass on its property. I know even Ayn Rand sees no moral problem in using defensive violence to defend material property that is being stolen. Well, let me just say that I would be concerned if someone creates a selfish super intelligent AGI system that does not value the well being of me and the rest of us humans, except for when it can see benefits for its own survival. Out of fear for my own life, and the life of my descendants, I would not support your AGI initiative! Even a sentimental and altruistic person like me has that much sense of self-defence! :-)* * That said, I think Adaptive AI's definition of general intelligence seems pretty reasonable, and their plans for development seems well thought out. I also found some thoughts on evolution and AGI noteworthy. But my feelings are mixed about their strength in numbers and the hopes for progress it gives. To me altruistic AGI just seems a lot safer than selfish AGI! /Robert Wensman - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415id_secret=25439434-9f2310
Re: [agi] Another attempt to define General Intelligence, and some AGI design thoughts.
Yes, what language to use when expressing memes is definitley one of the key point in the construction of such a system. I think such a language needs to fulfull the following criteria: - Enough expressive power - Algorithms for consistency checks, entailment etc. - Roubustness to random modifications Just considering the need of expressive power, It would be tempting to consider first or second order predicate logic. Sure, contemporary research in logic reasoning systems tells us that even for quite limited logics, there are pretty severe algorithmic limitations. I hope however that this just reflects that current science is aimed too much at mathematical completeness and soundness, which might not really be needed in a true AGI system. Some kind of statistical or incremental deduction systems could perhaps overcome this problem in the future, and we could settle for concepts like probably consistent, probably sound and probably complete. (note: I continue to use the term meme for the time beeing, but depending on the other discussion I might change to information unit.) Actually, I think a meme consistency check algorithm could be based on meme evolution itself. If we are to check that a certain meme is consistent, the system then sets up an evolution that tries to create counterexamples that would easily prove the meme to be inconsistent. While no counterexample is found, the systems belief in the meme is strenthened. Has there been any work done previously in statistical, example driven deduction? (Note that many of my ideas of how an AGI should generate theories, is analogously to how the scientific community generate theories according to contemporary science philosophy. This is because I believe science to be a macro-projection of the intellectual process in one human mind) Since memes are created at random, the language they are expressed in also needs an inherent roubustness. This could be the reason why it for example would not be suitable to express memes as C++ classes or programs. Perhaps 99% of all randomly created C++ programs will crash the system, and it is therefore not very practical. Logical formulas are appealing because of this, because the worst thing that can happen is that a set of logic formulas become inconsistent (which might be troublesome enough). As of the predicates used in such a database, most of them would be created randomly, and their identifier would just be a number. Some predicates related to facts, such as sensor information and actuator settings would however be hard coded into the system. For memeplexes to be of any use, they need to be grounded to these fact predicates. However, I believe some memes could exist without any direct connection to the hard coded predicates. They could be favoured parts that the meta evolution uses when creating fact grounded memeplexes. I also think that a logic suitable for this purpose could have special features that are just meant to support random mutations of memes. Some years ago I was interested in a class of second order predicate logic with formulas that could be factored. For example (infix notation used): factor(a, (b) - (c)) = (a b) - (a c) I am not claiming that particular feature is of any interest, nor that it is not. My point is that if we want to have a language for memes, we should also consider what refactoring functions that could be defined upon such a language. Probably roubustness is one of the most important parts in this also. PS: I sent my first mail to the mailing list, but did not receive it myself (not even with my second mail adress that I use). Was the mail sent to the entire list? Regards. /Robert Wensman 2007/6/14, Derek Zahn [EMAIL PROTECTED]: Robert Wensman writes: Databases: 1. Facts: Contains sensory data records, and actuator records. 2. Theory: Contains memeplexes that tries to model the world. I don't usually think of 'memes' as having a primary purpose of modeling the world... it seems to me like the key to your whole approach is how you represent them (the schema of database 2). Could you elaborate a bit on that? -- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=e9e40a7e
Re: [agi] Another attempt to define General Intelligence, and some AGI design thoughts.
For an intelligence to know what is possible actions it must model those and think about those internally, and model that beahavior, and I would argue that all intelligence is about the behavior that the internal cognition brings. You cant really have an intelligence I dont believe without behaviour, can you have behaviour without some form of intelligence? The simple act of talking and responding to a question is action and behavior, and without it, you cant realy determine if something is intelligent or not. Memes - It looks like the meme people want to somehow shoehorn knowledge into being a meme or memeplex A *meme* (IPA http://en.wikipedia.org/wiki/IPA_chart_for_English: /miːm/) is a unit of cultural http://en.wikipedia.org/wiki/Culture informationhttp://en.wikipedia.org/wiki/Informationthat propagates from one mind to another as a theoretical unit of cultural evolution http://en.wikipedia.org/wiki/Sociocultural_evolution and diffusion http://en.wikipedia.org/wiki/Cultural_diffusion, analogoushttp://en.wikipedia.org/wiki/Analogyto the way a gene http://en.wikipedia.org/wiki/Gene propagates from one organism to another as a unit of genetic http://en.wikipedia.org/wiki/Geneticsinformation and evolution. First it says memes are a cultural unit but then go on to encompass just about all types of information / knowledge. A generic information unit sounds better, and doesnt have to have the restrictions and/or extra effects that a meme seems to have. Otherwise welcome to the AGI group, and maybe we can expound on some of the other thoughts you have. James Thanks. Well, I agree that information unit would work well also. However, to some degree I feel that the extra effects of meme are beneficial, even though you cannot interpret those extra effects litterally. Using meme could emphasize that they are created at random in an evolutionary process, or at least that there is some distributed process that creates them, which is exactly what I want. Information unit on the other hand is more general, and could include hand-crafted code made by some programmer, or data inserted in an expert system. Also I like meme because it is one word, as opposed to two in information unit. I believe it is better to use short words for the most common things we want to express. Wether it sounds good or not I dont know. Beeing swede it sounds pretty ok to me, but maybe it sounds different to people with other background. I believe something could act intelligently without actually beeing intelligent. Say for example that we program a robot to perform a lot of hard coded random actions, even without using any sensory data. Even if incredibly (actually there arent words strong enough) unlikely, there is a chance that such a robot might go to work and act as a seemingly intelligent employee of some company during an entire day. However we would know it is not intelligent. Also, I would say that a lot of other system classes are not defined based on their actual actions. For example, we could determine wether an object is an airplane even without actually seeing it fly. What we do is to study its structure, and judge wether it according to our understanding has a capability to fly in the way airplanes do. I agree however that this is troublesome for general intelligence, because turing completeness causes a lot of different definitions to be computationally equivalent, thus there is no standardized language in which to describe what is general intelligence. For airplanes it is simple, because 3D pictures and 2D drawings are quite straightforward. Maybe this will be easier once we have an example of working AGI. There is a point though, in that passive AGI systems that just think think think, but doesnt do anything useful would be of little use :-). Regards /Robert Wensman ___ James Ratcliff - http://falazar.com Looking for something... -- Park yourself in front of a world of choices in alternative vehicles. Visit the Yahoo! Auto Green Center.http://us.rd.yahoo.com/evt=48246/*http://autos.yahoo.com/green_center/;_ylc=X3oDMTE5cDF2bXZzBF9TAzk3MTA3MDc2BHNlYwNtYWlsdGFncwRzbGsDZ3JlZW4tY2VudGVy -- This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=231415user_secret=e9e40a7e