RE: [agi] Consciousness vs. Intelligence
From: Dr. Matthias Heger [mailto:[EMAIL PROTECTED] The problem of consciousness is not only a hard problem because of unknown mechanisms in the brain but it is a problem of finding the DEFINITION of necessary conditions for consciousness. I think, consciousness without intelligence is not possible. Intelligence without consciousness is possible. But I am not sure whether GENERAL intelligence without consciousness is possible. In every case, consciousness is even more a white-box problem than intelligence. For general intelligence some components and sub-components of consciousness need to be there and some don't. And some could be replaced with a human operator as in an augmentation-like system. Also some components could be designed drastically different from their human consciousness counterparts in order to achieve more desirous effects in one area or another. ALSO there may be consciousness components integrated into AGI that humans don't have or that are almost non-detectable in humans. And I think that the different consciousness components and sub-components could be more dynamically resource allocated in the AGI software than in the human mind. John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Dr. Matthias Heger [mailto:[EMAIL PROTECTED] For general intelligence some components and sub-components of consciousness need to be there and some don't. And some could be replaced with a human operator as in an augmentation-like system. Also some components could be designed drastically different from their human consciousness counterparts in order to achieve more desirous effects in one area or another. ALSO there may be consciousness components integrated into AGI that humans don't have or that are almost non-detectable in humans. And I think that the different consciousness components and sub-components could be more dynamically resource allocated in the AGI software than in the human mind. Can neither say 'yes' nor 'no'. Depends on how we DEFINE consciousness as a physical or algorithm-phenomenon. Until now we each have only an idea of consciousness by intrinsic phenomena of our own mind. We cannot prove the existence of consciousness in any other individual because of the lack of a better definition. I do not believe, that consciousness is located in a small sub- component. It seems to me, that it is an emergent behavior of a special kind of huge network of many systems. But without any proper definition this can only be a philosophical thought. Given that other humans have similar DNA it is fair to assume that they are conscious like us. Not 100% proof but probably good enough. Sure the whole universe may still be rendered for the purpose of one conscious being, and in a way that is true, and potentially that is something to take into account. Consciousness has multiple definitions by multiple different people. But even without an exact definition you can still extract properties and behaviors from it and from those, extrapolations can be made and the beginnings of a model can be established. Even if it is an emergent behavior of a huge network of many systems doesn't preclude it from being described in a non-emergent way. And if it is only uniquely describable through emergent behavior it still has some general commonly accepted components or properties. John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Consciousness vs. Intelligence
On Saturday 31 May 2008 10:23:15 pm, Matt Mahoney wrote: Unfortunately AI will make CAPTCHAs useless against spammers. We will need to figure out other methods. I expect that when we have AI, most of the world's computing power is going to be directed at attacking other computers and defending against attacks. It is no different than evolution. A competitive environment makes faster rabbits and faster foxes. Without hostility, why would we need such large brains? In the biological world, big brains evolved to support reciprocal altruism, which requires recognizing individuals and knowing which ones owe you one and vice versa. http://en.wikipedia.org/wiki/Reciprocal_altruism Going back to Trivers' first studies: bats that practice R.A. have brains three times the size of ones that don't. Josh --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Matt Mahoney [mailto:[EMAIL PROTECTED] Unfortunately AI will make CAPTCHAs useless against spammers. We will need to figure out other methods. I expect that when we have AI, most of the world's computing power is going to be directed at attacking other computers and defending against attacks. It is no different than evolution. A competitive environment makes faster rabbits and faster foxes. Without hostility, why would we need such large brains? AI has a long way to go to thwart CAPTCHAs altogether. There are math CAPTCHAs (MAPTCHAs), 3-D CAPTCHAs, image rec CAPTCHAs, audio and I can think of some that are quite difficult for AI. Actually coming up with new CAPTCHAs would be a neat AI subproject, as well as thwarting existing ones. Does that have anything to do with consciousness? What's the test for consciousness? I think the way you are using the term, it is the Turing test. Are you sure? Isn't that supposed to just differentiate between computer and human? Not between unconscious and conscious? Unless you think consciousness is just belief... John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Mike Tintner [mailto:[EMAIL PROTECTED] You are - if I've understood you - talking about the machinery and programming that produce and help to process the movie of consciousness. I'm not in any way denying all that or its complexity. But the first thing is to define and model consciousness, before you work out that machinery etc. It may sound simple to you, because you're naturally enamoured of the programming/machinery part, and trying to work out sophisticated programming etc may sound much smarter and more exciting. But actually starting with that simple model is much more important. If you don't know, or only half know, or are radically confused about, what you're trying to explain, all the technical ideas you may come up with will be worthless. I understand what you are saying. But for me, I just need to get a mental visual on it. Like, say that you are an automotive engine designer or a class A engine mechanic. You have that engine running inside you head in that mental CAD system. You can change parameters like octane and you know the effects on running temperature. You can tweak the thing in your head, increase piston diameter size, you know what's going to happen. For consciousness I need to get a visual of the model and lock onto it then I can get the math and the code. Easy for an insecticidal level consciousness but when adding stuff I can't fit the whole model in my head so then it gets difficult to change variables and test. Kinda like, I'm trying ta think but nothin' happens!! HEH John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Matt Mahoney [mailto:[EMAIL PROTECTED] --- On Sun, 6/1/08, John G. Rose [EMAIL PROTECTED] wrote: AI has a long way to go to thwart CAPTCHAs altogether. There are math CAPTCHAs (MAPTCHAs), 3-D CAPTCHAs, image rec CAPTCHAs, audio and I can think of some that are quite difficult for AI. Actually coming up with new CAPTCHAs would be a neat AI subproject, as well as thwarting existing ones. I think security is going to be a driving force behind AI development. You can see which way this is heading. It is a scruffy approach, no grand theories, just a series of hacks and incremental improvements on both sides to get the job done. Ultimately all CAPTCHAs will fail and we will need AI to detect malicious activity downstream. Does that have anything to do with consciousness? What's the test for consciousness? I think the way you are using the term, it is the Turing test. Are you sure? Isn't that supposed to just differentiate between computer and human? Not between unconscious and conscious? Unless you think consciousness is just belief... Yes. A CAPTCHA is a cheap Turing test. Whether a machine thinks or is conscious is just an irrelevant distraction that Turing wanted to avoid. OK How about this. A CAPTCHA that combines human audio and visual illusion that evokes a realtime reaction only in a conscious physical human. Can audio visual illusion be used as a test for consciousness? Could it be used to evoke a specific conscious-only reaction in a human mind? Hmm... John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
--- On Sun, 6/1/08, John G. Rose [EMAIL PROTECTED] wrote: OK How about this. A CAPTCHA that combines human audio and visual illusion that evokes a realtime reaction only in a conscious physical human. Can audio visual illusion be used as a test for consciousness? Could it be used to evoke a specific conscious-only reaction in a human mind? Hmm... For example, a CAPTCHA plays an audio clip and you have to match it to one of several images. Or did you have something else in mind? In any case, solving a CAPTCHA means giving the right output for some input. There is no reason in principle that a machine could not solve it. What do you mean by consciousness? Do you agree that a human brain simulated by a computer at the neuron level would be functionally equivalent and indistinguishable from human? Or is there a mysterious force like consciousness that causes the machine to give a different output for some input? -- Matt Mahoney, [EMAIL PROTECTED] --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Matt Mahoney [mailto:[EMAIL PROTECTED] --- On Sun, 6/1/08, John G. Rose [EMAIL PROTECTED] wrote: OK How about this. A CAPTCHA that combines human audio and visual illusion that evokes a realtime reaction only in a conscious physical human. Can audio visual illusion be used as a test for consciousness? Could it be used to evoke a specific conscious-only reaction in a human mind? Hmm... For example, a CAPTCHA plays an audio clip and you have to match it to one of several images. Or did you have something else in mind? Ah, use common visual illusion techniques, build a software library. There may already be some. Then an audio illusion library. And then find a set of audio visual combinations that induce evocation of, say strings, like asking the user to recognize this string of alphanumeric characters. That's all one step. Then next step would be to find, using human volunteers, evocation results unique to human conscious experience. Example - this illusory sensory stimulation causes a human to sense this uniquely describable qualia. At a percentage success rate, like 90%. BUT the ultimate would be to be able to ring someone's bell at a 100% unique hit rate, based on DNA or something else or just their own signature uniqueness... In any case, solving a CAPTCHA means giving the right output for some input. There is no reason in principle that a machine could not solve it. The machine would have to replicate that specific human biochemical computational ability to be able to achieve surpassed pattern recognition fidelity. If you are able to dupe that, then you are human or better. What do you mean by consciousness? Do you agree that a human brain simulated by a computer at the neuron level would be functionally equivalent and indistinguishable from human? Or is there a mysterious force like consciousness that causes the machine to give a different output for some input? The machine, the human brain, will give a machine specific output, a.k.a., an illusion. Using audio visual illusion you may be able to create cognitory illusions whci are unique in a human brain. John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Mike Tintner [mailto:[EMAIL PROTECTED] You guys are seriously irritating me. You are talking such rubbish. But it's collective rubbish - the collective *non-sense* of AI. And it occurs partly because our culture doesn't offer a simple definition of consciousness. So let me have a crack at one. First off, let's remove consciousness-as-sentience. That's important, but it's secondary. The real issue of whether an AGI - a computer - should have consciousness, is, I suggest, should it have a world-movie ? IOW should it run a continuous sensory movie of the world around it? That's what every living creature from single cells upwards runs - a continuous sensory movie of the world around it, (although it took time to get to visual movies). That movie is clearly the central business of consciousness. And it is also clearly what evolution thought the absolute foundation of intelligence. Amazingly, it didn't listen to you guys, smart as you are - it didn't start with logic or mathematics or language, or anything that AI considers to be a sine qua non. That sensory movie is the absolute foundation of human intelligence too, the thing that never ever stops. Even you guys sometimes stop thinking in language, or maths, but your movie of the world around you never stops. And when you sleep, your mind keeps running movies of imagined worlds around you. Even then language and maths are at best only occasional participants in the movie. If you could stop thinking about just your computers, and start thinking - as you absolutely must here - about robots as well, then it is, I suggest, obvious that the first thing a robot needs is a world-movie . How can you survive in the world if you can't see the world, can't see where you need to go, or what's coming at you, (or which keys to push on your computer)? Even your fictional superAGI, if it were to be independent, would have to run a movie of the world around it, to protect itself from all the dangers of the world, like human programmers, bent on harming it, and ensure its supply of energy and other necessities. How too can you know about the world if you've never seen it, on-the- spot, firsthand, and in person through your free-roaming world-movie? (Sure I can. Wikipedia tells me everything I need to know about the world and life. Right). If you continue to think robotically, you also won't have any need to include high-falutin' forms of self-consciousness in your definitions of consciousness. Because that movie will obviously have to be an iworld-movie. Any robot or agent moving through the world must have a continuous sense of self - its integrated body, brain and sensors, rather than some homunculus - in relation to the world around it - and therefore a sense of itself watching the movie. There can be no consciousness WITHOUT self-consciousness - no world-movie without an I/eye/camera. You have to estimate continuously where things are in relation to yourself in order to shift your POV, or move this way or that, towards or away from things, as required. So no high-falutin' yes, but who am I really - I mean deep down kind of self-consciousness is required, just the most basic kind, that even schizophrenics have. It should be obvious that consciousness is your world-movie - it's what you're looking at right now, the show that never stops. How could you survive in the real world, without that world-movie? But when you're incredibly smart, and rational, and logical, like AI- ers, and are deeply prejudiced against anything to do with movies or images or imagination let alone bodies, you can't see the obvious - the couldn't-possibly-be-more-obvious. And you are prepared to settle for a totally sense-less, deaf-dumb-and-blind, brain-in-a-vat conception of intelligence. Then you end up resorting to the most desperately contorted arguments to justify your senselessness - I can see this one coming - ah but my cousin has been a total vegetable in a coma for the last twenty years, and he's conscious. Mike, The reason why people are thinking about all this stuff in terms of maths is because it is not all just fluffy philosophizing you have to have at least minimalistic math models in order to build software. So when you say iTheathre or iMovie I'm thinking bits per send, compression, color depth, Fourier transforms, object recognition probabilities,... sorry man that's how it is. Just saying that there is a movie projecting against the inside of your skull ain't gonna cut the mustard. Movie is too broad you have to define it more... John --- 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=103754539-40ed26 Powered by Listbox:
RE: [agi] Consciousness vs. Intelligence
From: Matt Mahoney [mailto:[EMAIL PROTECTED] What many people call consciousness is qualia, that which distinguishes you from a philosophical zombie, http://en.wikipedia.org/wiki/P-zombie There is no test for consciousness in this sense, but humans universally believe that they are conscious, and this belief is testable. Just ask someone. Do you really feel pain, or do you just behave as if you feel it? The belief in experiencing qualia is what I call recursive episodic memory. Episodic memory is the ability to recall a time sequence of events in the correct order. These events could include earlier acts of recall. For example, earlier today I recalled how yesterday a tune was playing in my head that I heard the day before (and so on). You probably do not remember any events that happened before you were 3 years old. You were clearly learning then, but it was not in episodic memory. A person without a hippocampus lacks episodic memory. He could learn new skills but wouldn't remember the lessons. Episodic memory has been demonstrated in birds, but we do not know if it is recursive. I don't know if recursive episodic memory is necessary for intelligence. When I need to come up with an algorithm when writing software, it is useful to go through the steps in my head and then be able to recall my thought process. It is also useful for databases to log read-only transactions. It is useful for computers to copy recently read data to cache. However, recursive episodic memory could also be an artifact of the brain's memory management system. Long term memory is written at a constant rate (about 2 bits per second, according to Landauer). During quiet times, it has to write something. People believe they are conscious. Why? Because they are. There is camaraderie between conscious beings, especially ones of similar consciousness like the same species. I believe that the guy walking down the street looking at me thinking about mugging me(say it's NYC) is conscious. Why? Because he believes that I am conscious and I might have a gun for defense. It's self reinforcing. We analyze each other's conscious thoughts. More than 2 agents and you get more effects and amplifications. Is there more than just a belief that we are conscious? Sure some rare individuals can block pain. But when they do so they are actually blocking the signals somehow or preventing their registration. There is a real pain that is blocked. It's real. Consciousness is a system related to structure, information flow, etc.. and there are different types and strengths. It's more than just a belief... and belief may be a product of consciousness. John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Consciousness vs. Intelligence
John, The reason why people are thinking about all this stuff in terms of maths is because it is not all just fluffy philosophizing you have to have at least minimalistic math models in order to build software. So when you say iTheathre or iMovie I'm thinking bits per send, compression, color depth, Fourier transforms, object recognition probabilities,... sorry man that's how it is. Just saying that there is a movie projecting against the inside of your skull ain't gonna cut the mustard. Movie is too broad you have to define it more... John, Thanks for response. I apologise to you and others for going over the top, but I was deeply frustrated and right to be frustrated - with the scientific/philosophical world (and not just AI) re their treatment of consciousness. No, I believe I'm right here. Maths is only quantification - the question is : what are you quantifying? Programs are only recipes to construct something or a sequence of behaviour. The question again is: what are you constructing? You have to START by providing a model of consciousness - of what it involves. It is grossly unscientific not to do so. And that movie is the sine qua non starting point for a model. Certainly intelligence has to be applied to the movie - to understand what is being reflected in the movie - the objects and world around you. And by all means quantify and program away - but first agree about what you are quantifying. Otherwise it's all basically hot air. And you guys - along with all other serious thinkers discussing this area - have NO AGREEMENT about what you are discussing, or whether any of you are talking about the same thing. That, if you think about it, is ridiculous. Damasio who is one of the best thinkers here, talks of consciousness as the movie in the mind - I believe that world movie is a step forward because it focusses on what the movie shows and is for. Some perspective here: words are absolutely wrong and misleading as a SOLE medium to discuss consciousness - they fragment whatever you are talking about. And consciousness is indeed a continuous movie, (for want of a still richer model) - and MUST be talked about with the aid of movies, as I tried to do. The reason people resist this model - is they are only comfortable talking in words. They are uncomfortable and ill-versed thinking visually and sensorily. Well, tough. If you are serious about consciousness, there is no alternative. You cannot discuss actual movies in cinemas seriously with just words. Nor can you discuss the movie-that-is-consciousness seriously with just words either. You guys, I am consistently arguing, have to learn respect for the brain. If the brain does things a certain way, then that is probably the ideal way to do it - in the technical, psychological sense - i.e. not the perfect way, not something that can't be improved, but a more or less inevitable and essential way, (in a very broad sense), to tackle the given problems. And the way the brain chooses throughout evolution to tackle the problems of survival is to run a movie. Your brain does not allow you, for example, to jump straight to imbibing logic and mathematics - as your computers can - it forces you to run a movie of the books you're reading, or computer screen you're looking at, and all those logical and mathematical figures have to be processed as IMAGES. Your brain insists that you SEE or otherwise sense what you're talking about. P.S. Qualia are a massive distraction here. People are ironically making the same mistake here that novice novelists make. When they start writing a story, they are obsessed with the FEELINGS of the events they describe. God, that love affair was so painful, you see. But actually you can't think straight about any events in your life, if you don't describe what you and other people are DOING... whether by actions, or thinking, or speech. Once you concentrate on those, the feelings automatically fall into place. The same is true with consciousness. Model first what consciousness does. It sees etc - runs a movie - of the world. P.P.S. As I pointed out, too, without a model of consciousness-as-movie the SELF wanders around, homeless, in any verbal discussion of consciousness. With the movie model, the self automatically has a place - it's the brain-body watching the movie, continually directing the movie, turning the camera this way and that. And then the contents of consciousness fall into place too - because every sight - every shot you see - including every photograph you look at - has a Point of View - inevitably implies a self watching at a distance. IOW every shot is a close-up, long distance, at an upward/downward angle from an implied viewer. And all this is true of course for all animals. So no there is no alternative to the movie model (only better or superior, modified versions) of it. What alternative are you or anyone else offering?
Re: [agi] Consciousness vs. Intelligence
--- On Sat, 5/31/08, Ben Goertzel [EMAIL PROTECTED] wrote: But in future, there could be impostor agents that act like they have humanlike subjective experience but don't ... and we could uncover them by analyzing their internals... What internal properties of a Turing machine distinguish one that has subjective experiences from an equivalent machine (implementing the same function) that only pretends to have subjective experience? -- Matt Mahoney, [EMAIL PROTECTED] --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Ben Goertzel [mailto:[EMAIL PROTECTED] If by conscious you mean having a humanlike subjective experience, I suppose that in future we will infer this about intelligent agents via a combination of observation of their behavior, and inspection of their internal construction and dynamics. An N-like subjective experience where N is human, animal, bug, space alien, god. I don't know if there needs to be an I since you could have a distributed, decentralized I or other forms. But in future, there could be impostor agents that act like they have humanlike subjective experience but don't ... and we could uncover them by analyzing their internals... If the imposters are good enough then they would be the same from a functional perspective. And if they were the same eventually they would improve their various attributes until we appeared as zombies to them and they as godlike to us even though they were just imitating our consciousness. This is under the assumption that subjective experience of an agent is correlated with (though not identical with) the patterns in the physical system serving as the substrate of that agent ... and that external behaviors only constitute a subset of these patterns... Yes unless there is that complexity layer disconnect... John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Mike Tintner [mailto:[EMAIL PROTECTED] That's correct. The model of consciousness should be the self [brain- body] watching and physically interacting with the movie [that is in a sense an open movie - rather than on a closed screen - projected all over the world outside, and on the inside of the body]. The self is an integrated brain-body unit, acting and responding with the whole body. But you missed out the all-important part which I believe you're all skipping over. What is your or anyone else's model of consciousness? Which model are you using? Or do you know anyone else using? Or do you not have a model? You've been talking about consciousness - *what* have you been talking about? Honestly? Mike, Just because you have movies and theatres and are experiencing a simuworld in your mind rolling around in the sand feeling all warm and fuzzy doesn't make it work. There are a few quantitative systems relationships that need to be strictly defined in order to have a model that isn't just a bunch of ideas slapped together that sound good to a philosophy student. The system has to come together in such a way that it functions and operates like a machine or a machine derivative so that you can actually build it within a lifetime. There are consciousness patterns. And there are consciousness inkblots. John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Consciousness vs. Intelligence
--- On Sat, 5/31/08, Ben Goertzel [EMAIL PROTECTED] wrote: I wrote: What internal properties of a Turing machine distinguish one that has subjective experiences from an equivalent machine (implementing the same function) that only pretends to have subjective experience? You're asking a different question. What I said was that internal properties could distinguish a) a machine having HUMANLIKE subjective experiences from b) a machine just claiming to have HUMANLIKE subjective experiences, but not really having them The reason I ask is because humans and human-like intelligence occupies a tiny region in the huge space of possible intelligence. If only humanlike subjective experience is important, then the definition is easy. Subjective experience is something humans have, and nothing else. End of argument. I am looking for a more general principle. On SL4 I proposed that an agent A receiving sensory input x at time t has subjective experience s(x|A) = K(A(t+)|A(t-)) where K is Kolmogorov complexity, A(t-) is the state of A immediately before input x and A(t+) is the state immediately afterwards. In other words, it the length (in bits) of the shortest program that takes as input a description of A prior to input x and outputs a description of A after input x. It measures how much A remembers about x, independent of K(x). In your book you mentioned the intensity of a pattern and gave a definition that included a lossy compression term (i.e. the part of x that A ignored). Note that in general, s(x|A) = K(x|A(t-)) where the difference is the number of bits that A ignored. This definition makes no distinction between having subjective experience and pretending to have it. It also makes no distinction between humanlike subjective experience and any other kind. By this definition, your computer can already have 10^12 bits of subjective experience, far more than the 10^9 bits of human long term memory estimated by Landauer. I do not mean to imply any ethical considerations by this. It is easy to confuse conscious entities (those having subjective experience) with entities that have rights or require compassion. The human ethical model has no such requirement. Ethics is an evolved function that selects for group fitness. You are compassionate to other humans because it increases the odds of passing on your genes. If all conscious entities were worthy of compassion, we would not have wars or eat meat. This leads us back to our original definition... -- Matt Mahoney, [EMAIL PROTECTED] --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
--- On Sat, 5/31/08, John G. Rose [EMAIL PROTECTED] wrote: From: Matt Mahoney [mailto:[EMAIL PROTECTED] I don't believe you are conscious. I believe you are a zombie. Prove me wrong. I am a zombie. Prove to me that I am not. Otherwise I will accuse you of being conscious. Exactly my point. -- Matt Mahoney, [EMAIL PROTECTED] --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Matt Mahoney [mailto:[EMAIL PROTECTED] --- On Sat, 5/31/08, John G. Rose [EMAIL PROTECTED] wrote: From: Matt Mahoney [mailto:[EMAIL PROTECTED] I don't believe you are conscious. I believe you are a zombie. Prove me wrong. I am a zombie. Prove to me that I am not. Otherwise I will accuse you of being conscious. Exactly my point. Just because you communicate through the limited bandwidth medium of text doesn't prove anything. You could make a smart zombie chatterbot and make it irritating enough where it is very terse and just fleeting enough where you never figure it out. So what? A problem though is eventually even a CAPTCHA won't work to filter these things. CAPTCHA's will evolve past character recognition to image recognition, it'll be interesting. What do you see in this image, a pony or a donkey. The bots will get smarter. And CAPTCHA's will ask even more humanlike questions. Does that have anything to do with consciousness? What's the test for consciousness? John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Mike Tintner [mailto:[EMAIL PROTECTED] you utterly refused to answer my question re: what is your model? It's not a hard question to start answering - i.e. either you do have some kind of model or you don't. You simply avoided it. Again. I have some models that I feel confident that would work yet they are basically insectoidal level consciousness. When I get to more sophisticated models the confidence that I have in them is not enough and the system is difficult to test just by thinking about it, and they are incomplete. So... still working on that. But I do have higher level intelligence models that I am confident in. And a major point I'm making is that - everyone is doing that. Everyone is picking some very limited aspect of consciousness that is important to them - experience, qualia, self-consciousness. And no one has or is offering a model of the whole. I haven't read up enough on other models to concur on this. But we can and must produce a model and point to what we're talking about. Here is someone who is conscious: http://www.bized.co.uk/images/man_remote.jpg We can observe consciousness from the outside with such a picture (or better still clip) - and point to those parts of him that are conscious and how and what parts of him like nerves, produce that consciousness . And we can model his consciousness from the inside: http://electrojusa.iespana.es/images/philips_25pt_7304_television__47429 .jpg http://www.engadgetmobile.com/media/2007/01/sch-u620-hands-on-2.jpg what he's seeing and hearing etc. A properly defined model - including my movie model - should include all those things. When you describe this you have to be careful how much computation your mind is doing and taking for granted. You make many assumptions just by looking at the pic and saying these are signs that this man is conscious. And saying that a handheld TV is some sort of model, ya that's making massive assumptions and shortcuts to the point of assuming that 99% of your model already exists where it doesn't; you're just pointing at data feeds and neglecting numerous other details which are the most important. John --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Consciousness vs. Intelligence
John:When you describe this you have to be careful how much computation your mind is doing and taking for granted. You make many assumptions just by looking at the pic and saying these are signs that this man is conscious. And saying that a handheld TV is some sort of model, ya that's making massive assumptions and shortcuts to the point of assuming that 99% of your model already exists where it doesn't; you're just pointing at data feeds and neglecting numerous other details which are the most important. John, You are - if I've understood you - talking about the machinery and programming that produce and help to process the movie of consciousness. I'm not in any way denying all that or its complexity. But the first thing is to define and model consciousness, before you work out that machinery etc. It may sound simple to you, because you're naturally enamoured of the programming/machinery part, and trying to work out sophisticated programming etc may sound much smarter and more exciting. But actually starting with that simple model is much more important. If you don't know, or only half know, or are radically confused about, what you're trying to explain, all the technical ideas you may come up with will be worthless. And that's the same mistake people are making with AGI generally - no one has a model of what general intelligence involves, or of the kind of problems it must solve - what it actually DOES - and everyone has left that till later, and is instead busy with all the technical programming that they find exciting - with the how it works side - without knowing whether anything they're doing is really necessary or relevant.. --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
--- On Sat, 5/31/08, John G. Rose [EMAIL PROTECTED] wrote: From: Matt Mahoney [mailto:[EMAIL PROTECTED] --- On Sat, 5/31/08, John G. Rose [EMAIL PROTECTED] wrote: From: Matt Mahoney [mailto:[EMAIL PROTECTED] I don't believe you are conscious. I believe you are a zombie. Prove me wrong. I am a zombie. Prove to me that I am not. Otherwise I will accuse you of being conscious. Exactly my point. Just because you communicate through the limited bandwidth medium of text doesn't prove anything. You could make a smart zombie chatterbot and make it irritating enough where it is very terse and just fleeting enough where you never figure it out. So what? My point is that you can't tell if a person is conscious (has experience) or a zombie (learns and has memory, but no qualia). I think you are talking about something else. By zombie, I mean http://en.wikipedia.org/wiki/Philosophical_zombie A problem though is eventually even a CAPTCHA won't work to filter these things. CAPTCHA's will evolve past character recognition to image recognition, it'll be interesting. What do you see in this image, a pony or a donkey. The bots will get smarter. And CAPTCHA's will ask even more humanlike questions. Unfortunately AI will make CAPTCHAs useless against spammers. We will need to figure out other methods. I expect that when we have AI, most of the world's computing power is going to be directed at attacking other computers and defending against attacks. It is no different than evolution. A competitive environment makes faster rabbits and faster foxes. Without hostility, why would we need such large brains? Does that have anything to do with consciousness? What's the test for consciousness? I think the way you are using the term, it is the Turing test. -- Matt Mahoney, [EMAIL PROTECTED] --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] Consciousness vs. Intelligence
From: Matt Mahoney [mailto:[EMAIL PROTECTED] --- John G. Rose [EMAIL PROTECTED] wrote: Consciousness with minimal intelligence may be easier to build than general intelligence. General intelligence is the one that takes the resources. A general consciousness algorithm, one that creates a consciousness in any environment may be simpler that a general intelligence algorithm that acquires intelligence in any environment. The two can go hand in hand but one can be minimized against the other. But I don't understand the relationship between consciousness and intelligence. I want to say that they are like disjoint vectors but that doesn't seem right... You need to define your terms. What properties of an algorithm make it conscious? What properties make it intelligent? To some people, the two terms are equivalent. To others, consciousness does not exist. How can the two terms be equivalent? Some may think that they are inseparable, or that one cannot exist without the other, I can understand that perspective. But there is a quantitative relationship between the two. When you get into strict definitions people get alienated... john --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Consciousness vs. Intelligence
On Thu, May 29, 2008 at 6:41 PM, John G. Rose [EMAIL PROTECTED] wrote: How can the two terms be equivalent? Some may think that they are inseparable, or that one cannot exist without the other, I can understand that perspective. But there is a quantitative relationship between the two. When you get into strict definitions people get alienated... For me, working meaning of consciousness is reflection, or a process of memory-formation about the processes going in the mind, which is the same thing as learning, since any kind of external information must first set in motion a process in the mind in order to be perceived. By intelligence, to separate it from knowledge, I understand the efficiency of learning. Thus, it can be said that in my definitions intelligence is a property of consciousness, but it's really unnecessarily confusing to use these way overloaded terms, and it's almost meaningless to use them without clarification of what's meant in a particular case. -- Vladimir Nesov [EMAIL PROTECTED] --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com
Re: [agi] Consciousness vs. Intelligence
--- John G. Rose [EMAIL PROTECTED] wrote: Consciousness with minimal intelligence may be easier to build than general intelligence. General intelligence is the one that takes the resources. A general consciousness algorithm, one that creates a consciousness in any environment may be simpler that a general intelligence algorithm that acquires intelligence in any environment. The two can go hand in hand but one can be minimized against the other. But I don't understand the relationship between consciousness and intelligence. I want to say that they are like disjoint vectors but that doesn't seem right... You need to define your terms. What properties of an algorithm make it conscious? What properties make it intelligent? To some people, the two terms are equivalent. To others, consciousness does not exist. -- Matt Mahoney, [EMAIL PROTECTED] --- 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=103754539-40ed26 Powered by Listbox: http://www.listbox.com