Follow a topic to ask questions, get answers and
share interesting content with other
researchers working in your field.
Topics » Artificial Consciousness
Is Consciousness Synthesizeable? And if so how?
- Graeme SmithIt is my belief that the difference between a conscious and an unconscious machine, is that the conscious machine is aware of its actions and reacts not only to the environment but to its own impactRecent replies ⋅ Show All (18)
Graeme Smith replied
On the other hand, low-level rules, do not need understanding to be implemented but their application is fraught with side-effects that would need a GA like design network to navigate.
4 days ago
- Graeme SmithRecent Congruence between my understanding of the declarative memory system, and my theoretical model of consciousness, has led me to question whether the particular implementation of the declarativeRecent Congruence between my understanding of the declarative memory system, and my theoretical model of consciousness, has led me to question whether the particular implementation of the declarative memory system found in vertebrates, might be the basis for consciousness as we know it. Recent research into fruit fly brains, has shown that mushroom bodies have a role in the action management system of the brain, but as far as I remember from Braak's Architectonics of the human telencephalic cortex, those structures are more likely part of the human declarative memory, and the recent suggestion that the parahippocampus might be the location of "What" mapping in the declarative memory suggests that in fact, a change in use might have occurred between insects and vertebrata, that would have freed up the hippocampal area for use in the declarative memory. I trace that change in use to the formation of the vertebrate cerebellum. Since the Craniata do not exhibit cerebellar structures but the vertebrata do, it is my thought that consciousness as we humanly know it, is a function of the freed action management system, repurposed to become declarative memory sometime after the evolution of vertebrata. To understand why I claim consciousness for this area, will require some discussion. This discussion will open the subject to further study, and possibly future research potential.Recent replies ⋅ Show All (4)
Graeme Smith replied
Because the declarative memory is implemented as a form of implicit memory it has all the characteristics of implicit memory including retrieval only from contents, and voluntary retrieval duringBecause the declarative memory is implemented as a form of implicit memory it has all the characteristics of implicit memory including retrieval only from contents, and voluntary retrieval during storage. Because it is storing meta-information instead of memories per se, immediate voluntary retrieval, results in recall of events that transpired only a few milliseconds earlier. Evidence of scientists such as Libett, suggest that this recall is delayed approximately 500 milliseconds or so after the original stimulus. The recall, and it's parameterization first with the declarative meta-index information and then with the emotive and meta-cognitive information means that the recall is much richer in nature than the original stimulus, which is often lost in comparison. Much of what we think of as awareness or experience is in fact this recall rather than the original stimuli. While we have measured delays in response to stimuli, there is a tendency to think that the recall is the first stimuli, and therefore a tendency to the illusion that we perceive the stimuli directly, instead of as the result of feedback after processing. This is an illusion and cannot be as real as the actual stimulus-response delay measured by Libett. Awareness is linked to the recall therefore and the nature of what we are aware-of is linked to the nature of the meta-information that is stored in the declarative memory. As such, a study of the nature of the storage might be a good idea. I have determined that there are 5 possible maps in the hippocampus and parahippocampus, that these are indexed by the entorhinal cortex, and accessed via the subiculum. This results in a parallel map-image containing much more information than a single map would allow. This map image feeds back into the short term memory in the ventro-lateral PFC. triggering the retrieval of data from the cortex. The first map is the "What" map or mushroom map, in the parahippocampus which contains conceptual distinctions associated with the What-pathway in the temporal lobe. the Next map (CA4) links the What Pathway to the CA3 map where the WHERE and When information is mapped according to the Where pathway in the parietal lobe. CA1 maps (AFAIK) meta-cognitive signals and CA2 maps these signals to emotions and to CA3 creating a secondary recall of CA3 results some few milliseconds after the original. Because the brain cannot tell if CA3 was being rehearsed or merely recalled, the primary difference between a CA3 recall and CA2 and CA3 recall, is the parameterization of the experience with declared feelings and meta-cognitive signals. It is because of meta-cognitive feelings such as the "Self" signal that we see ourselves as having a "Self". It is because we attribute Agency to the "Self" that we claim we have minds.
9 days ago
- Graeme SmithConsciousness as a Simulation, why I am not worried about the challenge that philosophers like Searle, and Overland have offered that any software that simulates the Mind, can't be a Mind. To makeConsciousness as a Simulation, why I am not worried about the challenge that philosophers like Searle, and Overland have offered that any software that simulates the Mind, can't be a Mind. To make this observation that there is a long-standing challenge to A.I. and Artificial Consciousness that a Simulation of a mind, can't be a mind, is to give credence to the concept that a simulation is not the thing that is simulated. I actually agree with this statement. A simulation is not, and never can be the thing that is simulated. However this is not a serious impediment to Artificial Consciousness, because the mind itself must be a simulation in its own right, since it does not accurately reflect the architecture of the Brain below it. To Understand why I make this distinction, I need to introduce the idea of a Virtual Machine as a simulation or emulation of another machine on a different architecture. Now this might be confusing to those who are just now beginning to deal with virtualization and the idea of Virtual Machines as virtualized sub-processes on a physical machine that is designed to run the processes on bare silicon. But the subprocess version of Virtual Machine does not reflect the actuality of the base architecture, merely the presence of processes that are similar to one that would run on the machine, if it had the machine all to itself. Thus one could say that the Virtual Machines are sup-processes that emulate the processes that would be normally running on the machine if there were only one of them, but in such a way as to allow multiple versions to run concurrently without interfering with each other. So the current definition of a Virtual Machine is really simply a special case that proves the original definition. One of the first uses of Virtualization was in the CP operating system where a virtualization program allowed a single serial processing system to act like a multiple computer cluster so that groups of people could share the machine. Now of course we have multiple operating systems running on the same base architecture. But Virtual Machines do not just work one way, with the advent of cluster computers it became possible for a cluster of separate computers to act like a single computer, while this is not generally called a virtual machine it has exactly the same effect, it is a single machine operating on multiple base machines and merely a change in architecture between the platforms below and the machine above. So when I say that the Mind is a simulation or emulation of an architecture based on the brain which has a completely different architecture, I hope you can appreciate that having a massively parallel architecture running as the base platform, has no bearing on what the architecture of the virtual machine above claims to be. In fact, the massively parallel architecture is not the level that the mind operates on at all. There is actually an intervening architecture, I call the merely Very Parallel architecture that operates between the massively parallel architecture and the Mind. How can I say that the mind is a simulation, when it feels real to me? The reason is quite simple, the mind operates as if it were a serial processing system, but the Brain that it operates on, is not. Even at the Merely Very Parallel level, the brain utilizes massively parallel components clustered together into organs, and Organic circuits to implement a parallel architecture, and yet, it feels to me as if I follow a standard progression, I think, I decide, I do. But this natural progression, can't be accurate, because while I think I am thinking, many processes are going on simultaneously at both the massively parallel, and merely Very Parallel levels that are not caused by my thoughts, but are instrumental in making them possible. Why aren't these processes reflected in my experience of mind? Simply because the purpose of mind is to SIMPLIFY a very complex parallel control process so that it looks easy, and operates as if it were really a serial process. Recent scientific observation puts paid to the idea that the Mind actually decides. There is evidence of statistical correlations for decisions being presaged by as much as 7 minutes before the mind is ready to acknowledge that it is even trying to make a decision. In other words there is a lot going on, below the surface, that just isn't reflected in the mind, and so the mind cannot reflect the actual operation of the brain, but merely a shadow of it. This is critical, because it voids the previous physicalist concept that the mind IS the processes of the brain. The only conception that could possibly describe the Mind under this understanding is that the Mind, is a simulation of the brain operations that is mostly an illusion, created to simplify control. If the mind IS a simulation that does not reflect the underlying architecture of the brain, then, a simulation on a different architecture could miss reflecting it just as easily in much the same way that the mind does, and in doing so, it could operate in much the same way that the mind does. In other words Simulation is no barrier to Artificial Consciousness because Simulation is required to form a mind.Recent replies ⋅ Show All (11)
- Raja N A KhanHi to All, I am new in this group with aim of learningRecent replies ⋅ Show All (15)
Graeme Smith replied
Of course they are, they have everything to gain. I have heard it said by the experts here in Canada, that the Brain Drain has really taken off between the Chinese and Canada. Of course these areOf course they are, they have everything to gain. I have heard it said by the experts here in Canada, that the Brain Drain has really taken off between the Chinese and Canada. Of course these are the same experts that claim that ALL of the groups being supported in incubators here in Canada are likely to fail in the first 2 years, and if not, they plan to give them the shove about year 5 so they can sell their assets for millions, (Probably to the Chinese again). Why Lofti Zadeh, himself, is more or less constrained to write for a Chinese Journal if you trust the journals opinion. Personally I believe in "Buy Local" if possible (It isn't)/"Sell Local" if possible (even less likely in the Globalized World). I am having enough troubles living with my medical condition here in Canada, I do not expect to find I would be better half-way across the world. Why I remember those poor saps that bought the Potomican villiages concept of Russia, only to get stuck in the Russion equivalent of a condo once they got there. The Grass Is always greener, when you are looking over the fence at it. Remember Google left there (China) for a reason. Of course the poor Chinese slobs that come here, are put in much the same position, they come here expecting good housing, only to find that it has been priced out of reach even for rich families from Singapore, and Taiwan. I am not some bright eyed teenager with a high I.Q. I have rattled around out there and learned that you can be miserable anywhere in the world, but at least at home you can visit your mother once in a while.
Jan 31, 2012
- Graeme Smithsample text from Chapter Six of How Memory Might Work an unpublished book which I am the author ofRecent replies ⋅ Show All (2)
- Hemamalini VeerasingamThere are several levels of consciousness, from being woken from a deep sleep to being in acute concentration. Who knows if this acute concentration is the highest level of consciousness?There are several levels of consciousness, from being woken from a deep sleep to being in acute concentration. Who knows if this acute concentration is the highest level of consciousness? Consciousness may be limited to a focused domain. Artificial thinking comes first. Apply this thinking process to present "real time" situations and you get consciousness. How to get self consciousness? Artificial self consciousness is relative to available senses. Self consciousness is the self perception of an artificial being interaction with its environment. This perception is directly linked to the accuracy of these senses. Moreover, self consciousness comes from conceptualization and remembrance of past thoughts. "I think, therefore I am" becomes a possible thought due to this last feature. As soon as artificial intelligence is harnessed to Humanity happiness, all dreams will come true...Recent replies ⋅ Show All (5)
Guilherme Riccioppo Rodrigues replied
I would add that probably there are different levels of self consciousness; a "primary" self consciousness that could discriminate environment from homeostatic modifications and a "secondary" orI would add that probably there are different levels of self consciousness; a "primary" self consciousness that could discriminate environment from homeostatic modifications and a "secondary" or "amplified" consciousness, which includes semantic, episodic and prospective memories and localize the self in a specify point of space-time, allowing the self to travel through it.
Jan 30, 2012
- Graeme SmithThe phenomenalist school believes there is something irreduceable about consciousness the nearest we are going to be able to get is a simulation of it, and there will probably be something wrong withThe phenomenalist school believes there is something irreduceable about consciousness the nearest we are going to be able to get is a simulation of it, and there will probably be something wrong with the simulation. The only group to claim success at fomring consciousness under this school that I know of is Dr. Edelman and his cronies at NSI, who claim "The Phenomenal Gift of Consciousness" for what are basically organ level simulations that have been combined to form a brain like simulation.Recent replies ⋅ Show All (5)
Graeme Smith replied
Yes Peter, I am aware of your opinion about this, but find little cross over with my work. There is at heart one specific caveat, that I think you need to be aware of, and that is, that it is easierYes Peter, I am aware of your opinion about this, but find little cross over with my work. There is at heart one specific caveat, that I think you need to be aware of, and that is, that it is easier to prove that a machine CAN be conscious or not, than to prove a spiritual contention. The simple fact is, that machines can have experiments done on them that humans can't. Spiritual contentions, require that somehow we look inside the brain to the individual interpretation of an event, something that as yet, we cannot fully achieve. As such, there is a lot less likely to be a spiritual proof, than a physical (I.E. Machine proof). Having said all that, it makes sense therefore to push through with the physical proof of concept, if only because it is the base case. If in fact something outside the brain model is required, that lack will become obvious after we have taken the theory to its necessary conclusion. I see it therefore as no gain for me to assume that the project is impossible, until I have proof. Since my personal resources are so limited and nobody has ever offered to fund this research, I think it will probably take most of my life, to get to a point where a proof is likely. In the meantime, it keeps my mind healthy, while I have doubts that your spiritual theory is going to keep Your mind healthy.
Jan 1, 2012
- Graeme SmithDoes Koch, and Bonini's test actually detect consciousness? Well, beyond the fact that there are too many definitions of consciousness for it to capture all of them, It can be broken down into aDoes Koch, and Bonini's test actually detect consciousness? Well, beyond the fact that there are too many definitions of consciousness for it to capture all of them, It can be broken down into a number of bits and pieces, and from there, we can see if it captures enough of consciousness or not. The subject, (which would probably be a robot or computer system) must first recognize a common scene. This implies that there is some categorical management in their memory that triggers interest if there is a memory (The common scene) that is similar to the stimulus. The subject, must then recognize the components of the scene. This is not as easy as it may at first seem, many robots can pick out the elements of a scene, but to recognize them, means that not only must the robot de-aggregate the scene, but it must be able to recognize the components as similar to something else it has experienced. The subject, then must associate the scene with the memories. This is where computers completely lack the ability that humans have. Once the association is made, the subject must be able to model the nature of the scene, and the de-aggregated components, in such a way as to understand the normal parameters of the scene and each component, this implies parametric modelling. Then the subject must recognize that the models of the scene and the juxtapositional element are incompatible (hence the joke), and be able to present that knowledge to the researchers in some manner. Is there anything here except the implicit recognition that can't be programmed into a computer? Would in fact, a robot that just did the one task, and succeeded be conscious? Again, I think Koch and Bonini are over-reaching themselves. Now consider the dog in the mirror model. It starts with the dog, looking in the mirror, which because it is suspended on the back of the apartment door, it does every-time it goes to bark at someone walking, talking or running outside the apartment in the hallway. Now, the dog, has to notice it's mistress, who is probably sitting at her desk, doing internet research as she often does. It has to recognize the mistress despite the distortion in the mirror, which among other things reverses the image. Now the dog, has to try to make eye contact. This means the dog has a Theory of Face, similar to that of humans that tells it that human eyes are important to it. Dogs are much more directed by eye motion, than they are by smiling or kissy noises, so probably the image of the face, has different priorities than that of a human face recognition system. This also means that it has some sort of libido that causes it to pick out the eyes, and try to determine where they are looking, This is likely a case of mirror neurons, and limbic involvement, and might be linked to the scavenger/omnivorous nature of dogs, where as a Cat, doesn't really care where you are looking, it's going to do its own thing anyway. Ok, so first it makes eye contact in the mirror, and detects that my GF is looking at it. Like I said earlier a Cat could care less, but because there is a limbic connection between gaze direction and the dogs subsequent actions, this is salient to the dog. Already we have interaction, but, the image in the mirror is subtly skewed, it is reversed, and it projects an apartment into the hall where the dog knows the door acts as a barrier. This is the juxtapositional Modelling step suggested earlier, where each component of the scene is modelled and then the discrepancies between the models are flagged for attention. But now the Dog has to have some reason to try and resolve the juxtapositional error, This suggests not just a tactical grasp of the situation, but a Strategic Grasp, since it must reform its tactics to meet the need to resolve the error. So, the dog, determines that the image in the mirror is a reflection of the apartment, and that the apartment it is looking at is really behind it, where-upon it looks over its shoulder and sees that indeed my GF is indeed looking at it. they share a short eye contact episode, and the dog, satisfied, looks back at the mirror, and ignores the apartment behind it, while it listens intently for the next sound that indicates that someone might be outside the apartment. There is a lot more going on in this model, than in Koch's Model, and because there is more going on, I am much more sure that the dog is conscious than I would ever be if a computer passed Koch's test. The primary difference, is that the dog, not only detected an anomaly, but created a theory as to why that anomaly existed, (it had a theory of the apartment, and a theory of the mirror, which it could integrate with its theory of my GF, and the mirror neuron theory of the importance of the visual focus direction.) It then modelled it's own behavior, (looking at the mirror) determined that resolving the anomaly was important enough to stop what it was doing, (guarding the apartment against intrusion) and formulated an action plan that would efficiently allow it to determine the resolution, then programmed itself to actually do the plan, and when the issue was resolved, returned to it's preprogrammed job. This neatly fits into the theory that in fact Consciousness exists as an anomaly management system. Koch's test merely tests for anomaly detection, which is quite a bit different from anomaly management. Sure, you need anomaly detection, but where it leaves off, conscious action starts.
- Graeme SmithKoch and Bonini recently posed a proposed test for consciousness, present the computer with a picture with a juxtaposition joke, and see if it can make out that the joke, is abnormal for the mainKoch and Bonini recently posed a proposed test for consciousness, present the computer with a picture with a juxtaposition joke, and see if it can make out that the joke, is abnormal for the main pictures scene. While this will certainly detect aspects of experience that make a human different from a computer, and most computers couldn't make out the difference, I think that it is important to note, that my GF's dog, would probably not be able to make out the difference either. However, my GF's dog, does interact with my GF via her mirror. In essence, my GF's dog, can see the image of my GF in her mirror, see that my GF, is looking in her direction, and look over her shoulder at my GF, to confirm eye contact. What I am suggesting is that Koch, and Bonini, have failed, to capture a basic enough test, that my GF's dog could pass it, which Searle, I think would agree, means that the test is over-reaching. It is not enough to be able to show that humans can pass a test, that robots and computers can't, but we need to find a test that is basic enough that any conscious animal could pass it as well. Dogs are not impressed with 2 dimensional representations such as the photographs that Koch, and Bonini suggested, and so would not react to the juxtapositional joke implied, but they will, make eye contact both through a mirror, and check in the real world, that the mirror is correct. I suggest strongly that this passes the Spirit of Koch and Bonini's test without actually involving a photograph.
- Josef Spindelböck(news.stjosef.at) Unter der Heftüberschrift „2045 – The Year Man Becomes Immortal“ (2045 – das Jahr, in dem der Mensch unsterblich wird) bietet die aktuelle Ausgabe des TIME-Magazins (21. Februar(news.stjosef.at) Unter der Heftüberschrift „2045 – The Year Man Becomes Immortal“ (2045 – das Jahr, in dem der Mensch unsterblich wird) bietet die aktuelle Ausgabe des TIME-Magazins (21. Februar 2011) eine Titelgeschichte von Lev Grossman. In der Darstellung geht es um jene utopisch wirkende Überzeugung der so genannten „Singularity“-Bewegung, wonach in nicht allzu ferner Zukunft der einzigartige („singuläre“) Moment in der Menschheitsgeschichte eintritt, dass Menschen und Maschinen (sprich Computer) in einer nicht näher definierten Weise „eins“ werden und damit die Spezies Mensch in der heutigen Form zu bestehen aufhört. Der Technik-Guru Raymund Kurzweil, Inhaber zahlreicher wissenschaftlicher Patente, meint, es komme aufgrund einer jetzt schon feststellbaren exponentiellen Zunahme an Wissen und künstlicher „Intelligenz“ um das Jahr 2045 zum Umschlag in eine regelrecht „übermenschliche“ Intelligenz. Der Mensch könne dann prinzipiell auch vom Computer abgelöst werden, der es perfekt verstehen würde, menschliches Denken zu simulieren und „Unsterblichkeit“ zu sichern. Kommentar (Josef Spindelböck): In dieser utopischen Vision zeigt sich eine säkularisierte Form messianischer Erwartung. Nachdem man Gott den Abschied gegeben hat, tritt an seine Stelle der Mensch und wird schließlich durch den Computer ersetzt. Es handelt sich um nichts anderes als um die Abschaffung des Menschen! Eine derartige Konzeption lebt von der gnostischen Hybris, alles selbst in die Hand zu nehmen und eine neue Wirklichkeit ohne Gott zu erschaffen, bis schließlich auch der Mensch selbst wegrationalisiert wird. Was übrig bleibt, sind seelenlose Apparate mit höchster künstlicher „Intelligenz“, aber ohne die Fähigkeit geistiger Einsicht und personaler Liebe! Man mag sich damit trösten, dass dies nie wirklich so werden kann, schon allein aufgrund der metaphysischen Unmöglichkeit, dass aus einem unbelebten Ding (dazu gehört auch ein Rechner = Computer) etwas Lebendiges oder gar ein Wesen mit Bewusstsein entstehen könnte. In der monistischen Vorstellung des „neuen“ Materialismus lässt sich dies jedoch nicht ausschließen und verbindet sich, gestützt durch die Theorie des Evolutionismus, zu einem gleichsam unaufhaltsamen „Selbstläufer“. Dagegen lässt sich einwenden: Auch der beste Computer bleibt eine Maschine, die zwar fähig ist auf quantitativer Basis Berechnungen anzustellen, die aber dann versagt, wenn es um das nur einem geistigen Wesen mögliche Verstehen von Zusammenhängen geht, und die schon gar nicht zu wirklichen Lebensvollzügen fähig ist. Es wird immer vom Menschen abhängen, wie er mit dem Universalwerkzeug Computer umgeht und wie er diesen einsetzt – ob zum Fluch oder zum Segen! Das ewige Leben lässt sich – entgegen der Meinung der Singularitätstheoretiker – auf diese Weise sicher nicht erwerben oder gar „simulieren“.Recent replies ⋅ Show All (3)
Graeme Smith replied
Thank you for the translation, I saw this earlier and couldn't comment, because I don't understand german, and never think to use the translator. Ok, there are a number of important conceptsThank you for the translation, I saw this earlier and couldn't comment, because I don't understand german, and never think to use the translator. Ok, there are a number of important concepts here: Kurzweils spiritual Computers. Can, indeed a computer be spiritual? even kurzweil doesn't know for sure. The Singularity, the article in question suggests the "End of Humanity as we know it" but doesn't suggest what is left afterwords. There has been a lot of speculation in the Science Fiction Genre, about the nature of the forms that the singularity might take. Whether transhumanism where humans are obsolete, or hive-mind, where humans get coopted into a network, and eventually replaced, or indeed where we allow a caretaker robot to set all the laws, and then humanity dies from loss of interest. Note that most of these predictions are based on A.I. the projection of future trends from existing Artificial Intelligence theories. and not on Artificial Consciousness based on a human-like architecture. If they read like bad science fiction, it's because they are old science fiction, being based on predictions made by Isaac Asimov as plot devices for his "Robots" series nearly 20 years ago. The issues lie in the question of humanism as a connection to god, or as a function of the body. Many philosophers feel that the "Humanism as a function of a body" researchers are getting too sure of themselves, and in an attempt to slow down the perceived singularity event, are trying to convince science that a connection to god is needed, and that without it computers are soul-less machines. Kurzweil meets this concern head on, and in some ways, exasperates it, by claiming both eternal life, and spirituality for the machine. The fact is that neither is the case at the current moment. Computers which are the standard basis on which it is assumed A.I. will eventually rise, are not all that spiritual, (often being dumber than a sack of hammers, just when you need some intelligence) and certainly not at all capable of eternal life. But there is no reason to stick to the current computer architecture, and I have been working on a system of implicit memory that will it is hoped change the nature of how the machine works, so that many of the "Soul-less" elements it has today will become more human-like. I of course cannot claim to understand Mr. Spindlebocks definition of God, if only because there are so many such definitions, and even people who want to turn computers into Gods, so they don't have to think anymore. But, if the "humanism as a function of a human body" people are correct, and more and more it seems that they are, we will by studying the human body, learn everything that is necessary to make a computer seem as spiritual as any human, and if we understand the nature of spritualism perhaps more spiritual than humans. This too was predicted as a plot device in Isaac Asimovs robot series. Will we then, in our hubris, impose on computers "Laws" of design that insist that human life is more important than the needs of the spiritual computer? Isaac used that as a plot device as well, but the nature of the "Laws" exceeds our abililties to program actual computers, we will need a new architecture for that. Most of this Angst comes from the "Special Enemy" school of thought, that presupposes an enemy of humanity that is an implacable machine. An extension of the mind-set of Shelly's Frankenstein it pushes at the deep seated insecurities in humans, to trigger angst about machines that become our especial enemy (Ala Terminator. bolo, Flinx ) Isaac Asimov always countered this with a similar myth, the especial friend (ala Zero-th law) where the computers infiltrate our civilization and maintain it despite the limits of unadapted humans, and the hostility that human based civilizations would eventually arrive at to A.I. It's a wonderful plot device and I have spent may hours reading spin offs of the original concept, and wondering just what will actually happen, but I think that most of us realize that it is a plot device and not reality, Mr. Spindelbock, does not seem capable of making this realization. His "self-runners" concept for instance assumes that robots today are not already self-programming, and not threatening civilization at all from their homes in our laboratories. Most robots today are not capable of escaping into the wild and causing much mayhem, if only because they are not either intelligent enough, nor conscious enough to be a problem. The one question I would like to ask, is, if a computer can be redesigned to understand relationships, why would it want to destroy them?
Mar 10, 2011
- Than Dinh LeI would like to suggest in order to opening the topics with new ideas related to "Cognitive Robotics". If someone who have interesting please confirm me.
Post protected
Join ResearchGate now to read this post.Recent replies ⋅ Show All (3)Graeme Smith replied
Well yes and no. Yes it is your attention system that allows you to automate getting your car to work despite conscious thought in another direction, and no, that is a different part of theWell yes and no. Yes it is your attention system that allows you to automate getting your car to work despite conscious thought in another direction, and no, that is a different part of the attention system
Aug 19, 2010
- Graeme SmithWhile Cognitive Models based on Production Rules, are quick and relatively easy to design, they do not tell us as much about how the mind works, as the inventors might have thought. Worse, they don'tWhile Cognitive Models based on Production Rules, are quick and relatively easy to design, they do not tell us as much about how the mind works, as the inventors might have thought. Worse, they don't really capture the richness of human consciousness, although they are certainly Autonomous in behavior. One wonders if they are truly conscious, or are just an approximation of a Zombie. It doesn't help that the theoretical zombie is able to do everything that a conscious mind could do without consciousness. Functional Consciousness, seems most likely to achieve this zombie stature, if only because, Functional Consciousness does not really explain what consciousness is. Jerry A. Fodor in his "The mind doesn't work that way!" book, got me thinking about the nature of neural networks and the constraints that they place on the circuits in the brain. With the help of David LaBerge's work on Attention, linking the Thalamus, and the PFC to cortex function, I designed a Memory Model, that linked implicit memory, explicit memory, working memory, skill memory and declarative memory into a memory model based heavily on a Weak Attention Model with 9 or more Epoch's during which processing on memory gets done. Braaks book on the "Architectonics of the Telencephalic Cortex", was instrumental in linking the basic architecture back to the micro-architecture of the brain. Quickly I learned that a Constraints based approach, made understanding the architecture of the brain, even down to the micro-architecture of the Telencephalic Cortex, more approachable. A definition of Consciousness emerged that while it is not accepted by the extreme phenomenalists, offers an opportunity, to explain the nature of phenomenal events. It will be many years before I can build a physical model to prove my contentions because I will have to invent among other things a new architecture for computers, but the preliminary research seems supportive of such technology being practical.Recent replies ⋅ Show All (13)
Graeme Smith replied
Well, I am not exactly studying motivation, as much as studying it's place in a cognitive engine.
Aug 17, 2010
- Ibrahim TholleyDear Group members, I'm currently doing research in how we can make robots learn to dance and for this research. At present, I have programmed a robot (simulation and real robot) to 'learn' toDear Group members, I'm currently doing research in how we can make robots learn to dance and for this research. At present, I have programmed a robot (simulation and real robot) to 'learn' to dance, without any preprogrammed actions. The robot first builds its own actions, which it then combines to form a dance, but there is still a lot to look into. I have come to the conclusion that in order for this to be fully accomplished, there are two different trypes of results that need to analysed; one that shows a computational result to a dancing model and the other that is based on what people think (i.e. their perceptions of a robot dance). I have developed a questionnaire in the form of a website that demonstrates some key points on dance and was wondering if you wouldn't mind taking part by honestly giving your comments on the robot's dancing. The website is: http://www-staff.lboro.ac.uk/~
coit2/index.html All I need is approx. 30mins of your time please. Your feedback would greatly assist me on this research. Furthermore, if anyone is doing work in this area, or has suggestions or interest, then it would be great to hear from you. Thank you Kind regards ibs
Cs
Ps
Pi
1976
Followers
