Jim, You can diagram each of your examples. That would be a good start. That way we can see how your system differs from any of the plethora of systems out there. Just a thought. ~PM.
Date: Mon, 8 Oct 2012 14:28:30 -0500 From: [email protected] To: [email protected] Subject: Re: [agi] Conceptual Structure? Jim, Well, language is the proverbial tip of the iceberg, concepts live below the level of language. Language just provides labels for easier access to what may be the same or similar concept. They live in the models, which can be represented as systems of relations, evidence evaluation systems, and such. What I'm asking is if you have an idea for the problem of modeling other minds. If everyone had the same concept models, there would be little point in talking to each other, and that is the purpose of language -- coordination (or influence) of action among people who may have different ideas about the same label (that is, the name of a concept). -- Dimitry On 10/8/2012 7:08 AM, Jim Bromer wrote: On Mon, Oct 8, 2012 at 2:31 AM, Dimitry Volfson <[email protected]> wrote: "John believes Mary Loves John." Ok, so as a surface structure relation, that's straightforward. But the deep structure is: Why does John believe that Mary Love John? Is it the way she talks to him, the way she touches him, the way she laughs at his jokes, etc. What evidence does John think means that Mary Loves John. "John believes Mary Loves John" implies John has a model of what Love looks like, and has seen some of that in Mary's relations to him. This model and evidence is the deep structure. And the evidence that produced the model is an even deeper structure. Did he get that model by watching his parents' behavior toward each other, or watching TV shows, or reading romance novels, or what his friends told him, etc.? There is no right answer to what is "love" -- it's a vague concept that different people can disagree on without a definitive answer. And many human concepts are like that (loyalty, bravery, cowardice, morality, etc). Dimitry, You are coming up with common answers to a question about a human experience but then you are claiming that because the reasons for the affect are vague and people disagree with the answers and there is no definitive answer then a network of conceptual structure won't work to produce AGI or something. OK, there are no definitive answers. Why would you think that is a problem? Do you believe that AGI has to be based on universally held truths or something? That is a historically regressive point of view and it never really was a sensible foundation for a model for human-like intelligence. I agree with you that intelligence is the ability to gather insights about a concept. So your example is ok. Your real criticism is based on the fact that while a structural analysis in linguistics can find a particular simple transformation for a statement, the reasons for the experience denoted by the statement are multiple and must be drawn from many experiences and from contemplation and from education. Let me answer one criticism that was implicit in your remarks. While the ability to learn from education is a mark of higher intelligence, the argument that a computer program is not therefore able to learn through some sort of education must be based on the assumption that computers are not capable of intelligence. This is obviously not a good reason to conclude that AGI is imposible or that the conceptual network and conceptual structure which I mentioned is not strong enough to produce intelligence. The contemporary complication is due to the fact that certain basic principles of AGI are elusive. That is, an automated program is able to derive some valid insights about the world, but it is missing so many foundational principles that even the simplest structures cannot be maintained. Yes that might be due to a lack of a method to discover basic truths, but it also might be due to an overly parsimonius methodology which simply will not provide the program with the ability to derive enough possibilties to build on. Perhaps the basis for natural intelligence might be better likened to a beaver dam than a metropolis. The conceptual network and conceptual structure theories would provide a mechanism to hold a variety of reasons and insights related to a concept that is being considered, and these insights would tend to be distributed. There is a problem with getting computers to explore multiple possible relations that build on multiple possible relations and I call that the complexity problem. However, this problem does not prove that the conceptual network and conceptual structure, as I am talking about it, is wrong. Jim Bromer On Mon, Oct 8, 2012 at 2:31 AM, Dimitry Volfson <[email protected]> wrote: Jim, For diagrams: Visio, or MindJet. Apart from that, are you considering deep structure relations, as opposed to surface structure (language). For Example, from the YKY "Concept Composition Logic" paper: "John believes Mary Loves John." Ok, so as a surface structure relation, that's straightforward. But the deep structure is: Why does John believe that Mary Love John? Is it the way she talks to him, the way she touches him, the way she laughs at his jokes, etc. What evidence does John think means that Mary Loves John. "John believes Mary Loves John" implies John has a model of what Love looks like, and has seen some of that in Mary's relations to him. This model and evidence is the deep structure. And the evidence that produced the model is an even deeper structure. Did he get that model by watching his parents' behavior toward each other, or watching TV shows, or reading romance novels, or what his friends told him, etc.? There is no right answer to what is "love" -- it's a vague concept that different people can disagree on without a definitive answer. And many human concepts are like that (loyalty, bravery, cowardice, morality, etc). On 10/7/2012 7:56 PM, Jim Bromer wrote: I don't have a diagram and I would not know how to draw one. The idea is that a great deal of information can be related to different kinds of concepts that might be considered central to some idea. I have figured out a way that I should be able to experiment with the idea using concrete examples expressed with simple language. This experiment will not be a true AGI program but it should allow me to see if the structural conceptual networks idea is feasible as a way to represent an AGI program. On Sun, Oct 7, 2012 at 7:14 PM, Piaget Modeler <[email protected]> wrote: Diagrams Jim, Diagrams. ~PM Date: Sun, 7 Oct 2012 18:07:38 -0400 Subject: Re: [agi] Conceptual Structure? From: [email protected] To: [email protected] I forgot about conceptual structure itself. Conceptual structure is based on the idea that structure in language is vital to understanding language, and that structure in ideas must also be understood to understand the ideas. For instance temporal structure is often important and so is positional structure. But when you think about it these two kinds of relationships are only concepts. While they seem to have a wide application to many different kinds of things they are still only concepts. This shows that concepts may play different kinds of roles when used with other concepts. This insight seems obvious to me but it also seems obviously important. If you can find that certain concepts can take on the role of an abstracting or generalizing agent then doesn't this imply that other concepts might also take on roles that go beyond their surface characteristics? For example, the position of an object is what it is. To recognize that position and relative position might be used to create highly generalized principles that have advanced mankind's understanding of matter and technology is to recognize that a seemingly dull feature of a concept can be used as an agent of insight. So then I am saying that by exploring the roles and structures of concepts I expect to find other activating principles of insight that may have eluded us so far. Jim Bromer AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription ____________________________________________________________ Refinance for 1.750%/2.926% APR Loans under 729K usually qualify for US GOV backed refinance programs theeasyloansite.com AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription ____________________________________________________________ Woman is 57 But Looks 27 Mom publishes simple facelift trick that angered doctors... ConsumerLifestyles.org ------------------------------------------- AGI Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-c97d2393 Modify Your Subscription: https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-2484a968 Powered by Listbox: http://www.listbox.com
