[agi] [WAS The Smushaby] The Logic of Creativity
Richard, Good. Glad you're replied. Let's have a go. Perhaps there are misunderstandings here - because you see if programming creative idea-hopping were trivially easy, as you suggest, even just in principle, you should have no problems designing a program that will make you a billionaire. Nothing is easy about what you did - for either AI or AGI. And no one in AGI has ever attempted creative problems.Perhaps you can show me wrong. 1. THE CENTRAL ISSUE - I suggest, to put it v. v. broadly at first, is this: *are there general logical procedures that can tackle creative problems, esp. high-level creative problems, ( like that of the engram)?. you say : yes, in principle easy; I say: impossible.(and actually worse - wrong even in principle) If there are, you should be able to show the merest outline of either: a) a logical procedure underlying your own creative thinking as discussed, OR b) a *normative* logical procedure for creative problems generally I'm going to make that challenge as reasonable and accessible as possible. for discussion, so let's define the two key terms here further 2. GENERAL LOGICAL PROCEDURES. I focus on a logical procedure to make this simple. A logical procedure should actually be capable of elaboration into a program, which can eventually determine every step and stage of an agent's thinking about a creative problem. But I'm not asking for a whole program, just the merest outline or even just a central element of the logical procedure at its heart. To clear up possible confusions, I take logic v. v broadly to incl not just the more formal kind, but. all the different types of program Jim B listed (and from what I've heard) yours, Ben's and, Pei's. Logic is basically a set of rules about how to combine certain objects. And a logical problem involves following those rules in a structured, inevitable way to construct new forms from those objects. So if A leads to B, and B leads to C then you can proceed logically and inevitably to conclude that A leads to C. Logic is continually being revolutionised and sophisticated, but none of the revolutions actually transcend logic. So those objects can now be fuzzy, probable, uncertain, complex and even comprise whole programs. And the rules can be evolutionary so that the objects and even the rules themselves keep changing. But however sophisticated it becomes, logic is still *structured, formal thinking. In the end, at any given time, A always leads to C, (in the appropriate kind of logic), even if's only with a probability of 0.4 Logic is completely opposed to illogic - *unstructured,* *free-form* thinking. A, above, cannot lead to A - because well it looks good. C cannot lead to A - because well I want to break the rules, and A cannot lead to AC - because well they look prettier that way. Those are illogical forms of thinking. False thinking. Wrong. Logic is not anything goes, no rules apply thinking. (I trust you'll agree). 2B. EXAMPLE OF LOGICAL PROCEDURE. So here's an example of the sort of thing I'm looking for, (and even just an element will do) - my first, very cackhanded, (so make allowances) attempt to define the central logical procedure in GA's:. 1.Take a set of known candidate solutions for a problem - (or,say candidate causes for an effect-to-be-explained) 2.Combine those solutions/causes according to certain rules. 3.Test the resulting solutions/causes to see which come closest to actually explaining the problem/effect. 4.Select the best subset, and recombine - possibly according to new rules, which will be recombinations of the old rules. 5.Test the resulting solutions and so on , until you arrive at a solution that actually works, or, a cause that really starts to explain the effect By all means redefine my ham attempt, but hopefully we can agree that that is in essence a general logical procedure, capable of elaboration into a comprehensive program. And it LOOKS at first sight, as if it might have promise for creativity - even if it requires work. (There have been a whole mix n'match family of AGI theories and psychological theories of creativity that resemble GA's - all inspired by the basic idea -take existing ideas about the creative problem and mix'em up a bit. - an idea which many clearly find appealing. Your idea as best I could understand it a while ago, involved taking whole sets of alternative candidate programs, and not just ideas) So I ask for just an outline or even just an element of a procedure like the above. But I also ask that you at least begin to show me how it applies to a particular creative problem or two. Anyone can produce convincing logical arguments in abstraction - we can prove logically that a hare can never overtake a tortoise who has a head start in a race. But as soon as you begin to test that logic - and apply it to real hare/tortoise races or similar, it's obviously nonsense..Science and technology demand that you test
Re: [agi] [WAS The Smushaby] The Logic of Creativity
I am reluctant to say this, but I am not sure if I actually understand what Mike is getting at. He described a number of logical (in the greater sense of being reasonable and structured) methods by which one could achieve some procedural goal, and then he declares that logic (in this greater sense that I believe acknowledged) was incapable of achieving it. Let's take a flying house. I have to say that there was a very great chance that I misunderstood what Mike was saying, since I believe that he effectively said that a computer program, using logically derived systems could not come to the point where it could creatively draw a picture of a flying house like a child might. If that was what he was saying then it is very strange. Obviously, one could program a computer to draw a flying house. So right away, his point must have been under stated, because that means that a computer program using computer logic (somewhere within this greater sense of the term) could follow a program designed to get it to draw a flying house. So right away, Mike's challenge can't be taken seriously. If we can use logical design to get the computer program to draw a flying house, we can find more creative ways to get it to the same point. Do you understand what I am saying? You aren't actually going to challenge me to write a rather insipid program that will draw a flying house for you are you? You accept the statement that I could do that if I wanted to right? If you do accept that statement, then you should be able to accept the fact that I could also write a more elaborate computer program to do the same thing, only it might, for example, do so only after the words house and flying were input. I think you understand that I could write a slightly more elaborate computer program to do the something like that. Ok, now I could keep making it more complicated and eventually I could get to the point where where it could take parts of pictures that it was exposed to and draw them in more creative combinations. If it was exposed to pictures of airplanes flying, and if it was exposed to pictures of houses, it might,. through quasi random experimentation try drawing a picture of the airplane flying past the house as if the house was an immense mountain, and then it might try some clouds as landscaping for the house and then it might try a cloud with a driveway, garbage can and a chimney, and eventually it might even draw a picture of a house with wings. All I need to do that is to use some shape detecting algorithms that have been developed for graphics programs and are used all the time by graphic artists that can approximately determine the shape of the house and airplane in the different pictures and then it would just be a matter of time before it could (and would) try to draw a flying house. Which step do you doubt, or did I completely misunderstand you? 1. I could (I hope I don't have to) write a program that could draw a flying house. 2. I could make it slightly more elaborate so, for example, that it would only draw the flying house if the words 'house' and 'flying' were input. 3. I could vary the program in many other ways. Now suppose that I showed you one of these programs. After that I could make it more complicated so that it went through a slightly more creative process than the program you saw the previous time. 4. I could continue to make the program more and more complicated. I could, (with a lot of graphics techniques that I know about but haven't actually mastered) write the program so that if it was exposed to pictures of houses and to pictures of flying, would have the ability to eventually draw a picture of a flying house (along with a lot of other creative efforts that you have not) even thought of. But the thing is, that I can do this without using advanced AGI techniques! So, I must retain the recognition that I may not have been able to understand you because what you are saying is not totally reasonable to me. Jim Bromer --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=126863270-d7b0b0 Powered by Listbox: http://www.listbox.com
Re: [agi] [WAS The Smushaby] The Logic of Creativity
I think what Mike is saying is that I could draw what I think a flying house would look like, and you could look at my picture and say it was a flying house, even though neither of us has ever seen one. Therefore, AGI should be able to solve the same kind of problems, and why aren't we designing and testing AGI this way? But don't worry about it. Mike doesn't know how to solve the problem either. -- Matt Mahoney, matmaho...@yahoo.com --- On Tue, 1/13/09, Jim Bromer jimbro...@gmail.com wrote: From: Jim Bromer jimbro...@gmail.com Subject: Re: [agi] [WAS The Smushaby] The Logic of Creativity To: agi@v2.listbox.com Date: Tuesday, January 13, 2009, 3:02 PM I am reluctant to say this, but I am not sure if I actually understand what Mike is getting at. He described a number of logical (in the greater sense of being reasonable and structured) methods by which one could achieve some procedural goal, and then he declares that logic (in this greater sense that I believe acknowledged) was incapable of achieving it. Let's take a flying house. I have to say that there was a very great chance that I misunderstood what Mike was saying, since I believe that he effectively said that a computer program, using logically derived systems could not come to the point where it could creatively draw a picture of a flying house like a child might. If that was what he was saying then it is very strange. Obviously, one could program a computer to draw a flying house. So right away, his point must have been under stated, because that means that a computer program using computer logic (somewhere within this greater sense of the term) could follow a program designed to get it to draw a flying house. So right away, Mike's challenge can't be taken seriously. If we can use logical design to get the computer program to draw a flying house, we can find more creative ways to get it to the same point. Do you understand what I am saying? You aren't actually going to challenge me to write a rather insipid program that will draw a flying house for you are you? You accept the statement that I could do that if I wanted to right? If you do accept that statement, then you should be able to accept the fact that I could also write a more elaborate computer program to do the same thing, only it might, for example, do so only after the words house and flying were input. I think you understand that I could write a slightly more elaborate computer program to do the something like that. Ok, now I could keep making it more complicated and eventually I could get to the point where where it could take parts of pictures that it was exposed to and draw them in more creative combinations. If it was exposed to pictures of airplanes flying, and if it was exposed to pictures of houses, it might,. through quasi random experimentation try drawing a picture of the airplane flying past the house as if the house was an immense mountain, and then it might try some clouds as landscaping for the house and then it might try a cloud with a driveway, garbage can and a chimney, and eventually it might even draw a picture of a house with wings. All I need to do that is to use some shape detecting algorithms that have been developed for graphics programs and are used all the time by graphic artists that can approximately determine the shape of the house and airplane in the different pictures and then it would just be a matter of time before it could (and would) try to draw a flying house. Which step do you doubt, or did I completely misunderstand you? 1. I could (I hope I don't have to) write a program that could draw a flying house. 2. I could make it slightly more elaborate so, for example, that it would only draw the flying house if the words 'house' and 'flying' were input. 3. I could vary the program in many other ways. Now suppose that I showed you one of these programs. After that I could make it more complicated so that it went through a slightly more creative process than the program you saw the previous time. 4. I could continue to make the program more and more complicated. I could, (with a lot of graphics techniques that I know about but haven't actually mastered) write the program so that if it was exposed to pictures of houses and to pictures of flying, would have the ability to eventually draw a picture of a flying house (along with a lot of other creative efforts that you have not) even thought of. But the thing is, that I can do this without using advanced AGI techniques! So, I must retain the recognition that I may not have been able to understand you because what you are saying is not totally reasonable to me. Jim Bromer --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your
Re: [agi] [WAS The Smushaby] The Logic of Creativity
Hi Jim Yes you have misunderstood the nature of a creative problem, but no criticism - I clearly need to spell this out v carefully - because I get that particular misunderstanding over and over from programmers. [I'll just answer briefly BTW because as I said, I want to do a much fuller, systematic argument.another time] The central part [though not the only part of this and every creative] problem for you and your computer and almost anyone is this: you have *no domain knowledge* of, and no rules for, flying houses. oh yes, you know about the domain of flying creatures say like birds, or flying planes, and rules that apply to them, and you know a lot about the domain of houses.and rules that apply to them But you have no domain knowledge - no semantic network - no rules about how to put the two together (or whether they do go together, or aren't a contradiction in terms] - or what the result should look like, or how it should function. Overall you have a v. *incomplete domain set.* And just to make life multiply difficult you don't have criteria of what a good flying house is. What IS a good flying house, Jim? And what makes one flying house better than another, because you're going to have to make such judgments, as you consider alternative possiblities - flat roof, sloped roof, cottage roof? Or could it be shaped like a chair? Any rules against flying chair houses? And yet, you, a human being, given that problem, which you have never seen before, and have no experience of, will have v. little difficulty crossing those two previously uncrossed domains and making up and drawing a flying house. (Try drawing one now) Although you will, as you go along, probably wonder and agonise if only for a few moments about a lot of things. Wings? Helicopter blades? Jets? Parachute? Balloon? And you'll agonise precisely because there is NO RULEBOOK about flying houses. Anywhere. Honest Jim. And there CAN'T be a definitive rulebook. That's life. Not logic. Now there is no computer anywhere in the world that can do what you can do - which is having an incomplete domain set, and incomplete set of rules, proceed to construct something in an altogether new domain, and make up the rules as you go. That's the problem for - and whole challenge of - AGI. - You're kind of illustrating my central thesis of creative block - you find it extremely difficult to concentrate on the central creative challenge of AGI. What's happening, I suggest, is this: faced with that creative challenge, you say: hey if you just show me a flying house, and tell me how to draw one,, I can program a computer to do that Yes you can. I know you can. That's called narrow AI. That's your computer basically being given the answer to the problem before it starts and not having to find one by itself. But that's what you guys have done all your lives - been given the answer to the problem, and then you just work out the details of how a computer can execute it.. That's narrow AI. This is AGI - the computer has to work out the answer, and the domains, and the rules, all by itself, WITHOUT sufficent prior instructions, WITHOUT a nice, complete rulebook - starting with at best a v. sketchy idea of what to do. Just like you. (You don't even don't know fully how to program - i.e. you're still learning - and yet you do it). --- Oh and just to answer Matt - if you want to keep doing narrow AI, like everyone else, then he's right - don't worry about it. Pretend it doesn't exist. Compress things :). But if you want to do real AGI, then yes worry about it. A lot. Because it's fun. And a far more exciting challenge. (Like I said, those other creative problems, which are unquestionably real rather than fantasy, and very abundant, have the same basic form). I am reluctant to say this, but I am not sure if I actually understand what Mike is getting at. He described a number of logical (in the greater sense of being reasonable and structured) methods by which one could achieve some procedural goal, and then he declares that logic (in this greater sense that I believe acknowledged) was incapable of achieving it. Let's take a flying house. I have to say that there was a very great chance that I misunderstood what Mike was saying, since I believe that he effectively said that a computer program, using logically derived systems could not come to the point where it could creatively draw a picture of a flying house like a child might. If that was what he was saying then it is very strange. Obviously, one could program a computer to draw a flying house. So right away, his point must have been under stated, because that means that a computer program using computer logic (somewhere within this greater sense of the term) could follow a program designed to get it to draw a flying house. So right away, Mike's challenge can't be taken seriously. If we can use logical
Re: [agi] [WAS The Smushaby] The Logic of Creativity
The notion of drawing a flying house based on knowledge of flying and houses is covered well in the psych literature on conceptual blending, which many AI theorists is paid attention to. Mechanisms for blending are specifically incorporated in the OpenCogPrime design, we just haven't gotten to implementing and testing them yet, because they rely on other things that are still being worked on. It is true that no current AI system is very good at conceptual blending. That doesn't mean that AGI theorists haven't thought deeply about the topic and don't have good ideas for addressing it. Adequate technology for implementing AGI has only very recently become available (or if I'm overoptimistic, may not yet quite be available), and working out the details of complex AGI designs via a combination of theory and experimentation just takes time and hard work. ben g On Tue, Jan 13, 2009 at 4:40 PM, Mike Tintner tint...@blueyonder.co.uk wrote: Hi Jim Yes you have misunderstood the nature of a creative problem, but no criticism - I clearly need to spell this out v carefully - because I get that particular misunderstanding over and over from programmers. [I'll just answer briefly BTW because as I said, I want to do a much fuller, systematic argument.another time] The central part [though not the only part of this and every creative] problem for you and your computer and almost anyone is this: you have *no domain knowledge* of, and no rules for, flying houses. oh yes, you know about the domain of flying creatures say like birds, or flying planes, and rules that apply to them, and you know a lot about the domain of houses.and rules that apply to them But you have no domain knowledge - no semantic network - no rules about how to put the two together (or whether they do go together, or aren't a contradiction in terms] - or what the result should look like, or how it should function. Overall you have a v. *incomplete domain set.* And just to make life multiply difficult you don't have criteria of what a good flying house is. What IS a good flying house, Jim? And what makes one flying house better than another, because you're going to have to make such judgments, as you consider alternative possiblities - flat roof, sloped roof, cottage roof? Or could it be shaped like a chair? Any rules against flying chair houses? And yet, you, a human being, given that problem, which you have never seen before, and have no experience of, will have v. little difficulty crossing those two previously uncrossed domains and making up and drawing a flying house. (Try drawing one now) Although you will, as you go along, probably wonder and agonise if only for a few moments about a lot of things. Wings? Helicopter blades? Jets? Parachute? Balloon? And you'll agonise precisely because there is NO RULEBOOK about flying houses. Anywhere. Honest Jim. And there CAN'T be a definitive rulebook. That's life. Not logic. Now there is no computer anywhere in the world that can do what you can do - which is having an incomplete domain set, and incomplete set of rules, proceed to construct something in an altogether new domain, and make up the rules as you go. That's the problem for - and whole challenge of - AGI. - You're kind of illustrating my central thesis of creative block - you find it extremely difficult to concentrate on the central creative challenge of AGI. What's happening, I suggest, is this: faced with that creative challenge, you say: hey if you just show me a flying house, and tell me how to draw one,, I can program a computer to do that Yes you can. I know you can. That's called narrow AI. That's your computer basically being given the answer to the problem before it starts and not having to find one by itself. But that's what you guys have done all your lives - been given the answer to the problem, and then you just work out the details of how a computer can execute it.. That's narrow AI. This is AGI - the computer has to work out the answer, and the domains, and the rules, all by itself, WITHOUT sufficent prior instructions, WITHOUT a nice, complete rulebook - starting with at best a v. sketchy idea of what to do. Just like you. (You don't even don't know fully how to program - i.e. you're still learning - and yet you do it). --- Oh and just to answer Matt - if you want to keep doing narrow AI, like everyone else, then he's right - don't worry about it. Pretend it doesn't exist. Compress things :). But if you want to do real AGI, then yes worry about it. A lot. Because it's fun. And a far more exciting challenge. (Like I said, those other creative problems, which are unquestionably real rather than fantasy, and very abundant, have the same basic form). I am reluctant to say this, but I am not sure if I actually understand what Mike is getting at. He described a number of logical (in the greater sense of being reasonable
Re: [agi] [WAS The Smushaby] The Logic of Creativity
--- On Tue, 1/13/09, Mike Tintner tint...@blueyonder.co.uk wrote: Oh and just to answer Matt - if you want to keep doing narrow AI, like everyone else, then he's right - don't worry about it. Pretend it doesn't exist. Compress things :). Now, Mike, it is actually a simple problem. 1. Collect about 10^8 random photos (about what we see in a lifetime). 2. Label all the ones of houses, and all the ones of things flying. 3. Train an image recognition system (a hierarchical neural network, probably 3-5 layers, 10^7 neurons, 10^11 connections) to detect these two features. You'll need about 10^19 CPU operations, or about a month on a 1000 CPU cluster. 4. Invert the network by iteratively drawing images that activate these two features and work down the hierarchy. (Should be faster than step 3). When you are done, you will have a picture of a flying house. Let me know if you have any trouble implementing this. And BTW the first 2 steps are done. http://images.google.com/images?q=flying+houseum=1ie=UTF-8sa=Xoi=image_result_groupresnum=5ct=title -- Matt Mahoney, matmaho...@yahoo.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=126863270-d7b0b0 Powered by Listbox: http://www.listbox.com
Re: [agi] [WAS The Smushaby] The Logic of Creativity
Jim, One more thing. This Smushaby of Flatway. It's creative - a creative - arrgh I don't know the word for it - it's not a full pun, nevertheless a novel double, if not quadruple entendre. - the sort of thing that Matt's statistical NLP simply won't get at all - and a fascinating insight into how the mind produces creative analogies, (and true AGI). Your first analogy/ thesis, as I understand, was that all these AGI approaches were actually producing rubbish/wrong answers which could be compared analogically with smush that had to be cleared away, and this was because they were following the Flat way, which can be interpreted as partly physically flat and/or partly musically flat/unpleasant sounding, which in turn are analogical/metaphorical for the limited nature of those AGI approaches. Actually I suggest you may have got that a bit wrong. It should perhaps have been Narrow Way (although Flatway *sounds* aurally better, which may be why you picked that instead) . The reason is that this first analogy obviously implies another counterpointed phrase/ analogy, namely The Lullaby of Broadway, which stands for sweet music, and that is presumably meant as an analogy for the creative harmonies you get when you follow the broad way (as opposed to narrow or flat way) re AGI approaches - (an open range of approaches rather than a restricted one). So that was a very creative, if not quite fully achieved, quadruple entendre - crossing and compressing (quite differently to Matt's sense) several domains in a novel way - the kind that literary critics love, can write whole chapters on., but are utterly alien to logic - and which, I suggest, was all there, give or take, in your brain, conscious or not. And thats reveal one of the keys, (ho ho - geddit -Lullaby) to AGI - namely, using concepts in an open broad way, so that they can embrace MULTIPLE SIMULTANEOUS MEANINGS and thus cross-associate multiple domains. Language's polysemy is central to AGI and domain-crossing. . But it's the complete reverse of logic and rationality and what Matt is doing. Logic could neither have produced such creative ambiguities/connections, nor understood them, nor evaluated them. In logic B = B, it has only one meaning, and YKY wouldn't know what to do if it didn't. In logic, a concept can have only one meaning. But in creative thinking and problemsolving, a concept can have multiple cross-associated meanings simultaneously - a B can in principle be a letter, a vitamin, an animal, a grade and a verb all at the same time. And that is central to being creative. Don't worry about it Matt. Just B a B. --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=126863270-d7b0b0 Powered by Listbox: http://www.listbox.com
Re: [agi] [WAS The Smushaby] The Logic of Creativity
Matt, Well little Matt, as your class teacher, in one sense this is quite clever of you. But you see, little Matt, when I gave you and the class that exercise, the idea was for you to show me what *you* could do - what you could produce from your own brain. I didn't mean you to copy someone else's flying house from a textbook. That's cheating Matt, - getting someone else to do the work for you - and we don't like cheats do we? So perhaps you can go away and draw a flying house all by yourself - a superduper one with lots of fabbo new bits that no one has ever drawn before, and all kinds of wonderful bells and whistles, that will be ten times better than that silly old foto. I know you can Matt, I have faith in you. And I know if you really, really try, you can understand the difference between creating your own drawing, and copying someone else's. Because, well frankly, Matt, every time I give you an exercise - ask you to write an essay, or tell me a story in your own words - you always, always copy from other people, even if you try to disguise it by copying from several people. Now that's not fair, is it Matt? That's not the American way. You have to get over this lack of confidence in yourself. Matt/Mike Tintner tint...@blueyonder.co.uk wrote: Oh and just to answer Matt - if you want to keep doing narrow AI, like everyone else, then he's right - don't worry about it. Pretend it doesn't exist. Compress things :). Now, Mike, it is actually a simple problem. 1. Collect about 10^8 random photos (about what we see in a lifetime). 2. Label all the ones of houses, and all the ones of things flying. 3. Train an image recognition system (a hierarchical neural network, probably 3-5 layers, 10^7 neurons, 10^11 connections) to detect these two features. You'll need about 10^19 CPU operations, or about a month on a 1000 CPU cluster. 4. Invert the network by iteratively drawing images that activate these two features and work down the hierarchy. (Should be faster than step 3). When you are done, you will have a picture of a flying house. Let me know if you have any trouble implementing this. And BTW the first 2 steps are done. http://images.google.com/images?q=flying+houseum=1ie=UTF-8sa=Xoi=image_result_groupresnum=5ct=title -- Matt Mahoney, matmaho...@yahoo.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=126863270-d7b0b0 Powered by Listbox: http://www.listbox.com
Re: [agi] [WAS The Smushaby] The Logic of Creativity
Mike, it's not cheating. It's called research :-) -- Matt Mahoney, matmaho...@yahoo.com --- On Tue, 1/13/09, Mike Tintner tint...@blueyonder.co.uk wrote: From: Mike Tintner tint...@blueyonder.co.uk Subject: Re: [agi] [WAS The Smushaby] The Logic of Creativity To: agi@v2.listbox.com Date: Tuesday, January 13, 2009, 7:38 PM Matt, Well little Matt, as your class teacher, in one sense this is quite clever of you. But you see, little Matt, when I gave you and the class that exercise, the idea was for you to show me what *you* could do - what you could produce from your own brain. I didn't mean you to copy someone else's flying house from a textbook. That's cheating Matt, - getting someone else to do the work for you - and we don't like cheats do we? So perhaps you can go away and draw a flying house all by yourself - a superduper one with lots of fabbo new bits that no one has ever drawn before, and all kinds of wonderful bells and whistles, that will be ten times better than that silly old foto. I know you can Matt, I have faith in you. And I know if you really, really try, you can understand the difference between creating your own drawing, and copying someone else's. Because, well frankly, Matt, every time I give you an exercise - ask you to write an essay, or tell me a story in your own words - you always, always copy from other people, even if you try to disguise it by copying from several people. Now that's not fair, is it Matt? That's not the American way. You have to get over this lack of confidence in yourself. Matt/Mike Tintner tint...@blueyonder.co.uk wrote: Oh and just to answer Matt - if you want to keep doing narrow AI, like everyone else, then he's right - don't worry about it. Pretend it doesn't exist. Compress things :). Now, Mike, it is actually a simple problem. 1. Collect about 10^8 random photos (about what we see in a lifetime). 2. Label all the ones of houses, and all the ones of things flying. 3. Train an image recognition system (a hierarchical neural network, probably 3-5 layers, 10^7 neurons, 10^11 connections) to detect these two features. You'll need about 10^19 CPU operations, or about a month on a 1000 CPU cluster. 4. Invert the network by iteratively drawing images that activate these two features and work down the hierarchy. (Should be faster than step 3). When you are done, you will have a picture of a flying house. Let me know if you have any trouble implementing this. And BTW the first 2 steps are done. http://images.google.com/images?q=flying+houseum=1ie=UTF-8sa=Xoi=image_result_groupresnum=5ct=title -- Matt Mahoney, matmaho...@yahoo.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=126863270-d7b0b0 Powered by Listbox: http://www.listbox.com
