Mike: No reply. We did. You didn't listen. That's why you're getting the dismissive attitude that has you so frustrated. When a working system is built we'll be able to show you which ones we settled on. Until then, we can all see that there's a wide range of choices, and you can't see any. This isn't going to change no matter how much either side argues their points. Can't we please drop it already? We get that you think we're all idiots. We disagree with you. You have no evidence at all to back up your claim. We have no evidence at all that is satisfactory according to your ludicrous standards. But we are actually busy learning what will work and what won't by *trying*, so we'll eventually find out the real answer, regardless of who is right. Until then, there's no further point in this conversation. If you persist, all I can say is, have fun preaching from your armchair. No one is going to listen because you have never tested or honed your theory with real world experience.
-- Sent from my Palm Pre On Oct 26, 2012 3:11 PM, Mike Tintner <[email protected]> wrote: Todor: my thinking machine is supposed to categorize them and give them a name without supervision, also to design any "infinitely varied blah-blah never ending" new kinds of chairs, I even have explained how it works. However I just note for the ones who does get generalizing - they will seem as "radically new, novel elements, blah-blah" for ones who don't understand generalization. In essence, their "chair-defining" properties would be the same, otherwise apparently they wouldn't be recognizable as chairs, except in - as mentined - arbitrary labeled sets, which are randomly labeled. Challenged to explain what are the uniform blocks - or elements – or semantic net units - of a chair, Todor, PM and Aaron, have all responded in the same “essential” way. “My machine will do it” “My program will do it – it’ll learn how”. How will the machine/program do it? What are the blocks they will identify? No reply. .Instead: How could I possibly be so stupid as not to understand what they have said? So cussed? So have-to-be-right? What are the blocks again? Oh that Mike, he has to be right, he can’t listen – there’s no point in explaining anything to him. What did you say were the blocks? God this man is impossible... Etc. etc.. It’s quite possible that never in history has there been a would-be (but not actually) creative field that has been populated by such a phenomenal percentage of bullshitters. It’s not just you guys – it runs through the field, and extends to the highest echelons present and past. If there are blocks, or elements, to a chair – or any concept, period - you can draw them. You ain’t drawing them. To paraphrase Feynman: if you can’t explain it simply, you almost certainly don’t understand it. When you try and draw the blocks, if you ever try and draw them, you’ll realise there are none. (And you might just begin to understand the formal beauty of blobs and patchworks). From: Todor Arnaudov Sent: Friday, October 26, 2012 8:13 PM To: AGI Subject: Re: [agi] The Fundamental Misunderstanding in AGI [was Superficiality] @Aaron, >Just give up, Todor. It doesn't matter if you're right or wrong. He's taken a stance that won't let him think he's wrong, even when he is. He'll re- >interpret what you've said to something else so he can pick it apart. He'll change the subject when he can't come up with a good argument. He'll >directly contradict the truth by denying you've done/said something you just did, or by relabeling it as something entirely different. He doesn't want >to find the truth; he wants to convince you he's already found it. He doesn't realize that there's a trade-off between feeling right and being right -- that >the humility to recognize your own failures and shortcomings as such is the very thing that makes you able to overcome them. It's more important to >him to look right than to actually be it. Thanks for the advice... You are right, and I always give up, then give another shot. :X I guess I'll give up in this iteration in a bit... :) Mike:>Wherever you look in the natural world, you DON’T see uniform blocks – you see groups of irregular, individual BLOBS – indeed PATCHWORKS of blobs – >rocks, earth, cells, tissues, living bodies, they’re all made of blobs, not blocks.. Todor: Have you ever heard of "particle physics", atoms, protons, neutrons, electrons, quantum physics, quants. A "blocks" means a "building blocks". Yes, in nature there are more of spheres than rectangles or cubes, it's a simpler form, requiring less variables to define, but it's a "building block either". The first *known* hypothesis for the atoms is made thousands of years ago, it's possible that people thought that tens of thousands years ago. What physics discovers is that world is ultimately made exactly of uniform "blocks", having absolutely identical properties. Mike:>Science always makes mathematically simplified models of the world. What you’re arguing is equivalent to saying Newton’s calculus proved that curves are made >of rectangles. They’re not – that’s a useful and brilliant formal simplification – in fact, translation. Todor: BTW, you're repeating a philosopher I've been arguing a decade ago, he was blaming science for "simplifying things" - well if you can understand things and can predict their behavior, you "simplify" them. If you cannot understand, you take is a whole and just copy it. Yet the ones who don't understand it (and can't simplify it in a workable fashion) cannot really understand the "simple things". Please design for me a simple digital computer, say equivalent of an Apple][, or not that complex, let it be like Whirlwind. Those kind of "simplified" concepts are simple only in theory, and if taken out of their complex and real application. Yes, "curves" and integrals are simplifications, the real world is made of bosons and fermions, of 10E+3894938493 particles - so how should you process it like that? How do you process electrons or quarks, and does your retina or brain knew initially what an electron is? "Curves" are just images on your retina, or sequences of neuronal activities when touching objects etc., and maths allows if you know some measures of those curves (in terms of the sensory matrices) to predict (compute) some other measures, which are not directly observable by the regular sensory matrix. Mike: >Rational technology INVENTED uniform blocks – it was one of the greatest, most imaginative inventions in history – and you “block-headsâ€? think they >*discovered* them - bricks, for example, and perfect rectangles, you think, *always* existed. Nonsense. Todor: Your notion of "imaginative" is pathetic, perhaps you think this is "the most imaginative art" are the rectangles of Malevic:http://en.wikipedia.org/wiki/File:Malevich.black-square.jpg The uniform blocks, especially triangles or rectangles are the less imaginative (the easiest, the simplest - I've explained this before), and Universe is built of uniform blocks, nature at larger scale, seen with naked eye either: insects (see a bunch of ants from the above), their eyes; bees's cells, nails, hair, grass; sand (!), leaves, fingers, eyes, trees in a forest; animals from the same species - see a flight of sparrows from a distance, or a pack of wolves, or look at the stars or the planets; humans either. Connect 4 points and you get a "rectangle", try to build a house with the fewest possible number of walls/parts and you get first a triangle, then make a compromise and use 4, which will give you twice the area for a small overhead - and there you go, that's a rectangle/tetragon. Why trying to do it with the fewest number of elements? Because it's a hard work, you get tired, have to hunt, to walk around in a search for food etc. That's why one would search for the simplest solution at the moment. Is it clear now how it can be "invented" from scratch, how that "new element" can be created... Of course, for humans who cannot generalize it, that's not rectangle, these are just "trees", they are "not uniform", "don't have exact dimensions" (in pre-stone age - right) etc., You reason so, because you cannot get the meaning of "generalization" and "resolution". For example, let's see "the greatest invention" - the wheel. Well, who said it's so great? It's absolutely trivial! Have you ever looked at the sky? In the stone age there wasn't TV or light bulbs, so you were supposed to see the moon and learn what a circle is from the very early age. In fact you learned even without the moon - have you seen your mother's eyes? They are circular, they have several circles, they are even spherical. Humans have probably seen the eyes of dead mammals either. The moon is a "wheel", the eyes are wheels, they roll. The trees are also circular and can roll. There's nothing ingenious in "inventing" the wheel. The problem is that modern people usually have a hard time thinking in terms of the context of the inventions - first they don't realize all those examples of wheels, and second they don't realize that the wheels require roads. How would you use a wheel for travelling, if you're surrounded by a dense forest and all around is muddy? For the wheel to make sense, roads are needed, or a dry place with flat areas where the wheels/cut stems of trees could roll. In case of trees - you must have technology and energy to cut trees that are big enough, and to slice or "delve" them to make wheels, or to have technology to dry and bend wood and connect pieces in order to make discus etc., and also have the energy to cut enough trees to make a road etc. I.e. the wheel itself is not the problem, the other technological difficulties are. Also you must have horses or other strong animals (first they had to be domesticated) to drive such heavy clumsy and probably not quite circular wheels (having a lot of friction), otherwise they are useless for transportation - not surprising that the first wheels were used for making pottery(?). Mike>The problem of AGI is always – if we’re talking formally, quasi-mathematically – to deal with blobs and patchworks of blobs. >But obviously, this kind of discussion is too high-falutin’ without examples, and specific analyses. Todor: A bullshit, I've analyzed your "blobs" too, and give more than explicit examples and analyses. Mike:>The general point is: you have not and will not show any uniform blocks to underlie those chairs, or the concept of CHAIR – or **any concept period**. In fact I already have shown the uniform blocks of the concept of "chair", regarding your last "blobs" - sure - those are not chairs - "coma". They are "chairs" only in your head, because you say they are chairs or because you have copied them from images labeled as "chairs". In fact it's just 2D black-white image - "exclamation sign", you can label any thing as a "chair", in fact it's just an image, in this case the rest is arbitrary label. The "blobs" on that particular image can be represented exactly in the resolution of this image as very simple functions - that means if using those functions the same looking "blobs" can be redrawn at another place, and by adjusting or transforming their free variables they can be drawn as bigger, smaller, inclined, rotated etc. If one see those and the other "blobs", and asked, he'll recognize them as similar. As of the real (3D) chars - I take the challenge, my thinking machine is supposed to categorize them and give them a name without supervision, also to design any "infinitely varied blah-blah never ending" new kinds of chairs, I even have explained how it works. However I just note for the ones who does get generalizing - they will seem as "radically new, novel elements, blah-blah" for ones who don't understand generalization. In essence, their "chair-defining" properties would be the same, otherwise apparently they wouldn't be recognizable as chairs, except in - as mentined - arbitrary labeled sets, which are randomly labeled. "Concepts" live in your head and in mind, in the real world yes, there are no "concepts", but mind cannot operate mentally with the real world, it can operate only with concepts derived from its sensory matrices space, and the sensory matrices are finite, "rectangular", "circular", "uniform" etc. Much more objective elements are "the particles" and measures based on their properties; if you prefer "wave functions", "number of hydrogen atoms diameters" etc. but unfortunately for your point, these measures seem to be exactly uniform, regular etc. Besides that brain or any system except the original Universe cannot process them at their original resolution. The experience have shown - brain or a machine don't have to model them at their original resolution, a TV with a low resolution or a video of 320x240 or half this is enough to see a lot. The mathematical physical models work ever finer (and their resolution grows, though), i.e. they produce correct predictions. Mike:>If anyone wants to talk specifics and analyse forms as I did, I’m delighted. Vague generalisations alone don’t cut it. You did what??? Yes, "vague" generalization that you can't get don't cut for you, it will "cut" your bullshit when we do create the thinking "terminators" to finish your confusions. (Relax, just the confusions.) --- Todor "Tosh" Arnaudov --- -- Twenkid Research: http://research.twenkid.com -- Self-Improving General Intelligence Conference: http://artificial-mind.blogspot.com/2012/07/news-sigi-2012-1-first-sigi-agi.html -- Todor Arnaudov's Researches Blog: http://artificial-mind.blogspot.com AGI | Archives | Modify Your Subscription AGI | Archives | Modify Your Subscription ------------------------------------------- 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
