Todor Arnaudov <[email protected]> wrote: I'm sick of this. Email lists are not friendly, they work for announcement and short Q/A-trouble-shootings-organization, those here are interrogation sessions - annoying, clumsy, lacking context, protracted. =====================================================
Well, they don't lack context. Some of the discussions here are useful to me because trying to figure out how someone is wrong can actually help me to form better responses and it also clarifies my thinking for me. Of course, I sometimes realize that one of my concepts is wrong and will make changes as I go. I am sorry that you are sickened by this. I beleive that the nature of the email lists is representing something about the way human beings work when no one knows how to get the job done. Of course, in a job where everything is decided by engineering (social, buisness or technical) the job specification is much easier and the only problems are in assigning the jobs. You might do better in job like that. My point of view is that AGI is very difficult and I don't say that I have all the answers. Almost all the people I talk to seem to be saying that either they have the secret sauce or they are joining the club where the authorities do. Well, I recognize that I don't have it all figured out and if you think you do or that someone else does then the best way to show us how it is done is by doing it. --------------------------------------------------------------------------------------------- Todor Arnaudov <[email protected]> wrote: Regarding the embodiment thing that's so obvious to B., A., or me, to you it's not... jUm, I've did already 1000 times, repeating things as for retarded ones. Indeed this "enlightening" is one of the sick things in those email lists of strangers, and it's hard to enlighten dark matter. ================================================= Naturally you didn't explain why embodiment is necessary for AI, instead you made a dim remark that it is obvious to B, A and you and that having explained it 1000 times, repeating it for the retarded ones, is sick because it is hard to enlighten dark matter. These kinds of remarks really are sloppy and deserve no respect. I think you might be happier if you tried to restrain yourself from making comments like these. I am going to assume that the "embodiment" thing that you mentioned has something to do with robotics. A computer processor does not get magic data from visual, audio, tactile and pproprioceptive sensors. It is just data to the processor (or electronic signals). It is true that such sensors would provide an AGI program with more data to work with but there is no evidence that higher levels of AGI have been achieved by working with robotics. If the embodiment was necessary for AGI theory was correct then we would expect to see robotic based AGI surpassing every other avenue of effort. While there have been some successes, like automated vehicles, they have been mostly based on specialized devices and a patch work of specialized programs that only work in a very narrow situation. Norvig and Thrun's online course had an example of something one of them wrote (worked on) that successfully piloted the car as it drove down a road with some kind of embankments (I think it was.) The program was definitely a narrow kind of AI because it seemed to work just because the embankment was for the most part well defined. The achievement was more clear when car successfully made it through sections where the road and the embankments were less well distinguished, but it still was not AGI. Most of the milestones in AI, however, did not have anything to do with robotic "embodiment". For example, the winning chess programs and IBM's Watson are definite milestones that were long predicted before they were actually realized and none of these rely on embodiment. One thing that is true, a fundamental truth of AI so to speak, is that an AGI program has to react to the Input-Output data environment. True AGI has to learn based on its interactions with the IO data environment. However, that data environment can be any kind of data environment as long as the data has certain kinds of interrelations that I cannot specify. I could make the attempt, but the present audience does not seem to interested. The idea that I am as dull as you have to claim because I make arguments like this is not credible. But the claim that you are right because I am so obviously wrong is ludicrous. Jim Bromer On Sun, Aug 26, 2012 at 7:48 PM, Todor Arnaudov <[email protected]> wrote: > Mike, jUm, > > >Tudor, > >You made a criticism about us and then you proceeded to do the very same > thing that you had criticized us for. Is it actually possible that you > were completely unaware that you did that, or, did you feel that >your > accusation and presentation of your case somehow transcended the > constraints on the behavior that you are so intolerant of. > > jUm, I was completely aware, and that was an illustration of this *, I'm > sick of this. Email lists are not friendly, they work for announcement and > short Q/A-trouble-shootings-organization, those here are interrogation > sessions - annoying, clumsy, lacking context, protracted. > > > >There was little substantial in your message but there was a lot of the > same kind of garbage that you were complaining about. > > But obviously you didn't get the message, even though it's screaming. > Regarding the embodiment thing that's so obvious to B., A., or me, to you > it's not. B., you must be very desperate to discuss it with logicians etc. > who don't get even this. > > > The fact is that Boris, Sergio, I and Adam and other people have been > talking about codes and models. > > Great, then next time you might make a substantial comment about it that > would enlighten us. > > jUm, I've did already 1000 times, repeating things as for retarded ones. > Indeed this "enlightening" is one of the sick things in those email lists > of strangers, and it's hard to enlighten dark matter. > > One knows something, others don't get it at all, one reason is that have > 1/10 of one's background or demonstrated cognitive capacities in general > domains, yet everybody believes he's the smartest one - even if the other > one is better than him in 20 domains in which the first one is absolutely > incapable and don't have a clue. No, AGI is something... different. Your > skills and demonstrated intellectual superiority in a, b, c, d, e, f, g, h > doesn't matter. It's ... different! :)) The reason is - everybody believes > he's the center of the Universe. > > One talks about neuroscience, others don't have a clue, but they believe > they understand it better. > > One has published papers and works that give more details about a point, a > lot of it, that answer all "creative" questions others ask - no one cares > or spent time to read it, ask the questions and wants personal answer, like > from a personal tutor, because the other one believes he's the best - he > should be "convinced", as some say - "convince me that you're right". > > Convincing is wrong - a court thing. And because you believe you're > smarter or know more than the rest, with no evidences. > > Everyone believes the other one is of lower rank to him, he doesn't > deserve respect and time (the confusing use of pronouns is deliberate) to > read his papers. > > "Enlightening" of the dark matter - a very little success. > > > >Todor: Embodiment is just coordinate spaces, interactivity and > modalities - not a > mystery<http://artificial-mind.blogspot.com/2011/12/embodiment-is-just-coordinate-spaces.html> > > >How do “coordinate spaces, interactivity and modalities “ *feel* and > have a unified sense of their body? > > Mike, didn't you get that qualia is not AGI, and you can't prove whether > anything has or doesn't have qualia, if it's immaterial? > > Well, understanding of some concepts require adequate cognitive > capacities, which if not met - understanding is impossible with any > explanations. > > Like you can't explain to a severely retarded person what "money" are - he > sees every piece of paper as "a piece of paper" - he can't read, can't > recognize digits, can't recognize images on the paper. > Let me tell you a story about him. > > How can? - Joe says. - You give "a piece of paper" to someone and then you > receive food or a candy, or something else, and the other person is even > smiling at you? > Why didn't she ate the food himself? What she would do with those papers? > He put the piece in that drawer... Why didn't she burn it at least? It's > more fun! > > Or look! That man puts that piece of cardboard there, moves it, it makes > "pi-pi-pi" sound, and they give him food! > > There's no reason! > > All those people are crazy! Their minds are scattered! > > > >How can they say “I feel small, like a worm”..”I feel tall like a > giant”.. “I feel like I’m on air”... ? > > Very easily, ask Mr. Kazachenko, he know also and would enjoy explaining > this to you, too. :) > > > > >How do they know how to imitate – create an embodied impression of - > Michael Caine, Tom Cruise, John Travolta and other figures, just by > watching a few seconds of film footage?.. > > Do you know how a tape recorder works? In computers and this case it's the > same, but the recording and the mapping has a vast amount of "channels" for > all the parameters, and first it has to recognize the mapping. I've talked > about this already. > > >and then create new imitations of these figures uttering new words and > making new moves? > > Alter some of the data in some of the channels. > > >How, essentially, one can ask: do the *parts* of a body form a unified > *whole*? > > They don't, the unified whole is an illusion as is decided from the top to > bottom, the limits mind has set to decide whether something is a part or > not. > I.e. that's a deliberative decision. "Wholes" are abstractions, for > example there are no molecules, but other "things" whose by-effects are > producing effects, which are useful to be accessed as "wholes" for > particular needs. > > >To science and our culture this is one of the greatest of mysteries. To > you – like everything else in AGI – apparently not. > > Qualia is not AGI. And "to the science and culture", dude, most of the > highest ranking dudes are not the highest IQ dudes, neither are the most > versatile. And the widely accepted believes etc. are often ones which are > accessible by an average-mind. If they are too complex or just impossible > to grasp for the capacities of an average high-ranked or whatever mind (in > given domain), they won't be "to our science and our culture", it would be > to "their science and their culture - of those weirdos, that no one > understands, because they are cranks, crazy". > > Read the story about the money again, or read the novel "The planet of the > apes" and watch the reaction of the apes when the main characters explain > that they fell down from space with a spaceship. That's impossible! We, the > apes, are the top of the nature! OK, you are. > > The "common conceptions" are ones which are accessible by some > IQ/cognitive capacity. > > > > But your ideas about embodiment – like almost all of your long posts – are > a set of wildly scattered pieces of ideas that do not cohere as a whole. > [Check others’ responses if you think I’m just being insulting]. > > Mike, what's "scattered" is your cognition, the "discoherence" is the > fault of your narrow working memory. You even confuse generalization and > differentiation and don't get elementary things (to me) explained like for > a retarded. And see you on the AGI products market - that's the real way > and place to prove an idea is working or not. Not in the "court". > > > -- > *--- 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 <https://www.listbox.com/member/archive/303/=now> > <https://www.listbox.com/member/archive/rss/303/10561250-164650b2> | > Modify<https://www.listbox.com/member/?&;>Your Subscription > <http://www.listbox.com> > ------------------------------------------- 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
