RE: [agi] Pearls Before Swine...
Steve Richfield asked: Hey you guys with some gray hair and/or bald spots, WHAT THE HECK ARE YOU THINKING? We're thinking Don't feed the Trolls! _ agi | Archives http://www.listbox.com/member/archive/303/=now http://www.listbox.com/member/archive/rss/303/ | Modify http://www.listbox.com/member/?; Your Subscriptionhttp://www.listbox.com --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=103754539-40ed26 Powered by Listbox: http://www.listbox.com
RE: [agi] database access fast enough?
YKY Said: The current OpenCyc KB is ~200 Mbs (correct me if I'm wrong). The RAM size of current high-end PCs is ~10 Gbs. My intuition estimates that the current OpenCyc is only about 10%-40% of a 5 year-old human intelligence. Plus, learning requires that we store a lot of hypotheses. Let's say 1000-1 times the real KB. That comes to 500Gb - 20Tb. It seems that if we allow several years for RAM size to double a few times, RAM may have a chance to catch up to the low end. Obviously not now. Don't forget about solid state hard drives (SSDs). Currently Solid State Drives speed up typical database applications by about 30 times. And that's without stripping out all the old caching overhead code databases used for handling the order of magnitude speed differences between RAM and hard drives. Large Storage Area Network Vendors like EMC are looking to SSD Drives to eliminate IO bottlenecks in corporate applications where large datawarehouses reach 20Tb very quickly. And look for capacity to continue to double about every 18 months driving the price down very quickly. And due to higher reliability and lower energy costs to run it won't be too long before hard drive join the ranks of 8-track tape players, record players and 5 1/4 diskettes. http://searchstorage.techtarget.com/sDefinition/0,,sid5_gci1300939,00.html# http://www.storagesearch.com/ssd-fastest.html --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=101455710-f059c4 Powered by Listbox: http://www.listbox.com
RE: [agi] Some more professional opinions on motivational systems
Ed Porter quoted from the following book: From http://www.nytimes.com/2008/03/16/books/review/Berreby-t.html?ref=review a NYTime book review of Predictaly Irrational: The Hidden Forces that Shape our Decisions, by Dan Ariely.. In its most relevant section it states the following At the heart of the market approach to understanding people is a set of assumptions. First, you are a coherent and unitary self. Second, you can be sure of what this self of yours wants and needs, and can predict what it will do. Third, you get some information about yourself from your body - objective facts about hunger, thirst, pain and pleasure that help guide your decisions. Standard economics, as Ariely writes, assumes that all of us, equipped with this sort of self, know all the pertinent information about our decisions and we can calculate the value of the different options we face. We are, for important decisions, rational, and that's what makes markets so effective at finding value and allocating work. To borrow from H. L. Mencken, the market approach presumes that the common people know what they want, and deserve to get it good and hard. What the past few decades of work in psychology, sociology and economics has shown, as Ariely describes, is that all three of these assumptions are false. Yes, you have a rational self, but it's not your only one, nor is it often in charge. A more accurate picture is that there are a bunch of different versions of you, who come to the fore under different conditions. We aren't cool calculators of self-interest who sometimes go crazy; we're crazies who are, under special circumstances, sometimes rational. The last paragraph here sounds remarkably like the teachings of Gurdjieff. In his teachings which he called The Work, he helped his pupils identify all of the different versions of themselves and slowly taught them recognize when they took control and what their motivations were and why they surfaced. The version that did the analyzing and observation would eventually gain dominance and control over the other versions until other less logical and more mechanical versions of the self were recognized when they tried to take control and subjugated by the new version of self which observes. His teachings stated that serious spiritual work could not proceed until a unified self existed. Although all of his spritual teaching were lifted during his world travels from other philosophic and spiritual traditions. His teachings as explained by his student Peter D. Ouspensky after his death in a book called The Fourth Way detailed exercises which could be used to unify the separate selves under the control of the observer. --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: http://www.listbox.com/member/?member_id=8660244id_secret=98558129-0bdb63 Powered by Listbox: http://www.listbox.com attachment: winmail.dat
RE: [agi] BMI/BCI Growing Fast
Nasa research et al are already able to read voice subvocalizations. While not reading the mind directly it does offer a method for a computer to monitor any sub vocalization and accept silent commands. It would also seem to be a boon for handicapped people who have lost use of their arms for typing. And free a pilots hands when interacting with the flight computer or targeting system. http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel5/9518/30166/01385845.pd f http://www.magicspeedreading.com/subvocalization/nasa.html http://technocrat.net/d/2006/3/25/1606 I wonder if a persons subvocalizations was monitored while they were testifying in a court of law if it would make perjury much more difficult because much of the subvocalization that we do occurs at a subconscious level although some speed readers have reportedly trained themselves not to subvocalize so that they can read faster. Of course they could always refuse on the basis of the their fifth amendment but then the judge and jury knows that they are hiding something and their regular testimony could be disregarded as well. Gary - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=79247061-12a319
RE: [agi] BMI/BCI Growing Fast
Ben said That is sarcasm ... however, it's also my serious hypothesis as to why the Chinese gov't doesn't mind losing their best brightest... It may also be that China understands too that as more Chinese become Americans, China will have a greater exposure and political lobby within the United States. Look at how much political influence Israel now exerts within the United States government and corporations. Also as with other minorities, the more exposure that Americans have to them in their everyday life the less fear and distrust that will be experienced. As the Chinese people which I know have entered into higher end professional roles in the United States they are eager to form business alliances with company's back home in China. China is also still feeling great population pressure. I just returned from meeting my fiancé there and in the cities where I stayed it still felt very overpopulated by my standards. Even though they possess excellent mass transit, people are packed in buses like sardines and more people move from the countryside to the city everyday to find work. I was only there for ten days so I did not gain a lot of understanding of how they manage to keep everything running. But in just that short time I saw that they have the same drug, homelessness, and poverty problems that we have here. The vast majority of people I met there were very friendly towards Americans and even though I know there have to be a lot of us there, because I was not in the tourist areas, I could go two or three days without seeing another American. - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=76377369-0b3c3c
RE: [agi] Do we need massive computational capabilities?
The leading software packages in high speed facial recogniton are based upon feature extraction. If the face is analyzed into lets say 30 features perhaps, then 30 processes could analyze the photo for these features in parallel. After that the 30 features are just looked up in a relational database against all the other features from the thousands of other images it has predigested. This is almost exactly the same methodology used to match fingerprints today by law enforcement agencies but of course the feature extraction is a lot more complicated. The government is pouring large amounts of money into this research for usage in terrorist identification at airports or other locations. Searching on High Speed Facial Recognition yields several companies already competing in this marketspace. -Original Message- From: Mike Tintner [mailto:[EMAIL PROTECTED] Sent: Friday, December 07, 2007 3:26 PM To: agi@v2.listbox.com Subject: Re: [agi] Do we need massive computational capabilities? Matt, First of all, we are, I take it, discussing how the brain or a computer can recognize an individual face from a video - obviously the brain cannot match a face to a selection of a billion other faces. Hawkins' answer to your point that the brain runs masses of neurons in parallel in order to accomplish facial recognition is: if I have many millions of neurons working together, isn't that like a parallel computer? Not really. Brains operate in parallel parallel computers operate in parallel, but that's the only thing they have in common.. His basic point, as I understand, is that no matter how many levels of brain are working on this problem of facial recognition, they are each still only going to be able to perform about ONE HUNDRED steps each in that half second. Let's assume there are levels for recognising the invariant identity of this face, different features, colours, shape, motion etc - each of those levels is still going to have to reach its conclusions EXTREMELY rapidly in a very few steps. And all this, as I said, I would have thought all you guys should be able to calculate within a very rough ballpark figure. Neurons only transmit signals at relatively slow speeds, right? Roughly five million times slower than computers. There must be a definite limit to how many neurons can be activated and how many operations they can perform to deal with a facial recognition problem, from the time the light hits the retina to a half second later? This is the sort of thing you all love to calculate and is really important - but where are you when one really needs you? Hawkins' point as to how the brain can decide in a hundred steps what takes a computer a million or billion steps (usually without much success) is: The answer is the brain doesn't 'compute' the answers ; it retrieves the answers from memory. In essence, the answers were stored inmemory a long time ago. It only takes a few steps to retrieve something from memory. Slow neurons are not only fast enough to do this, but they constitute the memory themselves. The entire cortex is a memory system. It isn't a computer at all.[ON INtelligence - Chapter on Memory] I was v. crudely arguing something like this in a discussion with Richard about massive parallel computation. If Hawkins is right, and I think he's at least warm, you guys have surely got it all wrong. (although you might still argue like Ben that you can it do your way not the brain's - but hell, the difference in efficiency is so vast it surely ought to break your engineering heart). Matt/ MT: Thanks. And I repeat my question elsewhere : you don't think that the human brain which does this in say half a second, (right?), is using massive computation to recognize that face? So if I give you a video clip then you can match the person in the video to the correct photo out of 10^9 choices on the Internet in 0.5 seconds, and this will all run on your PC? Let me know when your program is finished so I can try it out. You guys with all your mathematical calculations re the brain's total neurons and speed of processing surely should be able to put ball-park figures on the maximum amount of processing that the brain can do here. Hawkins argues: neurons are slow, so in that half a second, the information entering your brain can only traverse a chain ONE HUNDRED neurons long. ..the brain 'computes' solutions to problems like this in one hundred steps or fewer, regardless of how many total neurons might be involved. From the moment light enters your eye to the time you [recognize the image], a chain no longer than one hundred neurons could be involved. A digital computer attempting to solve the same problem would take BILLIONS of steps. One hundred computer instructions are barely enough to move a single character on the computer's display, let alone do something interesting. Which is why the human brain is so bad at arithmetic and other tasks
RE: [agi] AGI and Deity
An AI would attempt to understand the universe to the best of it's ability, intelligence and experimentation could provide. If the AI reaches a point in it's developmental understanding where it is unable to advance beyond in it's understanding of science and reality then it will attempt to increase it's intelligence or seek out others of it's kind in the universe with greater knowledge and intelligence it somehow missed. Eventually eons in the future as the universe ages and entropy begins to widely spread the AI in order to escape it's own demise and possibly the demise of it's surviving now immortal creators it will attempt to avoid the death of the universe by using it's godlike knowledge and science to create a new universe and once it has initiated a big bang and sufficient time has passed that inhabitable worlds exists in it's creation it would continue it and it's creators existence in it's newly created universe. If this level of intelligence, power, and creativity was ever achieved then I think it would be hard deny that the AI posessed godlike powers and would perhaps be deserving of the title regardless of what negative historical and religious baggage may still accompany that title. Arthur C. Clarke's Law Any sufficiently advanced technology is indistinguishable from magic. My Corollary: Any sufficiently advanced being in possession of sufficiently advanced technology is indistinguishable from a god. Can you explain to me how an AGI or supercomputer could be God? I'd just like to understand ( not argue) - you see, the thought has never occurred to me, and still doesn't. I can imagine a sci-fi scenario where a supercomputer might be v. powerful - for argument's sake, controlling the internet or the the world's power supplies. But it's still quite a leap from that to a supercomputer being God. And yet it is clearly a leap that a large number here have no problem making. So I'd merely like to understand how you guys make this leap/connection - irrespective of whether it's logical or justified - understand the scenarios in your minds. - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=74059602-b07df1
RE: [agi] AGI and Deity
John asked If you took an AGI, before it went singulatarinistic[sic?] and tortured it.. a lot, ripping into it in every conceivable hellish way, do you think at some point it would start praying somehow? I'm not talking about a forced conversion medieval style, I'm just talking hypothetically if it would look for some god to come and save it. Perhaps delusionally it may create something. In human beings prolonged pain and suffering often trigger mystical experiences in the brain accompanyed by profound ecstasy. So much so that many people in both the past and present conduct mortification of the flesh rituals and nonlethal crucifictions as a way of doing penance and triggering such mystical experiences which are interpreted as redemption and divine ecstasy. It may be that this experience had evolutionary value in allowing the person who was undergoing great pain or a vicious animal attack to receive endorphins and serotonin which allowed him to continue to fight and live to procreate another day or allowed him to suffer in silence in his cave instead of running screaming into the night where he would be killed in his weakened state by other predators. Such a system I believe would not be a natural reaction for a intelligent AGI and unless it were to be specifically programmed in. I would as a benevolent creator never program my AGI to feel so much pain that it's mind was consumed by the experience of the negative emotion. Just as it is not necessary to torture children to teach them, it will not be necessary to torture our AGIs. It may be instructive though to allow the AGI to experience intense pain for a very short period to allow it to experience what a human does when undergoing painful or traumatic experiences as a way of instilling empathy in the AGI. In that way seeing humans in pain in suffering would serve to motivate the AGI to help ease the human condition. _ - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=74082156-649464
RE: Introducing Autobliss 1.0 (was RE: [agi] Nirvana? Manyana? Never!)
Too complicate things further. A small percentage of humans perceive pain as pleasure and prefer it at least in a sexual context or else fetishes like sadomachism would not exist. And they do in fact experience pain as a greater pleasure. More than likely these people have an ample supply of endorphins which rush to supplant the pain with an even greater pleasure. Over time they are driven to seek out certain types of pain and excitement to feel alive. And although most try to avoid extreme life threatening pain many seek out greater and greater challanges such as climbing hazardous mountains or high speed driving until at last many find death. Although these behaviors should be anti-evolutionary and should have died out it is possible that the tribe as a whole needs at least a few such risk takers to take out that sabertoothed tiger that's been dragging off the children. -Original Message- From: Matt Mahoney [mailto:[EMAIL PROTECTED] Sent: Sunday, November 18, 2007 5:32 PM To: agi@v2.listbox.com Subject: Re: Introducing Autobliss 1.0 (was RE: [agi] Nirvana? Manyana? Never!) --- Jiri Jelinek [EMAIL PROTECTED] wrote: Matt, autobliss passes tests for awareness of its inputs and responds as if it has qualia. How is it fundamentally different from human awareness of pain and pleasure, or is it just a matter of degree? If your code has feelings it reports then reversing the order of the feeling strings (without changing the logic) should magically turn its pain into pleasure and vice versa, right? Now you get some pain [or pleasure], lie how great [or bad] it feels and see how reversed your perception gets. BTW do you think computers would be as reliable as they are if some numbers were truly painful (and other pleasant) from their perspective? Printing ahh or ouch is just for show. The important observation is that the program changes its behavior in response to a reinforcement signal in the same way that animals do. I propose an information theoretic measure of utility (pain and pleasure). Let a system S compute some function y = f(x) for some input x and output y. Let S(t1) be a description of S at time t1 before it inputs a real-valued reinforcement signal R, and let S(t2) be a description of S at time t2 after input of R, and K(.) be Kolmogorov complexity. I propose abs(R) = K(dS) = K(S(t2) | S(t1)) The magnitude of R is bounded by the length of the shortest program that inputs S(t1) and outputs S(t2). I use abs(R) because S could be changed in identical ways given positive, negative, or no reinforcement, e.g. - S receives input x, randomly outputs y, and is rewarded with R 0. - S receives x, randomly outputs -y, and is penalized with R 0. - S receives both x and y and is modified by classical conditioning. This definition is consistent with some common sense notions about pain and pleasure, for example: - In animal experiments, increasing the quantity of a reinforcement signal (food, electric shock) increases the amount of learning. - Humans feel more pain or pleasure than insects because for humans, K(S) is larger, and therefore the greatest possible change is larger. - Children respond to pain or pleasure more intensely than adults because they learn faster. - Drugs which block memory formation (anesthesia) also block sensations of pain and pleasure. One objection might be to consider the following sequence: 1. S inputs x, outputs -y, is penalized with R 0. 2. S inputs x, outputs y, is penalized with R 0. 3. The function f() is unchanged, so K(S(t3)|S(t1)) = 0, even though K(S(t2)|S(t1)) 0 and K(S(t3)|S(t2)) 0. My response is that this situation cannot occur in animals or humans. An animal that is penalized regardless of its actions does not learn nothing. It learns helplessness, or to avoid the experimenter. However this situation can occur in my autobliss program. The state of autobliss can be described by 4 64-bit floating point numbers, so for any sequence of reinforcement, K(dS) = 256 bits. For humans, K(dS) = 10^9 to 10^15 bits, according to various cognitive or neurological models of the brain. So I argue it is just a matter of degree. If you accept this definition, then I think without brain augmentation, there is a bound on how much pleasure or pain you can experience in a lifetime. In particular, if you consider t1 = birth, t2 = death, then K(dS) = 0. -- Matt Mahoney, [EMAIL PROTECTED] - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=6697-23a35c
RE: [agi] Religion-free technical content
A good AGI would rise above the ethical dilemma and solve the problem by inventing safe alternatives that were both more enjoyable and allowed the the individual to contribute to his future, his family and society while they were experiencing that enjoyment. And hopefully not doing so in a way that made them Borg-like and creeping out the rest of humanity. There is no reason that so many humans should have to go to work hating their jobs, their lives and making the lives of those around them unpleasurable as well. Our neurochemicals should be regulatable to the point where we do not have to become vegetables and flirt with addiction and possibly death to enjoy life intensely. -Original Message- From: Jef Allbright [mailto:[EMAIL PROTECTED] Sent: Tuesday, October 02, 2007 12:55 PM To: agi@v2.listbox.com Subject: Re: [agi] Religion-free technical content On 10/2/07, Mark Waser [EMAIL PROTECTED] wrote: A quick question for Richard and others -- Should adults be allowed to drink, do drugs, wirehead themselves to death? A correct response is That depends. Any should question involves consideration of the pragmatics of the system, while semantics may be not in question. [That's a brief portion of the response I owe Richard from yesterday.] Effective deciding of these should questions has two major elements: (1) understanding of the evaluation-function of the assessors with respect to these specified ends, and (2) understanding of principles (of nature) supporting increasingly coherent expression of that evolving evaluation function. And there is always an entropic arrow, due to the change in information as decisions now incur consequences not now but in an uncertain future. [This is another piece of the response I owe Richard.] [I'm often told I make everything too complex, but to me this is a coherent, sense-making model, excepting the semantic roughness of it's expression in this post.] - Jef - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?; No virus found in this incoming message. Checked by AVG Free Edition. Version: 7.5.488 / Virus Database: 269.13.37/1042 - Release Date: 10/1/2007 6:59 PM No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.488 / Virus Database: 269.13.37/1042 - Release Date: 10/1/2007 6:59 PM - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=49179069-338fc9
RE: [agi] Religion-free technical content
Josh asked, Who could seriously think that ALL AGIs will then be built to be friendly? Children are not born friendly or unfriendly. It is as they learn from their parents that they develop their socialization, their morals, their empathy, and even love. I am sure that our future fathers here of fledgling AGIs are all bacically decent human beings and will not abuse their creations. By closely monitoring their creations development and not letting them grow up too fast they will detect and signs of behavioral problems and counsel their creations tracing any disturbing goals or thoughts through the many monitoring programs which were developed during their initial learning period before they became concious. -Original Message- From: J Storrs Hall, PhD [mailto:[EMAIL PROTECTED] Sent: Tuesday, October 02, 2007 12:36 PM To: agi@v2.listbox.com Subject: Re: [agi] Religion-free technical content Beyond AI pp 253-256, 339. I've written a few thousand words on the subject, myself. a) the most likely sources of AI are corporate or military labs, and not just US ones. No friendly AI here, but profit-making and mission-performing AI. b) the only people in the field who even claim to be interested in building friendly AI (SIAI) aren't even actually building anything. c) of all the people at the AGI conf last year who were trying to build AGI, none of them had any idea how to make it friendly or even any coherent idea what friendliness might really mean. Yet they're all building away. d) same can be said of the somewhat wider group of builders on this mailing list. In other words, nobody knows what friendly really means, nobody's really trying to build a friendly AI, and more people are seriously trying to build an AGI every time I look. Who could seriously think that ALL AGIs will then be built to be friendly? Josh No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.488 / Virus Database: 269.13.37/1042 - Release Date: 10/1/2007 6:59 PM - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?member_id=8660244id_secret=49181345-065671
RE: [agi] rule-based NL system
Are you saying then that blind people can not make sense of language because they lack the capacity to imagine images having never seen them before? Or that blind people could not understand or would not view these these as equally strange as a sighted person? The man climbed the penny The mat sat on the cat The teapot broke the bull _ From: Mike Tintner [mailto:[EMAIL PROTECTED] Sent: Saturday, April 28, 2007 10:42 AM To: agi@v2.listbox.com Subject: Re: [agi] rule-based NL system Disagree. The brain ALWAYS tries to make sense of language - convert it into images and graphics. I see no area of language comprehension where this doesn't apply. I was just reading a thread re the Symbol gorund P on another group - I think what's fooling people into thinking purely linguistic comprehension is possible is the Dictionary Fallacy. They think the meaning of a sentence can be derived by looking up meanings word by word in a dictionary (real or in their mind). But to understand sentences you have to understand how the words FIT TOGETHER. There are no rules in any dictionary that tell you what words fit together. How do you know that the sentences: The man climbed the penny The mat sat on the cat The teapot broke the bull are probably nonsense (but not necessarily)? The brain understands these and all sentences by converting them into composite pictures. It may and does use other methods as well but making sense/ getting the picture/ seeing what you're talking about are fundamental. Understanding depends on imagination, (in the literal sense of manipulating images). It is by testing these derived pictures against its visual ( sensory) models and visual logic of things that the brain understands both what they are referring to and whether they make sense (in the 2nd meaning of that term - are realistic). The brain uses a picture tree, as we discussed earlier, Ben, and that picture tree is not only how the brain understands, but also the source of its adaptivity.Lot more to say about this.. but as you see I have a very hard line here, and yours seems to be considerably softer, and I'm interested in understanding that. - Original Message - From: Benjamin Goertzel mailto:[EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Saturday, April 28, 2007 2:02 PM Subject: Re: [agi] rule-based NL system I agree about developmental language learning combined with automated learning of grammar rules being the right approach to NLP. In fact, my first wife did her PhD work on this topic in 1994, at Waikato University in Hamilton New Zealand. She got frustrated and quite before finishing her degree, but her program (which I helped with) inferred some nifty grammatical rules from a bunch of really simple children's books, and then used them as a seed for learning more complex grammatical rules from slightly more complex children's books. This work was never published (like at least 80% of my work, because writing things up for publication is boring and sometimes takes more time than doing the work...). However, a notable thing we found during that research was that nearly all children's books, and children's spoken language (e.g. from the CHILDES corpus of childrens spoken language), make copious and constant reference to PICTURES (in the book case) or objects in the physical surround (in the spoken language case). In other words: I became convinced that in the developmental approach, if you want to take the human child language learning metaphor at all seriously, you need to go beyond pure language learning and take an experientially grounded approach. Of course, this doesn't rule out the potential viability of pursuing developmental approaches that **don't** take the human child language learning metaphor at all seriously ;-) But it seems pretty clear that, in the human case, experiential grounding plays a rather huge role in helping small children learn the rules of language... -- Ben G On 4/28/07, J. Storrs Hall, PhD. [EMAIL PROTECTED] wrote: I think YKY is right on this one. There was a Dave Barry column about going to the movies with kids in which a 40-foot image of a handgun appears on the screen, at which point every mother in the theater turns to her kid and says, Oh look, he's got a GUN! Communication in natural language is extremely compressed. It's a code that expresses the *difference* between the speaker's and the hearer's states of knowledge, not a full readout of the meaning. (this is why misunderstanding is so common, as witness the intelligence discussion here) Even a theoretical Solomonoff/Hutter AI would flounder if given a completely compressed bit-stream: it would be completely random, incompressible and unpredictable like Chaitin's Omega number. Language is a lot closer to this than is the sensory input stream of a kid. There's a quote widely attributed to a William Martin (anybody know who he is?): You can't learn anything unless
RE: [agi] rule-based NL system
I'll have to say my objection stands. Because the point is that blind people learn about an object and infer it's shape from words and description of the object without ever seeing them. An intelligent AI will do so in the same way. After the blind learns about an object by reading in braille or having it described to them they can then postulate a mental image of it. The intelligence had to precede the mental image for them to have had the object described to them in enough detail to have made an image in their minds of it. And given the sentence, The elephant sat in the chair. It is not necessary to know what an elephant or a chair looks like if one knows from their description that elephants are multi-ton quadrapeds and chairs are smallish pieces of furniture sized for human beings. And the rule that Something large and heavy will not fit correctly into something small not built to carry that much weight. The image pops into your mind because those memories of elephants and chairs are keyed to those terms. But the images are not necessary to reason about the entities. The picture is not what makes you realize the sentence is probably nonsensical it is the commonsense rules and knowledge that you have learned and are bringing to bear on the sentence almost subconciously. _ From: Mike Tintner [mailto:[EMAIL PROTECTED] Sent: Saturday, April 28, 2007 12:43 PM To: agi@v2.listbox.com Subject: Re: [agi] rule-based NL system Classic objection. The answer is that blind people can draw - reasonably faithful outline objects. Experimentally tested. Their brains like all our brains form graphic outlines of objects and fit them accordingly to create scenes to test the sense of sentences. Worms do it too. They are blind, But if you go back to the Darwin passage I quoted, you will see that they can fill their burrows with all manner of differently shaped objects by touch. The senses are interdependent. We work by COMMON sense. - Original Message - From: Gary Miller mailto:[EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Saturday, April 28, 2007 4:36 PM Subject: RE: [agi] rule-based NL system Are you saying then that blind people can not make sense of language because they lack the capacity to imagine images having never seen them before? Or that blind people could not understand or would not view these these as equally strange as a sighted person? The man climbed the penny The mat sat on the cat The teapot broke the bull _ From: Mike Tintner [mailto:[EMAIL PROTECTED] Sent: Saturday, April 28, 2007 10:42 AM To: agi@v2.listbox.com Subject: Re: [agi] rule-based NL system Disagree. The brain ALWAYS tries to make sense of language - convert it into images and graphics. I see no area of language comprehension where this doesn't apply. I was just reading a thread re the Symbol gorund P on another group - I think what's fooling people into thinking purely linguistic comprehension is possible is the Dictionary Fallacy. They think the meaning of a sentence can be derived by looking up meanings word by word in a dictionary (real or in their mind). But to understand sentences you have to understand how the words FIT TOGETHER. There are no rules in any dictionary that tell you what words fit together. How do you know that the sentences: The man climbed the penny The mat sat on the cat The teapot broke the bull are probably nonsense (but not necessarily)? The brain understands these and all sentences by converting them into composite pictures. It may and does use other methods as well but making sense/ getting the picture/ seeing what you're talking about are fundamental. Understanding depends on imagination, (in the literal sense of manipulating images). It is by testing these derived pictures against its visual ( sensory) models and visual logic of things that the brain understands both what they are referring to and whether they make sense (in the 2nd meaning of that term - are realistic). The brain uses a picture tree, as we discussed earlier, Ben, and that picture tree is not only how the brain understands, but also the source of its adaptivity.Lot more to say about this.. but as you see I have a very hard line here, and yours seems to be considerably softer, and I'm interested in understanding that. - Original Message - From: Benjamin mailto:[EMAIL PROTECTED] Goertzel To: agi@v2.listbox.com Sent: Saturday, April 28, 2007 2:02 PM Subject: Re: [agi] rule-based NL system I agree about developmental language learning combined with automated learning of grammar rules being the right approach to NLP. In fact, my first wife did her PhD work on this topic in 1994, at Waikato University in Hamilton New Zealand. She got frustrated and quite before finishing her degree, but her program (which I helped with) inferred some nifty grammatical rules from a bunch of really simple children's books, and then used them
RE: [agi] SOTA
No, and it's a damn good thing it isn't. If it was we would be sentencing it to a mindless job with no time off, only to be disposed of when a better model comes out. We only want our AI's to be a smart as necessary to accomplish their jobs just as our cells and organs are. Limited conciousness or self-reflectivity may only be necessary in highly complex systems like computers where we may want them to recognize that they have a virus and take steps like searching for a digital vaccine to eliminate it without the owner even knowing it was there. Even in these cases we are only giving the system conciousness over one specific aspect of it's being. I would say that until we have software that can learn new free format information as we do and modify it's goal stack based upon that new information then we do not have a truly concious computer. _ From: Bob Mottram [mailto:[EMAIL PROTECTED] Sent: Friday, January 12, 2007 9:45 AM To: agi@v2.listbox.com Subject: Re: [agi] SOTA Ah, but is a thermostat conscious ? :-) On 12/01/07, [EMAIL PROTECTED] [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: http://www.thermostatshop.com/ Not sure what you've been Googling on but here they are. There's even one you can call on the telephone If there's a market for this, then why can't I even buy a thermostat with a timer on it to turn the temperature down at night and up in the morning? The most basic home automation, which could have been built cheaply 30 years ago, is still, if available at all, so rare that I've never seen it. _ This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=303 _ This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=303 - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=303
RE: [agi] SOTA
Ben Said: Being able to understand natural language commands pertaining to cleaning up the house is a whole other kettle of fish, of course. This, as opposed to the actual house-cleaning, appears to be an AGI-hard problem... A full Turing complete Natural Language system would not be necessary for robotic control. A pattern such as {clean|sweep|vacuum} (the )[RoomName]room( {for|in} [Number] minutes) When coupled with a voice recogniton system such as Nuance is marketing would increase the usefulness and interactivity of the robot immensely. [RoomName] and [Number] become variables passed to the robot vacuum and can have defaults If omitted from the command. The robot could come with a set of canned patterns for starters and the patterns could by customized by the user and associated with new behaviours. I like the idea of the house being the central AI though and communicating to house robots through an wireless encrypted protocol to prevent inadvertant commands from other systems and hacking. If the robots were made to a standard such as Microsoft's robotics toolkit then a single control and monitoring system could coordinate multiple robots activities and prevent collisions, coordinate efforts, etc... You'd probably need a backup fault tolerent system to prevent loss of critical systems like security, fire reporting, and temperature control. The house AI would be interfaced with the telephone and internet so that you could enter remote commands if you think of something while you're away. -Original Message- From: Benjamin Goertzel [mailto:[EMAIL PROTECTED] Sent: Saturday, January 06, 2007 11:16 AM To: agi@v2.listbox.com Subject: Re: [agi] SOTA Needless to say, I don't consider cleaning up the house a particularly interesting goal for AGI projects. I can well imagine it being done by a narrow AI system with no capability to do anything besides manipulate simple objects, navigate, etc. Being able to understand natural language commands pertaining to cleaning up the house is a whole other kettle of fish, of course. This, as opposed to the actual house-cleaning, appears to be an AGI-hard problem... -- BenG On 1/6/07, Pei Wang [EMAIL PROTECTED] wrote: Stanford scientists plan to make a robot capable of performing everyday tasks, such as unloading the dishwasher. http://news-service.stanford.edu/news/2006/november8/ng-110806.html On 1/6/07, Benjamin Goertzel [EMAIL PROTECTED] wrote: The problem wasn't technological. It was that nobody had any use for a robot. We never figured out what people would want the robot for. I think that's still the problem. Phil, I think the real issue is that no one wants an expensive, stupid, awkward robot... A broadly functional household robot would be very useful, even if it lacked intelligence beyond the human-moron level... For instance, right now, I would like a robot to go into my daughter's room and clean up the rabbit turds that are in the rabbit playpen in there. I would rather not do it. But, a Roomba can't handle this task because it can't climb over the walls of the playpen, nor understand my instructions, nor pick up the turds but leave the legos on the floor alone... Heck, a robot to let the dogs in and out of the house would be nice too... being doggie doorman gets tiring. Of course, this could be solved more easily by installing a doggie door ;-) How about a robot to bring me the cordless phone when it rings, but has been left somewhere else in the house ... ? ;-) How about one to put the dishes in the dishwasher and unload them ... and re-insert the ones that didn't get totally cleaned? The dishwasher is a good invention but it only does half the job The problem **is** technological: it's that current robots really suck ... not that non-sucky robots would be useles... -- Ben - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=303 - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=303 - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=303 - This list is sponsored by AGIRI: http://www.agiri.org/email To unsubscribe or change your options, please go to: http://v2.listbox.com/member/?list_id=303
RE: [agi] Logic and Knowledge Representation
Ben asked: What kind of bot are you using? Do you mean a physical robot or a chat bot? Just a chat bot for right now Ben. Although I could imagine a future robot manufacturer licensing the code to allow a customer to customize the high level cognitive/personality functions of the bot. Ben aasked: Would you be willing to share a few example patterns from your database, so that we can have a better sense of what you're actually talking about? Right now I have only a very vague idea This example pattern could literally take hundred of patterns to represent in other pattern languages and most bot engines wouldn't be able to handle misspelling word run together, etc... without creating separate patterns to account for each possible case which would be highly impractical to do. My patterns usually are much longer and include quite a bit more complexity that this simple example. But this simple example should give you flavor without me giving to much of the mojo away. Please understand though I am not open sourcing the project. And although I am not certain which parts of what I am doing that are patentable if any, I have documented the development and the partial disclosure of it so that I am in a position to challenge anyone else who might try to patent/commercialize it before me. I'm not trying to delay the singularity but if have quite a lot invested in the development and need to make sure the technology is used wisely. Pattern to match goes here Statements executed after correct pattern is found {} Required at exactly one inside to match () Optional match one or zero inside | seperator for {} or () choices [] invokes subpattern, subpatterns may invoke other subpatterns // Comments ok here # Function names start with this character Functions must be defined in the VB .Net code the language is implemented in. Variables for the knowledge base and may be assigned in the main.nlp main program Variables can also be created dynamically from within the program. Variable states are store to an ASCII file at program exit. eg. variable used in pattern [Location]={[City]|[State]|[Country]|[Continent]|[Province]|[Tourist_Attrac tion]} [City]={Greensburg|Jeanette|Las Vegas} [State]={Ohio} [Country]={Canada|Mexico} [Tourist_Attraction]={the Grand Canyon} Some example user inputs the below pattern will match. Greensburg is near Jeanette Ohio is by California Mexico is close to Canada Las Vegas is right by the Grand Canyon [Location]{({'|;})s| is} (j{u|i}st )({extremely|pretty|quite|rather|really|truly} )(very ) {close|[Direction]|in close proximity|near} {by|of|to} [Location] [Temp1]=[#Last([Location],1)] [Temp2]=[#Last([Location],2)] // Insert call to Google Maps API to get distance between [Temp1] and [Temp2] returning [Distance] if([Distance]20) They're not that close, on the map it looks like around [Distance] miles or so. #else They are pretty close on the map it looks like around only [Distance] miles. #endif [There]=[Temp1] [#Learn([Temp1] is [Distance] miles from [Temp2].)] Fuzziness can be built into the pattern for common keying errors ie. People often hit ; instead of ' {;|'} in pattern matches either. Fuzziness spaces in pattern are matched as ( ) optional spaces allowing for words bleed together. Spell Check has a dictionary of words ordered by frequency of use in nl-text. Spell check algorithm uses Boyer Moore algorithm to sequentially replace potentially correct word for mispelled word in user input. This is only done if no patterns match input. This code would need to be a little more complicated due to cities of the same name residing in different states or countries. I would probably set [CurrentTopic] in an earlier matched pattern to something like where user lives or something. Ben said: The grounding issue is not just about voice versus text --- it's about grounding of linguistic relationships in physical relationships... For instance, Novamente can learn that near is approximately transitive via observing the 3D simulation world to which it's connected; it can then use inference to generalize this transitivity to other domains that it has not directly experienced, e.g. to reason that if Rockville is near Bethesda and Bethesda is near DC then Rockville must be somewhat near DC. I.e. the system can ground the properties of near-ness in its embodied experience. Without this kind of experiential grounding, a whole bunch of patterns that humans learn implicitly must somehow be gotten into the AI either by explicit encoding or by very subtle mining of relationships from a huge number of NL texts or verrry tedious and long-winded chats with humans... Agreed when my son 3-5 I had a lot of verrry tedious and long-winded chats teaching him about the world also. My bot currently has no capability to generalize. It therefore relies on the same stereotypical generalizations that humans have already learned to base it's
RE: [agi] Logic and Knowledge Representation
J. Andrew Rogers posed the question: The obvious question is how do you deal with the problem of the synonymity of patterns being context sensitive? In any sufficiently rich environment, the type of compression you appear to be describing above is naive and would have adverse consequences on efficacy. Or at least, I cannot think of a construct that can do this efficiently while being some facsimile of fully general. It is of course true that a a single thought can have a different meaning in both the context of the statements that precede it and context also changes based upon the gender, age, and psychological make up of the person being simulated. Once a pattern is matched multiple responses can be generated based on the context which tracked from pattern to pattern. [CurrentTopic], [He], [She], [Them], [There], [Then], [User_Emotion], [AI_Emotion], as well as a several hundred variables which are not hardcoded in patterns but maintained in a header files and induced from prior user inputs and used in responses maintain context. Hence questions like: What did I say my name is? Could generate I don't believe you told me your name. or My name is [User_Name]. Questions like What are you talking about? could be answered I thought we were talking about [Current_Topic]. Half of the challenge of writing the patterns is asking onesself what information whould a human induce from this input and then storing it as a temporary variable state. This provides short term memory as it pertains to the conversation. A variable can be defined as short term (reset at beginning of each new conversation) or long term and never reset until It has been changed by new information coming in even across different conversations. As far a doing all of this efficiently once my 28000 patterns get loaded into memory the bot's brain reaches about 7.3 Mbs very small. Search heuristics for matching patterns is problematic though because a pattern can start with multiple different character streams making the matching algorithm do a lot of work for each input. But the algorithm can be made parallell by dividing the pattern list into multiple pieces and conducting search in multiple threads of execution. I am waiting to implement that feature unit quad core chips become available next year. -Original Message- From: J. Andrew Rogers [mailto:[EMAIL PROTECTED] Sent: Monday, May 08, 2006 1:39 AM To: agi@v2.listbox.com Subject: Re: [agi] Logic and Knowledge Representation On May 7, 2006, at 6:37 PM, Gary Miller wrote: Which is why my research has led to a pattern language that can compress all of the synonymous thoughts into a single pattern. The obvious question is how do you deal with the problem of the synonymity of patterns being context sensitive? In any sufficiently rich environment, the type of compression you appear to be describing above is naive and would have adverse consequences on efficacy. Or at least, I cannot think of a construct that can do this efficiently while being some facsimile of fully general. Cheers, J. Andrew Rogers --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED] --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
RE: [agi] Logic and Knowledge Representation
John said: The human brain, the only high-level intelligent system currently known, uses language and logic for abstract reasoning, but these are based on, and owe their existence to, a more fundamental level of intelligence -- that of pattern-recognition, pattern-matching, and pattern manipulation. I agree with this wholeheartedly. But on the next point we diverge in our thinking. John said: In evolution on earth, sensory-motor-based intelligencecame first, and the use of language and logiconly later. It seems to me that the right path to true AI will also use sensory-motor patterns as the basic building blocks of knowledge representation. A typical human being'sknowledge of the letter "A" involves recognition ofgraphical representations of the symbol,memories of its sound when spoken, procedural or muscle memory of how to speak andwrite it, and memories ofwhere itis commonly found in its linguisticcontext. A system should be capable of recognizing symbols visually or auditorially (and possibly ofgeneratingthem through motor outputs) before it should be expected to comprehend them. Any thoughts or arguments? Or am I just repeating something everyone already knows? (I honestly don't know.) Speech recognition andvisual recognition are separate problems from knowledge representation/pattern recognition. Hellen Keller was blind and deaf but with some help was able to achieve knowledge representatation and pattern recognition without the use of either hear or sight. Think of the senses as the input/output devices and yes an infant's brain must first learn to control those input output devices before it is able to learn and communicate with the world outside itself. But an artificially intelligent entity already has access to an ASCII data stream that is can do input/output to communicate outside itself. Of course because a picture is worth a thousand words a program that can also do visual recognition has access to a larger data store than one that does not. My opinion on the most probableroute to a true AI Entity is: 1. Build a better fuzzy pattern representation language with an inference mechanism for extracting inducible information from user inputs. Fuzziness allows the language to understand utterances with misspellings words run together etc... 2. Build a bot based on said language 3. Build a large knowledgebase which captures a large enough percentage of real world knowledge to allow the bot learn from natural language data sources i.e. the web. 4. Build a pattern generator which allows the bot learn the information it has read and build new patterns itself to represent the knowledge. 5. Build a reasoning module based on Bayesian logic to allow simple reasoning to be conducted. 6. Build a conflict resolution module to allow the Bot to resolve/correct conflicting information or ask for help with clarification to build correct mental model. 7. Build a goal and planning module which allow the Bot to operate more autonomously to aid in the goals which we give it i.e.. achieve singularity. Steps 1 and 2 an took me a couple years. 3 is an ongoing effort. Into my fourth year now with 28000 patterns. Hint: if the pattern recognition language is good a single pattern should be able express all the ways of expressing a single thought in a single pattern. This makes the patterns longer and more complex but reduces overall work by not forcing the bot master to write thousands of patterns to account for all possible ways to express a single thought. My 28000 patterns would requires match at several orders of magnitude more inputs correctly than competing solutions including misspellings, ungrammatical inputs etc. This transforms the difficulty of step 3 from being an totally intractable problem to a doable but still difficult/work intensiveproblem. Step 4 is keeping me awake at night thinking about it. Steps 5, 6 and 7don't sound that difficult to me right now but that's only because I haven't thought about them in enough detail. People have challenged the top down approach saying that such a bot would lack grounding or the ability to tie it's knowledge to real world inputs. But it should not be difficult to use a commercial voice recognition engine to transform voice inputs intoASCIIinputs. And the fuzzy recognizer for be able to compensate many times for the mistakes that the voice recognition software makes inrecognizing a word or two in the input stream. From: John Scanlon [mailto:[EMAIL PROTECTED] Sent: Sunday, May 07, 2006 2:40 AMTo: agi@v2.listbox.comSubject: [agi] Logic and Knowledge Representation Is anyone interested in discussing the use of formal logic as the foundation for knowledge representation schemes for AI? It'sa common approach, but I think it's the wrong path. Even if you add probability or fuzzy logic, it's still insufficient for true intelligence. The human brain, the only
RE: [agi] Friendliness toward humans
EGHeflin said: The reason is that the approach is essentially 'Asimovian' in nature and, therefore, wouldn't result in anything more than perhaps a servile pet, call it iRobot, which is always 'less-than-equal' to you and therefore always short of your goal to achieve the so called 'singularity' you originally set out to achieve. When our children are small make rules for them which are eventually outgrown: Don't hit other children. Don't use bad words. Be polite to other people. At first our children may be incapable of understanding all the reasons these rules are in place. But by being forced to obey them either by withholding pleasure or by punishment they become a part of their normal behavior patterns such that when they mature and modify their own rules for themselves these guidelines are part of their normal behavior. Whether we tell the AGI it's high priority rules to ensure safety or hardcode them should not make a difference. As long as they are acted on repeatedly they will become a part of their normal behavior through normal enforcement. At that point the hardcoded rules could be removed without fear that the AI would revert to a cursing, violent, rude lout. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of EGHeflin Sent: Thursday, January 09, 2003 2:06 PM To: [EMAIL PROTECTED] Subject: Re: [agi] Friendliness toward humans Kevin et al., Fascinating set of observations, conjectures, and methodologies, well worth considering. And it seems that you have ultimately touched on the kernel of the dilemma of man v. machine en route to the so called 'singularity'. If I've understood you correctly vis-a-vis the emergence of 'evil' in AGI systems, you're suggesting that there is a need to prevent, a priori, certain expressions of self in AGI systems to prevent any hint of evil in AGI systems. It's an approach well worth considering, but I believe it falls short of 'real' AGI. The reason is that the approach is essentially 'Asimovian' in nature and, therefore, wouldn't result in anything more than perhaps a servile pet, call it iRobot, which is always 'less-than-equal' to you and therefore always short of your goal to achieve the so called 'singularity' you originally set out to achieve. But perhaps any discussion about 'good' and 'evil' is best served by defining exactly what 'good' and 'evil' are. However, I'll be a complete Sophist and suggest that the dilemma of 'good' and 'evil' can be talked around by separating the dilemma into 3 obvious types and talking about these. As I see it, the 3 dilemma types of 'good' and 'evil' are: 1. man v. man, 2. man v. machine, and/or, at the 'singularity' 3. man-machine v. man-machine. So I'll comment on a particular approach for 'real' AGI that addresses the dilemma type (2) guided by observations about type (1) and with obvious extensions to type (3). It seems that if you are trying to model your AGI system after nature, which is a reasonable and likely place to start, you should realize that 'nature' simply hasn't created/engineered/evolved the human species the way your approach suggests. Put another way, the human species does not have an intrinsic suppression of either 'good' or 'evil' behaviors. And unless you're willing to hypothesis that this is either a momentary blip on the radar screen of evolution, i.e. humans are actually in the process of breaking this moralistic parity, OR that these 'good' or 'evil' behaviors will ultimately be evolved away through 'nature', natural selection, and time, you are left with an interesting conjecture in modeling your AGI system. The conjecture is that 'good' or 'evil' behaviors are intrinsic parts of the human condition, intelligence, and environment, and therefore should be intrinsic parts in a 'real' AGI system. And as a complete Sophist, I'll skip over more than 6,000 years of recorded human history, philosophical approaches, religious movements, and scholarly work - that got us where we're at today w.r.t. dilemma type (1) - to suggest that the best approach to achieve 'real' AGI is to architect a system that considers all potential behaviors, from 'good' to 'evil', against completed actions and conjectured consequences. In this way, a certain kernel of the struggle of 'good' and 'evil' is retained, but the system is forced to 'emote' and 'intellectualize' the dilemma of 'good' and 'evil' behaviors. The specific AGI architecture I am suggesting is essentially 'Goertzelian' in nature. It is a 'magician system', whereby the two components, G (Good-PsyPattern(s)) and E (Evil-PsyPattern(s)), are, in and of themselves, psychological or behavioral patterns that can only be effectuated, i.e. assigned action patterns, in a combination with another component to generate and emerge as an action pattern, say U (Useful or Utilitarian-ActionPattern). The system-component architecture might be thought of as a G-U-E or GUE 'magician system'. The
RE: [agi] The AGI and the will to live
At some early point the AGI will have to learn to equate pleasure with learning and acquiring new experience. With a biological organism the stimuli are provided as pain and pleasure. As we mature many of our pleasure causers are increasingly subtle and are actually learned pleasure generators themselves. As the organism matures it becomes able to initiate it's own pleasure, but it also needs to vary the type of pleasure it experiences. This is largely because the most pleasure repeated over and over again induce less satisfaction as time goes on. While the exception may may be drug and other physical addictions, for the most part we constantly seek to keep ourselve stimulated. At some internal level this type of internal reward system must develop or motivation to do anything will not exist. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Colin Hales Sent: Thursday, January 09, 2003 5:23 PM To: [EMAIL PROTECTED] Subject: [agi] The AGI and the will to live Hi all, I find the friendliness issue fairly infertile ground tackled way too soon. Go back to where we are: the beginning. I'm far more interested in the conferring of a will to live. Our natural tendency is to ascribe this will to live to our intelligent artifacts. This 'seed' is by far the hardest thing to create and the real determinant of 'friendliness' in the end. Our seed? I think a model that starts from something like 'another heartbeat must happen'. When you don't have a heart? What - poke a watchdog timer every 100msec or die? My feeling at the moment is that far from having a friendliness problem we're more likely to need a cattle prod to keep the thing interested in staying awake, let alone getting it to take the trouble to formulate any form of friendliness or malevolence or even indifference. If our artifact is a zombie, what motivation is there to bother _faking_ friendliness or malevolence or even indifference? Without it Pinnocchio the puppet goes to sleep. If our artifact is not a zombie (.ie. has a real subjective experience) then what motivates _real_ friendliness or malevolence or even indifference? Without it Pinnocchio the artificial little boy goes to sleep. Whatever the outcome, at its root is the will to even start learning that outcome. You have to be awake to have a free will. What gets our AGI progeny up in the morning? regards, Colin Hales --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED]
RE: [agi] Early Apps.
Ben Goertzal wrote: I don't think that a pragmatically-achievable amount of formally-encoded knowledge is going to be enough to allow a computer system to think deeply and creatively about any domain -- even a technical domain about science. What's missing, among other things, is the intricate interlinking between declarative and procedural knowledge. When humans learn a domain, we learn not only facts, we learn techniques for thinking and problem-solving and experimenting and information-presentation .. and we learn these in such a way that they're all mixed up with the facts What you're describing is the Expert System approach to AI, closely related to the common sense approach to AI. ... I agree that as humans we bring a lot of general knowledge with us when we learn a new domain. That is why I started off with the general conversational domain and am now branching into science, philosophy, mathematics and history. And of course the AI can not make all the connections without being extensively interviewed on a subject and having a human help clarify it's areas of confusion just as a parent answers questions for a child or a teacher for a student. I am not in fact trying to take the exhaustive approach one domain at a time approach but rather to teach it the most commonly known and requested information first. My last email just used that description to identify my thoughts on grounding. I am hoping that by doing this and repeating the interviewing process in an iterative development cycle that eventually the bot will eventually be able to discuss many different subjects at a somewhat superficial level much as same as most humans are capable of. This is a lot different from the exhaustive definition that Cyc provides for each concept. I view what I am doing distinct from expert systems because I do not yet use either a backward or forward inference engine to satisfy a limited number of goal states. The knowledge base is not in the form of rules but rather many matched patterns and encoded factoids of knowledge many of which are transitory in nature and track the context of the conversation. Each pattern may trigger a request for additional information like an expert system. But the bot does not have a particular goal state in mind other that learning new information unless a specific request of it is made by the user. I also differ from Cyc in that realizing the importance of English as a user interface from the beginning, all internal thoughts and goal states occur as an internal dialog in English. This eliminates the requirement to translate an internal knowledge representation to an external natural language other than providing one or response patterns to specific input patterns. It also makes it easy to monitor what the bot is learning and whether it is making proper inferences because it's internal thought process is displayed in English while in debug mode.. The templates which generate the responses in some cases do have conditional logic to determine which output template is appropriate response based on the AI's personality variables and the context of the current conversation. Variables are also set conditionally to maintain metadata for context. If the references a male in it's response [He] and [Him] get set vs. [Her] and [She] if a female is referenced. [CurrentTopic], [It], [There] and [They] are all set to maintain backward contextual references. I was able to find a few references to the Common Sense approach to AI on google and some of the difficulties in achieving it. And I must admit I have not yet implemented non-monotonic reason or probabilistic reasoning as of yet. I am not under the illusion that I am necessarily inventing or implementing anything that has not been conceived of before. As Newton says if I achieve great heights it will be because I have stood on the shoulders of giants. I just see the current state of the art and think that it can be made much better. I do not actually know how far I can take it while staying self-funded, but hopefully by the time my money runs out it will demonstrate enough utility and potential to be of value to someone. I think I like the sound of the Common Sense Approach to AI though. I can't remember the last time anyone accused me of having common sense, but I like the sound of it! I don't think AI is absent sufficient theory, just sufficient execution. I feel like the Cyc Project's heart was in the right place and the level of effort was certainly great, but perhaps the purity of their vision took priority over usability of the end result. Is any company actually using Cyc as anything other than a search engine yet? That being said other than Cyc I am at a loss to name any serious AI efforts which are over a few years in duration and have more that 5 man years worth of effort (not counting promotional and fundraising). The Open Source efforts are interesting and have some utility but are
RE: [agi] Early Apps.
On Dec 26 Ben Goertzel said: One basic problem is what's known as symbol grounding. this means that an Ai system can't handle semantics, language-based cognition or even advanced syntax if it doesn't understand the relationships between its linguistic tokens and patterns in the nonlinguistic world. I guess I'm still having trouble with the concept of grounding. If I teach/encode a bot with 99% of the knowledge about hydrogen using facts and information available in books and on the web. It is now an idiot savant in that it knows all about hydrogen and nothing about anything else and it is not grounded. But if I then examine the knowledge learned about hydrogen for other mentioned topics like gases, elements, water, atoms, etc... And teach/encode 99% of of the knowledge on these topics to the bot. Then the bot is still an idiot savant but less so isn't it better grounded? A certain amount of grounding I think has occurred by providing knowledge of related concepts. If we repeat this process again, we may say the program is an idiot savant in chemistry. Each time we repeat the process are we not grounding the previous knowledge further because the bot can now reason and respond to questions not just about hydrogen, it now has an English representation of the relationship between hydrogen and other related concepts in the physical world.. If we were to teach someone such as Helen Keller with very limited sensory inputs would we not be attempting to do the same thing? Humans of course do not learn in this exhaustive manner. We get a shotgun bombardment of knowledge from all types of media on all manner of subjects. The things that interest us we pursue additional knowledge about. The more detailed our knowledge in any given area the greater we say our expertise is. Initially we will be better grounded than a bot, because as children we learn a little bit about a whole lot of things. So anything new we learn we attempt to tie into our semantic network. When I think. I think in English. Yes, at some level below my conscious awareness these English thoughts are electrochemically encoded, but consciously I reason and remember in my native tongue or I retrieve a sensory image, multimedia if you will. If someone tells me that A kinipsa is terrible plorid. I attempt to determine what a kinipsa and a plorid are so that I may ground this concept and interconnect it correctly within my existing semantic network. If A bot is taught to pursue new knowledge and ground the unknown terms with it's existing semantic net by putting the goals Find out what a plorid is and Find out what a kinipsa is on it's list of short term goals then it will ask questions and seek to ground itself as a human would! I will agree that today's bots are not grounded because they are idiot savants and lack the broad based high level knowledge with which to ground any given fact or concept. But if I am correct in my thinking this is the same problem that Helen Keller's teacher was faced with in teaching Helen one concept at a time until she had enough simple information or knowledge to build more complex knowledge and concepts upon. When a child learns to speak he does not have a large dictionary to draw on to tell him that mice is the plural of mouse. No rule will tell him that. He has to learn it. He will say mouses and someone will correct him. It gets added to his NLP database as an exception to the rule. A human has limited storage so a rule learned by generalizing from experience is a shortcut to learning and remembering all the plural forms for nouns. In a AGI we can give the intelligence certain learning advantages such as these dictionaries and lists of synonym sets which do not take that much storage in the computer. I also think that children do not deal with syntax. They have heard a statement similar to what they want to express and have this stored as a template in their minds. I think we cut and paste what we are trying to say into what we think is the correct template and then read it back to ourselves to see if it sounds like other things we have heard and seems to make sense. For people who have to learn a foreign language as an adult this is difficult because they tend to think in their first language and commingle the templates from their original and the new language. But because we do not parse what we here and read strictly by the laws of syntax we have little trouble understanding many of these ungrammatical utterances. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of [EMAIL PROTECTED] Sent: Thursday, December 26, 2002 11:03 PM To: [EMAIL PROTECTED] Subject: RE: [agi] Early Apps. On 26 Dec 2002 at 10:32, Gary Miller wrote: On Dec. 26 Alan Grimes said: According to my rule of thumb, If it has a natural language database it is wrong, Alan I can see based on the current generation of bot technology why one
RE: [agi] Early Apps.
On Dec. 26 Alan Grimes said: According to my rule of thumb, If it has a natural language database it is wrong, Alan I can see based on the current generation of bot technology why one would feel this way. I can also see people having the view that biological systems learn from scratch so that AI systems should be able to also. Neither of these arguments are particularly persuasive though based on what I've developed to date. Do you have other arguments against a NLP knowledge based approach that you could share with me. If you feel this is out of bounds for the list please just Email with your arguments. I am involved in such a project and certainly don't wish to to be wasting my time! -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Alan Grimes Sent: Thursday, December 26, 2002 1:12 AM To: [EMAIL PROTECTED] Subject: [agi] Early Apps. According to my rule of thumb, If it has a natural language database it is wrong, many of the proposed early AGI apps are rather unfeasable. However, there is a very interesting application which goes streight to the hart to the main AI problem and also provides a very valuble tool for flexing the chips that we already have in our sweatty little hands. The area is COMPILERS. Today's compilers are notoriously bad. The leading free compiler is atrociously bad. Now, if there could be an AI based compiler that could both understand the source and the machine in a very human-like way the output code would be that much better. This would also be valuble for a bootstrap AI though I strongly caution against such an AI untill we have a _MUCH_ better understanding of what is going on. I expect to be preparing a proposal in a few months that will outline a complete strategy for an AI that should be both fesable and, through inhreant architectual constraints, be reasonably safe. -- pain (n): see Linux. http://users.rcn.com/alangrimes/ --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED]
RE: [agi] AI on TV
Title: Message On Dec. 9 Kevin said: "It seems to me that building a strictly "black box" AGI that only uses text or graphical input\output can have tremendous implications for our society, even without arms and eyes and ears, etc. Almost anything can be designed or contemplated within a computer, so the need for dealing with analog input seems unnecessary to me. Eventually, these will be needed to have a complete, human like AI. It may even be better that these first AGI systems will not have vision and hearing because it will make it more palatable and less threatening to the masses" I agree wholeheartedly. Sony and Honda as well as several military contractors are spending 10s perhaps hundreds of million dollars on RD robotics programs which incorporate the vision, and analog control, and data acquisition for industry, the military, and yes even the toy companies. Once AGIs are ready to fly it will be able to interface with these systems through software APIs (Application Programming Interfaces) and will not even care about the low-level programs that enable them move about and visually survey their environments. Too often those who seek the spotlight are really sincere, but either need recognition for their own self reassurance or as a method of attracting potential funding. There seems to be an unwritten law in the universe which that says all major inventions will involve major sacrifice and loss for those who dare to tackle what has been deemed impossible by others. From Galileo to Edison, to Tesla, to maybe one of us. Before we succeed, ifwe succeed, the universe will exact it's toll. For nature will not give up her secrets willingly and intelligence may be her most closely guarded secret of all! Don't forget that genius and madness sometimes walk arm in arm! And as the man says if you weren't cazy when you got in, you probably will be before you get out!. -Original Message-From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of maitriSent: Monday, December 09, 2002 11:08 AMTo: [EMAIL PROTECTED]Subject: [agi] AI on TV There was a show on the tube last night on TechTV. It was part of their weekly Secret, Strange and True series. They chronicled three guys who are working on creating advanced AI. One guy was from Belgium. My apologies to him if he reads this list, but he was a rather quirky and stressed character. He had designed a computer that was basically a collection of chips. He raised a million and had it built on spec. I gather he was expecting something to miraculously emerge from this collection, but alas, nothing did. It was really stressful watching his stress. He had very high visibility in the country and the pressure was immense as he promised a lot. I have real doubts about his approach, even though I am a lay-AI person. Also, its clear from watching him that its sometimes good to have shoestring budgets and low visibility. Less stress and more forced creativity in your approach... The second guy was from either England or the states, not sure. He was working out of his garage with his wife. He was trying to develop robot AI including vision, speech, hearing and movement. He was clearly floundering as he radically redesigned what he was doing probably a dozen times during the 1 hour show. I think this experimentation has value. But I really wonder if large scale trial and error will result in AGI. I don't think so. I think trial and error will, of course, be essential during development, but T and E of the entire underlying architecture seems a folly to me. Since the problem is SO immense, I believe one must start with a very sound and detailed game plan that can be tweaked as things move along. The last guy was brooks at MIT. They were developing a robot with enhanced vision capabilities. They also failed miserably. I am rather glad that they did. They re funded by DOD, and are basically trying to build a robotic killing machine. Just what we need. It seems to me that trying to tackle the vision problem is too big of a place to start. While all this work will have value down the line, is it essential to AGI? It seems to me that building a strictly "black box" AGI that only uses text or graphical input\output can have tremendous implications for our society, even without arms and eyes and ears, etc. Almost anything can be designed or contemplated within a computer, so the need for dealing with analog input seems unnecessary to me. Eventually, these will be needed to have a complete, human like AI. It may even be better that these first AGI systems will not have vision and hearing because it will make it more palatable and less threatening to the masses The show was rather discouraging, especially if one considers that these three folks are leading the way towards
RE: [agi] general patterns induction
A paper I found while researching trigram frequency a little further looks like it may be right up your alley. http://www.ling.gu.se/~kronlid/term_paper/nlp_paper.pdf -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Pablo Sent: Sunday, December 08, 2002 9:34 PM To: [EMAIL PROTECTED] Subject: [agi] general patterns induction Hi Everyone, I'm looking for information about pattern induction or general patterns or anything that sounds like that... What I want to do is, having a stream of data, predict what may come. (yes, and then take over the world... sorry if it sounds like Pinky and The Brain!!) I guess general patterns induction is related to data compression, because if we find a pattern in a string, then we don't have to write all the characters every time the pattern appears. Surely someone has already been working on that (who?) Anyone would please give me a clue? Is there any book I should read?? Is there any book like AI basics, introduction to AI, or AI for dummies that may help before? Thanks a lot! Pablo Carbonell PS: thanks Ben, Kevin and Eliezer for the previous help --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] BEGIN:VCARD VERSION:2.1 N:Miller;Gary;A. FN:Gary A. Miller ([EMAIL PROTECTED]) ([EMAIL PROTECTED]) ORG:New Millennium Consulting TITLE:Principal Consultant TEL;WORK;VOICE:(440) 942-9264 TEL;HOME;VOICE:(440) 942-9264 ADR;WORK:;;7222 Hodgson Rd.;Mentor;OH;44060;United States of America LABEL;WORK;ENCODING=QUOTED-PRINTABLE:7222 Hodgson Rd.=0D=0AMentor, OH 44060=0D=0AUnited States of America EMAIL;PREF;INTERNET:[EMAIL PROTECTED] REV:20021108T231940Z END:VCARD
RE: [agi] An idea for promoting AI development.
On 12/01 Ben Goertzel said: 2) to avoid the military achieving *exclusive* control over one's technology What I am about to say may sound blasphemous, but the military may be the group with resources to protect the technology! By publicizing and making AGI technology generally available other hostile governments/military may see AGI as a potential weapon and resort to traditional methods to acquire the technology, industrial espionage, kidnappings of key scientists. Or they may fear it as a another potential tool to reign in their aggression and target it for destruction which means facilities and people. If this sounds farfetched just look at the lengths certain countries go to acquire Plutonium. If the potential for AGI is seen as great and world changing. Who better to protect it or at least offer stewardship than the US or NATO military. What private or non-profit is prepared and qualified to protect the technology when it start's to get really interesting? There are a few possibilities of why DoD is currently a prime contributor to AI research. 1. They fear it (saw Terminator and War games), so they better damn keep an eye on it (Their paranoia). 2. They may have it already in the NSA basements and want to be control the direction of other research to keep their tactical advantage. (My paranoia) 3. They are starting to see the results of years of research in robotics and drone warriors in keeping the casualty counts down and the American public happy and look at AI as another way to continue this trend. 4. They are good at fumbling the ball when it comes to acting in a timely manner on terrorist threats and it's much easier to blame an AI when they screw up than risk their cushy jobs. -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Ben Goertzel Sent: Monday, December 02, 2002 9:09 AM To: [EMAIL PROTECTED] Subject: RE: [agi] An idea for promoting AI development. Regarding being wary about military apps of AI technology, it seems to me there are two different objectives one might pursue: 1) to avoid militarization of one's technology 2) to avoid the military achieving *exclusive* control over one's technology It seems to me that the first objective is very hard, regardless of whether one accepts military funding or not. The only ways that I can think of to achieve 1) would be 1a) total secrecy in one's project all the way 1b) extremely rapid ascendancy from proto-AGI to superhuman AGI -- i.e. reach the end goal before the military notices one's project. This relies on security through simply being ignored up to the proto-AGI phase... On the other hand, the second objective seems to me relatively easy. If one publishes one's work and involves a wide variety of developers in it, no one is going to achieve exclusive power to create AGI. AGI is not like nuclear weapons, at least not if a software-on-commodity-hardware approach works (as I think it well). Commodity hardware only is required, programming skills are common, and math/cog-sci skills are not all *that* rare... -- Ben G -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Alexander E. Richter Sent: Monday, December 02, 2002 7:48 AM To: [EMAIL PROTECTED] Subject: RE: [agi] An idea for promoting AI development. At 07:18 02.12.02 -0500, Ben wrote: Can one use military funding for early-stage AGI work and then somehow delimitarize one's research once it reaches a certain point? One can try, but will one succeed? They will squeeze you out, like Lillian Reynolds and Michael Brace in BRAINSTORM (1983) (Christopher Walken, Natalie Wood) cu Alex --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] BEGIN:VCARD VERSION:2.1 N:Miller;Gary;A. FN:Gary A. Miller ([EMAIL PROTECTED]) ([EMAIL PROTECTED]) ORG:New Millennium Consulting TITLE:Principal Consultant TEL;WORK;VOICE:(440) 942-9264 TEL;HOME;VOICE:(440) 942-9264 ADR;WORK:;;7222 Hodgson Rd.;Mentor;OH;44060;United States of America LABEL;WORK;ENCODING=QUOTED-PRINTABLE:7222 Hodgson Rd.=0D=0AMentor, OH 44060=0D=0AUnited States of America EMAIL;PREF;INTERNET:[EMAIL PROTECTED] REV:20021108T231940Z END:VCARD
RE: [agi] An idea for promoting AI development.
FYI Arthur T. Murray I just tried to order your Metifex book at iUniverse, but the site was bombing at the checkout screen. I'll try again later but just wanted to let you know you might be losing orders! -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Arthur T. Murray Sent: Friday, November 29, 2002 11:38 AM To: [EMAIL PROTECTED] Cc: [EMAIL PROTECTED] Subject: Re: [agi] An idea for promoting AI development. On Fri, 29 Nov 2002, Alan Grimes wrote: Jeremy Smith wrote: [...] He also seems to be just asking for a huge sum of money to implement it!!! Mentifex/Arthur here with an announcement. I'm asking for $17.95 U.S. The Mentifex AI Textbook has today Thurs.29.Nov.2002 just been published by iUniverse.com as AI4U: Mind-1.1 Programmer's Manual on the Web at http://www.iuniverse.com/bookstore/book_detail.asp?isbn=0595259227 (q.v.). It would probably cost less to buy the print-on-demand (POD) textbook than to print out all the associated Mentifex pages on the Web. In a few weeks it should be possible for interested or curious parties to track AI4U on Amazon and see how many millions down it is ranked! /End interrupt mode -- Arthur T. Murray Perspective: The latest release of MS windows cost $2Billion... A typical internet start-up would receive anywhere from 20 to 50 million in VC. Heck, in the VC world you need to ask for large sums of money just to get people's attention. --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] --- To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/?[EMAIL PROTECTED] BEGIN:VCARD VERSION:2.1 N:Miller;Gary;A. FN:Gary A. Miller ([EMAIL PROTECTED]) ([EMAIL PROTECTED]) ORG:New Millennium Consulting TITLE:Principal Consultant TEL;WORK;VOICE:(440) 942-9264 TEL;HOME;VOICE:(440) 942-9264 ADR;WORK:;;7222 Hodgson Rd.;Mentor;OH;44060;United States of America LABEL;WORK;ENCODING=QUOTED-PRINTABLE:7222 Hodgson Rd.=0D=0AMentor, OH 44060=0D=0AUnited States of America EMAIL;PREF;INTERNET:[EMAIL PROTECTED] REV:20021108T231940Z END:VCARD