Re: [agi] The problem with AGI per Sloman
The space navigation system is a case in point. I listened to a talk by Peter Norvig in which he talked about looking at the moton of the heavens in an empirical way. Let's imagine angels and crystal spheres and see where we get. He rather decried Newton and the apple. In fact the apple story isn't true. Newton investigated gravity after discussion with Halley of comet fame. Newton was not generous enough to admit this and came up the *apple* story. In fact there is a 7th parameter - the planet's mass. This is needed when you think about the effect that a planet has on the others. Adams and Leverrier discovered Neptune in 1846 by looking at irregularities in the orbit of Uranus. Copernicus came up with the Heliocentric theory because Ptolomy's epicycles were so cumbersome. In fact a planet can be characterised by 6 things. Its major axis, its eccentricity, the plane of its orbit, the angle of the major axis and the position in orbit - time. These parameters are constant unless a third body is present. How do we get to Saturn? Cassini had a series of encounters. We view each encounter by calculating first a solar orbit and then a Venusian/Terrestrial/Jovian orbit. Deep in an encounter a planet is the main influence. 3 body problems are tricky and we do a numerical computation with time steps. We have to work out first of all the series of encounters and then the effects of inaccuracies. We need to correct, for example *before* we encounter a planet. How does this fit in with AGI? Well Peter I don't think you empirical approach will get you to Saturn. There is the need for theoretical knowledge. How is this knowledge represented in AGI? It is represented in an abstract mathematical form, a form which describes a general planet and incorporates the inverse square law. What in fact we need to know about space navigation is whether our system incorporates these abstract definitions. I would view AGI as something which will understand an abstract definition. This is something a little meta mathematical. It understands ellipses and planets in general and wants to know how this knowledge is incorporated in our system. In fact I would view AGI as something checking on the integrity of other programs. *Abstract Language* * * By this I mean an abstract grammatical description of a language, gender and morphology. This is vital. I don't think our Peter has ever learnt Spanish even, at least not properly. We can find out what morphology a word has once we have a few examples if*f* we have a morphological description built in. - Ian Parker A narrow embedded system, like say a DMV computer network is not an AGI. But that doesn't mean an AGI could not perform that function. In fact, AGI might arise out of these systems needing to become more intelligent. And an AGI system, that same AGI software may be used for a DMV, a space navigation system, IRS, NASDAQ, etc. it could adapt. .. efficiently. There are some systems that tout multi-use now but these are basically very narrow AI. AGI will be able to apply it's intelligence across domains and should be able to put its feelers into all the particular subsystems. Although I foresee some types of standard interfaces perhaps into these narrow AI computer networks; some sort of intelligence standards maybe, or the AGI just hooks into the human interfaces... An AGI could become a God but also it could do some useful stuff like run everyday information systems just like people with brains have to perform menial labor. John --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
RE: [agi] The problem with AGI per Sloman
It's just that something like world hunger is so complex AGI would have to master simpler problems. Also, there are many people and institutions that have solutions to world hunger already and they get ignored. So an AGI would have to get established over a period of time for anyone to really care what it has to say about these types of issues. It could simulate things and come up with solutions but they would not get implemented unless it had power to influence. So in addition AGI would need to know how to make people listen... and maybe obey. IMO I think AGI will take the embedded route - like other types of computer systems - IRS, weather, military, Google, etc. and we become dependent intergenerationally so that it is impossible to survive without. At that point AGI's will have power to influence. John From: Ian Parker [mailto:ianpark...@gmail.com] Sent: Saturday, June 26, 2010 2:19 PM To: agi Subject: Re: [agi] The problem with AGI per Sloman Actually if you are serious about solving a political or social question then what you really need is CRESS http://cress.soc.surrey.ac.uk/web/home . The solution of World Hunger is BTW a political question not a technical one. Hunger is largely due to bad governance in the Third World. How do you get good governance. One way to look at the problem is via CRESS and run simulations in second life. One thing which has in fact struck me in my linguistic researches is this. Google Translate is based on having Gigabytes of bilingual text. The fact that GT is so bad at technical Arabic indicates the absence of such bilingual text. Indeed Israel publishes more papers than the whole of the Islamic world. This is of profound importance for understanding the Middle East. I am sure CRESS would confirm this. AGI would without a doubt approach political questions by examining all the data about the various countries before making a conclusion. AGI would probably be what you would consult for long term solutions. It might not be so good at dealing with something (say) like the Gaza flotilla. In coing to this conclusion I have the University of Surrey and CRESS in mind. - Ian Parker On 26 June 2010 14:36, John G. Rose johnr...@polyplexic.com wrote: -Original Message- From: Ian Parker [mailto:ianpark...@gmail.com] How do you solve World Hunger? Does AGI have to. I think if it is truly G it has to. One way would be to find out what other people had written on the subject and analyse the feasibility of their solutions. Yes, that would show the generality of their AGI theory. Maybe a particular AGI might be able to work with some problems but plateau out on its intelligence for whatever reason and not be able to work on more sophisticated issues. An AGI could be hardcoded perhaps and not improve much, whereas another AGI might improve to where it could tackle vast unknowns at increasing efficiency. There are common components in tackling unknowns, complexity classes for example, but some AGI systems may operate significantly more efficiently and improve. Human brains at some point may plateau without further augmentation though I'm not sure we have come close to what the brain is capable of. John --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/? https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com agi | https://www.listbox.com/member/archive/303/=now Archives https://www.listbox.com/member/archive/rss/303/ | https://www.listbox.com/member/?; Modify 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/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
On 27 June 2010 21:25, John G. Rose johnr...@polyplexic.com wrote: It's just that something like world hunger is so complex AGI would have to master simpler problems. I am not sure that that follows necessarily. Computing is full of situations where a seemingly simple problem is not solved and a more complex one is. I remember posting some time ago on Cassini. Also, there are many people and institutions that have solutions to world hunger already and they get ignored. Indeed. AGI in the shape of a search engine would find these solutions. World Hunger might well be soluble *simply because so much work has already been done.* AGI might well start off as search and develop into feasibility ans solutions. So an AGI would have to get established over a period of time for anyone to really care what it has to say about these types of issues. It could simulate things and come up with solutions but they would not get implemented unless it had power to influence. So in addition AGI would need to know how to make people listen... and maybe obey. This is CRESS. CRESS would be an accessible option. IMO I think AGI will take the embedded route - like other types of computer systems - IRS, weather, military, Google, etc. and we become dependent intergenerationally so that it is impossible to survive without. At that point AGI's will have power to influence. Look! The point is this:- 1) An embedded system is AI not AGI. 2) AGI will arise simply because all embedded systems are themselves searchable. - Ian Parker *From:* Ian Parker [mailto:ianpark...@gmail.com] *Sent:* Saturday, June 26, 2010 2:19 PM *To:* agi *Subject:* Re: [agi] The problem with AGI per Sloman Actually if you are serious about solving a political or social question then what you really need is CRESShttp://cress.soc.surrey.ac.uk/web/home. The solution of World Hunger is BTW a political question not a technical one. Hunger is largely due to bad governance in the Third World. How do you get good governance. One way to look at the problem is via CRESS and run simulations in second life. One thing which has in fact struck me in my linguistic researches is this. Google Translate is based on having Gigabytes of bilingual text. The fact that GT is so bad at technical Arabic indicates the absence of such bilingual text. Indeed Israel publishes more papers than the whole of the Islamic world. This is of profound importance for understanding the Middle East. I am sure CRESS would confirm this. AGI would without a doubt approach political questions by examining all the data about the various countries before making a conclusion. AGI would probably be what you would consult for long term solutions. It might not be so good at dealing with something (say) like the Gaza flotilla. In coing to this conclusion I have the University of Surrey and CRESS in mind. - Ian Parker On 26 June 2010 14:36, John G. Rose johnr...@polyplexic.com wrote: -Original Message- From: Ian Parker [mailto:ianpark...@gmail.com] How do you solve World Hunger? Does AGI have to. I think if it is truly G it has to. One way would be to find out what other people had written on the subject and analyse the feasibility of their solutions. Yes, that would show the generality of their AGI theory. Maybe a particular AGI might be able to work with some problems but plateau out on its intelligence for whatever reason and not be able to work on more sophisticated issues. An AGI could be hardcoded perhaps and not improve much, whereas another AGI might improve to where it could tackle vast unknowns at increasing efficiency. There are common components in tackling unknowns, complexity classes for example, but some AGI systems may operate significantly more efficiently and improve. Human brains at some point may plateau without further augmentation though I'm not sure we have come close to what the brain is capable of. John --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/| Modifyhttps://www.listbox.com/member/?;Your Subscription http://www.listbox.com *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://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/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http
RE: [agi] The problem with AGI per Sloman
-Original Message- From: Ian Parker [mailto:ianpark...@gmail.com] So an AGI would have to get established over a period of time for anyone to really care what it has to say about these types of issues. It could simulate things and come up with solutions but they would not get implemented unless it had power to influence. So in addition AGI would need to know how to make people listen... and maybe obey. This is CRESS. CRESS would be an accessible option. Yes, I agree, it looks like that. IMO I think AGI will take the embedded route - like other types of computer systems - IRS, weather, military, Google, etc. and we become dependent intergenerationally so that it is impossible to survive without. At that point AGI's will have power to influence. Look! The point is this:- 1) An embedded system is AI not AGI. 2) AGI will arise simply because all embedded systems are themselves searchable. A narrow embedded system, like say a DMV computer network is not an AGI. But that doesn't mean an AGI could not perform that function. In fact, AGI might arise out of these systems needing to become more intelligent. And an AGI system, that same AGI software may be used for a DMV, a space navigation system, IRS, NASDAQ, etc. it could adapt. .. efficiently. There are some systems that tout multi-use now but these are basically very narrow AI. AGI will be able to apply it's intelligence across domains and should be able to put its feelers into all the particular subsystems. Although I foresee some types of standard interfaces perhaps into these narrow AI computer networks; some sort of intelligence standards maybe, or the AGI just hooks into the human interfaces... An AGI could become a God but also it could do some useful stuff like run everyday information systems just like people with brains have to perform menial labor. John --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
RE: [agi] The problem with AGI per Sloman
-Original Message- From: Ian Parker [mailto:ianpark...@gmail.com] How do you solve World Hunger? Does AGI have to. I think if it is truly G it has to. One way would be to find out what other people had written on the subject and analyse the feasibility of their solutions. Yes, that would show the generality of their AGI theory. Maybe a particular AGI might be able to work with some problems but plateau out on its intelligence for whatever reason and not be able to work on more sophisticated issues. An AGI could be hardcoded perhaps and not improve much, whereas another AGI might improve to where it could tackle vast unknowns at increasing efficiency. There are common components in tackling unknowns, complexity classes for example, but some AGI systems may operate significantly more efficiently and improve. Human brains at some point may plateau without further augmentation though I'm not sure we have come close to what the brain is capable of. John --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
Actually if you are serious about solving a political or social question then what you really need is CRESS http://cress.soc.surrey.ac.uk/web/home. The solution of World Hunger is BTW a political question not a technical one. Hunger is largely due to bad governance in the Third World. How do you get good governance. One way to look at the problem is via CRESS and run simulations in second life. One thing which has in fact struck me in my linguistic researches is this. Google Translate is based on having Gigabytes of bilingual text. The fact that GT is so bad at technical Arabic indicates the absence of such bilingual text. Indeed Israel publishes more papers than the whole of the Islamic world. This is of profound importance for understanding the Middle East. I am sure CRESS would confirm this. AGI would without a doubt approach political questions by examining all the data about the various countries before making a conclusion. AGI would probably be what you would consult for long term solutions. It might not be so good at dealing with something (say) like the Gaza flotilla. In coing to this conclusion I have the University of Surrey and CRESS in mind. - Ian Parker On 26 June 2010 14:36, John G. Rose johnr...@polyplexic.com wrote: -Original Message- From: Ian Parker [mailto:ianpark...@gmail.com] How do you solve World Hunger? Does AGI have to. I think if it is truly G it has to. One way would be to find out what other people had written on the subject and analyse the feasibility of their solutions. Yes, that would show the generality of their AGI theory. Maybe a particular AGI might be able to work with some problems but plateau out on its intelligence for whatever reason and not be able to work on more sophisticated issues. An AGI could be hardcoded perhaps and not improve much, whereas another AGI might improve to where it could tackle vast unknowns at increasing efficiency. There are common components in tackling unknowns, complexity classes for example, but some AGI systems may operate significantly more efficiently and improve. Human brains at some point may plateau without further augmentation though I'm not sure we have come close to what the brain is capable of. John --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?; Powered by Listbox: http://www.listbox.com --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
Colin, Thanks. Do you have access to any of the full articles? I can't make too informed comments about the quality of work of all the guys writing for this journal, but they're certainly raising v. important questions - and this journal appears to have been unjustly ignored by this group. Sloman, for example, seems to be exploring again the idea of a metaprogram (or I'd say, general program vs specialist program), wh. is the core of AGI, as Ben appears to be only v. recently starting to acknowledge: A methodology for making progress is summarised and a novel requirement proposed for a theory of how human minds work: the theory should support a single generic design for a learning, developing system From: Colin Hales Sent: Friday, June 25, 2010 4:30 AM To: agi Subject: Re: [agi] The problem with AGI per Sloman Not sure if this might be fodder for the discussion. The International Journal of Machine Consciousness (IJMC) has just issued Vol 2 #1 here: http://www.worldscinet.com/ijmc/02/0201/S17938430100201.html It has a Sloman article and invited commentary on it. cheers colin hales agi | Archives | Modify Your Subscription --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
But there is some other kind of problem. We should have figured it out by now. I believe that there must be some fundamental computational problem that is standing as the major obstacle to contemporary AGI. Without solving that problem we are going to have to wade through years of incremental advances. I believe that the most likely basis of the problem is efficient logical satisfiability. It makes the most senese given the nature of the computer and the nature of the best theories of mind. I think there must be a computational or physical/computational problem we have yet to clearly identify that goes along with an objection certain philosophers like Chalmers have made about neural correlates, roughly: why should one level of analysis or type of structure (eg neurons, brain regions, dynamically synchronized ensembles of neurons, or even the organism-environment system), have this magic property of consciousness? Since to me at least it seem obvious that the ecological level is the relevant level of analysis at which to find the meaning relevant to biological organisms, my sense is that we can reduce the above problem to a question about meaning/significance, that is: what is it about a system that makes it unified/integrated such that its relationship to other things constitutes a landscape of relevant meaning to the system as a whole. I think that if that an explanation of meaning-to-a-system is either the same as an explanation of first-hand subjectivity, or is closely tied to it, though if subjectivity turns out to be part of a physical problem and not a purely computational one, then we probably won't solve the above-posed problem without such a physical explanation being clarified (not necessarily explained though, just as we don't know what electricity really is for example). All computer software and situated robots that have ever been made are composed of actions or expressions that are meaningful to people, but software or robots have never been created that can refer to their own actions in a way that demonstrates skillful knowledge indicating that they are organized in a truly semantic way, as opposed to a merely programmatic way. --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
On Fri, Jun 25, 2010 at 7:35 PM, rob levy r.p.l...@gmail.com wrote: I think there must be a computational or physical/computational problem we have yet to clearly identify that goes along with an objection certain philosophers like Chalmers have made about neural correlates, roughly: why should one level of analysis or type of structure (eg neurons, brain regions, dynamically synchronized ensembles of neurons, or even the organism-environment system), have this magic property of consciousness? I don't think that it will be understood fully during our lifetimes. And I don't think that the unknown aspects of this is relevant to computer programming. However, the question of subjective meaning is very relevant. rob levy r.p.l...@gmail.com wrote: what is it about a system that makes it unified/integrated such that its relationship to other things constitutes a landscape of relevant meaning to the system as a whole. I think that if that an explanation of meaning-to-a-system is either the same as an explanation of first-hand subjectivity, or is closely tied to it, though if subjectivity turns out to be part of a physical problem and not a purely computational one, then we probably won't solve the above-posed problem without such a physical explanation being clarified (not necessarily explained though, just as we don't know what electricity really is for example). That is interesting. I wonder if there is a way to make that sense of subjectivity and subjective meaning a basic quality of a simple AGI program, and if it could be a valuable elemental method of analyzing the IO data environment. I think objectives are an important method of testing ideas (and idea-like impressions and reactions). And this combination of setting objectives to test ideas and further develop new ideas does seem to lend itself to developing a sense of subjective experience in relation to the 'objects' of the IO data environment. --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
[agi] The problem with AGI per Sloman
One of the problems of AI researchers is that too often they start off with an inadequate understanding of the problems and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty - which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over and over again in AI research is that the amazing advances in computer technology always seemed to suggest that similar advances in AI must be just off the horizon. And the reality is that there have been major advances in AI. In the 1970's a critic stated that he wouldn't believe that AI was possible until a computer was able to beat him in chess. Well, guess what happened and guess what conclusion he did not derive from the experience. One of the problems with critics is that they can be as far off as those whose optimism is absurdly unwarranted. If a broader multi-disciplinary effort was the obstacle to creating AGI, we would have AGI by now. It should be clear to anyone who examines the history of AI or the present day reach of computer programming that a multi-discipline effort is not the key to creating effective AGI. Computers have become pervasive in modern day life, and if it was just a matter of getting people with different kinds of interests involved, it would have been done by now. It is a little like saying that the key to safe deep sea drilling is to rely on the expertise of companies that make billions and billions of dollars and which stand to lose billions by mistakes. While that should make sense, if you look a little more closely, you can see that it doesn't quite work out that way in the real world. Jim Bromer On Thu, Jun 24, 2010 at 7:33 AM, Mike Tintner tint...@blueyonder.co.ukwrote: One of the problems of AI researchers is that too often they start off with an inadequate understanding of the *problems* and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://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/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
Yes... the idea underlying Sloman's quote is why the interdisciplinary field of cognitive science was invented a few decades ago... ben g On Thu, Jun 24, 2010 at 12:05 PM, Jim Bromer jimbro...@gmail.com wrote: Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over and over again in AI research is that the amazing advances in computer technology always seemed to suggest that similar advances in AI must be just off the horizon. And the reality is that there have been major advances in AI. In the 1970's a critic stated that he wouldn't believe that AI was possible until a computer was able to beat him in chess. Well, guess what happened and guess what conclusion he did not derive from the experience. One of the problems with critics is that they can be as far off as those whose optimism is absurdly unwarranted. If a broader multi-disciplinary effort was the obstacle to creating AGI, we would have AGI by now. It should be clear to anyone who examines the history of AI or the present day reach of computer programming that a multi-discipline effort is not the key to creating effective AGI. Computers have become pervasive in modern day life, and if it was just a matter of getting people with different kinds of interests involved, it would have been done by now. It is a little like saying that the key to safe deep sea drilling is to rely on the expertise of companies that make billions and billions of dollars and which stand to lose billions by mistakes. While that should make sense, if you look a little more closely, you can see that it doesn't quite work out that way in the real world. Jim Bromer On Thu, Jun 24, 2010 at 7:33 AM, Mike Tintner tint...@blueyonder.co.ukwrote: One of the problems of AI researchers is that too often they start off with an inadequate understanding of the *problems* and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://www.listbox.com/member/?;Your Subscription http://www.listbox.com/ *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://www.listbox.com/member/?;Your Subscription http://www.listbox.com -- Ben Goertzel, PhD CEO, Novamente LLC and Biomind LLC CTO, Genescient Corp Vice Chairman, Humanity+ Advisor, Singularity University and Singularity Institute External Research Professor, Xiamen University, China b...@goertzel.org “When nothing seems to help, I go look at a stonecutter hammering away at his rock, perhaps a hundred times without as much as a crack showing in it. Yet at the hundred and first blow it will split in two, and I know it was not that blow that did it, but all that had gone before.” --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
I have to agree that a big problem with the field is a lack of understanding of the problems and how they should be solved. I see too many people pursuing solutions to poorly defined problems and without defining why the solution solves the problem. I even see people pursuing solutions to the wrong problems altogether. I also believe that a strong knowledge of existing methods to solve problems is a hindrance in this research. It makes people not want to start from scratch. They just use a method that work to some degree, but is wrong at its foundation. Lately I've begun to consider these problems more carefully and directly. What I've found interesting is that even at an extremely simplified level, the solution is not immediately clear. In fact, there are many solutions. So, given so many solutions to even simplified problems, which one is the right one? This is the reason that AGI is so F-ing hard. The reason is that the right solution is not clear from the simplified problem or complete problem. The number of possible solutions increase in a sort of exponential manner as you add new constraints and complexity. What I've decided lately is to analyze several of these possible solutions in a sort of tree and pursue several for simplified versions. I hope to find a pattern and figure out which path to pursue more than others. I think many should be pursued though. Maybe this can be somewhat automated some day. Each solution has pros and cons. There aren't even a limited number of solutions. With creativity, you can probably find an infinite number of solutions to the same problem. This explains why the brain was able to accidentally come upon A solution. It works. Not necessarily the best, but it works quite well after being tested and refined over billions of years of evolution. Dave On Thu, Jun 24, 2010 at 7:33 AM, Mike Tintner tint...@blueyonder.co.ukwrote: One of the problems of AI researchers is that too often they start off with an inadequate understanding of the *problems* and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://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/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
Let me be very clear about this. Of course a multi-disciplinary approach is helpful! And when AGI becomes a reality, that will be even more obvious. I am only able to follow what I am able to follow thanks to the contemporary philosophers who note it and contribute to it. All that I am saying is that this is not the central problem that still needs to be solved. More and more people with different interests are using computers and their use is more than an electronic filing cabinet. And both pessimists and optimists will have aided in the study. The same thing goes for the discretionists and the weightednists. The rationalists and the intuitionists. The mystics and the supra-materialists. The hackers and the planners. The neural biologists and the ideationists. The dreamers and the pragmatists. But there is some other kind of problem. We should have figured it out by now. I believe that there must be some fundamental computational problem that is standing as the major obstacle to contemporary AGI. Without solving that problem we are going to have to wade through years of incremental advances. I believe that the most likely basis of the problem is efficient logical satisfiability. It makes the most senese given the nature of the computer and the nature of the best theories of mind. Jim Bromer On Thu, Jun 24, 2010 at 12:05 PM, Jim Bromer jimbro...@gmail.com wrote: Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over and over again in AI research is that the amazing advances in computer technology always seemed to suggest that similar advances in AI must be just off the horizon. And the reality is that there have been major advances in AI. In the 1970's a critic stated that he wouldn't believe that AI was possible until a computer was able to beat him in chess. Well, guess what happened and guess what conclusion he did not derive from the experience. One of the problems with critics is that they can be as far off as those whose optimism is absurdly unwarranted. If a broader multi-disciplinary effort was the obstacle to creating AGI, we would have AGI by now. It should be clear to anyone who examines the history of AI or the present day reach of computer programming that a multi-discipline effort is not the key to creating effective AGI. Computers have become pervasive in modern day life, and if it was just a matter of getting people with different kinds of interests involved, it would have been done by now. It is a little like saying that the key to safe deep sea drilling is to rely on the expertise of companies that make billions and billions of dollars and which stand to lose billions by mistakes. While that should make sense, if you look a little more closely, you can see that it doesn't quite work out that way in the real world. Jim Bromer On Thu, Jun 24, 2010 at 7:33 AM, Mike Tintner tint...@blueyonder.co.ukwrote: One of the problems of AI researchers is that too often they start off with an inadequate understanding of the *problems* and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://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/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
[BTW Sloman's quote is a month old] I think he means what I do - the end-problems that an AGI must face. Please name me one true AGI end-problem being dealt with by any AGI-er - apart from the toybox problem. As I've repeatedly said- AGI-ers simply don't address or discuss AGI end-problems. And they do indeed start with solutions - just as you are doing - re the TSP problem and the problem of combinatorial complexity, both of wh. have in fact nothing to do with AGI, and for neither of wh.. can you provide a single example of a relevant AGI problem. One could not make up this total avoidance of the creative problem, And AGI-ers are not just shockingly but obscenely narrow in their disciplinarity/ the range of their problem interests - maths, logic, standard narrow AI computational problems, NLP, a little robotics and that's about it - with by my rough estimate some 90% of human and animal real world problemsolving of no interest to them. That esp. includes their chosen key fields of language, conversation and vision - all of wh. are much more the province of the *arts* than the sciences, when it comes to AGI The fact that creative, artistic problemsolving presents a totally different paradigm to that of programmed, preplanned problemsolving, is of no interest to them - because they lack what educationalists would call any kind of metacognitive ( interdisciplinary) scaffolding to deal with it. It doesn't matter that programming itself, and developing new formulae and theorems - (all the forms IOW of creative maths, logic, programming, science and technology) - the very problemsolving upon wh. they absolutely depend.- also come under artistic problemsolving. So there is a major need for broadening AI AGI education both in terms of culturally creative problemsolving and true culture-wide multidisciplinarity. From: Jim Bromer Sent: Thursday, June 24, 2010 5:05 PM To: agi Subject: Re: [agi] The problem with AGI per Sloman Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over and over again in AI research is that the amazing advances in computer technology always seemed to suggest that similar advances in AI must be just off the horizon. And the reality is that there have been major advances in AI. In the 1970's a critic stated that he wouldn't believe that AI was possible until a computer was able to beat him in chess. Well, guess what happened and guess what conclusion he did not derive from the experience. One of the problems with critics is that they can be as far off as those whose optimism is absurdly unwarranted. If a broader multi-disciplinary effort was the obstacle to creating AGI, we would have AGI by now. It should be clear to anyone who examines the history of AI or the present day reach of computer programming that a multi-discipline effort is not the key to creating effective AGI. Computers have become pervasive in modern day life, and if it was just a matter of getting people with different kinds of interests involved, it would have been done by now. It is a little like saying that the key to safe deep sea drilling is to rely on the expertise of companies that make billions and billions of dollars and which stand to lose billions by mistakes. While that should make sense, if you look a little more closely, you can see that it doesn't quite work out that way in the real world. Jim Bromer On Thu, Jun 24, 2010 at 7:33 AM, Mike Tintner tint...@blueyonder.co.uk wrote: One of the problems of AI researchers is that too often they start off with an inadequate understanding of the problems and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) agi | Archives | Modify Your Subscription agi | Archives | Modify Your Subscription --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
RE: [agi] The problem with AGI per Sloman
I think some confusion occurs where AGI researchers want to build an artificial person verses artificial general intelligence. An AGI might be just a computational model running in software that can solve problems across domains. An artificial person would be much else in addition to AGI. With intelligence engineering and other engineering that artificial person could be built, or some interface where it appears to be a person. And a huge benefit is in having artificial people to do things that real people do. But pursuing AGI need not have to be pursuit of building artificial people. Also, an AGI need not have to be able to solve ALL problems initially. Coming out and asking why some AGI theory wouldn't be able to figure out how to solve some problem like say, world hunger, I mean WTF is that? John From: Mike Tintner [mailto:tint...@blueyonder.co.uk] Sent: Thursday, June 24, 2010 5:33 AM To: agi Subject: [agi] The problem with AGI per Sloman One of the problems of AI researchers is that too often they start off with an inadequate understanding of the problems and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty - which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) agi | https://www.listbox.com/member/archive/303/=now Archives https://www.listbox.com/member/archive/rss/303/ | https://www.listbox.com/member/?; Modify 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/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
Mike, I think your idealistic view of how AGI should be pursued does not work in reality. What is your approach that fits all your criteria? I'm sure that any such approach would be severely flawed as well. Dave On Thu, Jun 24, 2010 at 12:52 PM, Mike Tintner tint...@blueyonder.co.ukwrote: [BTW Sloman's quote is a month old] I think he means what I do - the end-problems that an AGI must face. Please name me one true AGI end-problem being dealt with by any AGI-er - apart from the toybox problem. As I've repeatedly said- AGI-ers simply don't address or discuss AGI end-problems. And they do indeed start with solutions - just as you are doing - re the TSP problem and the problem of combinatorial complexity, both of wh. have in fact nothing to do with AGI, and for neither of wh.. can you provide a single example of a relevant AGI problem. One could not make up this total avoidance of the creative problem, And AGI-ers are not just shockingly but obscenely narrow in their disciplinarity/ the range of their problem interests - maths, logic, standard narrow AI computational problems, NLP, a little robotics and that's about it - with by my rough estimate some 90% of human and animal real world problemsolving of no interest to them. That esp. includes their chosen key fields of language, conversation and vision - all of wh. are much more the province of the *arts* than the sciences, when it comes to AGI The fact that creative, artistic problemsolving presents a totally different paradigm to that of programmed, preplanned problemsolving, is of no interest to them - because they lack what educationalists would call any kind of metacognitive ( interdisciplinary) scaffolding to deal with it. It doesn't matter that programming itself, and developing new formulae and theorems - (all the forms IOW of creative maths, logic, programming, science and technology) - the very problemsolving upon wh. they absolutely depend.- also come under artistic problemsolving. So there is a major need for broadening AI AGI education both in terms of culturally creative problemsolving and true culture-wide multidisciplinarity. *From:* Jim Bromer jimbro...@gmail.com *Sent:* Thursday, June 24, 2010 5:05 PM *To:* agi agi@v2.listbox.com *Subject:* Re: [agi] The problem with AGI per Sloman Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over and over again in AI research is that the amazing advances in computer technology always seemed to suggest that similar advances in AI must be just off the horizon. And the reality is that there have been major advances in AI. In the 1970's a critic stated that he wouldn't believe that AI was possible until a computer was able to beat him in chess. Well, guess what happened and guess what conclusion he did not derive from the experience. One of the problems with critics is that they can be as far off as those whose optimism is absurdly unwarranted. If a broader multi-disciplinary effort was the obstacle to creating AGI, we would have AGI by now. It should be clear to anyone who examines the history of AI or the present day reach of computer programming that a multi-discipline effort is not the key to creating effective AGI. Computers have become pervasive in modern day life, and if it was just a matter of getting people with different kinds of interests involved, it would have been done by now. It is a little like saying that the key to safe deep sea drilling is to rely on the expertise of companies that make billions and billions of dollars and which stand to lose billions by mistakes. While that should make sense, if you look a little more closely, you can see that it doesn't quite work out that way in the real world. Jim Bromer On Thu, Jun 24, 2010 at 7:33 AM, Mike Tintner tint...@blueyonder.co.ukwrote: One of the problems of AI researchers is that too often they start off with an inadequate understanding of the *problems* and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://www.listbox.com/member/?;Your Subscription http://www.listbox.com/ *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://www.listbox.com/member/?;Your
Re: [agi] The problem with AGI per Sloman
John, You're making a massively important point, wh. I have been thinking about recently. I think it's more useful to say that AGI-ers are thinking in terms of building a *complete AGI system* (rather than person) wh. could range from a simple animal robot to fantasies of an all intelligent brain-in-a-box. No AGI-er has (and no team of supercreative AGI-ers could have) even a remotely realistic understanding of how massively complex a feat this would be. I've changed recently to thinking that realistic AGI in the near future will have to concentrate instead (or certainly have one major focus) on what might be called local AGI as opposed to global AGI - getting a robot able to do just *one* or two things in a truly general way - with a very well-defined goal - rather than a true all-round AGI robot system. (more of this another time). Look at Venter - he is not trying to build a complete artificial cell in one - that would be insane, and yet not a tiny fraction of the insanity of present AGI systembuilders' goals. He is taking it one narrow step at a time - one relatively narrow part at a time. That is a law of both natural and machine evolution to wh. I don't think there are any exceptions - from simple to complex in gradual, progressive stages. From: John G. Rose Sent: Thursday, June 24, 2010 6:20 PM To: agi Subject: RE: [agi] The problem with AGI per Sloman I think some confusion occurs where AGI researchers want to build an artificial person verses artificial general intelligence. An AGI might be just a computational model running in software that can solve problems across domains. An artificial person would be much else in addition to AGI. With intelligence engineering and other engineering that artificial person could be built, or some interface where it appears to be a person. And a huge benefit is in having artificial people to do things that real people do. But pursuing AGI need not have to be pursuit of building artificial people. Also, an AGI need not have to be able to solve ALL problems initially. Coming out and asking why some AGI theory wouldn't be able to figure out how to solve some problem like say, world hunger, I mean WTF is that? John From: Mike Tintner [mailto:tint...@blueyonder.co.uk] Sent: Thursday, June 24, 2010 5:33 AM To: agi Subject: [agi] The problem with AGI per Sloman One of the problems of AI researchers is that too often they start off with an inadequate understanding of the problems and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty - which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) agi | Archives | Modify Your Subscription agi | Archives | Modify Your Subscription --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
On Thu, Jun 24, 2010 at 12:52 PM, Mike Tintner tint...@blueyonder.co.ukwrote: [BTW Sloman's quote is a month old] Are you sure it was A. Sloman who wrote or said that? From where I'm sitting it looks like it was Margaret Boden who wrote it. But then again, I am one of those people who sometimes make mistakes. Jim Bromer On Thu, Jun 24, 2010 at 12:52 PM, Mike Tintner tint...@blueyonder.co.ukwrote: [BTW Sloman's quote is a month old] I think he means what I do - the end-problems that an AGI must face. Please name me one true AGI end-problem being dealt with by any AGI-er - apart from the toybox problem. As I've repeatedly said- AGI-ers simply don't address or discuss AGI end-problems. And they do indeed start with solutions - just as you are doing - re the TSP problem and the problem of combinatorial complexity, both of wh. have in fact nothing to do with AGI, and for neither of wh.. can you provide a single example of a relevant AGI problem. One could not make up this total avoidance of the creative problem, And AGI-ers are not just shockingly but obscenely narrow in their disciplinarity/ the range of their problem interests - maths, logic, standard narrow AI computational problems, NLP, a little robotics and that's about it - with by my rough estimate some 90% of human and animal real world problemsolving of no interest to them. That esp. includes their chosen key fields of language, conversation and vision - all of wh. are much more the province of the *arts* than the sciences, when it comes to AGI The fact that creative, artistic problemsolving presents a totally different paradigm to that of programmed, preplanned problemsolving, is of no interest to them - because they lack what educationalists would call any kind of metacognitive ( interdisciplinary) scaffolding to deal with it. It doesn't matter that programming itself, and developing new formulae and theorems - (all the forms IOW of creative maths, logic, programming, science and technology) - the very problemsolving upon wh. they absolutely depend.- also come under artistic problemsolving. So there is a major need for broadening AI AGI education both in terms of culturally creative problemsolving and true culture-wide multidisciplinarity. *From:* Jim Bromer jimbro...@gmail.com *Sent:* Thursday, June 24, 2010 5:05 PM *To:* agi agi@v2.listbox.com *Subject:* Re: [agi] The problem with AGI per Sloman Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over and over again in AI research is that the amazing advances in computer technology always seemed to suggest that similar advances in AI must be just off the horizon. And the reality is that there have been major advances in AI. In the 1970's a critic stated that he wouldn't believe that AI was possible until a computer was able to beat him in chess. Well, guess what happened and guess what conclusion he did not derive from the experience. One of the problems with critics is that they can be as far off as those whose optimism is absurdly unwarranted. If a broader multi-disciplinary effort was the obstacle to creating AGI, we would have AGI by now. It should be clear to anyone who examines the history of AI or the present day reach of computer programming that a multi-discipline effort is not the key to creating effective AGI. Computers have become pervasive in modern day life, and if it was just a matter of getting people with different kinds of interests involved, it would have been done by now. It is a little like saying that the key to safe deep sea drilling is to rely on the expertise of companies that make billions and billions of dollars and which stand to lose billions by mistakes. While that should make sense, if you look a little more closely, you can see that it doesn't quite work out that way in the real world. Jim Bromer On Thu, Jun 24, 2010 at 7:33 AM, Mike Tintner tint...@blueyonder.co.ukwrote: One of the problems of AI researchers is that too often they start off with an inadequate understanding of the *problems* and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://www.listbox.com/member/?;Your Subscription http://www.listbox.com/ *agi* | Archives https://www.listbox.com/member
Re: [agi] The problem with AGI per Sloman
Dave, Re my first point there is no choice whatsoever - you (any serious creative) *have* to start by addressing the creative problem - in this case true AGI end-problems. You have to start, e.g.,, addressing the problem part of your would-be plane, the part that's going to give you take-off, because that then affects all the other parts of the plane/machine. Start anywhere else the odds of irrelevancy would be around IMO 100%, It is truly staggering that this primary law of serious creativity is being flouted - with inevitably zero results. Re the idea of a truly broad education, sure that's v. idealistic. But AGI education should be at the v. least open-minded - willing to consider any form of problemsolving. People should be willing to think, for example, about an alternative form of machine to the TM because again there is no choice - the TM does *not* incorporate or address the creative problemsolving of the programmer upon which it depends. But there is no willingness to think outside the box/frame here. Re flawed goals - yes, I think you I might agree, that almost any goals you set for your AGI will be overambitious. However, if they are insanely overambitious, as in trying to build an entire AGI system, then you can't really learn much from your mistakes. If they are basic, local AGI goals, as I mentioned to John - (and altho I disagree with your particular goals, your overall philosophy seems to be broadly consistent with this idea) - then you can learn from your mistakes, and make your targets more realistic still. From: David Jones Sent: Thursday, June 24, 2010 6:22 PM To: agi Subject: Re: [agi] The problem with AGI per Sloman Mike, I think your idealistic view of how AGI should be pursued does not work in reality. What is your approach that fits all your criteria? I'm sure that any such approach would be severely flawed as well. Dave On Thu, Jun 24, 2010 at 12:52 PM, Mike Tintner tint...@blueyonder.co.uk wrote: [BTW Sloman's quote is a month old] I think he means what I do - the end-problems that an AGI must face. Please name me one true AGI end-problem being dealt with by any AGI-er - apart from the toybox problem. As I've repeatedly said- AGI-ers simply don't address or discuss AGI end-problems. And they do indeed start with solutions - just as you are doing - re the TSP problem and the problem of combinatorial complexity, both of wh. have in fact nothing to do with AGI, and for neither of wh.. can you provide a single example of a relevant AGI problem. One could not make up this total avoidance of the creative problem, And AGI-ers are not just shockingly but obscenely narrow in their disciplinarity/ the range of their problem interests - maths, logic, standard narrow AI computational problems, NLP, a little robotics and that's about it - with by my rough estimate some 90% of human and animal real world problemsolving of no interest to them. That esp. includes their chosen key fields of language, conversation and vision - all of wh. are much more the province of the *arts* than the sciences, when it comes to AGI The fact that creative, artistic problemsolving presents a totally different paradigm to that of programmed, preplanned problemsolving, is of no interest to them - because they lack what educationalists would call any kind of metacognitive ( interdisciplinary) scaffolding to deal with it. It doesn't matter that programming itself, and developing new formulae and theorems - (all the forms IOW of creative maths, logic, programming, science and technology) - the very problemsolving upon wh. they absolutely depend.- also come under artistic problemsolving. So there is a major need for broadening AI AGI education both in terms of culturally creative problemsolving and true culture-wide multidisciplinarity. From: Jim Bromer Sent: Thursday, June 24, 2010 5:05 PM To: agi Subject: Re: [agi] The problem with AGI per Sloman Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over and over again in AI research is that the amazing advances in computer technology always seemed to suggest that similar advances in AI must be just off the horizon. And the reality is that there have been major advances in AI. In the 1970's a critic stated that he wouldn't believe that AI was possible until a computer was able to beat him in chess. Well, guess what happened and guess what conclusion he did not derive from the experience. One of the problems with critics is that they can be as far off as those whose optimism is absurdly unwarranted. If a broader multi-disciplinary effort
Re: [agi] The problem with AGI per Sloman
Are you sure it was A. Sloman who wrote or said that? From where I'm sitting it looks like it was Margaret Boden who wrote it. But then again, I am one of those people who sometimes make mistakes. Jim Bromer And this is indeed another of your mistakes: http://onthehuman.org/2010/05/can-computer-models-help-us-to-understand-human-creativity/ see [ironically] his: CREATIVITY AS A RESPONSE TO COMBINATORICS and nb. his previous paras [wh. I hadn't fully noted] and wh. also agree with me: All this is a brief introduction to the study of the many ways in which biological evolution was under pressure to provided humans and other animals with information-processing mechanisms that are capable of acquiring many different kinds of information and then developing novel ways of using that information to solve any of millions of different problems without having to learn solutions by trial and error, without having to be taught, and without having to imitate behaviour of others. I.e. they are P-creative solutions. I conjecture that these highly practical forms of creativity, which are obviously important in crafts, engineering, science, and many everyday activities at home or at work, are closely related to the mechanisms that also produce artistic forms of creativity. From: Jim Bromer Sent: Thursday, June 24, 2010 6:57 PM To: agi Subject: Re: [agi] The problem with AGI per Sloman On Thu, Jun 24, 2010 at 12:52 PM, Mike Tintner tint...@blueyonder.co.uk wrote: [BTW Sloman's quote is a month old] Are you sure it was A. Sloman who wrote or said that? From where I'm sitting it looks like it was Margaret Boden who wrote it. But then again, I am one of those people who sometimes make mistakes. Jim Bromer On Thu, Jun 24, 2010 at 12:52 PM, Mike Tintner tint...@blueyonder.co.uk wrote: [BTW Sloman's quote is a month old] I think he means what I do - the end-problems that an AGI must face. Please name me one true AGI end-problem being dealt with by any AGI-er - apart from the toybox problem. As I've repeatedly said- AGI-ers simply don't address or discuss AGI end-problems. And they do indeed start with solutions - just as you are doing - re the TSP problem and the problem of combinatorial complexity, both of wh. have in fact nothing to do with AGI, and for neither of wh.. can you provide a single example of a relevant AGI problem. One could not make up this total avoidance of the creative problem, And AGI-ers are not just shockingly but obscenely narrow in their disciplinarity/ the range of their problem interests - maths, logic, standard narrow AI computational problems, NLP, a little robotics and that's about it - with by my rough estimate some 90% of human and animal real world problemsolving of no interest to them. That esp. includes their chosen key fields of language, conversation and vision - all of wh. are much more the province of the *arts* than the sciences, when it comes to AGI The fact that creative, artistic problemsolving presents a totally different paradigm to that of programmed, preplanned problemsolving, is of no interest to them - because they lack what educationalists would call any kind of metacognitive ( interdisciplinary) scaffolding to deal with it. It doesn't matter that programming itself, and developing new formulae and theorems - (all the forms IOW of creative maths, logic, programming, science and technology) - the very problemsolving upon wh. they absolutely depend.- also come under artistic problemsolving. So there is a major need for broadening AI AGI education both in terms of culturally creative problemsolving and true culture-wide multidisciplinarity. From: Jim Bromer Sent: Thursday, June 24, 2010 5:05 PM To: agi Subject: Re: [agi] The problem with AGI per Sloman Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over and over again in AI research is that the amazing advances in computer technology always seemed to suggest that similar advances in AI must be just off the horizon. And the reality is that there have been major advances in AI. In the 1970's a critic stated that he wouldn't believe that AI was possible until a computer was able to beat him in chess. Well, guess what happened and guess what conclusion he did not derive from the experience. One of the problems with critics is that they can be as far off as those whose optimism is absurdly unwarranted. If a broader multi-disciplinary effort was the obstacle to creating AGI, we would have AGI by now. It should be clear to anyone who examines the history of AI or the present day reach of computer programming
Re: [agi] The problem with AGI per Sloman
Mike, start by addressing the creative problem. this phrase doesn't mean anything to me. You haven't properly defined what you mean by creative to me. What do you think the true AGI end-problems are? Try not to use the word creative so much. There possible algorithms that produce high level creative behavior but which are not any more creative themselves than many existing algorithm. I think your over emphasis on creativity is unfounded. Dave On Thu, Jun 24, 2010 at 2:54 PM, Mike Tintner tint...@blueyonder.co.ukwrote: Dave, Re my first point there is no choice whatsoever - you (any serious creative) *have* to start by addressing the creative problem - in this case true AGI end-problems. You have to start, e.g.,, addressing the problem part of your would-be plane, the part that's going to give you take-off, because that then affects all the other parts of the plane/machine. Start anywhere else the odds of irrelevancy would be around IMO 100%, It is truly staggering that this primary law of serious creativity is being flouted - with inevitably zero results. Re the idea of a truly broad education, sure that's v. idealistic. But AGI education should be at the v. least open-minded - willing to consider any form of problemsolving. People should be willing to think, for example, about an alternative form of machine to the TM because again there is no choice - the TM does *not* incorporate or address the creative problemsolving of the programmer upon which it depends. But there is no willingness to think outside the box/frame here. Re flawed goals - yes, I think you I might agree, that almost any goals you set for your AGI will be overambitious. However, if they are insanely overambitious, as in trying to build an entire AGI system, then you can't really learn much from your mistakes. If they are basic, local AGI goals, as I mentioned to John - (and altho I disagree with your particular goals, your overall philosophy seems to be broadly consistent with this idea) - then you can learn from your mistakes, and make your targets more realistic still. *From:* David Jones davidher...@gmail.com *Sent:* Thursday, June 24, 2010 6:22 PM *To:* agi agi@v2.listbox.com *Subject:* Re: [agi] The problem with AGI per Sloman Mike, I think your idealistic view of how AGI should be pursued does not work in reality. What is your approach that fits all your criteria? I'm sure that any such approach would be severely flawed as well. Dave On Thu, Jun 24, 2010 at 12:52 PM, Mike Tintner tint...@blueyonder.co.ukwrote: [BTW Sloman's quote is a month old] I think he means what I do - the end-problems that an AGI must face. Please name me one true AGI end-problem being dealt with by any AGI-er - apart from the toybox problem. As I've repeatedly said- AGI-ers simply don't address or discuss AGI end-problems. And they do indeed start with solutions - just as you are doing - re the TSP problem and the problem of combinatorial complexity, both of wh. have in fact nothing to do with AGI, and for neither of wh.. can you provide a single example of a relevant AGI problem. One could not make up this total avoidance of the creative problem, And AGI-ers are not just shockingly but obscenely narrow in their disciplinarity/ the range of their problem interests - maths, logic, standard narrow AI computational problems, NLP, a little robotics and that's about it - with by my rough estimate some 90% of human and animal real world problemsolving of no interest to them. That esp. includes their chosen key fields of language, conversation and vision - all of wh. are much more the province of the *arts* than the sciences, when it comes to AGI The fact that creative, artistic problemsolving presents a totally different paradigm to that of programmed, preplanned problemsolving, is of no interest to them - because they lack what educationalists would call any kind of metacognitive ( interdisciplinary) scaffolding to deal with it. It doesn't matter that programming itself, and developing new formulae and theorems - (all the forms IOW of creative maths, logic, programming, science and technology) - the very problemsolving upon wh. they absolutely depend.- also come under artistic problemsolving. So there is a major need for broadening AI AGI education both in terms of culturally creative problemsolving and true culture-wide multidisciplinarity. *From:* Jim Bromer jimbro...@gmail.com *Sent:* Thursday, June 24, 2010 5:05 PM *To:* agi agi@v2.listbox.com *Subject:* Re: [agi] The problem with AGI per Sloman Both of you are wrong. (Where did that quote come from by the way. What year did he write or say that.) An inadequate understanding of the problems is exactly what has to be expected by researchers (both professional and amateurs) when they are facing a completely novel pursuit. That is why we have endless discussions like these. What happened over
Re: [agi] The problem with AGI per Sloman
I think there is a great deal of confusion between these two objectives. When I wrote that if you had a car accident due to a fault in AI/AGI and Matt wrote back talking about downloads this was a case in point. I was assuming that you had a system which was intelligent but was *not* a download in any shape or form. Watsonhttp://learning.blogs.nytimes.com/2010/06/23/waxing-philosophical-on-watson-and-artificial-intelligence/is intelligent. I would be interested to know other peoples answers to the 5 questions. 1) Turing test - Quite possibly with modifications. Watson needs to be turned into a chatterbox. This can be done fairly trivially by allowing Watson to store conversation in his database. 2) Meaningless question. Watson could produce results of thought and feed these back in. Watson could design a program by referencing other programs and their comment data. Sinilarly for engineering. 3,4,5 Absolutely not. How do you solve World Hunger? Does AGI have to. I think if it is truly G it has to. One way would be to find out what other people had written on the subject and analyse the feasibility of their solutions. - Ian Parker On 24 June 2010 18:20, John G. Rose johnr...@polyplexic.com wrote: I think some confusion occurs where AGI researchers want to build an artificial person verses artificial general intelligence. An AGI might be just a computational model running in software that can solve problems across domains. An artificial person would be much else in addition to AGI. With intelligence engineering and other engineering that artificial person could be built, or some interface where it appears to be a person. And a huge benefit is in having artificial people to do things that real people do. But pursuing AGI need not have to be pursuit of building artificial people. Also, an AGI need not have to be able to solve ALL problems initially. Coming out and asking why some AGI theory wouldn't be able to figure out how to solve some problem like say, world hunger, I mean WTF is that? John *From:* Mike Tintner [mailto:tint...@blueyonder.co.uk] *Sent:* Thursday, June 24, 2010 5:33 AM *To:* agi *Subject:* [agi] The problem with AGI per Sloman One of the problems of AI researchers is that too often they start off with an inadequate understanding of the *problems* and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/| Modifyhttps://www.listbox.com/member/?;Your Subscription http://www.listbox.com *agi* | Archives https://www.listbox.com/member/archive/303/=now https://www.listbox.com/member/archive/rss/303/ | Modifyhttps://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/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
On Thu, Jun 24, 2010 at 3:21 PM, Mike Tintner tint...@blueyonder.co.ukwrote: Are you sure it was A. Sloman who wrote or said that? From where I'm sitting it looks like it was Margaret Boden who wrote it. But then again, I am one of those people who sometimes make mistakes. Jim Bromer And this is indeed another of your mistakes: http://onthehuman.org/2010/05/can-computer-models-help-us-to-understand-human-creativity/ see [ironically] his: CREATIVITY AS A RESPONSE TO COMBINATORICS and nb. his previous paras [wh. I hadn't fully noted] and wh. also agree with me: All this is a brief introduction to the study of the many ways in which biological evolution was under pressure to provided humans and other animals with information-processing mechanisms that are capable of acquiring many different kinds of information and then developing novel ways of using that information to solve any of millions of different problems without having to learn solutions by trial and error, without having to be taught, and without having to imitate behaviour of others. I.e. they are P-creative solutions. I conjecture that these highly practical forms of creativity, which are obviously important in crafts, engineering, science, and many everyday activities at home or at work, are closely related to the mechanisms that also produce *artistic* forms of creativity. But he goes on to say, Of course, none of this will impress people who don't WANT to believe that machines can be creative. They just need to learn to think more creatively, and that is another one of your mistakes. --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com
Re: [agi] The problem with AGI per Sloman
I suggest we form a team for this purpose ..and I am willing to join From: Mike Tintner tint...@blueyonder.co.uk To: agi agi@v2.listbox.com Sent: Thu, June 24, 2010 2:33:01 PM Subject: [agi] The problem with AGI per Sloman One of the problems of AI researchers is that too often they start off with an inadequate understanding of the problems and believe that solutions are only a few years away. We need an educational system that not only teaches techniques and solutions, but also an understanding of problems and their difficulty — which can come from a broader multi-disciplinary education. That could speed up progress. A. Sloman ( who else keeps saying that?) agi | Archives | Modify Your Subscription --- agi Archives: https://www.listbox.com/member/archive/303/=now RSS Feed: https://www.listbox.com/member/archive/rss/303/ Modify Your Subscription: https://www.listbox.com/member/?member_id=8660244id_secret=8660244-6e7fb59c Powered by Listbox: http://www.listbox.com