Re: [agi] Comments from a lurker...
Mark, On 4/16/08, Mark Waser [EMAIL PROTECTED] wrote: True, but this is inherent with ALL less than perfectly understood systems and is not in any way peculiar to Dr. Eliza. Extrapolations are inherently hazardous, sometimes without reasonable limit. Correct. Part of the point to AGI is to automatically create knowledge bases that are as complete as possible. Dr. Eliza seems to be a reasonable attempt to use a small amount of cherry-picked knowledge to solve a wide but not complete range of unsolved problems of a given type -- and has all of the standard inherent advantages and disadvantages of that approach. Wouldn't you agree? Yes. There were a bunch of them and I don't claim to be a historian. As I understood those methods they used two kinds of expertise - one of which was similar to the symptoms and conditions that I use, and another that guided the repair process. Dr. Eliza does without the guidance. This has the advantage that it works with inept experts, and the disadvantage that it can be less efficient than if it had good guidance. I had to find a grand heuristic to replace expert-entered probabilities and the rest of that guidance. After lots of experimenting, that grand heuristic turned out to be incredibly simple, buried in the symptom weighting for various conditions, being that you count the first potential symptom (or its verified absence) as 80%, the next one as 80% * 20% = 16%, the third as 80% * 4% = 3%, etc. This gives a lot of weighting to the leading symptoms, but nonetheless seemed to work well. Wow! That's a *really* wicked tail-off. Seems really counter-intuitive. Yes - it surprised me too, and it took a bunch of effort for me to get a good handle on why it worked, because I REALLY don't like my software to depend on things that I don't understand. It comes from Shannon's information theory. The amount of information in a datum is most dependent on the attendant noise. If you had a perfect symptom that exactly tracked a cause-and-effect chain link, then you would do best to ignore all other symptoms, regardless of whether they supported or contradicted the perfect symptom. In our less-than-perfect world, the list of potentially useful symptoms is usually short, and the noise comes from other cause-and-effect chain links that may exhibit substantially identical symptoms. If you have two symptoms, one with high noise and one with low noise, you do best by substantially ignoring the noisy symptom. The key to separating links using noisy symptoms is to use more than noisy symptom that hopefully has uncoupled noise. When your knowledge composer KNOWS about the 80% roll-off, then they CAREFULLY select which symptoms to use and which to ignore, for a secondary human effect of keeping the knowledge composer from throwing in everything but the kitchen sink along with the dirty wash water. Note further that unmentioned symptoms are NOT significantly considered in computing the result, only those that are affirmed or denied. This means that if ONLY the third symptom in the list that would only have a 3% effect if among others, has a 100% effect if it is alone. This results in noisy results - Dr. Eliza reports 100% interim probability, but fails to mention the 50% noise factor, and continues to press the user to answer questions about the two symptoms that precede the 3% symptom that is currently driving everything. Note also that the 3% symptom is probably also driving other potential conditions where it may be earlier in the list, and those conditions may are also be inserting their own questions. To separate the various 100%s in interim results, I added a heuristic to slightly reduce the 100% results proportionately to how far down the list that the first confirmed/denied symptom is. In typical use, there are often as many negative results (from denied symptoms) than positive results! What could a negative probability possibly mean? Not only do we have no believable evidence of the associated condition, but if natural forces were to try to force it, that those forces would probably fail approximately the indicated percentage of the time. I'm not sure what you mean by guided the repair process Where the expert's model of a decision tree, questioning, significance of symptoms, etc., is used instead of the engine's own generated one that may annoy the knowledge composer. It is interesting to watch others composing for Dr. Eliza, because they have their own ideas how to proceed in the presence of certain symptoms that may be of wide variance to Dr. Eliza's approach. So far, discussing this with them at length has yielded that that there really isn't any good reason for doing it their way, and by letting Dr. Eliza do its own thing, that inputting is a LOT easier. Note that there are NO expert-entered percentages in the Knowledge.mdb, which seems to result in BETTER operation because experts almost as often lead things astray with myths as guide
Re: [agi] Comments from a lurker...
Steve Richfield said (regarding his Dr Eliza system): I cannot see ANY argument that it is NOT novel AI. Certainly no one expressed any such doubts at the WORLDCOMP conferences where it was presented and demonstrated. Maybe, but WORLDCOMP doesn't appear to be a particularly serious conference. Your papers included many pages of off-topic ranting, that simply would not be accepted at a legitimate venue. Novelty isn't a matter of consensus opinion. It doesn't matter how your system has been recieved by others in the past. Even though your explanation of the system is sketchy, I see a lot in common with Expert Systems and I'm not the only one on this list saying the same thing. Expert systems are more than just fault trees - visit the library of your local university and carefully check out a book on the topic (it doesn't matter if the book is 15 years old). Surely it is in your best interest and in the spirit of honest scientific research to follow up this lead in detail to discover whether you are reinventing the wheel, rather than attempting to ignore our suggestions and argue the point? I've read your additional comments on this list and still stand by my claims that as far as I can tell from your description of Dr Eliza, it probably isn't novel AI. If there is any innovation, my feeling is that it lies in having found a user-interface that may lead to better user acceptance of an Expert System. -Ben --- 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] Comments from a lurker...
Ben, On 4/16/08, Benjamin Johnston [EMAIL PROTECTED] wrote: Steve Richfield said (regarding his Dr Eliza system): I cannot see ANY argument that it is NOT novel AI. Certainly no one expressed any such doubts at the WORLDCOMP conferences where it was presented and demonstrated. Maybe, but WORLDCOMP doesn't appear to be a particularly serious conference. Your papers included many pages of off-topic ranting, that simply would not be accepted at a legitimate venue. Some were accepted and presented. Some were accepted and then dropped for various reasons, some were never accepted, and some are just rough drafts. Novelty isn't a matter of consensus opinion. It doesn't matter how your system has been recieved by others in the past. Even though your explanation of the system is sketchy, I see a lot in common with Expert Systems and I'm not the only one on this list saying the same thing. Expert systems are more than just fault trees - visit the library of your local university and carefully check out a book on the topic (it doesn't matter if the book is 15 years old). Surely it is in your best interest and in the spirit of honest scientific research to follow up this lead in detail to discover whether you are reinventing the wheel, rather than attempting to ignore our suggestions and argue the point? I've read your additional comments on this list and still stand by my claims that as far as I can tell from your description of Dr Eliza, it probably isn't novel AI. If there is any innovation, my feeling is that it lies in having found a user-interface that may lead to better user acceptance of an Expert System. I have been discussing this in depth with Mark. I suggest lurking and jumping in at the end if you disagree with Mark's handling. In a nutshell, as I see it, the fundamental departure is in the cause-and-effect chain link handling. Since this is deeply embedded in the structure of the knowledge base (take a look at the relationships diagram, the Link Types table, etc. in the Knowledge.mdb file), it underlies the operation, much of which is very similar to old Expert Systems. Steve Richfield --- 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] Comments from a lurker...
Steve Josh, You guys ought to get a kick out of this: http://www.physorg.com:80/news127452360.html. We don't need no stinking gigahertz circuits when we can have terahertz guided-wave circuits. That's 1000 times faster than gigahertz (but, of course, you know that). Based on the terahertz radiation portion of the infrared spectrum. Some guys in Utah made it work (I.e., were able to split and exchange signals at terahertz speeds). It's an interesting read. They figure about 10 years to commercial deployment. That might just be in time to save Moore's ass once again. ;-) Enjoy! Cheers, Brad - Original Message - From: Steve Richfield To: agi@v2.listbox.com Sent: Tuesday, April 15, 2008 3:28 PM Subject: Re: [agi] Comments from a lurker... Josh, On 4/15/08, J Storrs Hall, PhD [EMAIL PROTECTED] wrote: On Monday 14 April 2008 04:56:18 am, Steve Richfield wrote: ... My present efforts are now directed toward a new computer architecture that may be more of interest to AGI types here than Dr. Eliza. This new architecture should be able to build new PC internals for about the same cost, using the same fabrication facilities, yet the processors will run ~10,000 times faster running single-thread code. This (massively-parallel SIMD) is perhaps a little harder than you seem to think. I did my PhD thesis on it and led a multi-million-dollar 10-year ARPA-funded project to develop just such an architecture. I didn't see any attachments. Perhaps you could send me some more information about this? Whenever I present this stuff, I always emphasize that there is NOTHING new here, just an assortment of things that are decades old. Hopefully you have some good ideas in there, or maybe even some old ideas that I can attribute new thinking to. The first mistake everybody makes is to forget that the bottleneck for existing processors isn't computing power at all, it's memory bandwidth. All the cruft on a modern processor chip besides the processor is there to ameliorate that problem, not because they aren't smart enough to put more processors on. Got this covered. Each of the ~10K ALUs has ~8 memory banks to work with, for a total of ~80K banks, so there should be no latencies except for inter-ALU communication. Have I missed something here? The second mistake is to forget that processor and memory silicon fab use different processes, the former optimized for fast transistors, the latter for dense trench capacitors. You won't get both at once -- you'll give up at least a factor of ten trying to combine them over the radically specialized forms. Got that covered. Once multipliers and shift matrices are eliminated and only a few adders, pipeline registers, and a little random logic remain, then the entire thing can be fabricated with MEMORY fab technology! Note that memories have been getting smarter (and even associative), e.g. cache memories, and when you look at their addressing, row selection, etc., there is nothing more complex than I am proposing for my ALUs. While the control processor might at first appear to violate this, note that it needs no computational speed, so its floating point and other complex instructions can be emulated on slow memory-compatible logic. The third mistake is to forget that nobody knows how to program SIMD. This is a long and complicated subject. I spent a year at CDC digging some of the last of the nasty bugs out of their Cyber-205 FORTRAN compiler's optimizer and vectorizer, whose job it was to sweep these issues under the rug. There are some interesting alternatives, like describing complex code skeletons and how to vectorize them. When someone writes a loop whose structure is new to the compiler, someone else would have to explain to the computer how to vectorize it. Sounds kludgy, but co0nsidering the man-lifetimes that it takes to write a good vectorizing compiler, this actually works out to much less total effort. I absolutely agree that programmers will quickly fall into two groups - those who get it and make the transition to writing vectorizable code fairly easily, and those who go into some other line of work. They can't even get programmers to adopt functional programming, for god's sake; the only thing the average programmer can think in is BASIC, I can make a pretty good argument for BASIC, as its simplicity makes it almost ideal to write efficient compilers for. Add to that the now-missing MAT statements for simple array manipulations, and you have a pretty serious competitor for all other approaches. or C which is essentially machine-independent assembly. C is only SISD machine independent. When you move to more complex architectures, its paradigm breaks down. Not even LISP. APL, which is the closest approach to a SIMD language, died a decade or so back
Re: [agi] Comments from a lurker...
On Wed, Apr 16, 2008 at 11:31 AM, Steve Richfield [EMAIL PROTECTED] wrote: I have been discussing this in depth with Mark. I suggest lurking and jumping in at the end if you disagree with Mark's handling. In a nutshell, as I see it, the fundamental departure is in the cause-and-effect chain link handling. Since this is deeply embedded in the structure of the knowledge base (take a look at the relationships diagram, the Link Types table, etc. in the Knowledge.mdb file), it underlies the operation, much of which is very similar to old Expert Systems. Steve, Do you have a paper that focuses on conceptual aspects of this cause-and-effect chain thing, and not technical details of specific medical application (or maybe you can summarize it)? How basically this mechanism works, how it learns, does it form new concepts (events), what is the teaching procedure, what capabilities does it have (what computations it can implement), etc. The point is not whether it's similar or dissimilar to old expert systems of whatever, but the implications on its capabilities. If it's in fact not similar to those things, it doesn't answer the real question. -- Vladimir Nesov [EMAIL PROTECTED] --- 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] Comments from a lurker...
True, but this is inherent with ALL less than perfectly understood systems and is not in any way peculiar to Dr. Eliza. Extrapolations are inherently hazardous, sometimes without reasonable limit. Correct. Part of the point to AGI is to automatically create knowledge bases that are as complete as possible. Dr. Eliza seems to be a reasonable attempt to use a small amount of cherry-picked knowledge to solve a wide but not complete range of unsolved problems of a given type -- and has all of the standard inherent advantages and disadvantages of that approach. Wouldn't you agree? There were a bunch of them and I don't claim to be a historian. As I understood those methods they used two kinds of expertise - one of which was similar to the symptoms and conditions that I use, and another that guided the repair process. Dr. Eliza does without the guidance. This has the advantage that it works with inept experts, and the disadvantage that it can be less efficient than if it had good guidance. I had to find a grand heuristic to replace expert-entered probabilities and the rest of that guidance. After lots of experimenting, that grand heuristic turned out to be incredibly simple, buried in the symptom weighting for various conditions, being that you count the first potential symptom (or its verified absence) as 80%, the next one as 80% * 20% = 16%, the third as 80% * 4% = 3%, etc. This gives a lot of weighting to the leading symptoms, but nonetheless seemed to work well. Wow! That's a *really* wicked tail-off. Seems really counter-intuitive. I'm not sure what you mean by guided the repair process. --- 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] Comments from a lurker...
asked why I didn't ... instead of proclaiming that since I didn't use methods applicable to an Apple, that it is bad engineering. It is pretty obvious to me why you don't have current work experience to exhibit, as any organization faced with your interpersonal approaches would promptly fire you. ;-) Maybe my interpersonal skills are good enough that they do exactly what I want them to do and maybe I'm smart enough to constantly tailor them to the environment as appropriate. Steve OK, I'll send you that long paper I mentioned. There are several others, but I hesitate to jam up this forum giving you stuff to read that you will only gloss over and make snide remarks about. Actually, I appreciated this paper. You're clearly a systems guy and are frustrated by people who don't see that almost *everything* is a system and should be dealt with that way. I sympathize fully. My only negative observations are that you don't seem to fully understand the systems yourself before criticizing them -- i.e. you seem to have put in absolutely NO THOUGHT as to why the medical and political systems have *evolved* to the place that they are currently. You need to understand a system *BEFORE* you attempt to change it. Instead, just as with your programming, you assume that -- just because you think you understand the top-most general level -- you understand the entire system. And it is quite clear that you don't take the time to understand systems or people all the way down to the metal. - Original Message - From: Steve Richfield To: agi@v2.listbox.com Sent: Monday, April 14, 2008 4:17 PM Subject: Re: [agi] Comments from a lurker... Mark, On 4/14/08, Mark Waser [EMAIL PROTECTED] wrote: Steve Perhaps you can relate your own experiences in this area. Argument from Authority . . . . but what the heck . . . . :-) Earliest scientific computing papers (one from the science side, one from the computing side) Computer Modeling of Muscle Phosphofructokinase Kinetics Journal of Theoretical Biology, Volume 103, Issue 2, 21 July 1983, Pages 295-312 Mark R. Waser, Lillian Garfinkel, Michael C. Kohn and David Garfinkel A Computer Program for Analyzing Enzyme Kinetic Data Using Graphical Display and Statistical Analysis Computers and Biomedical Research, Volume 17, Issue 3, June 1984, Pages 289-301 Serge D. Schremmer, Mark R. Waser, Michael C. Kohn and David Garfinkel Hardware Integration Project - True Omni-font OCR device (1983-1984) Developed software turning any Apple IIe and any fax machine into a true Omni-font OCR reader pages were solved as cryptograms so even *random* fonts were interpretable used 6502 assembly; unloaded the Apple IIe operating system as necessary (memory problems? what memory problems?) AI Project - Case Method Credit Expert System Shell Builder (1984-1985) Developed in Pascal for Citicorp's FastFinance Leasing System Used by technophobic executives without any problems AI Project - Expert System for Army Logistics Procurement (1986-1987) Developed for/Deployed at Fort Belvoir, VA; Presented at Army Logistics Conference in Williamsburg Part of the Project Manager's Support System AI Project - Project Impact Advisor (1986-1987) Rewrote boss's prototype system implemented in Lisp on special hardware as a PC-based Prolog system Part of the Project Manager's Support System AI/Hardware Project - Neural Network for Diagnosing Thallium Images of the Heart (1987-1988) Successfully convinced top Air Force brass that Air Force doctors were misdiagnosing test pilot check-up images Used Sigma Neural Network hardware boards Hardware Project - Fax Network Switch (1990-1991) Developed for/Deployed by the Australian Government/Embassy for all traffic between Canberra and Washington Subsequently sold to Sony Created multiple terminate-and-stay-resident programs to provide simultaneous 16-fax and dual T1-modem capability under MS-DOS Used Brooktrout 4-port fax boards Hardware Project - Secure Telephone Unit (1991-1992) Developed initial prototype marrying COTS 80286 motherboard, modem, and TI TMS C32000 FPU with custom hardware and software Enhanced and integrated commercially available TI TMS C32000 software for various voice codecs Developed all control software (80286 assembly) Developed all software for debugging custom integrating hardware developed by other company employees Hmmm . . . that's not even ten years with over fifteen to go . . . and I'm boring *myself* to tears despite skipping a bunch of non-relevant stuff . . . . ;-) VERY impressive, but, the subject was the challenges of running real-time AI on non-real-time
Re: [agi] Comments from a lurker...
On Monday 14 April 2008 04:56:18 am, Steve Richfield wrote: ... My present efforts are now directed toward a new computer architecture that may be more of interest to AGI types here than Dr. Eliza. This new architecture should be able to build new PC internals for about the same cost, using the same fabrication facilities, yet the processors will run ~10,000 times faster running single-thread code. This (massively-parallel SIMD) is perhaps a little harder than you seem to think. I did my PhD thesis on it and led a multi-million-dollar 10-year ARPA-funded project to develop just such an architecture. The first mistake everybody makes is to forget that the bottleneck for existing processors isn't computing power at all, it's memory bandwidth. All the cruft on a modern processor chip besides the processor is there to ameliorate that problem, not because they aren't smart enough to put more processors on. The second mistake is to forget that processor and memory silicon fab use different processes, the former optimized for fast transistors, the latter for dense trench capacitors. You won't get both at once -- you'll give up at least a factor of ten trying to combine them over the radically specialized forms. The third mistake is to forget that nobody knows how to program SIMD. They can't even get programmers to adopt functional programming, for god's sake; the only thing the average programmer can think in is BASIC, or C which is essentially machine-independent assembly. Not even LISP. APL, which is the closest approach to a SIMD language, died a decade or so back. Now frankly, a real associative processor (such as described in my thesis -- read it) would be very useful for AI. You can get close to faking it nowadays by getting a graphics card and programming it GPGPU-style. I quit architecture and got back into the meat of AI because I think that Moore's law has won, and the cycles will be there before we can write the software, so it's a waste of time to try end-runs. Associative processing would have been REALLY useful for AI in the 80's, but we can get away without it, now. Josh --- 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] Comments from a lurker...
Josh, On 4/15/08, J Storrs Hall, PhD [EMAIL PROTECTED] wrote: On Monday 14 April 2008 04:56:18 am, Steve Richfield wrote: ... My present efforts are now directed toward a new computer architecture that may be more of interest to AGI types here than Dr. Eliza. This new architecture should be able to build new PC internals for about the same cost, using the same fabrication facilities, yet the processors will run ~10,000 times faster running single-thread code. This (massively-parallel SIMD) is perhaps a little harder than you seem to think. I did my PhD thesis on it and led a multi-million-dollar 10-year ARPA-funded project to develop just such an architecture. I didn't see any attachments. Perhaps you could send me some more information about this? Whenever I present this stuff, I always emphasize that there is NOTHING new here, just an assortment of things that are decades old. Hopefully you have some good ideas in there, or maybe even some old ideas that I can attribute new thinking to. The first mistake everybody makes is to forget that the bottleneck for existing processors isn't computing power at all, it's memory bandwidth. All the cruft on a modern processor chip besides the processor is there to ameliorate that problem, not because they aren't smart enough to put more processors on. Got this covered. Each of the ~10K ALUs has ~8 memory banks to work with, for a total of ~80K banks, so there should be no latencies except for inter-ALU communication. Have I missed something here? The second mistake is to forget that processor and memory silicon fab use different processes, the former optimized for fast transistors, the latter for dense trench capacitors. You won't get both at once -- you'll give up at least a factor of ten trying to combine them over the radically specialized forms. Got that covered. Once multipliers and shift matrices are eliminated and only a few adders, pipeline registers, and a little random logic remain, then the entire thing can be fabricated with *MEMORY* fab technology! Note that memories have been getting smarter (and even associative), e.g. cache memories, and when you look at their addressing, row selection, etc., there is nothing more complex than I am proposing for my ALUs. While the control processor might at first appear to violate this, note that it needs no computational speed, so its floating point and other complex instructions can be emulated on slow memory-compatible logic. The third mistake is to forget that nobody knows how to program SIMD. This is a long and complicated subject. I spent a year at CDC digging some of the last of the nasty bugs out of their Cyber-205 FORTRAN compiler's optimizer and vectorizer, whose job it was to sweep these issues under the rug. There are some interesting alternatives, like describing complex code skeletons and how to vectorize them. When someone writes a loop whose structure is new to the compiler, someone else would have to explain to the computer how to vectorize it. Sounds kludgy, but co0nsidering the man-lifetimes that it takes to write a good vectorizing compiler, this actually works out to much less total effort. I absolutely agree that programmers will quickly fall into two groups - those who get it and make the transition to writing vectorizable code fairly easily, and those who go into some other line of work. They can't even get programmers to adopt functional programming, for god's sake; the only thing the average programmer can think in is BASIC, I can make a pretty good argument for BASIC, as its simplicity makes it almost ideal to write efficient compilers for. Add to that the now-missing MAT statements for simple array manipulations, and you have a pretty serious competitor for all other approaches. or C which is essentially machine-independent assembly. C is only SISD machine independent. When you move to more complex architectures, its paradigm breaks down. Not even LISP. APL, which is the closest approach to a SIMD language, died a decade or so back. Yes. This is a political/psychological issue, as there were its practitioners who learned its hieroglyphics, and the rest of the mere mortals who simply ignored it. No one (that I know of)ever made the obvious simple step of producing a humanized front-end to the language. BTW, APL is still alive in some financial modeling applications. Now frankly, a real associative processor (such as described in my thesis -- read it) would be very useful for AI. You can get close to faking it nowadays by getting a graphics card and programming it GPGPU-style. I quit architecture and got back into the meat of AI because I think that Moore's law has won, and the cycles will be there before we can write the software, so it's a waste of time to try end-runs. Not according to Intel, who sees the ~4GHz limit as being a permanent thing. I sat on my ideas for ~20 years, just waiting for this to happen and blow
Re: [agi] Comments from a lurker...
J Storrs Hall, PhD wrote: ... The third mistake is to forget that nobody knows how to program SIMD. They can't even get programmers to adopt functional programming, for god's sake; the only thing the average programmer can think in is BASIC, or C which is essentially machine-independent assembly. Not even LISP. APL, which is the closest approach to a SIMD language, died a decade or so back. ... Josh Actually I believe that Prograf (a dataflow language) had a programming model that was by far the most SIMD. Much more so than APL. It also died awhile back, trying to transition from the Mac to MSWind95. It did, however, convince me that reasonable programming idioms from SIMD were reasonable. (Actually, I think Prograf could have been implemented as MIMD. Since it was running on a single processor system, though, the actual implementation was serial.) P.S.: versions of APL still exist. The last time I checked the language was called, I believe, J. http://en.wikipedia.org/wiki/J_programming_language (Such a nice searchable name!) They eliminated the special symbols, but I don't remember what they replaced them with. Don't know if the implementation is SIMD. --- 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] Comments from a lurker...
Steve Perhaps you can relate your own experiences in this area. Argument from Authority . . . . but what the heck . . . . :-) Earliest scientific computing papers (one from the science side, one from the computing side) Computer Modeling of Muscle Phosphofructokinase Kinetics Journal of Theoretical Biology, Volume 103, Issue 2, 21 July 1983, Pages 295-312 Mark R. Waser, Lillian Garfinkel, Michael C. Kohn and David Garfinkel A Computer Program for Analyzing Enzyme Kinetic Data Using Graphical Display and Statistical Analysis Computers and Biomedical Research, Volume 17, Issue 3, June 1984, Pages 289-301 Serge D. Schremmer, Mark R. Waser, Michael C. Kohn and David Garfinkel Hardware Integration Project - True Omni-font OCR device (1983-1984) Developed software turning any Apple IIe and any fax machine into a true Omni-font OCR reader pages were solved as cryptograms so even *random* fonts were interpretable used 6502 assembly; unloaded the Apple IIe operating system as necessary (memory problems? what memory problems?) AI Project - Case Method Credit Expert System Shell Builder (1984-1985) Developed in Pascal for Citicorp's FastFinance Leasing System Used by technophobic executives without any problems AI Project - Expert System for Army Logistics Procurement (1986-1987) Developed for/Deployed at Fort Belvoir, VA; Presented at Army Logistics Conference in Williamsburg Part of the Project Manager's Support System AI Project - Project Impact Advisor (1986-1987) Rewrote boss's prototype system implemented in Lisp on special hardware as a PC-based Prolog system Part of the Project Manager's Support System AI/Hardware Project - Neural Network for Diagnosing Thallium Images of the Heart (1987-1988) Successfully convinced top Air Force brass that Air Force doctors were misdiagnosing test pilot check-up images Used Sigma Neural Network hardware boards Hardware Project - Fax Network Switch (1990-1991) Developed for/Deployed by the Australian Government/Embassy for all traffic between Canberra and Washington Subsequently sold to Sony Created multiple terminate-and-stay-resident programs to provide simultaneous 16-fax and dual T1-modem capability under MS-DOS Used Brooktrout 4-port fax boards Hardware Project - Secure Telephone Unit (1991-1992) Developed initial prototype marrying COTS 80286 motherboard, modem, and TI TMS C32000 FPU with custom hardware and software Enhanced and integrated commercially available TI TMS C32000 software for various voice codecs Developed all control software (80286 assembly) Developed all software for debugging custom integrating hardware developed by other company employees Hmmm . . . that's not even ten years with over fifteen to go . . . and I'm boring *myself* to tears despite skipping a bunch of non-relevant stuff . . . . ;-) Mark Good thing that you're smarter than that and know how to trash a machine so your stuff will work. Steve Given that apparently no one else has been able to make commercial speech-to-text work with real-time AI, I'll accept that as a complement. You shouldn't have. It was pure sarcasm. You need to look harder at what is available out there. Real-time speech-to-text is not the problem (though the accuracy rate is still below what is to be preferred -- a problem which your solution does *NOT* address). Fitting real-time speech-to-text into a small enough, friendly enough footprint to work with real-time AI is not the problem (although *you* do seem to be having problems doing it with a *GOOD* engineering solution). Coming up with a worthwhile AI is the problem BUT I haven't seen any sign of such a thing from you. Steve It is unclear what happened for you to make your comments in the tone that you used. On first glance it appears that you simply didn't carefully read the article. For example, did you notice that Nuance actually has a patent on how they suck up 100.0% of the CPU, leaving nothing for concurrent AI programs? How about constructively addressing the technical ISSUES instead of sounding like an idiot by making snide comments. If you can't prevent a program from sucking up 100% of your CPU, you aren't competent to be working at this level. There are *all sorts* of ways to stop evil behavior like this to include: a.. pre-allocating memory to yourself (or your AI) before firing up the offending programming b.. replacing the operating system pointers to the memory allocation routines to your routines which will then lie to the offender about the amount of memory available c.. working on multiple linked boxes The kludges that you are resorting to are just plain *BAD* engineering. There are *ALWAYS* clean work-arounds -- if you're competent enough to find them. Steve Then there is the fact that Dr. Eliza
Re: [agi] Comments from a lurker...
Well, that's embarrassing . . . . flame somebody and realize that you got part of it wrong yourself . . . . ;-) __ Mark If you can't prevent a program from sucking up 100% of your CPU, you aren't competent to be working at this level. There are *all sorts* of ways to stop evil behavior like this to include: a.. pre-allocating memory to yourself (or your AI) before firing up the offending programming b.. replacing the operating system pointers to the memory allocation routines to your routines which will then lie to the offender about the amount of memory available c.. working on multiple linked boxes Duh. Nothing like proposing memory solutions for a CPU problem . . . . ;-) How about the easily applicable solutions (without any work on your part) of running multiple virtual machines on the same box OR (as proposed before) multiple linked boxes. - Original Message - From: Mark Waser To: agi@v2.listbox.com Sent: Monday, April 14, 2008 10:48 AM Subject: Re: [agi] Comments from a lurker... Steve Perhaps you can relate your own experiences in this area. Argument from Authority . . . . but what the heck . . . . :-) Earliest scientific computing papers (one from the science side, one from the computing side) Computer Modeling of Muscle Phosphofructokinase Kinetics Journal of Theoretical Biology, Volume 103, Issue 2, 21 July 1983, Pages 295-312 Mark R. Waser, Lillian Garfinkel, Michael C. Kohn and David Garfinkel A Computer Program for Analyzing Enzyme Kinetic Data Using Graphical Display and Statistical Analysis Computers and Biomedical Research, Volume 17, Issue 3, June 1984, Pages 289-301 Serge D. Schremmer, Mark R. Waser, Michael C. Kohn and David Garfinkel Hardware Integration Project - True Omni-font OCR device (1983-1984) Developed software turning any Apple IIe and any fax machine into a true Omni-font OCR reader pages were solved as cryptograms so even *random* fonts were interpretable used 6502 assembly; unloaded the Apple IIe operating system as necessary (memory problems? what memory problems?) AI Project - Case Method Credit Expert System Shell Builder (1984-1985) Developed in Pascal for Citicorp's FastFinance Leasing System Used by technophobic executives without any problems AI Project - Expert System for Army Logistics Procurement (1986-1987) Developed for/Deployed at Fort Belvoir, VA; Presented at Army Logistics Conference in Williamsburg Part of the Project Manager's Support System AI Project - Project Impact Advisor (1986-1987) Rewrote boss's prototype system implemented in Lisp on special hardware as a PC-based Prolog system Part of the Project Manager's Support System AI/Hardware Project - Neural Network for Diagnosing Thallium Images of the Heart (1987-1988) Successfully convinced top Air Force brass that Air Force doctors were misdiagnosing test pilot check-up images Used Sigma Neural Network hardware boards Hardware Project - Fax Network Switch (1990-1991) Developed for/Deployed by the Australian Government/Embassy for all traffic between Canberra and Washington Subsequently sold to Sony Created multiple terminate-and-stay-resident programs to provide simultaneous 16-fax and dual T1-modem capability under MS-DOS Used Brooktrout 4-port fax boards Hardware Project - Secure Telephone Unit (1991-1992) Developed initial prototype marrying COTS 80286 motherboard, modem, and TI TMS C32000 FPU with custom hardware and software Enhanced and integrated commercially available TI TMS C32000 software for various voice codecs Developed all control software (80286 assembly) Developed all software for debugging custom integrating hardware developed by other company employees Hmmm . . . that's not even ten years with over fifteen to go . . . and I'm boring *myself* to tears despite skipping a bunch of non-relevant stuff . . . . ;-) Mark Good thing that you're smarter than that and know how to trash a machine so your stuff will work. Steve Given that apparently no one else has been able to make commercial speech-to-text work with real-time AI, I'll accept that as a complement. You shouldn't have. It was pure sarcasm. You need to look harder at what is available out there. Real-time speech-to-text is not the problem (though the accuracy rate is still below what is to be preferred -- a problem which your solution does *NOT* address). Fitting real-time speech-to-text into a small enough, friendly enough footprint to work with real-time AI is not the problem (although *you* do seem to be having problems doing it with a *GOOD* engineering solution). Coming up with a worthwhile AI
Re: [agi] Comments from a lurker...
ROTFLMAO! Excellent! Thank you. - Original Message - From: Steve Richfield To: agi@v2.listbox.com Sent: Sunday, April 13, 2008 6:09 PM Subject: Re: [agi] Comments from a lurker... Mark, On 4/13/08, Mark Waser [EMAIL PROTECTED] wrote: I then asked if anyone in the room had a 98.6F body temperature, and NO ONE DID. Try this in a room with normal people. ~3/4 of the general population reaches ~98.6F sometime during the day. The remaining 1/4 of the population have a varying assortment of symptoms generally in the list of hypothyroid symptoms, even though only about 1/4 of those people have any thyroid-related issues. Then look at the patients who enter the typical doctor's practice. There, it is about 50% each way. Then, look at the patients in a geriatric practice, where typically NONE of the people reach 98.6F anytime during the day. You'll get almost the same answer. 98.6 is just the Fahrenheit value of a rounded Celsius value -- not an accurate gauge. Wrong. Healthy people quickly move between set points at ~97.4F, ~98.0F, and 98.6F. However, since medical researchers aren't process control people, they have missed the importance of this little detail. My standard temperature is 96.8 -- almost two degrees low -- and this is perfectly NORMAL. Thereby demonstrating the obsolescence of your medical information. NOW I understand! Simply resetting someone from 97.something temperature to 98.6F results in something like another ~20 IQ points. People usually report that it feels like waking up, perhaps for the first time in their entire lives. I can hardly imagine the level of impairment that you must be working though. NO WONDER that you didn't see the idiocy of making your snide comments. Any good medical professional understands this. Only if they have gray hair. This all comes from an old American Thyroid Association study that was published in JAMA to discredit Wilson's Thyroid Syndrome (Now Wilson's Temperature Syndrome, which has since been largely discredited for other reasons) that my article references. There, many healthy people had their temperatures taken at 8:00AM, and they found three groups: 1. People who were ~97.4F 2. People who were ~98.6F 3. People who were somewhere in between. However, if you take a healthy person and plot their temperature through the day, you find that they sleep at 97.4F, and pop up to 98.6F sometime during the first 3 hours after waking up. In short, the ATA study was ENTIRELY consistent with my model and observations. However, inexplicably, the authors concluded that people don't have any set temperature, without providing any explanation as to how they reached that conclusion. However, YOUR temperature is REALLY anomalous and WAY outside the range of the ATA's study, and possibly consistent with serious hypothyroidism. Have you had your TSH tested yet? If not, then fire your present incompetent doctor and find a board-certified endocrinologist. Don't criticize others for your assumptions of what they believe. Why not, when I have read the articles, tested dozens of healthy (and many more unhealthy) people myself, and seen that in light of the observable facts, that some conventional medical dogma absolutely MUST be wrong. Please, please get your temperature fixed before making any more snide postings here. I find your snide comments to be painful, and I strongly suspect that you too will see the errors of your ways and correct them when you finally wake up as discussed above. Steve Richfield --- 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
Re: [agi] Comments from a lurker...
--- Steve Richfield [EMAIL PROTECTED] wrote: Why go to all that work?! I have attached the *populated* Knowledge.mdb file that contains the knowledge that powers the chronic illness demo of Dr. Eliza. To easily view it, just make sure that any version of MS Access is installed on your computer (it is in Access 97 format) and double-click on the file. From there, select the Tables tab, and click on whatever table interests you. I looked at your file. Would I be correct that if I described a random health problem to Dr. Eliza that it would suggest that my problem is due to one of: - Low body temperature - Fluorescent lights - Consuming fructose in the winter - Mercury poisoning from amalgam fillings and vaccines - Aluminum cookware - Hydrogenated vegetable oil - Working a night shift - Aspirin (causes macular degeneration) - Or failure to accept divine intervention? Is that it, or is there a complete medical database somewhere, or the capability of acquiring this knowledge? Do you have a medical background, or have you consulted with doctors in building the database? BTW, regarding processes that use 100% of CPU in Windows. Did you try Ctrl-Alt-Del to bring up the task manager, then right click on the process and change its priority? -- Matt Mahoney, [EMAIL PROTECTED] --- 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] Comments from a lurker...
On Mon, Apr 14, 2008 at 4:17 PM, Steve Richfield [EMAIL PROTECTED] wrote: You've merely been a *TROLL* and gotten the appropriate response. Thanks for playing but we have no parting gifts for you. Who is the we you are referencing? Do you have a mouse in your pocket, or is that the Royal we? YOU are the only snide asshole/troll whom I have had the displeasure of observing on this forum. Can you point to anyone ELSE here who acts as you do? I don't want to participate in calling anyone a Troll. What I have observed of Matt's online presence, he was giving you an opportunity to disprove the Troll status rather than transparently ignoring you. I'm guessing he'll simply give up soon. I have little interest in downloading your software and tables and arcane howto for making it all work. In my opinion, you really can't call your product AGI until I can converse with it directly - either via it's own email address or (for a 'real-time' Turing test) an IRC channel. How difficult would it be for you to extend the Dr Eliza interface with an IRC bot frontend? If it is as accurate as you claim, it might help a lot more people by dispensing see a REAL doctor to get X checked out than as ... well, whatever it is now. Even with an accuracy rate that exceeds average doctors, I'll be as likely to dismiss it as I would dismiss a real doctor - but the machine doesn't need to play golf or drive expensive cars so it can devote the time that people can't (or won't). [I had a doctor say, Your iron level is too low, eat more red meat. followed immediately with, Your cholesterol is too high, eat less red meat. I was thinking, Your diagnosis is unusable, I want my co-pay back ] --- 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] Comments from a lurker...
Mike, On 4/14/08, Mike Dougherty [EMAIL PROTECTED] wrote: I have little interest in downloading your software and tables and arcane howto for making it all work. In my opinion, you really can't call your product AGI until I can converse with it directly - either via it's own email address or (for a 'real-time' Turing test) an IRC channel. I will concede that AGI has little interest in Dr. Eliza, and I have little interest in AGI as it seems to be individually defined here. Hence, I plead no contest to this statement. How difficult would it be for you to extend the Dr Eliza interface with an IRC bot frontend? I have looked extensively at this. There are a number of issues: 1. It won't be widely useful without a LOT more knowledge. Remember, my choice of CHRONIC illness - those conditions that doctors can do little/nothing for, yet with the advancement of various sorts of alternative health care approaches, many of these DO have effective interventions. People typically having these conditions fall into some particular social categories: a. The elderly, many of whom won't talk with anyone who doesn't have an MD. If they had only gone down to the nearest clinic and saw the naturopath on duty, many of them wouldn't have their chronic condition. b. The poor, who can only qualify for mainstream MD care without paying for it themselves, and they don't have the money for such risky investments. c. In any case, most people with chronic health conditions do NOT have Internet access! 2. conversing with Dr. Eliza can be frustrating, because it insists on talking about whatever it sees as pivotal, and has no internal ability to converse about whatever it is that the patient thinks is important. More often than not, some passing indirect mention of a seemingly irrelevant symptom will turn out to be the clue that puts it all together, so Dr. Eliza may start asking about that symptom to make sure that it is real since so much depends on it. I really can't imagine Dr. Eliza ever competing for ANY Turing-related prize, because it so completely lacks the personal touch. 3. My present front runner plan is to lurk on many health-related sites, analyze every posting, and wait until it sees enough to be really sure about saying something (has seen enough to propose a complete cure), and then post the questions or reply as appropriate. Alternatively, service emails, which encourages people to write carefully thought out problem statements. If it is as accurate as you claim, Obviously, it is no better than its knowledge base. it might help a lot more people by dispensing see a REAL doctor to get X checked out than as ... well, whatever it is now. I agree. An alternative plan that might be worth a LOT of money is to forge a relationship with a nationwide medical provider like Group Health. Dr. Eliza is pretty good at dragging out the details even if you don't look at its opinions about them. If you like the advice and it requires medication, then just click the button and show up at the Group Health pharmacy, show your ID, and pick up your meds. If you reject its advice, at least your doctor can read the health statement a LOT faster than he can listen to you talk. No matter what happens, the provider would come out ahead. There are a number of political pitfalls in this, but I am still looking for just the right provider to do this with. Even with an accuracy rate that exceeds average doctors, I'll be as likely to dismiss it as I would dismiss a real doctor - but the machine doesn't need to play golf or drive expensive cars so it can devote the time that people can't (or won't). The whole thing hinges around *difficult* problems, *chronic* illnesses, etc.If you doctor can fix a problem, then you don't need Dr. Eliza, though the price is certainly right. However, when your doctor tells you to cancel your magazine subscriptions, as mine once did, then at least some people open their minds to alternative advice. [I had a doctor say, Your iron level is too low, eat more red meat. followed immediately with, Your cholesterol is too high, eat less red meat. Please excuse me for a moment while I change hats... Iron (a pure free radical) levels are regulated by your central metabolic control system to keep the total free radical level where it wants it to be. Most doctors make such opinions without testing, and sometimes the levels are low FOR A GOOD REASON. One fellow from Australia was downwind from some British A-bombs that were tested, so he was full of free radicals from the fallout. His iron levels were regulated to be low. He could (and did for a while) eat iron pills like candy and his levels didn't move a bit. Most of the iron hype is obsolete by decades, and comes from old Geritol ads. I presume that your doctor had his/her share of gray hair? Drug companies have literally bought and paid for laboratories to lower the normal range for cholesterol in order to sell more pills.
Re: [agi] Comments from a lurker...
Most people haven't a clue what it takes to make real-time AI really work. Wow. I'm really unimpressed. First off, your list of kludges that *you* have to go through to make *your* program work is truly frightening. You are obviously very invested in the idea that you are much smarter than the entire rest of the software industry. Most people would assume that newer versions of software are generally better (except for too many added features). Good thing that you're smarter than that and know how to trash a machine so your stuff will work. Then there is the fact that Dr. Eliza operates according to principles that aren't taught in any school and would be unfamiliar without some external education. Sounds like voodoo to me -- unless you have all this stuff written up so that you can provide this education (and the education can be validated). Didn't think so. Hence, I see a LOT of frustration and probably no benefit from such a posting. Or a LOT of laughter and no benefit to you.;-) - Original Message - From: Steve Richfield To: agi@v2.listbox.com Sent: Sunday, April 13, 2008 4:12 PM Subject: Re: [agi] Comments from a lurker... Mark, On 4/12/08, Mark Waser [EMAIL PROTECTED] wrote: So . . . . are you willing to immediately release your current Dr. Eliza code to Open Source and let us see it and help humanity together? Most people haven't a clue what it takes to make real-time AI really work. I have attached an article that explains some of it. You must first remove a lot of problematical software from a new target computer, install a bunch of obsolete versions of software, go through ~2 hours of install process, and often spend another couple of hours (and sometimes a couple of days) making small code changes to deal with some new environmental factor, e.g. an unfamiliar version of Windows. Then there is the fact that Dr. Eliza operates according to principles that aren't taught in any school and would be unfamiliar without some external education. The way that I show its operation in person, is to call up the maintenance panel and start explaining how the dozens of tables work together to make it all work. Sometimes I make small changes to demonstrate resulting changes in operation. There are ~100K lines of code to do NOTHING other than what the tables tell it to do, and sweep the irregularities of all underlying systems under the rug. When you understand the tables, you then understand Dr. Eliza. Without that understanding, the code would be meaningless; and with that understanding, the code would be redundant. To illustrate, a friend made much of it work in German, including things that are uniquely German like run-on words, by simply making appropriate table entries. Hence, I see a LOT of frustration and probably no benefit from such a posting. Steve Richfield - Original Message - From: Steve Richfield To: agi@v2.listbox.com Sent: Saturday, April 12, 2008 1:53 AM Subject: Re: [agi] Comments from a lurker... Vladimir, On 4/11/08, Vladimir Nesov [EMAIL PROTECTED] wrote: Hence, simulational System Dynamics must be confined to systems whose operation can be observed or instrumented. Unfortunately, this lets out most of the REALLY important real-world problems, especially medicine, from simulated solution. That reasoning new cures for medical conditions that are unknown to the computer at once appears to be SO difficult, yet is relatively easy given the right approach, is why I/we chose chronic illness, the hardest part of medicine, as our demo. Why does it follow? There is only a difference of degree. If you've got a messy real-world problem, you know little, if you have an algorithm giving the solution, you know all. The trick is to be able to benefit from many intermediate grades of specification. This has two different answers: #1: When your doctor has just told you that you have something incurable (I have been there) is a really bad time to start a large research project, ESPECIALLY when the answers are already out there, but in small fragments that must be strung together. My own illness took me 4 months to locate the pieces and string them together. This should have only taken a few minutes with something like Dr. Eliza. Why bother simulating something when the research has already been done? #2: The entire world is working on thousands of important research problems. Yes, you CAN apply SD principles and develop a simulation that may help with one of those problems, despite its imperfections. Many of the millions of people in the world are applying SD principles to the thousands of problems right now. Yes, anything that can help with such efforts would be very useful
Re: [agi] Comments from a lurker...
Mike, On 4/12/08, Mike Tintner [EMAIL PROTECTED] wrote: Steve:If you've got a messy real-world problem, you know little, if you have an algorithm giving the solution, you know all. This is the bit where, like most, you skip over the nature of AGI - messy real-world problems. What you're saying is: hey if you've got a messy problem, it's great, nay perfect if you have a neat solution. Contradiction in terms and reality. If it's messy, there isn't a neat solution. However, there are MANY interesting points in between these two extremes. Typically, given the best experts (quotes used to highlight the fact that claiming expertise in something that is poorly understood, as doctors routinely do, is a bit of an oxymoron) available, you can identify several cause-and-effect chain links that are contributing to your problem, even though there remains most of the problem that you still do NOT understand. If you can ONLY identify a cure to a single link between the root cause and the self-sustaining loop at the end, and identify any way at all to temporarily interrupt (doctors call this a treatment) any link in the self-sustaining loop at the end, you can permanently cure the difficult problem, even though most of it remains a complete mystery. That this simple fact has remained hidden has misled AI and AGI, and will continue to mislead it until everyone involved understands this. Take most cancers. If you have one, what do you do? Well, there are a lot of people out there offering you a lot of v. conflicting treatments and proposals, and there is no neat, definitive answer to your problem. Only because various misdirected interests are misleading the process. To illustrate, about a year ago I delivered a presentation to a roomfull of cancer survivors (and people who were trying to survive it). I explained the complex part that body temperature apparently played, and exactly why it was almost unknown for a cancer patient to have a normal 98.6F=37C body temperature. I then asked if anyone in the room had a 98.6F body temperature, and NO ONE DID. THERE is a pretty definitive answer, but getting it out to the experts is probably impossible because they have other dysfunctional models to use. I have an article about this if you would like it. There is a safe and simple one-day cure for erroneous body temperature, yet no cancer sufferer that I know of has ever done it!!! That's the kind of problem a human general intelligence has to deal with, and was designed to deal with. Above is a simple case where even when presented with the answer, there is no way of propagating it to the rest of the human race. I have a friend who is the Director of Research for the Medical Center of a major University, whose own personal surgical experiences supported everything I said so he openly accepted it. I spent 4 hours discussing various approaches to getting this message out. His take - there was no path that he could identify to accomplish this. The detailed explanations of the paths that we considered would fill a small book. Places like Wikipedia have a filtering process that is guaranteed to block any such postings. In short, I wouldn't look at human general intelligence too closely, as except for some rare cases, it too is an oxymoron. It would be MUCH easier to build a really intelligent system than to build a humanly intelligent system. Not the neat ones. (And how do I communicate that to you - get you other AGI-ers to focus on that? Because what you'll do is say: Oh sure it's messy, but there's gotta be a neat solution. You won't be able to stay with the messiness. It's too uncomfortable. My communication problem is in itself a messy one - like most problems of communicating to other people, e.g. how do you sell your AGI system or get funding?) YES, there IS a topic of mutual interest. There used to be people called venture capitalists, but people doing this function no longer exist. There are now people calling themselves venture capitalists whom people used to call investment bankers. There are angel investors who do the initial seed investing, but who lack the resources to follow up with major investments once the seed investment has succeeded. In short, I have sort of given up on finding anyone who has the CAPACITY to invest in any sort of AI/AGI, as all investors have money raised on a prospectus which, upon careful reading, guaranteed that they will NOT invest in AI/AGI. Some of the common exclusional reasons include: 1. Where are your paying customers? 2. What prior University research is this built upon? 3. Where is your intellectual property protection? 4. Where am I going to find other investors with whom to share the risk? Steve Richfield --- agi Archives: http://www.listbox.com/member/archive/303/=now RSS Feed: http://www.listbox.com/member/archive/rss/303/ Modify Your Subscription:
Re: [agi] Comments from a lurker...
Josh, On 4/12/08, J Storrs Hall, PhD [EMAIL PROTECTED] wrote: On Friday 11 April 2008 03:17:21 pm, Steve Richfield wrote: Steve: If you're saying that your system builds a model of its world of discourse as a set of non-linear ODEs (which is what Systems Dynamics is bout) then I (and presumably Richard) are much more likely to be interested... No it doesn't. Instead, my program is designed to work on systems that are not nearly enough known to model. THAT is the state of the interesting (at least to me) part of the real world. If the programmer builds the model of the world beforehand, and the system uses it, it's just standard narrow AI. If the system builds the model itself from unstructured inputs, it's AGI. ... and if the computer can work with a very incomplete portion of a model, then it is USEFUL AGI. In some sense, we know how to do that: it's called the scientific method. However, as normally explained, it leaves a lot to intuition. Form a theory isn't too far from and then a miracle occurs. In other words, we need to be a little more explicit in how our system will form a theory. Perhaps a good way to characterize any given AGI is to specify: (a) what form are its hypotheses in (b) how are they generated (c) how are they tested (d) how are they revised Would it be fair to say that Dr. Eliza tries to form a causal net / influence diagram type structure? Sort of. What it DOES do is to identify isolated cause-and-effect chain links and assign probabilities to their existence in the present problem, WITHOUT presently attempting to thread them all together. In actual operation, most probabilities are either 90% or 10% after the first few questions. Yes, there would be some benefit to such threading and that is planned for the future, but it is common/usual to be able to solve problems with some tiny number of the links being identified. Two must be identified (one near the root cause and one in the self-sustaining loop in the cause and effect chain) to effect a cure, and given that the lengths of the chains leading to self-sustaining loops is typically ~twice as long as the self-sustaining loops themselves, it typically takes ~4 links to be identified to effect a permanent cure, though this number can be as low as 2 or high without limit. In short, I proclaim your definition of AGI as being nearly useless, because it requires FAR more than necessary information to operate. Of course, if you happen to have all that much information, it would sure be nice to be able to fully utilize it - something that the Dr. Eliza approach was never intended to do. Steve Richfield --- 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
Re: [agi] Comments from a lurker...
Mark, On 4/13/08, Mark Waser [EMAIL PROTECTED] wrote: Most people haven't a clue what it takes to make real-time AI really work. Wow. I'm really unimpressed. Perhaps you can relate your own experiences in this area. First off, your list of kludges that *you* have to go through to make *your* program work is truly frightening. Yes, I agree. You are obviously very invested in the idea that you are much smarter than the entire rest of the software industry. No, I was only pointing out that once you take something like speech-to-text and try to embed it into a REAL-TIME AI application, that unexpected problems crop up that the original designers never did consider, and STILL haven't considered. Also, that paper is now 2 years old and RAMs are now much larger, so things have changed a bit. Most people would assumMark,e that newer versions of software are generally better (except for too many added features). Yes, I agree, as I also initially made that same assumption. Good thing that you're smarter than that and know how to trash a machine so your stuff will work. Given that apparently no one else has been able to make commercial speech-to-text work with real-time AI, I'll accept that as a complement. However... It is unclear what happened for you to make your comments in the tone that you used. On first glance it appears that you simply didn't carefully read the article. For example, did you notice that Nuance actually has a patent on how they suck up 100.0% of the CPU, leaving nothing for concurrent AI programs? How about constructively addressing the technical ISSUES instead of sounding like an idiot by making snide comments. Then there is the fact that Dr. Eliza operates according to principles that aren't taught in any school and would be unfamiliar without some external education. Sounds like voodoo to me -- unless you have all this stuff written up so that you can provide this education (and the education can be validated). Didn't think so. Perhaps you missed the fact that I already posted that I have several articles that I would gladly send to anyone who requested them. However, there ARE limits to just how much can be packed into a published article. One of them even secured special permission to exceed the maximum length limit, when the WORLDCOMP conference committee couldn't suggest ANY part of it that could be omitted without damaging the rest of it. Hence, I see a LOT of frustration and probably no benefit from such a posting. Or a LOT of laughter and no benefit to you.;-) I am new here, having only made one posting and answered queries to that posting. However, if this were MY group, I would remove you as a member for making such snide comments rather than simply explaining your issues and asking for anything you see is missing, like explanatory articles. People working in AI/AGI get LOTS of derision from the rest of CS (and you certainly sound like you come from that extraction) and we certainly don't need any more here, on what should be a safe forum to express our ideas. Ben Goertzel, are you awake here? This one is in YOUR court. Steve Richfield -- Original Message - *From:* Steve Richfield [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, April 13, 2008 4:12 PM *Subject:* Re: [agi] Comments from a lurker... Mark, On 4/12/08, Mark Waser [EMAIL PROTECTED] wrote: So . . . . are you willing to immediately release your current Dr. Eliza code to Open Source and let us see it and help humanity together? Most people haven't a clue what it takes to make real-time AI really work. I have attached an article that explains some of it. You must first remove a lot of problematical software from a new target computer, install a bunch of obsolete versions of software, go through ~2 hours of install process, and often spend another couple of hours (and sometimes a couple of days) making small code changes to deal with some new environmental factor, e.g. an unfamiliar version of Windows. Then there is the fact that Dr. Eliza operates according to principles that aren't taught in any school and would be unfamiliar without some external education. The way that I show its operation in person, is to call up the maintenance panel and start explaining how the dozens of tables work together to make it all work. Sometimes I make small changes to demonstrate resulting changes in operation. There are ~100K lines of code to do NOTHING other than what the tables tell it to do, and sweep the irregularities of all underlying systems under the rug. When you understand the tables, you then understand Dr. Eliza. Without that understanding, the code would be meaningless; and with that understanding, the code would be redundant. To illustrate, a friend made much of it work in German, including things that are uniquely German like run-on words, by simply making appropriate table entries. Hence
Re: [agi] Comments from a lurker...
Mark, On 4/13/08, Mark Waser [EMAIL PROTECTED] wrote: I then asked if anyone in the room had a 98.6F body temperature, and NO ONE DID. Try this in a room with normal people. ~3/4 of the general population reaches ~98.6F sometime during the day. The remaining 1/4 of the population have a varying assortment of symptoms generally in the list of hypothyroid symptoms, even though only about 1/4 of those people have any thyroid-related issues. Then look at the patients who enter the typical doctor's practice. There, it is about 50% each way. Then, look at the patients in a geriatric practice, where typically NONE of the people reach 98.6F anytime during the day. You'll get almost the same answer. 98.6 is just the Fahrenheit value of a rounded Celsius value -- not an accurate gauge. Wrong. Healthy people quickly move between set points at ~97.4F, ~98.0F, and 98.6F. However, since medical researchers aren't process control people, they have missed the importance of this little detail. My standard temperature is 96.8 -- almost two degrees low -- and this is perfectly NORMAL. Thereby demonstrating the obsolescence of your medical information. NOW I understand! Simply resetting someone from 97.something temperature to 98.6F results in something like another ~20 IQ points. People usually report that it feels like waking up, perhaps for the first time in their entire lives. I can hardly imagine the level of impairment that you must be working though. NO WONDER that you didn't see the idiocy of making your snide comments. Any good medical professional understands this. Only if they have gray hair. This all comes from an old American Thyroid Association study that was published in JAMA to discredit Wilson's Thyroid Syndrome (Now Wilson's Temperature Syndrome, which has since been largely discredited for other reasons) that my article references. There, many healthy people had their temperatures taken at 8:00AM, and they found three groups: 1. People who were ~97.4F 2. People who were ~98.6F 3. People who were somewhere in between. However, if you take a healthy person and plot their temperature through the day, you find that they sleep at 97.4F, and pop up to 98.6F sometime during the first 3 hours after waking up. In short, the ATA study was ENTIRELY consistent with my model and observations. However, inexplicably, the authors concluded that people don't have any set temperature, without providing any explanation as to how they reached that conclusion. However, YOUR temperature is REALLY anomalous and WAY outside the range of the ATA's study, and possibly consistent with serious hypothyroidism. Have you had your TSH tested yet? If not, then fire your present incompetent doctor and find a board-certified endocrinologist. Don't criticize others for your assumptions of what they believe. Why not, when I have read the articles, tested dozens of healthy (and many more unhealthy) people myself, and seen that in light of the observable facts, that some conventional medical dogma absolutely MUST be wrong. Please, please get your temperature fixed before making any more snide postings here. I find your snide comments to be painful, and I strongly suspect that you too will see the errors of your ways and correct them when you finally wake up as discussed above. Steve Richfield == - Original Message - *From:* Steve Richfield [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Sunday, April 13, 2008 4:42 PM *Subject:* Re: [agi] Comments from a lurker... Mike, On 4/12/08, Mike Tintner [EMAIL PROTECTED] wrote: Steve:If you've got a messy real-world problem, you know little, if you have an algorithm giving the solution, you know all. This is the bit where, like most, you skip over the nature of AGI - messy real-world problems. What you're saying is: hey if you've got a messy problem, it's great, nay perfect if you have a neat solution. Contradiction in terms and reality. If it's messy, there isn't a neat solution. However, there are MANY interesting points in between these two extremes. Typically, given the best experts (quotes used to highlight the fact that claiming expertise in something that is poorly understood, as doctors routinely do, is a bit of an oxymoron) available, you can identify several cause-and-effect chain links that are contributing to your problem, even though there remains most of the problem that you still do NOT understand. If you can ONLY identify a cure to a single link between the root cause and the self-sustaining loop at the end, and identify any way at all to temporarily interrupt (doctors call this a treatment) any link in the self-sustaining loop at the end, you can permanently cure the difficult problem, even though most of it remains a complete mystery. That this simple fact has remained hidden has misled AI and AGI, and will continue to mislead it until everyone involved understands
Re: [agi] Comments from a lurker...
Regardless of the fact I may not understand all of the goings on of Dr. Eliza, I would still love to see the source. And perhaps you are unaware, but many good programmers can picture what tables and databases would look like with amazing accuracy with nothing but the code that generates or accesses it. I'm sure there is at least one individual on this list that could look at the code and accurately see the tables. Obviously not the content, but the structure and reason behind its design. On Sun, Apr 13, 2008 at 3:12 PM, Steve Richfield [EMAIL PROTECTED] wrote: Mark, On 4/12/08, Mark Waser [EMAIL PROTECTED] wrote: So . . . . are you willing to immediately release your current Dr. Eliza code to Open Source and let us see it and help humanity together? Most people haven't a clue what it takes to make real-time AI really work. I have attached an article that explains some of it. You must first remove a lot of problematical software from a new target computer, install a bunch of obsolete versions of software, go through ~2 hours of install process, and often spend another couple of hours (and sometimes a couple of days) making small code changes to deal with some new environmental factor, e.g. an unfamiliar version of Windows. Then there is the fact that Dr. Eliza operates according to principles that aren't taught in any school and would be unfamiliar without some external education. The way that I show its operation in person, is to call up the maintenance panel and start explaining how the dozens of tables work together to make it all work. Sometimes I make small changes to demonstrate resulting changes in operation. There are ~100K lines of code to do NOTHING other than what the tables tell it to do, and sweep the irregularities of all underlying systems under the rug. When you understand the tables, you then understand Dr. Eliza. Without that understanding, the code would be meaningless; and with that understanding, the code would be redundant. To illustrate, a friend made much of it work in German, including things that are uniquely German like run-on words, by simply making appropriate table entries. Hence, I see a LOT of frustration and probably no benefit from such a posting. Steve Richfield - Original Message - *From:* Steve Richfield [EMAIL PROTECTED] *To:* agi@v2.listbox.com *Sent:* Saturday, April 12, 2008 1:53 AM *Subject:* Re: [agi] Comments from a lurker... Vladimir, On 4/11/08, Vladimir Nesov [EMAIL PROTECTED] wrote: Hence, simulational System Dynamics must be confined to systems whose operation can be observed or instrumented. Unfortunately, this lets out most of the REALLY important real-world problems, especially medicine, from simulated solution. That reasoning new cures for medical conditions that are unknown to the computer at once appears to be SO difficult, yet is relatively easy given the right approach, is why I/we chose chronic illness, the hardest part of medicine, as our demo. Why does it follow? There is only a difference of degree. If you've got a messy real-world problem, you know little, if you have an algorithm giving the solution, you know all. The trick is to be able to benefit from many intermediate grades of specification. This has two different answers: #1: When your doctor has just told you that you have something incurable (I have been there) is a really bad time to start a large research project, ESPECIALLY when the answers are already out there, but in small fragments that must be strung together. My own illness took me 4 months to locate the pieces and string them together. This should have only taken a few minutes with something like Dr. Eliza. Why bother simulating something when the research has already been done? #2: The entire world is working on thousands of important research problems. Yes, you CAN apply SD principles and develop a simulation that may help with one of those problems, despite its imperfections. Many of the millions of people in the world are applying SD principles to the thousands of problems right now. Yes, anything that can help with such efforts would be very useful, however... A machine that tracks what EVERYONE is doing, collects the fragments of wisdom that come from every project and has the entire world's wisdom to apply to ANY stated problem, whether or not the person stating the problem has any clue at all what lies inside the computer or what to ask. THAT would be thousands of times more valuable than any one SD tool, however successful it might be. THAT is what Dr. Eliza was designed to do. YES, something like Dr. Eliza would be more powerful if people had better SD and other tools to perform their research. As things now stand, the Internet is only a library with absolutely NO ability to take fragments from here
Re: [agi] Comments from a lurker...
Hi Steve, I've read the papers you've posted on this list, and looked at your Access database... Text-to-speech, speech-to-text and the simple textual preprocessing can enhance usability, but they tend to make a system *appear* more intelligent rather than adding real depth to a system. So if we ignore that aspect of Dr Eliza, then it seems to me that what you've got is a simple expert system that is cued by keywords in text (as opposed to a stereotypical expert system that engages in a dialogue with the user directly by asking preprogrammed questions). Am I understanding this correctly? I'm not sure how you could go so far as claiming that this is AGI, or even that this is novel AI. More realistically, I get the impression that your innovation might be in the user-interface (i.e., a textually cued expert system, that prompts the user to refine input text) - potentially interesting as a way of improving user acceptance of expert systems, but better suited to a user interaction forum rather than an AI forum. -Ben James, On 4/13/08, *James Hill* [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: Regardless of the fact I may not understand all of the goings on of Dr. Eliza, I would still love to see the source. And perhaps you are unaware, but many good programmers can picture what tables and databases would look like with amazing accuracy with nothing but the code that generates Currently, everything is entered by hand. or accesses it. I'm sure there is at least one individual on this list that could look at the code and accurately see the tables. Obviously not the content, but the structure and reason behind its design. Why go to all that work?! I have attached the populated Knowledge.mdb file that contains the knowledge that powers the chronic illness demo of Dr. Eliza. To easily view it, just make sure that any version of MS Access is installed on your computer (it is in Access 97 format) and double-click on the file. From there, select the Tables tab, and click on whatever table interests you. There is little hidden here, as magic codes are English words indicating what they do, etc. Note that some tables drive the sentence chopper that chops compound and complex sentences into two or more simple sentences. It isn't perfect, but it GREATLY improves the accuracy of following operations. BTW, one continuing minor challenge is that Dr. Eliza sometimes complains about things that you didn't actually write - at least not as in its complaint. What happens is that the input is rewritten in many passes to become a longer list of simpler sentences, with idioms restated, some spelling errors corrected, etc. By the time that things get around to actually complaining about what came it, it may be somewhat different from what actually came in. This hasn't been a significant problem, and is sometimes a benefit because it shows people how the computer sees the messes that they write. Dr. Eliza makes very few complaints, but when it sees something like If ... then ... it has a pretty good idea that you are NOT describing a problem, but rather you are trying to educate it. Since that is outside of its paradigm of operation, it does complain. Happy reading. With that and about a half of a man-decade of development, you too will have your very own independently developed version of Dr. Eliza! Steve Richfield --- 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
Re: [agi] Comments from a lurker...
Vladimir, On 4/11/08, Vladimir Nesov [EMAIL PROTECTED] wrote: Hence, simulational System Dynamics must be confined to systems whose operation can be observed or instrumented. Unfortunately, this lets out most of the REALLY important real-world problems, especially medicine, from simulated solution. That reasoning new cures for medical conditions that are unknown to the computer at once appears to be SO difficult, yet is relatively easy given the right approach, is why I/we chose chronic illness, the hardest part of medicine, as our demo. Why does it follow? There is only a difference of degree. If you've got a messy real-world problem, you know little, if you have an algorithm giving the solution, you know all. The trick is to be able to benefit from many intermediate grades of specification. This has two different answers: #1: When your doctor has just told you that you have something incurable (I have been there) is a really bad time to start a large research project, ESPECIALLY when the answers are already out there, but in small fragments that must be strung together. My own illness took me 4 months to locate the pieces and string them together. This should have only taken a few minutes with something like Dr. Eliza. Why bother simulating something when the research has already been done? #2: The entire world is working on thousands of important research problems. Yes, you CAN apply SD principles and develop a simulation that may help with one of those problems, despite its imperfections. Many of the millions of people in the world are applying SD principles to the thousands of problems right now. Yes, anything that can help with such efforts would be very useful, however... A machine that tracks what EVERYONE is doing, collects the fragments of wisdom that come from every project and has the entire world's wisdom to apply to ANY stated problem, whether or not the person stating the problem has any clue at all what lies inside the computer or what to ask. THAT would be thousands of times more valuable than any one SD tool, however successful it might be. THAT is what Dr. Eliza was designed to do. YES, something like Dr. Eliza would be more powerful if people had better SD and other tools to perform their research. As things now stand, the Internet is only a library with absolutely NO ability to take fragments from here and there and string them all together to solve a problem. Dr. Eliza's methods makes the information ACTIVE and able to interrelate. I certainly wouldn't discourage anyone from developing better research tools. However, the vast majority of existing research is presently inaccessible unless you know exactly what to ask for, and adding more to this inaccessible lot seems to me to be of diminishing value until something like Dr. Eliza is on everyone's desktop to string the bits of wisdom together to solve everyone's real-world problems. THEN would be a good time to switch efforts as you suggest, when the tools are in place to fully utilize the sorts of things that you are looking to develop. Steve Richfield --- 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
Re: [agi] Comments from a lurker...
On Friday 11 April 2008 03:17:21 pm, Steve Richfield wrote: Steve: If you're saying that your system builds a model of its world of discourse as a set of non-linear ODEs (which is what Systems Dynamics is bout) then I (and presumably Richard) are much more likely to be interested... No it doesn't. Instead, my program is designed to work on systems that are not nearly enough known to model. THAT is the state of the interesting (at least to me) part of the real world. If the programmer builds the model of the world beforehand, and the system uses it, it's just standard narrow AI. If the system builds the model itself from unstructured inputs, it's AGI. In some sense, we know how to do that: it's called the scientific method. However, as normally explained, it leaves a lot to intuition. Form a theory isn't too far from and then a miracle occurs. In other words, we need to be a little more explicit in how our system will form a theory. Perhaps a good way to characterize any given AGI is to specify: (a) what form are its hypotheses in (b) how are they generated (c) how are they tested (d) how are they revised Would it be fair to say that Dr. Eliza tries to form a causal net / influence diagram type structure? Josh --- 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
Re: [agi] Comments from a lurker...
Steve:If you've got a messy real-world problem, you know little, if you have an algorithm giving the solution, you know all. This is the bit where, like most, you skip over the nature of AGI - messy real-world problems. What you're saying is: hey if you've got a messy problem, it's great, nay perfect if you have a neat solution. Contradiction in terms and reality. If it's messy, there isn't a neat solution. Take most cancers. If you have one, what do you do? Well, there are a lot of people out there offering you a lot of v. conflicting treatments and proposals, and there is no neat, definitive answer to your problem. That's the kind of problem a human general intelligence has to deal with, and was designed to deal with. Not the neat ones. (And how do I communicate that to you - get you other AGI-ers to focus on that? Because what you'll do is say: Oh sure it's messy, but there's gotta be a neat solution. You won't be able to stay with the messiness. It's too uncomfortable. My communication problem is in itself a messy one - like most problems of communicating to other people, e.g. how do you sell your AGI system or get funding?) --- 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
Re: [agi] Comments from a lurker...
So . . . . are you willing to immediately release your current Dr. Eliza code to Open Source and let us see it and help humanity together? - Original Message - From: Steve Richfield To: agi@v2.listbox.com Sent: Saturday, April 12, 2008 1:53 AM Subject: Re: [agi] Comments from a lurker... Vladimir, On 4/11/08, Vladimir Nesov [EMAIL PROTECTED] wrote: Hence, simulational System Dynamics must be confined to systems whose operation can be observed or instrumented. Unfortunately, this lets out most of the REALLY important real-world problems, especially medicine, from simulated solution. That reasoning new cures for medical conditions that are unknown to the computer at once appears to be SO difficult, yet is relatively easy given the right approach, is why I/we chose chronic illness, the hardest part of medicine, as our demo. Why does it follow? There is only a difference of degree. If you've got a messy real-world problem, you know little, if you have an algorithm giving the solution, you know all. The trick is to be able to benefit from many intermediate grades of specification. This has two different answers: #1: When your doctor has just told you that you have something incurable (I have been there) is a really bad time to start a large research project, ESPECIALLY when the answers are already out there, but in small fragments that must be strung together. My own illness took me 4 months to locate the pieces and string them together. This should have only taken a few minutes with something like Dr. Eliza. Why bother simulating something when the research has already been done? #2: The entire world is working on thousands of important research problems. Yes, you CAN apply SD principles and develop a simulation that may help with one of those problems, despite its imperfections. Many of the millions of people in the world are applying SD principles to the thousands of problems right now. Yes, anything that can help with such efforts would be very useful, however... A machine that tracks what EVERYONE is doing, collects the fragments of wisdom that come from every project and has the entire world's wisdom to apply to ANY stated problem, whether or not the person stating the problem has any clue at all what lies inside the computer or what to ask. THAT would be thousands of times more valuable than any one SD tool, however successful it might be. THAT is what Dr. Eliza was designed to do. YES, something like Dr. Eliza would be more powerful if people had better SD and other tools to perform their research. As things now stand, the Internet is only a library with absolutely NO ability to take fragments from here and there and string them all together to solve a problem. Dr. Eliza's methods makes the information ACTIVE and able to interrelate. I certainly wouldn't discourage anyone from developing better research tools. However, the vast majority of existing research is presently inaccessible unless you know exactly what to ask for, and adding more to this inaccessible lot seems to me to be of diminishing value until something like Dr. Eliza is on everyone's desktop to string the bits of wisdom together to solve everyone's real-world problems. THEN would be a good time to switch efforts as you suggest, when the tools are in place to fully utilize the sorts of things that you are looking to develop. Steve Richfield -- agi | Archives | Modify Your Subscription --- 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
Re: [agi] Comments from a lurker...
On Friday 11 April 2008 01:59:42 am, Steve Richfield wrote: Your experience with the medical community is not too surprising: I believe that the Expert Systems folks had similar troubles way back when. IMO the Expert Systems people deserved bad treatment! Actually, the medical expert systems of the 80s I had any conection with, such as the glaucoma expert from Rutgers, beat out human doctors in diagnoses within their field of expertise. (And still weren't adopted...) BTW, the attached paper included some remarks about Jay Forrester System Dynamics. Forrester came out of exactly the same background as Cybernetics -- working on automatic radar-directed fire-control systems, at MIT, during WWII. And both his stuff and Cybernetics consists basically of applying feedback and control theory (and general differential analysis) to things ranging from neuroscience to economics. Steve: If you're saying that your system builds a model of its world of discourse as a set of non-linear ODEs (which is what Systems Dynamics is bout) then I (and presumably Richard) are much more likely to be interested... Josh ps -- of course, you know that if you're using Excel to integrate dynamical systems, you are in a state of sin. --- 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
Re: [agi] Comments from a lurker...
Josh, On 4/11/08, J Storrs Hall, PhD [EMAIL PROTECTED] wrote: Actually, the medical expert systems of the 80s I had any conection with, such as the glaucoma expert from Rutgers, beat out human doctors in diagnoses within their field of expertise. (And still weren't adopted...) Yes. There was a presentation at the last WORLDCOMP AI conference where a computer was routinely beating people at identifying details in jaw X-rays, that were needed to make key measurements to fit prosthetics. This sort of an approach works well in closed systems that are well understood, but poorly in wide open systems where the goal is to get some handle on a bad situation, from which to leverage some sort of cure - and possibly even something that has never before been proposed for that particular ailment. How can a (simple) computer possibly do this? Because new (to the computer) cause and effect chains usually have some known (to the computer) links. If just one link in the lead up to the self-sustaining loop in the cause and effect chain is known and has a cure, and if just one link in the self-sustaining loop has any sort of treatment at all, then the condition can easily be cured for once and for all, even though the overall condition and diagnosis is completely unknown, and remains unknown to the computer. BTW, the attached paper included some remarks about Jay Forrester System Dynamics. Forrester came out of exactly the same background as Cybernetics -- working on automatic radar-directed fire-control systems, at MIT, during WWII. And both his stuff and Cybernetics consists basically of applying feedback and control theory (and general differential analysis) to things ranging from neuroscience to economics. Like Joe Weizenbaum, Jay Forrester had a REALLY good idea that he put into his writings, but only as a side note and not any sort of main theme. However, I'll take a good idea from anywhere that I can find it. That idea was that when you analyze many similar systems, like corporations, that you typically find a few common modes of failures from among the countless possible modes of failure. Once you have gotten to that point, you can quickly correct most of the failures by simply screening for the common modes of failure, with no simulation needed. Jay was fixing corporations that are each unique, though they do operate in a standard world. However, people (and circuit boards and most of the other things that people might like computers to fix) are MUCH more alike than corporations, so that most experts already know most of the known modes of failure - with no computerized analysis needed to identify those potential modes of failure. Even military science has remained much as it was millennia ago, as exemplified by writings like *The Art of War*, despite the emergence of modern weapons. For example, in Roman times it took ~2% of a population to occupy it, and that percentage remains true today. That we are attempting to do this with ~1% of the population of Iraq underlies our failures there. This was pointed out early in the Iraq conflict, but those generals were promptly discharged for having a bad attitude. Steve: If you're saying that your system builds a model of its world of discourse as a set of non-linear ODEs (which is what Systems Dynamics is bout) then I (and presumably Richard) are much more likely to be interested... No it doesn't. Instead, my program is designed to work on systems that are not nearly enough known to model. THAT is the state of the interesting (at least to me) part of the real world. In short, I am apparently going where no one has gone before - applying new methods to solving difficult problems in poorly understood systems. I'll gladly leave the easy stuff (modeling well-understood systems) to others. BTW, there is a generally unrecognized principle (except to some experienced System Dynamics types), that the cause and effect chains are LONG and usually involve some lack of understanding among those who designed the systems that we must now deal with. Only the most arrogant would presume their own perfection in comparison with those who designed the world in which we live. Correcting that arrogance is THE primary benefit of System Dynamics, which forces people to code how the systems REALLY work (to make the simulations play like reality) and not just how they THINK that those systems work. Hence, simulational System Dynamics must be confined to systems whose operation can be observed or instrumented. Unfortunately, this lets out most of the REALLY important real-world problems, especially medicine, from simulated solution. That reasoning new cures for medical conditions that are unknown to the computer at once appears to be SO difficult, yet is relatively easy given the right approach, is why I/we chose chronic illness, the hardest part of medicine, as our demo. ps -- of course, you know that if you're using Excel to integrate dynamical systems, you
Re: [agi] Comments from a lurker...
On Fri, Apr 11, 2008 at 11:17 PM, Steve Richfield [EMAIL PROTECTED] wrote: Steve: If you're saying that your system builds a model of its world of discourse as a set of non-linear ODEs (which is what Systems Dynamics is bout) then I (and presumably Richard) are much more likely to be interested... No it doesn't. Instead, my program is designed to work on systems that are not nearly enough known to model. THAT is the state of the interesting (at least to me) part of the real world. In short, I am apparently going where no one has gone before - applying new methods to solving difficult problems in poorly understood systems. I'll gladly leave the easy stuff (modeling well-understood systems) to others. It's what AGI is arguably about: simulating complex processes that are not directly and completely specified. BTW, there is a generally unrecognized principle (except to some experienced System Dynamics types), that the cause and effect chains are LONG and usually involve some lack of understanding among those who designed the systems that we must now deal with. Only the most arrogant would presume their own perfection in comparison with those who designed the world in which we live. Correcting that arrogance is THE primary benefit of System Dynamics, which forces people to code how the systems REALLY work (to make the simulations play like reality) and not just how they THINK that those systems work. Hence, simulational System Dynamics must be confined to systems whose operation can be observed or instrumented. Unfortunately, this lets out most of the REALLY important real-world problems, especially medicine, from simulated solution. That reasoning new cures for medical conditions that are unknown to the computer at once appears to be SO difficult, yet is relatively easy given the right approach, is why I/we chose chronic illness, the hardest part of medicine, as our demo. Why does it follow? There is only a difference of degree. If you've got a messy real-world problem, you know little, if you have an algorithm giving the solution, you know all. The trick is to be able to benefit from many intermediate grades of specification. -- Vladimir Nesov [EMAIL PROTECTED] --- 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
Re: [agi] Comments from a lurker...
- Original Message From: Steve Richfield [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Thursday, April 10, 2008 2:58:09 PM Subject: [agi] Comments from a lurker... [snip] BTW, the principles behind Dr. Eliza are rather unique. I'd be glad to send some papers to anyone who is interested. Briefly, Joe Weisenbaum's original Eliza was built on two concepts, one good and one bad, that no one previously separated. The good concept was that individual links in complex cause and effect chains could be recognized by the occurrence of slightly variable but easily describable snippets of text/speech. The bad concept was that text/speech could be usefully manipulated by juggling words around. Joe then wrote a book discrediting his own Eliza (with its unseparated concepts), thereby causing AI research to take a wrong turn 40 years ago from which it never recovered. However, the internals of Dr. Eliza aren't really the subject of this posting, other than to demonstrate that AGI now already exists, at least in this one potentially useful form. Any thoughts on all this? Hi Steve, thanks for speaking up. Is what you call Dr. Eliza, the same software program as the well known Eliza? In any case I would like to see example transcripts of the doctor-beating medical dialog you mention. As perhaps you know, I am building an English dialog system to bootstrap AGI, and if successful in realizing its design, Texai should easily win the Chatterbot Challenge (some year) and also perhaps the Loebner Prize (some year). I studied existing chatbots of course. If you are familiar with A.L.I.C.E., I would be interested to learn your opinion on how it compares with Dr. Eliza. -Steve Reed Stephen L. Reed Artificial Intelligence Researcher http://texai.org/blog http://texai.org 3008 Oak Crest Ave. Austin, Texas, USA 78704 512.791.7860 agi | Archives | Modify Your Subscription __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.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=98558129-0bdb63 Powered by Listbox: http://www.listbox.com
Re: [agi] Comments from a lurker...
I'd be interested in looking at a paper. However, I'll be honest: your claim of AGI sounds over-inflated, mainly because it sounds like your algorithm is text-specific and wouldn't help with things like vision, robot control, etc. Nonetheless, a good 'chatbot' is still something of interest (I hope that term isn't taken as derogatory). On Thu, Apr 10, 2008 at 4:39 PM, Stephen Reed [EMAIL PROTECTED] wrote: - Original Message From: Steve Richfield [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Thursday, April 10, 2008 2:58:09 PM Subject: [agi] Comments from a lurker... [snip] BTW, the principles behind Dr. Eliza are rather unique. I'd be glad to send some papers to anyone who is interested. Briefly, Joe Weisenbaum's original Eliza was built on two concepts, one good and one bad, that no one previously separated. The good concept was that individual links in complex cause and effect chains could be recognized by the occurrence of slightly variable but easily describable snippets of text/speech. The bad concept was that text/speech could be usefully manipulated by juggling words around. Joe then wrote a book discrediting his own Eliza (with its unseparated concepts), thereby causing AI research to take a wrong turn 40 years ago from which it never recovered. However, the internals of Dr. Eliza aren't really the subject of this posting, other than to demonstrate that AGI now already exists, at least in this one potentially useful form. Any thoughts on all this? Hi Steve, thanks for speaking up. Is what you call Dr. Eliza, the same software program as the well known Eliza? In any case I would like to see example transcripts of the doctor-beating medical dialog you mention. As perhaps you know, I am building an English dialog system to bootstrap AGI, and if successful in realizing its design, Texai should easily win the Chatterbot Challenge (some year) and also perhaps the Loebner Prize (some year). I studied existing chatbots of course. If you are familiar with A.L.I.C.E., I would be interested to learn your opinion on how it compares with Dr. Eliza. -Steve Reed Stephen L. Reed Artificial Intelligence Researcher http://texai.org/blog http://texai.org 3008 Oak Crest Ave. Austin, Texas, USA 78704 512.791.7860 agi | Archives | Modify Your Subscription __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com agi | Archives | Modify Your Subscription --- 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
Re: [agi] Comments from a lurker...
Claims of having created an impressive AI - sans any credible evidence - are a dime a dozen. I've lost track of how many times I've read similar claims being made over the last decade or so, which often lead to a brief flap of excitement. However, I have a feeling that one of these days someone is going to make such a claim and then it turn out to be genuine. On 10/04/2008, Abram Demski [EMAIL PROTECTED] wrote: I'd be interested in looking at a paper. However, I'll be honest: your claim of AGI sounds over-inflated, mainly because it sounds like your algorithm is text-specific and wouldn't help with things like vision, robot control, etc. Nonetheless, a good 'chatbot' is still something of interest (I hope that term isn't taken as derogatory). On Thu, Apr 10, 2008 at 4:39 PM, Stephen Reed [EMAIL PROTECTED] wrote: - Original Message From: Steve Richfield [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Thursday, April 10, 2008 2:58:09 PM Subject: [agi] Comments from a lurker... [snip] BTW, the principles behind Dr. Eliza are rather unique. I'd be glad to send some papers to anyone who is interested. Briefly, Joe Weisenbaum's original Eliza was built on two concepts, one good and one bad, that no one previously separated. The good concept was that individual links in complex cause and effect chains could be recognized by the occurrence of slightly variable but easily describable snippets of text/speech. The bad concept was that text/speech could be usefully manipulated by juggling words around. Joe then wrote a book discrediting his own Eliza (with its unseparated concepts), thereby causing AI research to take a wrong turn 40 years ago from which it never recovered. However, the internals of Dr. Eliza aren't really the subject of this posting, other than to demonstrate that AGI now already exists, at least in this one potentially useful form. Any thoughts on all this? Hi Steve, thanks for speaking up. Is what you call Dr. Eliza, the same software program as the well known Eliza? In any case I would like to see example transcripts of the doctor-beating medical dialog you mention. As perhaps you know, I am building an English dialog system to bootstrap AGI, and if successful in realizing its design, Texai should easily win the Chatterbot Challenge (some year) and also perhaps the Loebner Prize (some year). I studied existing chatbots of course. If you are familiar with A.L.I.C.E., I would be interested to learn your opinion on how it compares with Dr. Eliza. -Steve Reed Stephen L. Reed Artificial Intelligence Researcher http://texai.org/blog http://texai.org 3008 Oak Crest Ave. Austin, Texas, USA 78704 512.791.7860 agi | Archives | Modify Your Subscription __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com agi | Archives | Modify Your Subscription --- 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/?; Powered by Listbox: http://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=98558129-0bdb63 Powered by Listbox: http://www.listbox.com
Re: [agi] Comments from a lurker...
Steve Richfield wrote: Hi there, I am coming at AGI from an apparently unique perspective. Back in 2001 I contracted an incurable illness (idiopathic atrial fibrillation). Having been involved in a couple of medical research projects in the long distant past, I simply took this as another project and dived in to find a cure. After 4 months of researching during every conscious moment (my AF left me unconscious most afternoons), I did indeed find a one-day cure (followed by a year of recovery). The cure (resetting my body temperature back to 98.6F) isn't what is interesting here. What IS interesting here is why this took me 4 months instead of ~1 hour. Certainly there was SOMETHING wrong with the Internet as it presently exists. Continuing my story... Just days later, I got a project repairing some unrepairable circuit boards for ~$20M military aircraft simulators. I brought my 18-year-old daughter into the project as my apprentice to perform some of the lengthy and boring testing. She had observed my curing of my AF and seen some other successful medical research projects, and after a week on the job commented that the process of repairing circuit boards is just like curing illnesses, only the very specific activities (e.g. screening for errant metabolic parameters vs. screening for nodes with errant dynamic impedances) were different. We talked about this for several weeks, and she was exactly right. I then decided to write an AI program to solve very difficult problems, which gradually morphed into the Dr. Eliza that has been presented and demonstrated by my kids at various past WORLDCOMP AI conferences. Dr. Eliza does NOT deal in direct questions, but rather takes problem statements and drills down into whatever it was that the author needed NOT to know to have such a problem. After all, the only reason that we have problems is that there is something important that we don't already understand or haven't already applied. Meanwhile, I was participating heavily in Yahoo's WS-Forum, helping others with similar internal regulatory problems. I loaded Dr. Eliza with my own essential knowledge about errant body temperatures and idiopathic atrial fibrillation and started throwing unretouched postings into Dr. Eliza. Surprisingly, Dr. Eliza often noticed subtle indirect references to contributory factors that I had missed on my own readings of the postings. In short, with only ~200 of my own knowledge records in its knowledge base, it was serious competition to ME. I then added the ability to exchange knowledge via USENET with other incarnations of Dr. Eliza, so that many authors could contribute to a knowledge base that would WAY outperform any of the individual authors. OK, so why aren't I rich? For the same reasons that most smart guys aren't rich. People simply don't trust anyone or anything that may be smarter than they are. I have a good friend who is the Director of Research for a major University's medical center. I discussed Dr. Eliza at length with him, and he flatly stated that there was NO POLITICAL WAY to integrate such a product into any major medical environment, simply because no doctor is going to stand aside and watch a computer run circles around them. BTW, the principles behind Dr. Eliza are rather unique. I'd be glad to send some papers to anyone who is interested. Briefly, Joe Weisenbaum's original Eliza was built on two concepts, one good and one bad, that no one previously separated. The good concept was that individual links in complex cause and effect chains could be recognized by the occurrence of slightly variable but easily describable snippets of text/speech. The bad concept was that text/speech could be usefully manipulated by juggling words around. Joe then wrote a book discrediting his own Eliza (with its unseparated concepts), thereby causing AI research to take a wrong turn 40 years ago from which it never recovered. However, the internals of Dr. Eliza aren't really the subject of this posting, other than to demonstrate that AGI now already exists, at least in this one potentially useful form. Extrapolating into the future: I see no hope for this group to mitigate the impact of AGI, nor should it, any more than the Luddites were able to blunt the entry of modern mechanized manufacturing. The one thing that writing Dr. Eliza drilled into me is that people, even PhDs, even me, are REALLY REALLY STUPID compared with computationally combining the expertise of many people. That the Dr. Eliza project included the discovery of Reverse Reductio ad Absurdum reasoning that is crucial to solving apparently irresolvable disputes hammers this home, since society's failure to understand RRAA underlies nearly every dispute world history, yet this somehow went undiscovered during millennia of wars and other disputes. Providing some mechanized intelligence is a service and a hope for mankind's future sanity, and
Re: [agi] Comments from a lurker...
Stephen, On 4/10/08, Stephen Reed [EMAIL PROTECTED] wrote: - Original Message From: Steve Richfield [EMAIL PROTECTED] To: agi@v2.listbox.com Sent: Thursday, April 10, 2008 2:58:09 PM Subject: [agi] Comments from a lurker... [snip] BTW, the principles behind Dr. Eliza are rather unique. I'd be glad to send some papers to anyone who is interested. Briefly, Joe Weisenbaum's original Eliza was built on two concepts, one good and one bad, that no one previously separated. The good concept was that individual links in complex cause and effect chains could be recognized by the occurrence of slightly variable but easily describable snippets of text/speech. The bad concept was that text/speech could be usefully manipulated by juggling words around. Joe then wrote a book discrediting his own Eliza (with its unseparated concepts), thereby causing AI research to take a wrong turn 40 years ago from which it never recovered. However, the internals of Dr. Eliza aren't really the subject of this posting, other than to demonstrate that AGI now already exists, at least in this one potentially useful form. Any thoughts on all this? Hi Steve, thanks for speaking up. Is what you call Dr. Eliza, the same software program as the well known Elizahttp://en.wikipedia.org/wiki/ELIZA ? Not at all. Then, why the name Dr. Eliza? Our guiding light was: Suppose that the wrong turn 40 years ago had NOT been made. What would Eliza be now, what with the Internet, modern languages and coding methods, relational databases, dropping the bad half of the original Eliza and substituting an engine to reconstruct cause-and-effect chains, etc., etc.? Add to that, our first strong demo was finding cures for real-world chronic illnesses. The first demo was written in Access with a generous sprinkling of SQL, but we then yanked the computationally intensive part out and compiled it under Visual Basic when we married Dragon NaturallySpeaking to it, for a real-time conversational speech interface. BTW, speech recognition works GREAT because it reliably recognizes long words and specific phrases, even when it messes up everything else. However, people with chronic illnesses wanted nothing to do with it, as they were unwilling to concede ANY chance of a speech recognition error that might cause Dr. Eliza to miss a cure for their illness. In any case I would like to see example transcripts of the doctor-beating medical dialog you mention. As perhaps you know, I am building an English dialog system to bootstrap AGI, and if successful in realizing its design, Texai should easily win the Chatterbot Challengehttp://www.chatterboxchallenge.com/(some year) and also perhaps the Loebner Prize http://en.wikipedia.org/wiki/Loebner_prize (some year). I studied existing chatbots http://en.wikipedia.org/wiki/Chatbot of course. If you are familiar with A.L.I.C.Ehttp://en.wikipedia.org/wiki/Artificial_Linguistic_Internet_Computer_Entity., I would be interested to learn your opinion on how it compares with Dr. Eliza. Dr. Eliza is NOT intended to be a chatterbot and will never win any of those prizes. It rather stubbornly IGNORES where you might quite intentionally be leading the conversation and persistently asks questions that seem to be quite irrelevant to the problem that you have presented. Of course, the answers to those questions are EXACTLY what is needed to sort out the various possibilities of links in the cause and effect chain. Note that real-world problem cause and effect chains usually look like a figure 6, with a root cause, several subsequent links, and a self-sustaining loop at the end. To completely cure a (medical, electronic, political, personal, etc.) problem, you must attack the closest link possible to the root cause AND you must interrupt the self-sustaining loop at ANY point. Steve Richfield --- 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
Re: [agi] Comments from a lurker...
Steve Richfield wrote: Abram, On 4/10/08, *Abram Demski* [EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote: I'd be interested in looking at a paper. How are papers handled on this forum? Do I post it as an attachment, or just send it to you? What is the netiquette here? However, I'll be honest: your claim of AGI sounds over-inflated, mainly because it sounds like your algorithm is text-specific and wouldn't help with things like vision, robot control, etc. It all depends on your definition of AGI. My belief is that the definition generally accepted here lacks the God/Santa Clause consideration. A _REAL_ AGI would have simultaneous presence at thousands of locations, and would leverage on countless indigenous biological mechanisms (people) for most if not all of the I/O. BTW, perhaps you have seen the movie Colossus, the Forbin Project? Did you know that was the movie version of the first book of a trilogy, and that the subsequent books then explored a world with a version of AGI that I think may be beyond the one being considered here? Anyway, Dr. Eliza is very close to THAT concept of AGI. Nonetheless, a good 'chatbot' is still something of interest (I hope that term isn't taken as derogatory). As I explained in an earlier email, this is NOT a chatbot and will never win a chatbot competition. The ONLY reason it has a conversational interface is because that is the only thing that we could see working in the real world. Very early designs utilized lengthy questionnaires, but it soon became obvious that to work well in even a single subject domain, the questionnaire would have to be a bound volume. The concept of AGI is really more about being completely general (hence the G), so it would be premature to use it of your system: a true AGI would be able to cope with a very broad range of circumstances (or be capable of learning how to cope with them). That definition would imply that an AGI would ipso facto be able to act as a chatbot. That is not to say that you may not have devised an interesting and useful system, however, with mechanisms that could be a valuable basis for something more general. It all depends on he details. The best protocol for broadcasting your paper, btw, would probably be to just send it directly to anyone who asks for a copy. That would be the first step. Richard Loosemore --- 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