From comsat@vpics1 Thu Oct 3 16:20:20 1985 Date: Thu, 3 Oct 85 16:20:11 edt From: comsat@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id aa02322; 1 Oct 85 22:46 EDT Date: Sun 29 Sep 1985 16:27-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #130 To: AIList@SRI-AI Received: from rand-relay by vpi; Thu, 3 Oct 85 16:03 EST AIList Digest Sunday, 29 Sep 1985 Volume 3 : Issue 130 Today's Topics: Literature - New Citation/Abstract Distribution Service & Leff Bibliographies & Recent Articles ---------------------------------------------------------------------- Date: Fri, 27 Sep 85 13:38:21 cdt From: Laurence Leff Subject: New Citation/Abstract Distribution Service I have volunteered to organize an electronic mechanism for the distribution of technical report lists from Universities and R&D labs. Some (and hopefully all) of the people producing technical reports would send a copy of the list to me. I would then send these to a moderated group on USENET as well as a mailing list for those sites on the INTERNET who do not get news (ARPANET, CSNET, etc.). I need two things from you: 1) if your organization prepares technical reports and sends them out to interested parties (perhaps for a fee), please arrange to have electronically readable copy of your lists sent to trlist%smu@csnet-relay. 2) if people at your organization would like to receive lists of tech reports produced by universities and R&D labs, please provide me an electronic address to send them to (if you are not on USENET). Send such administrative mail to trlist-request%smu@ csnet-relay. Some frequently asked questions: 1. What are the advantages of sending my lists to you? a. Most of the people to whom you are sending printed lists will be receiving this list, either through the INTERNET as a mailing list or as a moderated news group on the USENET distributed bulletin board system. Thus you can save the postage and printing costs in mailing these lists. I would be happy to provide you with a list of institutions receiving this list as a mailing list as well as those institutions on USENET who would be receiving it that way. You can use this to prune the mailing list you use to send out printed copies of your technical report lists. b. Many people at the Universities are not aware of technical report lists. I have been sending out lists of AI tech reports to the AIList, an electronic newsletter on AI, for some time. Every time I do so, my electronic mailbox fills up with requests on how to obtain the tech reports. Many of these requests come from the most prestigious AI organizations in the country. c. Many companies, particularly those on the USENET, would not otherwise be aware of your research. There are hundreds of small companies on USENET who have no other access to the wealth of information represented by University and other tech reports. 2. What is a technical report? Most universities and big company R&D labs publish reports about their research. Some are highly research oriented (like new results in automata theory). Others are manuals for their public domain software or tutorials. ... 3. What format should the tech report lists be in? Please see to it that there is some info indicating how people can order the tech reports (whether sending you a check to cover costs, requests via electronic mail or the reports can be electronically available for Arpanet FTP transfer). If you are already producing the list in some format, feel free to use that format. If you are preparing the list just for this purpose, I would prefer that you use the input format for bib/refer, a common bibliography tool. This way people can dump the lists into a file on their machine and be able to do keyword searches. Also bib/refer will automatically include and format references in documents to be formatted or typeset. However, I would prefer the material in some weird format than not to have it at all! For those not familiar with bib/refer, here is a brief tutorial. Each report or other item should be a sequence of records which are not separated by blank lines. Each report should be separated by the others by one or more blank lines. Each report entry consists of a label consisting of a % followed by a capital letter and then a space. Then include the information. If the information for a field (such as an abstract) requires more than one line, just continue the field on a new line with no initial space. The labels needed for tech reports are: %A Author's name (this field should be repeated for each author). %T Title of report %R report number %I issuer, this will be the name of your institution. This may be ommited if implied by the report number %C City where published (not essential) %D Date of publication %X Abstract Here is an example of some tech report listings in the appropriate format: %A D. Rozenshtein %A J. Chomicki %T Unifying the Use and Evolution of Database Systems: A Case Study in PROLOG %R LCSR-TR-68 %I Laboratory for Computer Science Research, Rutgers University %K frame control %A C. V. Srinivasan %T CK-LOG, A Calculus for Knowledge Processing in Logic %R DCS-TR-153 %I Laboratory for Computer Research, Rutgers University %K MDS 4. I already have exchange agreements with other Universities. How does this affect them? The only change would be how the information on what technical reports you have for them to request gets transferred. [...] 5. I need to charge for my tech reports to cover costs. Fine. Just include the prices for your reports next to each report (you can use the %X field for that too). At the beginning of the list you send me, state where checks should be sent and to whom they should be made payable. 6. What about non-CS reports? I am happy to handle reports for other departments. If the volume of non-CS reports becomes significant, I will split the list into tr-cs, tr-math, tr-ee etc. I would suspect that the majority of the people receiving this list would be CS researchers since CS departments are quick to join networks, etc. However, some CS researchers (myself included) are working in applications of computers and would like to receive information in those areas as well. 7. I am already on USENET. What should I do? I anticipate a USENET moderated group in a time frame of one to two weeks which will contain the same information as the technical report lists. If you indicate that you will get the information via USENET, I will remove your name when the list is established. If you want to wait a week or two to see if the list comes up, that is OK too. I can send back copies of the TR Lists that get sent out in the first few batches of the mailing. I will also send out on the USENET group, everything that got sent out in the mailing list so you won't miss anything either way. 8. I am on Arpanet, BITNET, etc. I can get to Arpanet sites through csnet-relay so there is no problem there. Otherwise, send me your address as best you know it. I will get through to you if at all possible. ------------------------------ Date: 27 Sep 1985 20:03-CST From: leff%smu.csnet@CSNET-RELAY.ARPA Subject: Bibliographies [Forwarded from a message to AIList-Request.] You have probably noticed the announcement of the new tech report list. [...] The thing that started me on this was the response of AIList readers to my lists of tech reports. >From what filled up my mailbox, it was obvious that many if not most of your readers were not seeing the tech report lists and a substantial fraction of those did not even know that tech reports existed! Hopefully this list will serve a useful function for everyone. It was something that should have been done a long time ago. [...] I have increased the number of magazines from which my bibliographies (type 1) are drawn. We now have added ComputerWorld as well as a few minor magazines. ComputerWorld did a very good job on IJCAI-85 and I found material there that was no place else. [...] According to bib, we now have 430 documents sent to you since I changed formats to machine readable. This does not include information sent to you in other formats. [I would like to thank Laurence for providing his services to AIList and the net community. It's a heck of a hobby, but he does a great job. -- KIL] ------------------------------ Date: 27 Sep 1985 20:00-CST From: leff%smu.csnet@CSNET-RELAY.ARPA Subject: Recent Articles %A Ruth E. Davis %T Logic Programming and Prolog: A Tutorial %J IEEE Software %D SEP 1985 %P 53-62 %V 2 %N 5 %T Advertisement %A Texas Instruments %J IEEE Spectrum %D SEP 1985 %V 22 %N 9 %P 22-23 %X announces a TV satellite symposium that can be received by various companies using a satellite dish. (On November 13, 1985) %A T. A. Marsland %A F. Popowich %T Parallel Game-Tree Search %J IEEE Transactions on Pattern Analysis and Machine Intelligence %V PAMI-7 %N 4 %D JUL 1985 %P 442-452 %K alpha-beta %A K. C. Drake %A E. S. McVey %A R. M. Inigo %T Sensing Error for a Mobile Robot Using Line Navigation %J IEEE Transactions on Pattern Analysis and Machine Intelligence %V PAMI-7 %N 4 %D JUL 1985 %P 485-490 %A A. Lansner %A O. Ekeberg %T Reliability and Speed of Recall in an Associate Network %J IEEE Transactions on Pattern Analysis and Machine Intelligence %V PAMI-7 %N 4 %D JUL 1985 %P 490-498 %A W. A. Gale %T Book Review: The AI Business: The Commerical Uses of Artificial Intelligence- P. Winston and K. Prendergast, Eds %J IEEE Transactions on Pattern Analysis and Machine Intelligence %V PAMI-7 %N 4 %D JUL 1985 %P 499 %T FMC Invests $3.5 Million in Knowledge-Based Systems %J IEEE Software %D JUL 1985 %V 2 %N 4 %P 101 %K Teknowledge %X FMC has invested $3.5 million in Teknowledge. %T US, Japan AI firms enter joint ventures %J IEEE Software %D JUL 1985 %V 2 %N 4 %P 101 %K Carnegie Group Intelligent Technology Knowledge Craft Language Craft Jack Geer McDonnell Douglas %X Carnegie Group and Intelligent Technology have signed a joint venture agreement where Intelligent Technology will distribute Knowledge Crat and Language Craft throughout the far east. They will be creating Japanese language versions of these products. Carnegie Group has appointed Jack Geer, formally of the Knowledge Engineering Division of McDonnell Douglas Information Systems Group, as director of marketing. %A Eric Bender %T AI Firms Outgrow Seat-of-the-Pants Style %J ComputerWorld %V 19 %N 35 %D SEP 2, 1985 %P 10+ %K Golden Common Lisp %X general article on Gold Hill Computers, the author of Golden Common Lisp, and Arity computers, the author of a Prolog compiler/interpreter. Golden Common Lisp has sold approximately 3000 copies. They employ 18 people and have a "monthly run rate" of $200,000. They anticipate selling a large memory version for the PC/AT which can use 16M of memory. The largest user of Arity products is software vendors with classic DP shoppers a closer second. Arity Prolog tools were used by one software vendor to develop a system to consult on software installation. %T Vendors Fuel AI Language Debate %J ComputerWorld %V 19 %N 35 %D SEP 2, 1985 %X discusses battle between Prolog and Lisp as standard for AI. %A Charles Babcock %T Experts Beat out Expert Systems at Financial Firm %J ComputerWorld %V 19 %N 35 %D SEP 2, 1985 %P 12+ %K business insurance financial Metropolital Life Roger Jones expert system medical %X Rojer Jones, planning manager in Metropolital Life's corporate systems planning division, said that it is very difficult to encode expert thinking and in many cases the bank prefers training human experts to developing expert systems. He said that experts might lie, be wrong or the business might change. He emphasized that some experts are so possessive of their knowledge that they would covertly sabotage the expert system development process. In the case of one project, it cost more to transcribe the 30 pages of medical information to provide to the expert system than to have an underwriter evaluate the information. Insurance companies are using larger pools of data to determine actuarial tables so that expert systems based on segregated pools (e. g. with men separated from women) would be obsolete. He claims that the systems that were successful (Prospector, Dendral) mapped a very broad knowledge base of simple facts. %T Artificial? Or Intelligent? %J ComputerWorld %V 19 %N 35 %D SEP 2, 1985 %P 18 %X editorial on AI, saying that there is a legitimate demand for AI technology but it will be frought with hard work and that DP staffs can't neglect their day to day work to pursue the interesting AI interests. %A Eric Bender %T Lively Discussions Highlight AI Meet %J ComputerWorld %V 19 %N 35 %D SEP 2, 1985 %P 49+ %K IJCAI %X quotes overhead at IJCAI-85 Beau Sheil of Xerox Artificial Intellgience Systems - AI works best with a lot of information that can be manipulated at a shallow level Xerox has received an order for 1000 of its 1185 work stations and is flooded with requests at similar volume. Alan Kay of Apple said "we need to do problem finding," not problem solving. He also griped about logic programming and parallel processing. He made a comment that if the Intel 80286 is a weak architecture, what is it when you have 16 of them? [probably a veiled reference to Intel's Cosmic Cube.] Larry Levesque of Carnegie Group said that AI products and demos are tools and they don't scale up. %T Secure Xerox Workstation Out %J ComputerWorld %V 19 %N 35 %D SEP 2, 1985 %P 63+ %X Lisp Machine %K Xerox announced the 1108-105T, an AI work station that meeets standards for release of electronic radiation that can be tapped for use in National Security and other places. It also announced an 1108 series of AI work stations that can interface with IBM and Multibus equipment. %T Systems and Peripherals %J ComputerWorld %V 19 %N 35 %D SEP 2, 1985 %K digitization machine vision Industrial Vision Systems %X Industrial Vision Systems has announced a 400 dot/in digitizing scanner capable of handling up to 36 in wide paper. It costs $79,000. %A John Gallant %T Will IBM Take AI by Storm %J Computer World %D SEP 9, 1985 %V 19 %N 36 %P 61+ %K IJCAI-85 %X general article on IBM's role in AI %T Inference Enhances ART Development Environment %J Computer World %D SEP 9, 1985 %V 19 %N 36 %P 62 %X improvements to Inference Corp's Automated Reasoning Tool include: 1) improved color graphics including a away of attaching graphics primitives to rules, 2) a pseudo-natural language development enviornment which allows a developer to build his knowledge-base using an English-like syntax 3) a mixed initiative processing environment allowing the expert system to prompt for information while reacting to user inquiriers 4) separately compilable rule files %T Software and Services %J ComputerWorld %D SEP 9, 1985 %V 19 %N 36 %P 69 %K Lucid Sun Common Lisp %X Sun Microsystem has announced a version of LUCID, a Common LISP implementation for its work stations costing $375.00. %A Sol Libes %T Bytelines %J Byte %D SEP 1985 %P 420 %V 10 %N 9 %X Kurzweil Applied Intelligence speech recognition KVS-3000 %X Kurzweil has introduced a KVS-3000 that can handle 1000 words continuous speech with 100 per cent accuracy. It is selling at $3000.00 in quantity and comes in PC, multibus and RS232C versions. It is speaker adaptive and its performance increases the more it talks with the same user. %A Jean Renard Ward %%A Barry Blesser %T Interactive recognition of Handprinted Characters for Computer Input %J IEEE Computer Graphics and Applications %V 5 %N 9 %P 24-37 %K Character Recogniton Pencept %X discusses the human interface issues once you have character recognition on your computer; i. e. how best to interface handwritten character recognition with your product. %T Hitachi to Spend $40 million in U. S. on High-Tech Gear %J Electronic News %D SEP 2, 1985 %V 31 %N 1565 %P 10 %K Tektronix %X Hitachi has spent $700,000 at Tektronix which includes an undisclosed amount of "artificial intelligence products." %T Raytheon Acquires Stake in Lisp Machines %J Electronic News %D AUG 26, 1985 %P 20+ %V 31 %N 1564 %K energy venture capital military electronics %X Raytheon invested 4.5 million dollars into Lisp Machine Inc. through its new venture capital subsidiary. Lisp Machine Inc. has raised a total of twelve million dollars in a fourth round of financing. Raytheon hopes to be using Lisp Machine products in its military electronics and energy business. Ti currently owns 9 percent of LMI. other investors include Abingworth plc, InterVEN, Genesis Venture Capital, Manufacturers Hanover Trust. %A Eric Bender %T Artificial Intelligence: On the Road to Reality %J ComputerWorld %D AUG 26, 1985 %V 19 %N 34 %K IJCAI-85 %X summary and analysis of events at IJCAI-85 %X Many vendors continued development work up until the demonstrations themselves and there were more bugs than typical at a computer conference vendor exhibition. Beau Sheil of Xerox stated that in order for AI to work at a company, the company has to have long term horizons, a real clear idea about its needs and a history of applying technology to solve those problems. %A Eric bender %T IJCAI sees HP, Intellicorp moves in AI programming %J ComputerWorld %D AUG 26, 1985 %V 19 %N 34 %P 10+ %K Hewlett-Packard Common Lisp HP 9000 Xerox Palladian Software financial %X Hewlett-Packard will have a Common Lisp development environment running on its HP 9000 series 300 family of workstations. It will have a system called Browsers "that will automatically have appropriate tools for the task at hand." Xerox will be announcing a Common Lisp in second quarter 186. Intellicorp announced a system to allow personal computers to act as delivery vehicles for expert systems. Initial versions will use VAX systems as hosts and IBM PC's and the Macintosh as delivery vehicles. Palladian Software announced its financial advisor expert system. It runs on Symbolics and Texas Instruments Lisp Machines and communicates with IBM and DEC systems. A system with four work stations runs for $95,000. %A Eric Bender %T Symbolics, Xerox offer enhanced AI workstations %J ComputerWorld %D AUG 26, 1985 %V 19 %N 34 %P 11 %K Lisp Machine %X Xerox announced an 1185 which costs $9.995 and runs Interlisp-D software and serves as a delivery vehicle. They also announced an 1186 development system with 3.6M of internal memory and 80M of hard disk storage for $15,865. %T Aion offers AI development system %J ComputerWorld %D AUG 26, 1985 %V 19 %N 34 %P 55 %K expert system microcomputer venture capital %X Aion announced a new expert system for the IBM Pc written in Pascal. They anticipate selling an IBM-370 version for first-quarter 1986. They are focussing on the traditional DP environment. They received 2.4 million in venture capital. %T Microcomputers %J ComputerWorld %D AUG 26, 1985 %V 19 %N 34 %P 61 %K The Institute for Scientific Analysis Small-X microcomputer expert system %X The Instuitute for Scientific Analysis introduced Small-X, an expert system development tool for the IBM-PC. It costs $249.95 and can control or exchange data with other applications running under Microsoft. %A S. Jerrold Kaplan %T Designing a Portable Natural Language Database Query System %J ACM Transactions on Database Systems %V 9 %N 1 %D MAR 1984 %P 1-19 %A C. Hornsby %A H. C. Leung %T The Design and Implementation of a Flexible Retrieval Language for a Prolog Database System %J SIGPLAN %V 20 %N 9 %D SEP 1985 %P 43-51 %X implementation of a database management system in PROLOG %A Donna Raimondi %T Ansa offers Paradox IBM-compatible relational DBMS %J ComputerWorld %V 19 %N 38 %P 12 %D SEP 23, 1985 %K data base system interface microcomputer heuristic query optimization Ansa Software Paradox Sevin Rosen Management %X A data base management package which uses machine reasoning to evaluate user requests and write programs for the user. It uses query by example, program synthesis and heuristic query optimization techniques %A Jeffry Beeler %T Symantec package out %J ComputerWorld %V 19 %N 38 %P 12 %D SEP 23, 1985 %K Symantec natural language microcomputer data base system interface %X a new product which uses natural language to interface with a database %T Software and Services %J ComputerWorld %V 19 %N 38 %D SEP 23, 1985 %P 50 %K Franz Common Lisp flavors %X a new release of Franz Lisp, Opus 42, is out which supports Lisp flavors, functions returning multiple values, multiple name spaces in the Lisp environment, hash table objects, history mechanism. It is available for Apollo, Sun, Cadmus, Masscomp, Tektronix, Harris and Digital equipment Corp. $5,000 first copy, $1,000 subsequent copies %A Craig D. Rose %T R&D Race Tightens for Fifth-Generation Computers %J Electronics %D SEP 23, 1985 %V 58 %N 38 %P 30-31 %X Attempts to compare and contrast the Fifth Generation efforts of Europe, Japan and the U. S. The Strategic Computing Initiative (SCI) of DARPA will spend one billion dollars over ten years. Europe's Esprit program is being funded with 1.2 billion for five years and Britain is spending $455 million over the same time span for the Alvey project. Japan's Fifth Generation Project is funded at twenty to thirty million dollars per year. Japanese companies are spending in total about five times as much money as ICOT. The Canadian Society for Fifth Generation computing has requested fourty million dollars over three years. %T Machine Vision Maker Raises Three Million Dollars %J Electronics %D SEP 23, 1985 %V 58 %N 38 %P 41 %X Itran venture capital inspection General Motors %X Itran Corp has raised three million dollars in a new round of venture financing. Itran markets systems that inspects parts on factory floors. It sells systems to General Motors. %A Tobias Naegele %T How Rensing Got His Robot Working %J Electronics %D SEP 23, 1985 %V 58 %N 38 %P 42 %K Renco Electronics %X describes experiences of a small firm in installing robots in their factory. %T AI Tools Automate Software Translation %J Electronics %D SEP 23, 1985 %V 58 %N 38 %P 59-61 %K Lexeme Michael Shamos computer language translation conversion expert system %X Lexeme sells an expert system that translates from one computer language to another. It supports Ada and C as target language and accepts input of Fortran, PL/1, Bliss and SPL. They are developing COBOL, BASIC, Algol, Jovial and CMS-2 versions. They also handle conversions from one language to another. There is a separate page on the personalities and stories of the founders. [Michael Shamos, the president, is also well known for his work in computational geometry. --Leff] He managed to pick up a law degree as well as a Ph. D. in computer science! ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Thu Oct 3 16:30:30 1985 Date: Thu, 3 Oct 85 16:30:14 edt From: comsat@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a003405; 2 Oct 85 3:10 EDT Date: Sun 29 Sep 1985 16:38-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #131 To: AIList@SRI-AI Received: from rand-relay by vpi; Thu, 3 Oct 85 16:16 EST AIList Digest Sunday, 29 Sep 1985 Volume 3 : Issue 131 Today's Topics: Seminars - KSL/Symbolic Systems Resources Group (SU) & Qualitative Simulation of Mechanisms in Diagnosis (UT) & Purpose-Directed Analogy (GTE) & Connectionist Parallel Distributed Processing (UCB) & Processes, Simultaneity and Causality (SRI) & Theory of Declarative Knowledge (UPenn), Seminar Series - Software Environments (CSLI), Conferences - Society for Computer Simulation & SIGIR/SIGDOC Workshop ---------------------------------------------------------------------- Date: Tue 24 Sep 85 11:10:07-PDT From: Ana Haunga Subject: Seminar - KSL/Symbolic Systems Resources Group (SU) KSL/Symbolic Systems Resources Group Tom Rindfleisch and Bill Yeager Stanford University This is the first of several SIGLunches this fall that will summarize work in each of the five sublabs of the Stanford Knowledge Systems Laboratory (KSL), including the Heuristic Programming Project, HELIX Group, Medical Computer Science Group, Logic Group, and Symbolic Systems Resources Group (SSRG). This week's talk will consist of a brief overview of the KSL as an AI laboratory and a survey of SSRG research and development activities. Since 1980, the computing environment for KSL research has been moving slowly away from central time-shared mainframes (like the SUMEX 2060 and VAX) toward networked Lisp workstations. Improvements in workstation performance, falling prices, better packaging, and a wider vendor selection are now accelerating this transition. Over the next five years, we are proposing to phase out the SUMEX research mainframes so that all KSL computing will be workstation-based -- not only research program development but common tasks like text processing, mail, file management, and budgeting. This raises several important issues that will require a community system software effort comparable to that in the 1970's that led to the current TOPS-20 and UNIX environments. How can the user computing environment be improved using workstation bitmapped graphics and AI methods for more intelligent systems/applications programs? How can user displays connect flexibly to workstations -- from home, over remote networks like ARPANET, and locally over Ethernet? How can the considerable computing power distributed among many workstations be combined to support individual user tasks? What are the impacts on network protocols and services (file servers, gateways, printing, etc.) of large numbers of workstations? ------------------------------ Date: Tue, 24 Sep 85 15:50:05 cdt From: rajive@sally.UTEXAS.EDU (Rajive Bagrodia) Subject: Seminar - Qualitative Simulation of Mechanisms in Diagnosis (UT) Qualitative Simulation of Mechanisms and Causal Models in Medical Diagnosis Ben Kuipers Friday, 27th September, Pai 3.38 12 pm Researchers in the AIM (Artificial Intelligence in Medicine) community have concluded that expert medical diagnosis requires knowledge in the form of causal models, to support reasoning about how physiological mechanisms work and interact. One form of causal reasoning is qualitative simulation of descriptions of the structure of mechanisms to yield predictions of their behavior. Qualitative simulation has a number of interesting mathematical properties, and a fast algorithm. The knowledge base of mechanism descriptions makes it possible to model both a healthy and a broken physiological mechanism with minor perturbations of the same structural description. This talk will review recent results and open problems in both qualitative simulation and its application to expert systems in medicine. ------------------------------ Date: Thu, 26 Sep 85 10:35:40 EDT From: Bernard Silver Subject: Seminar - Purpose-Directed Analogy (GTE) GTE LABORATORIES INCORPARATED 40 Sylvan Rd, Waltham, MA 02254 TIME: October 3, 10AM SPEAKER: Smadar Kedar-Cabelli Laboratory for Computer Science Rutgers University, NJ TITLE: PURPOSE-DIRECTED ANALOGY Existing techniques for analogical reasoning are based on mapping some underlying causal network of relations between analogous situations. However, causal relations relevant for the purpose of one analogy may be irrelevant for another. We describe here a technique which uses an explicit representation of the analogy to automatically create the relevant causal network. NOTE: If you wish to attend this seminar, please contact Bernard Silver on (617) 466-2663 ------------------------------ Date: Wed, 25 Sep 85 14:09:11 PDT From: chertok%ucbcogsci@Berkeley.EDU (Paula Chertok) Subject: Seminar - Connectionist Parallel Distributed Processing (UCB) BERKELEY COGNITIVE SCIENCE PROGRAM Fall 1985 Cognitive Science Seminar -- IDS 237A TIME: Tuesday, October 1, 11:00 - 12:30 PLACE: 240 Bechtel Engineering Center (followed by) DISCUSSION: 12:30 - 1:30 in 200 Building T-4 SPEAKER: David Rumelhart, Institute for Cognitive Science, UCSD TITLE: ``Parallel Distributed Processing: Explora- tions in the Microstructure of Cognition'' Parallel Distributed Processing (PDP) is the name which I and my colleagues at San Diego have given to the class of neurally-inspired models of cognition we have been studying. We have applied this class of "connectionist" models to a variety of domains including perception, memory, language acquisition and motor control. I will briefly present a gen- eral framework for the class of PDP models, show how these models can be applied in the case of acquisiton of verb mor- phology, and show how such macrostructural concepts as the schema can be seen as emerging from the microstructure of PDP models. Implications of the PDP perspective for our under- standing of cognitive processes will be discussed. ------------------------------ Date: Thu 26 Sep 85 16:58:54-PDT From: LANSKY@SRI-AI.ARPA Subject: Seminar - Processes, Simultaneity and Causality (SRI) PROCESSES, SIMULTANEITY AND CAUSALITY Michael P. Georgeff Artificial Intelligence Center SRI International 11:00 AM, MONDAY, September 30 SRI International, Building E, Room EJ228 (new conference room) The notion of process is essential for reasoning about the behavior of multiple agents or single agents in dynamic worlds. In this talk, we show why reasoning about process is so important, and contrast this with other approaches in AI which are primarily based on the allowable behaviors of agents. An algebra of processes based on events is given. We then show how events can be represented as changes of world state, and how state properties can be inferred from the model. Interestingly, no STRIPS-like assumption is involved in the definition of events, thus allowing a proper model-theoretic semantics. One of the most important features of the model is a hiding operation. This provides an abstraction capability that can be used to avoid the combinatorial explosion typical of other AI approaches. Finally, we introduce a notion of causality between events and processes. This, together with the notion of simultaneous actions and hiding operations, allows us to avoid most of the problems associated with the frame problem. ------------------------------ Date: Sat, 28 Sep 85 18:06 EDT From: Tim Finin Subject: Seminar - Theory of Declarative Knowledge (UPenn) TOWARDS A THEORY OF DECLARATIVE KNOWLEDGE Krzysztof R. Apt, LITP, Universite Paris, IBM Thomas J. Watson Research Center 3:00pm Tuesday 1 Oct, CIS, University of Pennsylvania We study logic programming with negation from the point of view of its use for building expert system shells. We achieve a separation between the declarative and procedural meaning of the programs. We do this by defining a class of stratified programs which disallow certain combination of recursion and negation and to which we restrict our study. We develop a fixed point theory of non-monotonic operators and apply it to provide a declarative meaning of the programs based on model theory. We also define a backchaining interpretor and show that in the absence of function symbols it computes a selected model of a stratified program. ------------------------------ Date: Tue 24 Sep 85 10:35:38-PDT From: Terry Winograd Subject: Seminar Series - Software Environments (CSLI) [Forwarded from the CSLI bboard by Laws@SRI-AI.] New project meeting on environments Mondays 1-2 in the trailer classroom, Ventura [Future meetings will be from 12 to 1:15.] Beginning Monday, Sept. 30 there will be a weekly meeting on environments for working with symbolic structures (this includes programming environments, specification environments, document preparation environments, "linguistic workstations", and grammar- development environments). As a part of doing our research, many of us at CSLI have developed such environments, sometimes as a matter of careful design, and sometimes by the seat of the pants. In this meeting we will present to each other what we have done, and also look at work done elsewhere (both through guest speakers and reading discussions). The goal is to look at the design issues that come up in building environments and to see how they have been approached in a variety of cases. We are not concerned with the particular details ("pop-up menus are/aren't better than pull-down menus") but with more fundamental problems. For example: What is the nature of the underlying structure the environment supports: chunks of text? a data-base of relations? a tree or graph structure? How is this reflected in the basic mode of operation for the user? How does the user understand the relation between objects (and operations on them) that appear on the visible representation (screen and/or hardcopy) and the corresponding objects (and operations) on some kind of underlying structure? How is this maintained in a situation of multiple presentations (different views and/or multiple windows)? How is it maintained in the face of breakdown (system failure or catastrophic user error in the middle of an edit, transfer, etc.)? Does the environment deal with a distributed network of storage and processing devices? If so, does it try to present some kind of seamless "information space" or does it provide a model of objects and operations that deals with moving things (files, functions, etc.) from one "place" to another, where different places have relevant different properties (speed of access, security, shareability, etc.)? How is consistency maintained between separate objects that are conceptually linked (source code and object code, formatter source and printer-ready files, grammars and parse-structures generated from them, etc.)? To what extent is this simply left to user convention, supported by bookkeeping tools, or automated? What is the model for change of objects over time? This includes versions, releases, time-stamps, reference dates, change logs, etc., How is information about temporal and derivational relationships supported within the system? What is the structure for coordination of work? How is access to the structures regulated to prevent "stepping on each other's toes"? to facilitate joint development? to keep track of who needs to do what when? Lurking under these are the BIG issues of ontology, epistemology, representation, and so forth. Hopefully our discussions on a more down- to-earth level will be guided by a consideration of the larger picture and will contribute to our understanding of it. The meeting is open to anyone who wishes to attend. Topics will be announced in advance in the newsletter. The first meeting will be devoted to a general discussion of what should be addressed and to identifying the relevant systems (and corresponding people) within CSLI, and within the larger (Stanford, Xerox, SRI) communities in which it exists. ------------------------------ Date: 24 Sep 1985 09:24-CST From: leff%smu.csnet@CSNET-RELAY.ARPA Subject: Conference - Society for Computer Simulation Sponsor: Society for Computer Simulation Dates: October 24-25 1985 Location: General Dynamics Recreation Area Fee: $25 both day; $15 one day; $10 both days full time students AI related talks listed below: Thursday, October 24 8:45 AM Session I "Adaptive Sequencing Rules in a Shop Floor Control System", Chris Gill, General Dyanamics Session II "Shape Memory Alloy fo Robot Muscles", MIke Zerkus, Jeff Akus, Martin Spizale, Louisiana Tech University "Advanced Data and Picture Transformation System", Dr. V. Devarajan, Y. P. Chen, LTV Corporation 10:30 AM Keynote Address Keyote address: "Robotics and Intelligent Systems: An Overview" Dr. George Bekey, University of Southern California 1:30 PM Session IV "A Symbolic Expert System for the Design of Digital Controllers for Space Vehicles", Dr. Wolf Kohn, Robert Norsworthy, Lockheed EMSCO 3:15 PM Session VI "Ultra Sonic Ranging for Robot Sensing: Dr. Troy Henson, Louisiana Tech University "A Presentation of Expert Systems at NASA Johnson Space Center, Dr. Wade Webster, Lockheed EMSCO Friday, October 25 Session VII "Exploiting Artificial Intelligence in Simulation" Walter Strucely, Texas Instruments "Application of an Expert System in Process Control in Aerospace Manufactuirng Bill Skelton, LTV Corporation 9:45 AM Session X "General Dynamics Simulation Systems and Artificial Intelligence: Rich Teichgraeber, General Dynamics ------------------------------ Date: Sat, 7 Sep 85 18:28:58 edt From: Michael Lesk Subject: SIGIR/SIGDOC Workshop [Excerpted from the IRList Digest by Laws@SRI-AI.] The following is a proposal for a workshop which, although not yet formally approved, [Note: Diana Patterson of SIGDOC has signed - Ed] is very likely to take place in Snowbird, Utah, June 30-July 2, 1986. Chair: Michael Lesk; Local Arrangements: Lee Hollaar; Treasurer: Karen Kukich. Attendance will be limited to 75; there will be no formal proceedings, but a report will be written for some ACM publication; a number of prominent people (Karen Sparck Jones, David McDonald, Donald Walker, Patricia Wright, etc.) have indicated interest in attending. Comments on the workshop, or indications of interest, are welcome. Please notify the chair at: bellcore!lesk, or lesk%bellcore@csnet-relay, or (if you have current routing tables) lesk@bellcore. Phone: 201-829- 4070. NOTE: I will be on vacation Sept 9 - Oct 4; failure to reply during those dates merely means your message has not been read!! -- Thanks, Michael Lesk Writing to be Searched: A Workshop on Document Generation Principles As computers learn to write English, and others improve at searching it, they ought to benefit from people who know how to do these jobs. We're proposing a workshop bringing together AI special- ists in document generation, information retrieval experts, people who know how to write manuals, and those who write programs to evaluate writing. In recent years there has been a surge of interest in the use of computer programs that write English.[1,2,3] Expert systems, for exam- ple, need to explain what they are doing. Programs are making increasing strides in fluency, domain coverage, and expressive power.[4,5] In fact, it is remarkable that there has been a long dis- cussion over the last ten years about whether or not apes have mastered language, based on utterances such as ``Please tickle more, come Roger tickle''[6] while computer programs saying things like ``The market crept upward early in the session yesterday, but stumbled shortly before trading ended''[7,8] have not impressed the public nearly as much. But even supposing that computers can now write English, what should they write? [...] [There followed a long essay about having computers write computer manuals. -- KIL] Workshop Specifics. In this workshop we will bring together subject specialists in four main areas: * Artificial intelligence researchers working in natural language generation; * Documentation specialists interested in writing style and qual- ity, and in the definition of a `good' document; * Text analysis developers, building programs that analyze text automatically and try to make value judgments about it; and * Retrieval experts, who know how to build systems for keyword matching and retrieval. Another major area that should be represented, but possibly not until a later meeting, is computer graphics. The value of illustrations, diagrams, and charts is unquestioned but it is not clear how we can integrate graphics with text today. [...] Our best possible outcome, of course, is that the participants will find something which is not quite a conventional reference manual, but serves the same purpose and does it better. Whether this will be a structured document still written in English, or a question-answering database with an explanation generator, it is impossible to say. But unless the various groups start talking to one another, we'll never find out. Michael Lesk Bell Communications Research 435 South St., Rm. 2A-385 Morristown, NJ 07960 August 9, 1985 References 1. E. Conklin and D. McDonald, "Salience: The Key to the Selection Problem in Natural Language Generation," Proc. 20th Meeting ACL, pp. 129-135, 1982. 2. K. R. McKeown, "The TEXT System for Natural Language Generation: An Overview," Proc. 20th Meeting ACL, pp. 113-120, Toronto, Ont., 1982. 3. R. E. Cullingford, M. W. Krueger, M. Selfridge, and M. A. Bien- kowski, "Automated Explanations as a Component of a Computer- Aided Design System," IEEE Trans. Sys., Man & Cybernetics, pp. 168-181, 1982. 4. W. C. Mann, "An Overview of the NIGEL Text Generation Grammar," Proc. 21st ACL Meeting, pp. 79-84, 1983. 5. A. K. Joshi and B. L. Webber, "Beyond Syntactic Sugar," Proc. 4th Jerusalem Conf. on Information Technology, pp. 590-594, 1984. 6. S. Chevalier-Skolnikoff, "The Clever Hans Phenomenon, Cuing and Ape Signing: A Piagetan Analysis of Methods for Instructing Animals," in The Clever Hans Phenomenon: Communication with Horses, Whales, Apes and People, ed. Thomas Sebeok and Robert Rosenthal, vol. 364, pp. 60-93, New York Academy of Sciences, 1981. 7. Karen Kukich, Knowledge-Based Report Generation: A Knowledge- Engineering Approach to Natural Language Report Generation. Ph.D Thesis, University of Pittsburgh, 1983 8. Karen Kukich, "ANA's First Sentences: Sample Output from a Natural Language Stock Report Generator," Proc. Nat'l Online Meeting, pp. 271-80, 1983. 9. G. Salton and M. McGill, Introduction to Modern Information Retrieval, McGraw-Hill, 1983. 10. Among sellers of free text retrieval systems are ``Cucumber Information Systems'' (5611 Kraft Drive, Rockville, MD 20852) and ``Knowledge Systems, Inc.'' (12 Melrose St., Chevy Chase, MD 20815). 11. G. Salton, The SMART Retrieval System -- Experiments in Automatic Document Processing, Prentice-Hall, 1971. 12. G. W. Furnas, T. K. Landauer, L. M. Gomez, and S. T. Dumais, "Statistical Semantics: Analysis of the potential performance of key-word information systems," Bell Sys. Tech. J., vol. 62, no. 6, pp. 1753-1806, 1983. 13. Marion O. Harris, "Thoughts on an All-Natural User Interface," Proc. Summer USENIX Conf., pp. 343-347, Portland, Oregon, June 1985. 14. L. M. Bernstein and R. E. Williamson, "Testing of a Natural Language Retrieval System for a Full Text Knowledge Base," J. Amer. Soc. Inf. Sci, vol. 35, no. 4, pp. 235-247, 1984. 15. R. E. Williamson, "ANNOD -- A Navigator of Natural-Language Organized (Textual) Data," Proc. 8th SIGIR Meeting, pp. 252-266, Montreal, Quebec, 1985. 16. M. E. Lesk, "Programming Languages for Text and Knowledge Pro- cessing," Ann. Rev. Inf. Sci. and Tech., vol. 19, pp. 97-128, 1984. 17. Janet Asteroff, "On Technical Writing and Technical Reading," Information Technology and Libraries, vol. 4, no. 1, pp. 3-8, March 1985. 18. Christine Borgmann, "The User's Mental Model of an Information Retrieval System," Proc. 8th SIGIR Meeting, pp. 268-273, Mont- real, Quebec, 1985. 19. Marilyn Mantel and Nancy Haskell, "Autobiography of a First-Time Discretionary Microcomputer User," Human Factors in Computing Systems: Proc. CHI '83 Conference, pp. 286-290, 1983. 20. Bill Swartout, "GIST English Generator," Proc. AAAI-82, pp. 404- 409, Pittsburgh, Penn., 1982. 21. Ariel Shattan and Jenny Hecker, "Documenting UNIX: Beyond Man Pages," Proc. Summer USENIX meeting, pp. 437-454, Portland, Ore., 1985. 22. Karen Kukich, "Design of a Knowledge-Based Report Generator," Proc. 21st Meeting ACL, pp. 145-50, 1983. 23. E. Voorhees and G. Salton, "Automatic Assignment of Soft Boolean Operators," Proc. SIGIR Conf., pp. 54-69, 1985. 24. L. L. Cherry and N. H. Macdonald, "The Unix Writer's Workbench Software," Byte, vol. 8, no. 10, pp. 241-248, Oct. 1983. 25. G. E. Heidorn, K. Jensen, L. A. Miller, and R. J. Byrd, "The Epistle Text-Critiquing System," IBM Systems J., vol. 21, no. 3, pp. 305-326, 1982. 26. M. O. Harris, Howto: An Amateur System for Program Counseling, 1983. private communication. 27. J. R. Cowie, "Automatic Analysis of Descriptive Texts," Conf. on Applied Natural Language Processing, pp. 117-123, Santa Monica, Cal., Feb. 1-3, 1983. 28. M. S. Tuttle, D. D. Sherertz, M. S. Blois, and S. Nelson, "Expertness from Structured Text? Reconsider: A Diagnostic Prompting System," Conf. on Applied Natural Language Processing, pp. 124-131, Santa Monica, Cal., Feb. 1-3, 1983. 29. Patricia Wright, "Manual Dexterity: a user-oriented approach to creating computer documentation," Human Factors in Computing Sys- tems: Proc. CHI '83 Conference, pp. 11-18, 1983. 30. T. G. Sticht, "Comprehending Reading at Work," in Cognitive Processes in Comprehension, ed. M. A. Just and P. A. Carpenter, Lawrence Erlbaum, 1977. ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Fri Oct 4 18:31:40 1985 Date: Fri, 4 Oct 85 18:31:34 edt From: comsat@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a000194; 4 Oct 85 0:04 EDT Date: Mon 30 Sep 1985 23:06-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #132 To: AIList@SRI-AI Received: from rand-relay by vpi; Fri, 4 Oct 85 13:04 EST AIList Digest Tuesday, 1 Oct 1985 Volume 3 : Issue 132 Today's Topics: Opinion - AI Hype ---------------------------------------------------------------------- Date: Tue, 24 Sep 1985 05:25 EDT From: "David D. Story" Subject: AI hype or .02+.02=.04 Depends on whether you like worth?less gadgets I guess. SEC apparently doesn't according to a recent article in Computerworld regarding Paradyne bidding in 1981. Dave ------------------------------ Date: Tue, 24 Sep 85 08:56:41 EDT From: Herb Lin Subject: SDI/AI/Free and open Debate Date: Sun, 22 Sep 85 19:44:48 PDT From: Richard K. Jennings SDI (ie. Space Development Initiative) is laying the ground work for the commercialization of space which we will all take for granted in 2000 or so. This is most emphatically NOT the function of the President's SDI, or even that of the DoD. Maybe we would wish it to be (I would certainly prefer such a goal to the current one), but it is not. I base my observation on official statements from the President, General Abrahmson, Caspar Weinberger and others. If you discount these statements, then the claim that SDI is for space commercialization is essentially opinion. If you are willing to stay off the interstate higways, the inland waterways, airplanes and other fruits of technology ripened by close association (computers, and computer networks as has been pointed out) -- worry about the military and AI and SDI. But upon close inspection, I think it is better that the military have the technology and work the bugs out on trivial things like autonomous tanks BEFORE it is an integral part of an artificial life support system. Every study that has investigated the funding of technical R&D has concluded that spin-off is an economically unsound way to fund it. If you want to develop spin-offs, then you fund the spin-off directly, not indirectly. Technology development is a good thing, and it must be debugged before it gets into wide public use, but using THAT to justify military spending is to romanticize the military R&D process more than appropriate. Maybe the military is the only institution powerful enough and rich enough to pay for risky R&D. True enough. But that is a social choice that the nation has made; in my view that is inappropriate, and it does not have to be that way. Herb ------------------------------ Date: 24 Sep 1985 10:38 PST From: Mike Kane Reply-to: PRODMKT@ACC.ARPA Subject: AI/SDI Hype I have followed the evolution of AI for several years now, from a mere academic curiosity, to where it is today. True, I am not a participant, just an interested observer. The recent exchanges on the net regarding the commercialization of AI and AI's role in the SDI, have indeed been stimulating, and are too much to let pass without comment. The first point I wish to make is for Capt. Jennings to reread his history of American technology. Case in point: Aircraft. The airplane existed for years before the military leaders in this country viewed this technology as anything more than pure circus. Aviation technology in this country, prior to World War II, was funded and promoted as a purely commercial entity. Remember Billy Mitchell? [The Wright brothers did have Army funding for much of their work, though. -- KIL] True, after WW II, the military began to completely dominate the aeronautical industries in this country. But this occurred only after people like Douglas, Lindbergh, et. al, had proved the technology and commercial viability. There is a direct paralell with AI here. J. Cugini's comments were directly applicable I think. Before AI takes it's place beside data communications, DBMS, etc, as industry commodity segments, it must find a practical purpose in life. I seriously doubt whether SDI or other DOD related applications fulfill this requirement. It may in fact extend the technology, but will AI ever grow wings and fly away, so to speak, without someone finding a practical, dollar breeding, reason for it to exist? This is not intended as a flame directed at academia, but AI must find a path to the marketplace, if it is to survive. You can't expect DARPA funding forever. M. Kane ------------------------------ Date: Wed 25 Sep 85 00:18:07-PDT From: Gary Martins Subject: What does it mean ? In dousing a recent anti-"AI" flame [AIList V3 #126], Prof. Minsky asserts, among other things: To my knowledge, ONLY AI systems, so far, can drive cars, carry on conversations, and debug electronic systems. They don't do these jobs very well yet, but they're coming along -- and have no competition in those areas from any other kind of software. The boldness and economy of this sort of response to criticisms of "AI" have not lost their charm over the years! But, at the risk of falling into flaming ourselves, let's take a closer look. The following clauses are of special interest: (A) ONLY AI systems, so far, can drive cars, carry on conversations, ... (B) They don't do these jobs very well yet (C) but they're coming along (D) [they] have no competition in those areas from any other kind of software and, from earlier in Prof. Minsky's message: (E) AI systems are better than other kinds of software On hearing an authority of Prof. Minsky's stature assert (A), the average intelligent citizen (e.g., business magazine editor, R&D funding officer, corporate manager, housewife, ...) might well conclude: (F) There exist AI systems which can drive cars, and carry on conversations Would Prof. Minsky be comfortable with this inference? Perhaps the conclusion should be qualified, in a manner familiar to real-world systems engineers: (F') There exist AI systems which, while they CANNOT drive cars or carry on conversations (in the ordinary meaning of these phrases), CAN now perform the essentials of these tasks in such a way that they can be straightforwardly scaled up to the real tasks Do you buy this? Well, then, how about: (F") There are no AI systems today which can really drive cars or carry on conversations, but we are keenly hopeful that SOMEDAY such systems may exist Setting these quibbles aside, let's zoom in for an even closer look at the word "ONLY" in (A) [only AI systems ...]. Our intelligent citizen might take this to mean: (G) AI is the ONLY reasonable hope of achieving sophisticated goals like these On the face of it, this would seem to conflict with (B) [but not well ...], given the long history of "AI" research in these areas! But those with short (long-term) memories may be soothed by the time-honored refrain (C) [coming along ...], even without quantification. But we could be flaming up the wrong tree. Maybe (A) is really just a factual boast in modest disguise: (H) The nation's biggest AI Labs have been pretty successful in monopolizing R&D funds in areas like these. [I.e., only AI systems do these jobs because researchers in other disciplines have not been funded to attempt them. -- KIL] Whatever the other merits of this interpretation, it certainly helps us to see what (D) [no competition ...] really means. We are left with (E) [AI is better ...]. Should our intelligent citizen believe it? Like Marxist economics, (E) may be a very difficult thesis to sustain on the factual public record. Like it or not, we live in a world (the so-called "real world") that surrounds us with utterly non-"AI" software that keeps track of payrolls, arranges airline reservations, manages power distribution grids, guides missiles, allocates resources, monitors inventories, analyzes radar signals, does computer animation, assists in mechanical design and fabrication, manipulates spreadsheets, controls space vehicles, drives robots, integrates CAT scans, and performs lots of other mundane tasks. Of course, both (B) [but not well ...] and (C) [coming along ...] still apply to some extent in many of these areas, but the existing accomplishments are genuine and valuable. In fact, there are some nicely engineered non-"AI" systems that play world-class chess, and drive both commercial and military high-performance aircraft in daily operations! Even though "AI" has been around for about as long as the rest of computing, its record of real-world deployment is hardly consistent with (E) [AI is better ...], even now at the crest of the latest "AI" boom. On the contrary, this record seems rather skimpy and inconsequential sometimes, doesn't it? Could it be true that (E) is a kind of modern cult shibboleth, stimulating to believers but mostly just mystifying to the uninitiated? ------------------------------ Date: Wed, 25 Sep 85 08:21 EDT From: Attenber%ORN.MFENET@LLL-MFE.ARPA Subject: AI hype Here is an outsider's opinion regarding AI hype. The mood of a field and the tone of the technical presentations are shaped more by political pressures and events than by the morals of the researchers. As a grad student in particle physics I felt that people were a little devious in presenting results or proposals. Fortunately the audience is always on guard and the question-and-answer sessions tend to be very spirited. Probably several decades of spectacular successes have encouraged people to be optimistic, and fierce competition for machine time and funds encourages people to present their results "in the best possible light". As a researcher in particle physics I observe a much more open, even pessimistic, attitude in oral presentations and publications. This may be due to disappointing results in the early years of the field. Speakers tend to rush to present features of the data which they don't yet understand, and the audience asks questions which are intended to be constructive. And yet the competition for funds is very intense. In fact I feel that the current funding squeeze in plasma physics is partly due to underselling the current encouraging results. I would encourage people in AI to be enthusiastic about prospects for future programs (without, of course, getting caught making a statement which can't be defended.) ------------------------------ Date: 26 Sep 1985 01:39-EST From: Todd.Kueny@G.CS.CMU.EDU Subject: Observations on Expert Systems o The existence of expert systems implies practice (refinement) and physiological learning are not necessarily prerequisites for becomming an expert. o An expert system ignores the transmission loss of the expert => knowledge engineer => program => user data path. o Expert systems filter out ``feel'' (both phsycial and mental), ``intuition'', and other ill defined, illogical quantities experts use when making decsions. The motivation for these observations is derived from the following real world experience: Instead of a computer and some expert system software I will use me. I will presume I am at least as ``intelligent'' as the computer. I now select my domain: conoeing in whitewater. I will also act as my own ``knowledge engineer'' and mentally transcribe the instructions of the expert (the canoe instructor) into my memory; again I assume I am at least as good as a computer knowledge representation and a knowledge engineer. So, I should be well prepared to canoe down some whitewater rapids. I launch my canoe and within the first 100 yards or so I am dumped unceremoniously into the river by a nasty current. . . Anyone who has been involved in a situation such as this realizes the fallacy of attempting to become an expert in a relatively short time without actually experiencing, learning, and practicing within the domain. If the expert system model of ``becoming an expert'' were valid I should be able to become an expert merely by studying an expert system. -Todd K. ------------------------------ Date: Thu, 26 Sep 85 20:27:34 edt From: Brad Miller Subject: Re: AI hype and Marvin Minsky's reply In defense of my friend Bill Anderson: Compare Marvin's posting to Weizenbaum's book "Computer Power and Human Reason". He makes the point that not only is AI hype, but folks like Dr. Minsky may be fundamentally deluded [i.e. their world view that a computer can do anything a person can is incorrect]. Brad Miller miller@rochester.arpa miller!rochester University of Rochester CS Dept. Lab Manager ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Fri Oct 4 18:30:44 1985 Date: Fri, 4 Oct 85 18:30:36 edt From: comsat@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a000885; 4 Oct 85 2:26 EDT Date: Thu 3 Oct 1985 11:10-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #133 To: AIList@SRI-AI Received: from rand-relay by vpi; Fri, 4 Oct 85 17:59 EST AIList Digest Thursday, 3 Oct 1985 Volume 3 : Issue 133 Today's Topics: Opinion - AI Hype & System Verification ---------------------------------------------------------------------- Date: Fri, 27 Sep 85 18:52:47 PDT From: Hibbert.pa@Xerox.ARPA Subject: Re: "SDI/AI/Free and open Debate" in AIList Digest V3 #128 Date: Sun, 22 Sep 85 19:44:48 PDT From: Richard K. Jennings Subject: SDI/AI/Free and open Debate ... For those interested in the history of technology, most of the things we take for granted (microelectronics, automobiles, planes, interstate highway system) were gestated and field tested by the US Military. ... If you are willing to stay off the interstate higways, the inland waterways, airplanes and other fruits of technology ripened by close association (computers, and computer networks as has been pointed out) -- worry about the military and AI and SDI. But upon close inspection, I think it is better that the military have the technology and work the bugs out on trivial things like autonomous tanks BEFORE it is an integral part of an artificial life support system. Agreed, the military was responsible for most of the advances you cite. This doesn't do anything towards convincing me that that's the only possible way for that outcome to have come about, or even that better things wouldn't have happened in the absence of all the money going for these ostensibly military purposes. As a matter of fact, given my belief that ends NEVER justify means, I don't even agree that having those things is good. (Considering that people who didn't consent were forced to help pay for them.) P.S. Why do you think field debugging of autonomous tanks will be less costly/dangerous than of artificial life support systems? In neither case will all the bugs be found in simulations and I'd expect programmers to do better debugging in the midst of doctors practicing than in the middle of a tank battle. Chris ------------------------------ Date: Saturday, 28 September 1985 06:53:19 EDT From: Duvvuru.Sriram@cive.ri.cmu.edu Subject: AI/ES Hype : Some Discussions Gary Martin posed some interesting questions in AILIST # 126. One problem lies in the definition of Artificial Intelligence. In recent issues of AILIST some interesting points were raised as to what constitutes an Expert System. Similar discussion on AI would be useful to the AILIST readers. I understand that Gary Martin worked with the development of many "so called AI" systems and he feels that these systems did not do what they were expected to do. One must note that AI has entered the commercial market only recently and it will, probably, be a few years before we see any successes/failures. The commercialization of AI started after the Japanese initiated the fifth generation computing project. The American press also helped in this endeavor. With all the publicity given to the subject, people working in the area decided to sell their (research) products. I guess they have the right to advertise their products. I view expert system techniques as a programming philosophy. I like it and advocate its use. However, I do not claim that these techniques are AI - just that they evolved from research in AI. As a marketing strategy many tool builders use different techniques to sell their product. If people feel that the AI products are sheer crap, they can always speak up in conferences (such as the one in ES in Govt. symposium), magazines (I saw one such article in Datamation sometime ago, written by Gary Martin) - and even write to Consumer Report. Just because a few people take advantage of the situation and make tall claims does not mean that the whole field should be criticized. I have seen some recent Expert System Tools that claim to induce from examples. These systems are nothing but very good decision table evaluators. I guess the phrase "induce from examples" is used as a marketing strategy. This kind of stuff happens in all fields. It is left to the consumer to decide what to buy (I guess that there are a lot of consultants who are willing to advise you on this subject). Sriram ------------------------------ Date: Thu, 26 Sep 85 20:29:22 -0100 From: Gideon Sahar Subject: Hype, AI, et al. Recently a representative of a well known International company gave a seminar here in Edinburgh, in which we were promised the moon, earth, and virtually the whole milky way. It was claimed that with a little work, and some money, a literal pot of gold is just around the corner. Now I do not disagree with the basic premise that there is enormous promise in AI, and I realize that profits are (one) driving force behind any progress. But I do object to glowing accounts of successes in AI, which turn out (always!) to be X1/XCONN. I find hype exasparating and tiring, but in this case it is also dangerous for the continuing well-being of AI. The glowing promises are going to be proven impossible to fulfill, and the holders of the purse strings will draw them tight and plunge AI into another winter. So please, dampen down the enthusiasm, and inject some realism into your reports and salesmanship. Don't forget to mention the blood, the sweat, and the tears. Gideon Sahar (gideon%uk.ac.ed.edai@ucl-cs.arpa) Dept. of AI University of Edinburgh. ------------------------------ Date: 30-Sep-85 16:00:23-PDT From: jbn@FORD-WDL1.ARPA Subject: SIFT Verification The ``AI hype'' problem has been around for some time in the field of program verification by proof of correctness techniques. In that field, there has been little progress in the last two or three years, despite very impressive claims made in the late 1970s and early 1980s. However, we are now getting some hard evidence as to part of the cause of the trouble. I have just obtained a copy of ``Peer Review of a Formal Verification/Design Proof Methodology'', (NASA Conference Publication 2377, NASA Langley Research Center, Scientific and Technical Information Branch, 1985), which is highly critical of SRI International's work in the area. The work being evaluated is SRI's verification of the Software Implemented Fault Tolerance system, a multiprocessor system intended for use in future aircraft flight control systems. SIFT was a major effort to utilize mathematical proof of correctness technology, including automatic theorem provers, on a real problem. This was a multi-year effort, occupying most of a decade. Some quotes from the report: [p. 22] ``Scientific workers are expected to describe their accomplishments in a way that will not mislead or misinform. Members of the peer review panel felt that many publications and conference presentations of the SRI International verification work have not accurately presented the accomplishments of the project; several panel members, as a result of the peer review, felt that much of what they though had been done had indeed not been done.'' ``The research claims that the panel considered to be unjustified are primarily in two categories; the first concerns the methodology purportedly used by SRI International to validate SIFT, and the second concerns the degree to which the validation had actually been done. Many publications and conference presentations concerning SIFT appear to have misrepresented the accomplishments of the project.'' [p. 23] ``The incompleteness of the SIFT verification exercise caused concern at the peer review. Many panel members who expected (from the literature) a more extensive proof were disillusioned. It was the consensus of the panel that SRI's acomplishment claims were strongly misleading.'' This sort of thing cannot be tolerated. When someone publishes papers that make it look as if a hard problem has been cracked, but the results are not usable by others, it tends to stop other work in the field. The pure researchers avoid the field because the problems appear to be solved and someone else has already taken the credit for the solution, and the system builders avoid the field because the published results don't tell them enough to build systems. This is exactly what has happened to verification. I'm singling out SRI here because this is one of the few cases where a funding agency has called for a formal review of a research project by an outside group and the review resulted in findings as strong as the ones quoted above. Some people at SRI who have seen this note have complained that I am quoting the report out of context, so if you are really interested, you should call NASA Langley (VA) and get the entire report; it's free. NASA permitted SRI to insert a rebuttal of sorts as an appendix after the review took place (in 1983), so the 1985 report gives both the SRI position and that of the review committee. John Nagle [As moderator, I felt it my duty to ask for comments from someone at SRI who knew of this project. I thank John Nagle for acknowledging that the review committee's findings are disputed, but feel that additional points in our behind-the-scenes discussion are worth passing along. John's points about the effect of hype on AI research are important and well within the scope of AIList discussion. The current state of the art in automatic verification is also appropriate for AIList (or for Soft-Eng@MIT-XX or ARMS-D@MIT-MC). I do not know whether the merits and demerits of this particular verification project and its peer review are worthy of discussion; I am simply attempting to pass along as balanced and complete a presentation as possible. I hope that my editorial decisions do not reflect undue bias on behalf of my employer, SRI International. -- KIL] Date: Thu 26 Sep 85 15:05:12-PDT From: MELLIAR-SMITH@SRI-AI.ARPA Subject: John Nagle's AIList Request The report refered to by John Nagle (Peer Review of a Formal Verification/Design Proof Methodology, NASA Conference Publication 2377, NASA Langley Research Center, 1985) is a 50-odd page evaluation of an attempt to provide a proof for a realistic system. The project in its entirety was, and still is, beyond the capabilities of our currently available technology, and thus the proof exercise was incomplete; the proofs achieved were design level proofs, down to the level of prepost conditions, rather than proofs to the level of code. The report reviewed what was done, and not entirely negatively. The main cause for concern was that some descriptions of the work did not make sufficiently clear precisely what had been done and what the limitations of the work were. In particular, the first quotation cited by John Nagle continues: "In many cases, the misrepresentation stems from the omission of facts rather than from inclusion of falsehood." [...] A more recent, and very painstaking, peer review of SRI's work in verification, together with the corresponding work at GE Research Labs, SDC, and University of Texas, will be published by the DOD Computer Security Center next month. Michael Melliar-Smith Date: 26-Sep-85 16:35:35-PDT From: jbn@FORD-WDL1.ARPA Subject: Re: SIFT [...] Unfortunately, previous statements have contained rather strong claims. The peer review quotes from Meillar-Smith's and Richard Schwartz's ``Hierchical Specification of the SIFT Fault-tolerant Flight Control System,'' CSL-123, SRI International, March 1981: ``This paper describes the methodology employed to demonstrate rigorously that the SIFT computer meets its reliability requirements.'' And from the same authors, in ``Formal Specification and Mechanical Verification of SIFT'', IEEE Trans on Computers, C-31, #7, July, 1982: ``The formal proof, that a SIFT system in a `safe' state operates correctly despite the presence of arbitrary faults, has been completed all the way from the most abstract specification to the PASCAL program. This proof has been performed using STP, a new specification and verification system, designed and developed at SRI International. Except where explicitly indicated, every specification exhibited in this paper has been mechanically proven consistent with the requirements on SIFT and with the Pascal implementation. The proof remains to be completed that the SIFT executive performs an appropriate, safe, and timely reconfiguration in the presence of faults.'' Those are strong statements. And they just aren't true. [... you have selected two statements from reports of mine about SIFT and stated that they are not true. YOU ARE WRONG! ... -- Michael Melliar-Smith] Unless we make it clear where we are today, future work will founder. When you fail and admit it, the field moves forward; everyone learns what doesn't work and what some of the problems are. But when failure is hidden, new workers are led to make the same mistakes again. I'd like to see verification work and be used. Excessive claims have seriously damaged this field. Only recently, though, has solid information debunking these claims become available. Dissemination of the information in this report will help those in the field make progress that is real. [...] John Nagle ------------------------------ End of AIList Digest ******************** From csvpi@vpics1 Fri Oct 4 22:23:08 1985 Date: Fri, 4 Oct 85 22:23:02 edt From: csvpi@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a004894; 4 Oct 85 13:46 EDT Date: Thu 3 Oct 1985 20:20-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #134 To: AIList@SRI-AI Received: from rand-relay by vpi; Fri, 4 Oct 85 22:13 EST AIList Digest Friday, 4 Oct 1985 Volume 3 : Issue 134 Today's Topics: Queries - Micro LISP & DEC20 LISP & Expert System Tools, LISP - Franz Functions, Survey - AI Project Planning, Bibliography - Connectionism and Parallel Distributed Processing ---------------------------------------------------------------------- Date: Thu 3 Oct 85 17:30:30-PDT From: Robert Blum Subject: LISP on micros: need info and refs I teach a tutorial every year at SCAMC (the largest medical computing meeting in the US) on AI in medicine. My students invariably want to know how to run LISP on their micros. I'd like to request information on the various currently marketed LISPs for micros (their cost, their speed, quality, facilities). Pointers to review articles would also be helpful. Thanks, Bob Blum (BLUM@sumex) [Has anyone saved past AIList micro-LISP messages in a convenient form? My archive is not in a convenient form for multimessage topical searches. -- KIL] ------------------------------ Date: 4 Oct 1985 10:06-EST From: (Steinar Kj{rnsr|d) Subject: Common Lisp for DEC20/TOPS20 I have a paper in front of me dated July 1984 describing avaiable Common Lisps. I'm aware of several new implementations since July 1984(!), and especially I'm interested in versions for DEC20/TOPS20 and VAX/UNIX. Does anyone know *IF* and *WHERE* these versions are avaiable ? Steinar Kjaernsroed, Institute of Informatics, University of Oslo, NORWAY ------------------------------ Date: Fri 4 Oct 85 08:19:08-PDT From: Mark Richer Subject: AI software tools Several months ago I wrote a report on commercial AI software tools (KEE, ART, S.1, DUCK, and SRL+ ... now called Language Craft). The paper was basically in two parts : 1) general criteria for evaluating tools, 2) a description and partial evaluation of the five tools. I am now revising this paper and would welcome input from people. I am especially interested in comments from people that have used of these software tools OR another tool not described in my paper, but which you feel is another good candidate for discussion. ------------------------------ Date: Wed, 2 Oct 85 10:36 EST From: Atul Bajpai Subject: QUERY: Expert System Building Tools Any ideas about what would be the implications (advantages/problems) of using two (or more) different Expert System Tools (eg. KEE, ART, S.1, Knowledge Carft etc.) to a single application. Has anyone done this before? Is it practical to try such a thing? Any and all comments are welcome. Thanks. --Atul Bajpai-- (bajpai%gmr.csnet@csnet-relay OR atul@sushi.arpa) ------------------------------ Date: Wed, 2 Oct 85 09:35:06 edt From: Walter Hamscher Subject: Franz Functions (1) (getenv '|USER|) will return the user ID (Franz manual section 6). Then you can either write a C routine that uses _getpwent (Unix manual section 3) and link it in using cfasl (Franz section 8.4), or franz code that opens and scans through the file /etc/passwd. (2) (status ctime). Franz manual section 6. (3) `plist'. Section 2. (4) Try closing the file after you write to it, then reopening it in "a" mode. Avoid filepos. Incidentally, the mailing list "franz-friends@Berkeley.ARPA" is a much more appropriate place for questions like this. ------------------------------ Date: 1 Oct 1985 15:12:53-EDT From: kushnier@NADC Subject: AI Project Planning Survey AI PROJECT PLANNING SURVEY In order to develop a better planning methodology for AI project management, the Artificial Intelligence group at the Naval Air Development Center is conducting a survey among readers of AILIST. The answers to the survey will provide guidelines and performance metrics which will aid in the ability to "scope out" future AI projects. A summary of the results of this survey will be provided to the respondents. Although complete entries are desired, any information will help. Response to this survey must be made at no cost to the Government. Please respond by 18 Oct 1985. General Project Information 1. PROJECT NAME: (MYCIN, R1, PROSPECTOR, . . .) 2. NAME OF DEVELOPMENT GROUP: 3. NAME AND ADDRESS OF CONTACT: 4. SHORT DESCRIPTION OF PROJECT: 5. TYPE OF SYSTEM: (Interpretation, Diagnosis, Monitoring, Prediction, Design, Planning, Control, Debugging, Repair, Instruction) 6. DEVELOPMENT DATES: 7. CONTACT FOR SOFTWARE AND DOCUMENTATION: 8. CURRENT LEVEL OF PROGRESS: (Fully successful, still developing, demonstrated but not in use, . . .) Implementation Specifics 9. METHOD OF KNOWLEDGE REPRESENTATION: (Frames, production rules,...) 10. IMPLEMENTATION LANGUAGE: (Lisp, Prolog, Ops5, . . .) 11. COMMERCIAL TOOLS USED (if any): (KEE, ART, M.1, . . .) 12. NUMBER OF MAN-YEARS INVESTED: Knowledge representation: Knowledge acquisition: Implementation: 13. HOST COMPUTER: (Vax, Symbolics, IBM-PC, . . .) 14. NUMBER AND AVERAGE SIZE IN BYTES OF RULES, FRAMES, WFFs OR OBJECTS (where applicable): Performance Criterion 15. PROGRAM EXECUTION TIME: (Avg. rule firings/task, rules fired/sec, . . .) 16. AMOUNT OF MEMORY IN BYTES FOR: Short term memory: Long term memory: Procedural memory: 17. IN WHAT WAYS DID THE SCOPE OF THE DOMAIN CHANGE? (Full scope, rescoping needed,...) Knowledge Acquisition 18. KNOWLEDGE ACQUISITION EFFORT IN MAN-YEARS: 19. NUMBER OF EXPERTS USED: 20. EXPERT SELECTION METHOD: 21. EXPERT INTERVIEWING METHODOLOGY: Please send your responses to: Ralph Fink, rfink@NADC.ARPA Code 5021 Naval Air Development Center Warminster, PA 18974 ------------------------------ Date: Tue, 1 Oct 85 00:51:02 est From: "Marek W. Lugowski" Subject: A reading list for a "Connectionism and Parallel Dist. Proc." course After months of brainstorming (er, inaction), here comes at last the promised reading list for Indiana U.'s planned connectionism course, 78 items long--just right for a semester. IU's a liberal arts school, after all. :-) -- Marek A list of sources for teaching a graduate AI course "Connectionism and Parallel Distributed Processing" at Indiana University's Computer Science Department. Compiled by Marek W. Lugowski (marek@indiana.csnet) and Pete Sandon (sandon@wisc-ai.arpa), Summer 1985. Ackley, D. H., "Learning Evaluation Functions in Stochastic Parallel Networks," CMU Thesis Proposal, December 1984. Amari, S-I., "Neural Theory of Association and Concept-Formation," Biological Cybernetics, vol. 26, pp. 175-185, 1977. Ballard, D. H., "Parameter Networks: Toward a Theory of Low-Level Vision," IJCAI, vol. 7, pp. 1068-1078, 1981. Ballard, D. H. and D. Sabbah, "On Shapes," IJCAI, vol. 7, pp. 607-612, 1981. Ballard, D. H., G. E. Hinton, and T. J. Sejnowski, "Parallel Visual Computation," Nature, vol. 103, pp. 21-26, November 1983. Ballard, D. H., "Cortical Connections: Structure and Function," University of Rochester Technical Report #133, July 1984. Block, H. D., "A Review of "Perceptrons: An Introduction to Computational Geometry"," Information and Control, vol. 17, pp. 501-522, 1970. Bobrowski, L., "Rules of Forming Receptive Fields of Formal Neurons During Unsupervised Learning Processes," Biological Cybernetics, vol. 43, pp. 23-28, 1982. Bongard, M., "Pattern Recognition", Hayden Book Company (Spartan Books), 1970. Brown, C. M., C. S. Ellis, J. A. Feldman, T. J. LeBlanc, and G. L. Peterson, "Research with the Butterfly Multicomputer," Rochester Research Review, pp. 3-23, 1984. Christman, D. P., "Programming the Connection Machine," MIT EECS Department Masters Thesis, 1984. Csernai, L. P. and J. Zimanyi, "Mathematical Model for the Self-Organization of Neural Networks," Biological Cybernetics, vol. 34, pp. 43-48, 1979. Fahlman, S. E., NETL, a System for Representing and Using Real Knowledge , MIT Press, Cambridge, Massachusetts, 1979. Fahlman, S. E., G. E. Hinton, and T. J. Sejnowski, "Massively Parallel Architectures for AI: NETL, Thistle and Boltzmann Machines," Proceedings of the National Conference on Artificial Intelligence, 1983. Feldman, J. A., "A Distributed Information Processing Model of Visual Memory," University of Rochester Technical Report #52, December 1979. Feldman, J. A., "Dynamic Connections in Neural Networks," Biological Cybernetics, vol. 46, pp. 27-39, 1982. Feldman, J. A. and D. H. Ballard, "Computing with Connections," in Human and Machine Vision, ed. J. Beck, B. Hope and A. Rosenfeld (eds), Academic Press, New York, 1983. Feldman, J. A. and L. Shastri, "Evidential Inference in Activation Networks," Rochester Research Review, pp. 24-29, 1984. Feldman, J. A., "Connectionist Models and Their Applications: Introduction," Special Issue of Cognitive Science, vol. 9, p. 1, 1985. Fry, G., ed., Nanotechnology Notebook, an unpublished collection of published and unpublished material on molecular computing. Contact either the editor (cfry@mit-prep.arpa) or Scott Jones (saj@mit-prep.arpa) at MIT Artificial Intelligence Laboratory for distribution and/or bibliography information. Contains material by Eric Drexler, Richard Feynman, Kevin Ulmer and others. Fukushima, K., "Cognitron: A Self-organizing Multilayered Neural Network," Biological Cybernetics, vol. 20, pp. 121-136, 1975. Fukushima, K., "Neocognitron: A Self-organizing Neural Network Model for a Mechanism of Pattern Recognition Unaffected by Shift in Position," Biological Cybernetics, vol. 36, pp. 193-202, 1980. Hebb, D. O., The Organization of Behavior, Wiley, New York, 1949. Hewitt, C., "Viewing Control Structures as Patterns of Passing Messages", Artificial Intelligence: An MIT Perspective, Winston and Brown, Editors, MIT Press, Cambridge, Massachusetts, 1979. Hewitt, C. and P. de Jong, "Open Systems", MIT Artificial Laboratory Memo #691, December 1982. Hewitt, C. and H. Lieberman, "Design Issues in Parallel Architectures for Artificial Intelligence", MIT Artificial Intelligence Lab Memo #750, November 1983. Hewitt, C., an article on asynchronous parallel systems not simulable by Turing Machines or Omega-order logic, BYTE, April 1985. Hillis, D. W., "New Computer Architectures and Their Relationship to Physics or Why Computer Science Is No Good", International Journal of Theoretical Physics, vol. 21, pp. 255-262, 1982. Hinton, G. E., "Relaxation and its Role in Vision," University of Edinburgh Ph.D. Dissertation, 1977. Hinton, G. E. and J. A. Anderson, Parallel Models of Associative Memory, Lawrence Erlbaum Associates, Hillsdale, New Jersey, 1981. Hinton, G. E. and T. J. Sejnowski, "Optimal Perceptual Inference," Proceedings of the IEEE Computer Society Conference on CV and PR, pp. 448-453, June 1983. Hinton, G. E., T. J. Sejnowski, and D. H. Ackley, "Boltzmann Machines: Constraint Satisfaction Networks that Learn," CMU Department of Computer Science Technical Report No. 84-119, May 1984. Hinton, G. E., "Distributed Representations", CMU Department of Computer Science Technical Report No. 84-157, October 1984. Hirai, Y., "A New Hypothesis for Synaptic Modification: An Interactive Process between Postsynaptic Competition and Presynaptic Regulation," Biological Cybernetics, vol. 36, pp. 41-50, 1980. Hirai, Y., "A Learning Network Resolving Multiple Match in Associative Memory," IJCPR, vol. 6, pp. 1049-1052, 1982. Hofstadter, D. R., "Who Shoves Whom Inside the Careenium?, or, What is the Meaning of the Word 'I'?", Indiana University Computer Science Department Technical Report #130, Bloomington, Indiana, 1982. Hofstadter, D, "The Architecture of Jumbo", Proceedings of the International Machine Learning Workshop, Monticello, Illinois, June 1983. Hofstadter, D. R., "The Copycat Project", MIT Artificial Intelligence Lab Memo #755, Cambridge, Massachusetts, January 1984. Hofstadter, D. R., Metamagical Themas, Basic Books, New York, 1985. Hopfield, J.J., "Neural Networks and Physical Systems with Emergent Collective Computational Abilities," Proceedings of the National Academy of Sciences, vol. 79, pp. 2554-2558, 1982. Jefferson, D. E., "Virtual Time", UCLA Department of Computer Scienced Technical Report No. 83-213, Los Angeles, California, May 1983. Jefferson, D. E. and H. Sowizral, "Fast Concurrent Simulation Using the Time Warp Mechanim, Part 1: Local Control", Rand Corporation Technical Report, Santa Monica, California, June 1983. Jefferson, D. E. and H. Sowizral, "Fast Concurrent Simulation Using the Time Warp Mechanim, Part 2: Global Control", Rand Corporation Technical Report, Santa Monica, California, August 1983. John, E. R., "Switchboard versus Statistical Theories of Learning and Memory," Science, vol. 177, pp. 850-864, September 1972. Kandel, E. R., "Small Systems of Neurons," Scientific American, vol. 241, pp. 67-76, September 1979. Kanerva, P., Self-Propagating Search: A Unified Theory of Memory, Center for Study of Language and Information (CSLI) Report No. 84-7 (Stanford Department of Philosophy Ph.D. Dissertation), Stanford, California, 1984. To appear as a book by Bradford Books (MIT Press). Kanerva, P., "Parallel Structures in Human and Computer Memory", Cognitiva 85, Paris, June 1985. Kirkpatrick, S., and C. D. Gelatt, Jr. and M. P. Vecchi, "Optimization by Simulated Annealing", Science, vol. 220, no. 4598, 13 May 1983. Kohonen, T., "Self-Organized Formation of Topologically Correct Feature Maps," Biological Cybernetics, vol. 43, pp. 59-69, 1982. Kohonen, T., "Analysis of a Simple Self-Organizing Process," Biological Cybernetics, vol. 44, pp. 135-140, 1982. Kohonen, T., "Clustering, Taxonomy, and Topological Maps of Patterns," IJCPR, vol. 6, pp. 114-128, 1982. Kohonen, T., Self-Organization and Associative Memory, 2nd edition, Springer-Verlag, New York, 1984. Lieberman, H., "A Preview of Act1", MIT Artificial Intelligence Lab Memo #626. Malsburg, C. von der, "Self-Organization of Orientation Sensitive Cells in the Striate Cortex," Kybernetik, vol. 14, pp. 85-100, 1973. McClelland, J. L., "Putting Knowledge in its Place: A Scheme for Programming Parallel Processing Structures on the Fly," Cognitive Science, vol. 9, pp. 113-146, 1985. McClelland, J. L. and D. E. Rumelhart, "Distributed Memory and the Representation of General and Specific Information," Journal of Experimental Psychiatry: General, vol. 114, pp. 159-188, 1985. McClelland, J. L. and D. E. Rumelhart, Editors, Parallel Distributed Processing: Explorations in the Microstructure of Cognition, Volume 1, Bradford books (MIT Press), Cambridge, Massachusetts, 1985 (in press). McCullough, W. S., Embodiments of Mind, MIT Press, Cambridge, Massachusetts, 1965. Minsky, M. and S. Papert, Perceptrons: An Introduction to Computational Geometry, MIT Press, Cambridge, Massachusetts, 1969. Minsky, M., "Plain Talk about Neurodevelopmental Epitemology", MIT Artificial Intelligence Lab Memo #430, June 1977. Minsky, M., "K-Lines: A Theory of Memory", MIT Artificial Intelligence Lab Memo #516, June 1979. Minsky, M., "Nature Abhors an Empty Vaccum", MIT Artificial Intelligence Lab Memo #647, August 8, 1981. Mozer, M. C., "The Perception of Multiple Objects: A Parallel, Distributed Processing Approach", UCSD Thesis Proposal, Institute for Cognitive Science, UCSD, La Jolla, California, August 1984. Nass, M. M. and L. N. Cooper, "A Theory for the Development of Feature Detecting Cells in Visual Cortex," Biological Cybernetics, vol. 19, pp. 1-18, 1975. Norman, "Categorization of Action Slips", Psychological Review, vol. 88, no. 1, 1981. Palm, G., "On Associative Memory," Biological Cybernetics, vol. 36, pp. 19-31, 1980. Plaut, D. C., "Visual Recognition of Simple Objects by a Connection Network," University of Rochester Technical Report #143, August 1984. Rosenblatt, F., Principles of Neurodynamics, Spartan Books, New York, 1962. Rumelhart, D. E. and D. Zipser, "Feature Discovery by Competitive Learning," Cognitive Science, vol. 9, pp. 75-112, 1985. Sabbah, D., "Design of a Highly Parallel Visual Recognition System," IJCAI, vol. 7, pp. 722-727, 1981. Sabbah, D., "Computing with Connections in Visual Recognition of Origami Objects," Cognitive Science, vol. 9, pp. 25-50, 1985. Smolensky, P., "Schema Selection and Stochastic Inference in Modular Environments", Proceedings of the AAAI-83, Washington, 1983. Theriault, D. G., "Issues in the Design and Implementation of Act2", MIT Artificial Intelligence Lab Technical Report #728, June 1983. Touretzky, D. S. and G. E. Hinton, "Symbols Among the Neurons: Details of a Connectionist Inference Architecture," IJCAI, vol. 9, pp. 238-243, 1985. Uhr, L., "Recognition Cones and Some Test Results," in Computer Vision Systems, ed. A. R. Hanson and E. M. Riseman, Academic Press, New York, 1978. Uhr, L. and R. Douglass, "A Parallel-Serial Recognition Cone System for Perception: Some Test Results," Pattern Recognition, vol. 11, pp. 29-39, 1979. Van Lehn, K., "A Critique of the Connectionist Hypothesis that Recognition Uses Templates, not Rules," Proceedings of the Sixth Annual Conference of the Cognitive Science Society, 1984. Willshaw, D. J., "A Simple Network Capable of Inductive Generalization," Proceedings of the Royal Society of London, vol. 182, pp. 233-247, 1972. A list of contacts known to us that have taught or are interested in teaching a similar course/seminar: J. Barnden, Indiana U. Computer Science G. Hinton, CMU Computer Science D. Hofstadter, U. of Michigan Psychology J. McClelland, CMU Psychology D. Rumelhart, UCSD Psychology P. Smolensky, U. of Colorado Computer Science D. Touretzky, CMU Computer Science ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Tue Oct 8 04:41:52 1985 Date: Tue, 8 Oct 85 04:41:48 edt From: comsat@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a019826; 7 Oct 85 2:50 EDT Date: Sun 6 Oct 1985 22:55-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #135 To: AIList@SRI-AI Received: from rand-relay by vpi; Tue, 8 Oct 85 04:33 EST AIList Digest Monday, 7 Oct 1985 Volume 3 : Issue 135 Today's Topics: Query - TIMM Expert System Tool, Psychology & Logic - Probabilistic Counterexample to Modus Ponens ---------------------------------------------------------------------- Date: Fri 4 Oct 85 23:49:55-EDT From: Richard A. Cowan Subject: Expert System Tool Does anyone know anything about TIMM, an "expert system tool" put out by General Research Corp? I hear it costs $50,000 or so. I'd be interested to hear what a tool could do that costs that much money, especially in comparison to KEE. Don't want to hear much; just send a "thumbs up/thumbs down" reply directly to cowan@mit-xx. Thanks, Rich ------------------------------ Date: Tue, 1 Oct 85 20:07:27 PDT From: Hibbert.pa@Xerox.ARPA Subject: Re: A Counterexample to Modus Ponens In a recent issue of AIList, John McLean cited an article about inconsistencies in public opinion that apparently said: Before the 1980 presidential election, many held the two beliefs below: (1) If a Republican wins the election then if the winner is not Ronald Reagan, then the winner will be John Anderson. (2) A Republican will win the election. Yet few if any of these people believed the conclusion: (3) If the winner is not Reagan then the winner will be Anderson. I would say the problem with this analysis is that people believed, instead of statement 1, the following similar statement: (1a) If a Republican wins the election and the winner is not Ronald Reagan, then the winner will be John Anderson. People may have been willing to say that they believed #1, but that's only because they didn't know the difference bewtween 1 and 1a. Chris ------------------------------ Date: Wed, 2 Oct 85 14:11:34 edt From: John McLean Subject: Re: A Counterexample to Modus Ponens Chris Hibbert says that the purported counterexample to modus ponens I reported rests on a distinction between (1) If a Republican wins the election then if the winner is not Ronald Reagan, then the winner will be John Anderson. and (1a) If a Republican wins the election and the winner is not Ronald Reagan, then the winner will be John Anderson. He says that People may have been willing to say that they believed #1, but that's only because they didn't know the difference bewtween 1 and 1a. I would like to see some further discussion of this since I'm afraid that I don't see the difference between (1) and (1a) either. Certainly there is no difference with respect to inferential power as far as classical logic is concerned. John ------------------------------ Date: Thu, 3 Oct 85 13:24:56 PDT From: Hibbert.pa@Xerox.ARPA Subject: Re: A Counterexample to Modus Ponens Oops. I didn't think through very clearly what I had in mind. (And definitely didn't say clearly what I was thinking.) I'll try again. People's thinking may have been closer to probabilities rather than classical logic. (Since most people don't understand either in any formal way.) I'd like to claim that people believed something more like the following: (1b) If a Republican wins the election, then if the winner is not Ronald Reagan, then the winner will probably be John Anderson. (2b) A Republican will probably win the election. without believing: (3b) If the winner is not Reagan then the winner will probably be Anderson. And these beliefs are entirely consistent. (Given that peoples standards for what constitutes a high probability are remarkably inconsistent.) Thanks John for making me think out more clearly what I had in mind. Chris ------------------------------ Date: 2 Oct 85 08:33:00 EDT From: "CUGINI, JOHN" Reply-to: "CUGINI, JOHN" Subject: A (supposed) Counterexample to Modus Ponens > Date: Fri, 27 Sep 85 10:57:18 edt > From: John McLean > > In the most recent issue of The_Journal_of_Philosophy, there is an > article by Vann McGee that presents several counterexamples to modus > ponens. I am not sure whether to count them as counterexamples or as > cases where we hold inconsistent beliefs. If the latter view is right, > it should be of interest to those who model belief systems. [...] I think this is just a problem with imprecisely stated beliefs: surely the reasonable interpretation is that one believes (1) in the simple sense that it's (almost certainly) true, but one believes: (2') it is *probable* that a Republican will win the election. After all, if you believed it was *certain* that "a Republican wins the election" was true (maybe because you woke up the morning after the election and someone told you: a Republican has won), you would then believe (3). "Normal" rules of logic don't work with probabilistic statements in lots of cases, eg: I believe: (1) the die will not come up as a 1. (2) the die will not come up as a 2. ... (6) the die will not come up as a 6. but I certainly don't believe the conjunction of (1)-(6). Of course, (1)-(6) should really be stated as: (1) the die will (probably) not come up as 1, etc. So, I consider these cases as neither true counterexamples to modus ponens, nor as examples of inconsistent beliefs. John Cugini Institute for Computer Sciences and Technology National Bureau of Standards Bldg 225 Room A-265 Gaithersburg, MD 20899 phone: (301) 921-2431 ------------------------------ Date: Thu, 3 Oct 85 16:40:49 edt From: John McLean Subject: Modus Ponens I agree that probability is the way to go in modeling beliefs. It is easy to construct a distribution where the probabilities for (1) and (2) are both quite high but where the probability of (3) is quite low even though (3) follows from (1) and (2). Hence if believing p means that one assigns a high probability to it, it is possible to believe (1) and (2) without believing their logical consequence. John ------------------------------ Date: Thu, 3 Oct 85 18:28:03 PDT From: albert@UCB-VAX.Berkeley.EDU (Anthony Albert) Subject: Re: Counterexample to modus ponens (1) If a Republican wins the election then if the winner is not Ronald Reagan, then the winner will be John Anderson. (2) A Republican will win the election. Yet few if any of these people believed the conclusion: (3) If the winner is not Reagan then the winner will be Anderson. I don't think the given example is a counter-example of modus-ponens or of contradictory belief systems. It is not modus-ponens because the statements are not true or false, but are beliefs. You cannot make a statement that is true or false about the future. As far as beliefs, a non-contradictory set could be: 1) If a Republican wins then Reagan will win. 2) If Reagan doesn't win then a Republican won't win. 3) Unless a Democrat wins, a Republican will win. A belief X can always be qualified, usually by an all things being equal type of default qualification. The given example ignores these implicit conditions and so leads to a contradiction. ------------------------------ End of AIList Digest ******************** From csvpi@vpics1 Tue Oct 8 04:50:16 1985 Date: Tue, 8 Oct 85 04:50:10 edt From: csvpi@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a020313; 7 Oct 85 4:16 EDT Date: Sun 6 Oct 1985 23:26-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #136 To: AIList@SRI-AI Received: from rand-relay by vpi; Tue, 8 Oct 85 04:35 EST AIList Digest Monday, 7 Oct 1985 Volume 3 : Issue 136 Today's Topics: Seminars - Higher-Order Logic Features in Prolog (UPenn) & Cognitive Science and Computers (UCB) & The Programmer's Apprentice: KBEmacs (CMU) & Belief, Awareness, and Limited Reasoning (SU) & Planning Under Uncertainty using Simulation (SU) & Aggregation in Qualitative Simulation (MIT) & Conflict in Problem Solving (MIT) & Introspection (SU) & Compact Lisp Machine (SMU), Seminar Series - IBM Yorktown Projects (Rutgers) ---------------------------------------------------------------------- Date: Sun, 29 Sep 85 10:44 EDT From: Dale Miller Subject: Seminar - Higher-Order Logic Features in Prolog (UPenn) [Forwarded from the Prolog Digest by Restivo@SU-SCORE.] A student of mine is holding a seminar at the University of Pennsylvania that might be of interest to the Prolog bboard readers. -Dale Miller Joint Mathematics / Computer Science LOGIC COLLOQUIUM Introducing Higher-Order Logic Features into Prolog Gopalan Nadathur Monday 30th September 1985 4:40 p.m., DRL 4E17 This talk reports work being undertaken towards a doctoral dissertation under the supervision of Prof. Dale Miller. This work is motivated by a desire to examine whether certain features afforded by higher-order logics are useful in a computational setting. In this talk we shall present a language that is similar to that of Horn Clauses of first-order logic except that first-order terms are now replaced by typed lambda-calculus terms. We shall discuss a theorem-prover based on higher-order unification for this logic. We shall also attempt to motivate the usefulness of this language for specifying and performing computations. We are interested in extending this language by permitting suitably restricted occurrences of predicate variables, and we shall conclude our talk by a brief discussion of the issues involved in doing so. ------------------------------ Date: Wed, 2 Oct 85 12:46:31 PDT From: admin@ucbcogsci.Berkeley.EDU (Cognitive Science Program) Subject: Seminar - Cognitive Science and Computers (UCB) BERKELEY COGNITIVE SCIENCE PROGRAM Fall 1985 Cognitive Science Seminar -- IDS 237A TIME: Tuesday, October 8, 11:00 - 12:30 PLACE 240 Bechtel Engineering Center DISCUSSION: 12:30 - 1:30 in 200 Building T-4 SPEAKER: Terry Winograd, Computer Science, Stanford University TITLE: "What Can Cognitive Science Tell Us About Computers?" Much work in cognitive science rests on the assumption that there is a common form of "information processing" that under- lies human thought and language and that also corresponds to the ways we can program digital computers. The theory should then be valid both for explaining the functioning of the machines (at whatever level of "intelligence") and for under- standing how they can be integrated into human situations and activities. I will argue that theories like those of current cognitive science are based on a "rationalistic" tradition, which is appropriate for describing the mechanics of machine operation, but is inadequate for understanding human cognitive activity and misleading as a guide to the design and application of computer technology. The emphasis will be on looking at alternatives to this tradition, as a starting point for under- standing what computers really can do. ------------------------------ Date: 1 Oct 1985 1410-EDT From: Sylvia Brahm Subject: Seminar - The Programmer's Apprentice: KBEmacs (CMU) SPEAKER: Richard C. Waters (AIL, MIT) TOPIC: The Programmer's Apprentice: A Session with KBEmacs WHEN: Friday, October 18, 1985 WHERE: Wean Hall 4605 TIME: 1:30 P.M. The Knowledge-Based Editor in Emacs (KBEmacs) is the current demonstra- tion system implemented as part of the Programmer's Apprentice project. KBEmacs is capable of acting as a semi-expert assistant to a person who is writing a program -- taking over some parts of the programming task. Using KBEmacs, it is possible to construct a program by issuing a series of high level commands. This series of commands can be as much as an order of magnitude shorter than the program it describes. KBEmacs is capable of operating on Ada and Lisp programs of realistic size and complexity. Although KBEmacs is neither fast enough nor robust enough to be considered a true prototype, both of these problems could be overcome if the system were to be reimplemented. ------------------------------ Date: Tue 1 Oct 85 08:40:19-PDT From: Anne Richardson Subject: Seminar - Belief, Awareness, and Limited Reasoning (SU) DAY October 1, 1985 EVENT Computer Science Colloquium PLACE Skilling Auditorium TIME 4:15 TITLE Belief, Awareness, and Limited Reasoning PERSON Dr. Joe Halpern FROM IBM Corporation BELIEF, AWARENESS, AND LIMITED REASONING Classical possible-worlds models for knowledge and belief suffer from the problem of logical omniscience: agents know all tautologies and their knowledge is closed under logical consequence. This unfortunately is not a very accurate account of how people operate! We review possible-worlds semantics, and then go on to introduce three approaches towards solving the problem of logical omniscience. In particular, in our logics, the set of beliefs of an agent does not necessarily contain all valid formulas. One of our logics deals explicitly with awareness, where, roughly speaking, it is necessary to be aware of a concept before one can have beliefs about it, while another gives a model of local reasoning, where an agent is viewed as a society of minds, each with its own cluster of beliefs, which may contradict each other. The talk will be completely self-contained. ------------------------------ Date: Tue 1 Oct 85 14:27:13-PDT From: Alison Grant Subject: Seminar - Planning Under Uncertainty using Simulation (SU) Medical Information Sciences Colloquium Thursday, October 3, 1985 Stanford University Medical Center Room M-114 1:15-2:15 P.M. Speaker: Curt Langlotz, MIS Program Title: Planning under uncertainty using probabilistic and symbolic simulation Artificial intelligence research has largely concentrated on solving two kinds of planning problems: (1) problems for which there is certainty about the consequences of action and for which the planning goals can be met completely (e.g., robot movement between rooms in a building), and (2) problems for which explicit guidelines exist for the construction of plans (e.g., the ONCOCIN, MOLGEN, and ATTENDING programs). However, many planning problems are characterized by a lack of explicit plan construction guidelines, goals that are difficult to satisfy completely, and actions whose consequences cannot be predicted with certainty. This talk will describe an architecture for planning in such situations and will outline the motivations behind its design. One key component of this new architecture is the ability to predict the consequences of plans. A simulation architecture is currently under development to make these predictions. It will also be described, along with the motivations for rejecting existing simulation techniques (both qualitative and deterministic) in the domain of cancer therapy planning. ------------------------------ Date: Mon, 30 Sep 1985 16:21 EDT From: Peter de Jong Reply-to: Cog-Sci-Request%MIT-OZ Subject: Seminar - Aggregation in Qualitative Simulation (MIT) [Forwarded from the MIT bboard by SASW@MIT-MC.] Thursday 3, October 4:00pm Room: NE43- 8th floor Playroom The Artificial Intelligence Lab Revolving Seminar Series "The Use of Aggregation in Qualitative Simulation" Daniel S. Weld MIT AI Lab. I introduce a technique called aggregation which has several applications to the problem of qualitative simulation and envisioning: - It can simplify reasoning by dynamically creating more abstract process descriptions of the types of change occurring in a system. - It can enable the application of powerful continuous analytic techniques such as limit analysis to systems whose descriptions include discrete processes. - It can direct the reformulation of quantities to more abstract representations. Aggregation works by searching the simulation history structure to find cycles of repeating processes. Once cycles have been detected, a more abstract continuous process, equivalent to the net effect of the cycle, is created. Analysis now proceeds on the continuous process. Aggregation correctly handles cycles that contain other cycles. A program called PEPTIDE has been written to test these ideas in the domain of molecular genetics. Paper tests have also been done in the domains of digital electronics and xerography. ------------------------------ Date: Mon 30 Sep 85 18:18:03-EDT From: Michael Eisenberg Subject: Seminar - Conflict in Problem Solving (MIT) [Forwarded from the MIT bboard by Laws@SRI-AI.] Andre Boder is scheduled to give a talk at the next ideas seminar, TOMORROW, Tuesday Oct. 1, at 4:30. The talk is scheduled to be held at the 3rd floor conference room in NE43. Title: What Is a Conflict in Problem Solving? I will address the question of why people have difficulty in problem-solving, arguing that in most cases, a conflict between incompatible ideas may be evoked. The conjecture is based on the analysis of familiar-schemes brought to bear in the problem. I will show that the relation between these schemes may generate incompatible representations of the same situation. Conflicts result from difficulty in reducing these incompatibilities. ------------------------------ Date: Thu 3 Oct 85 07:50:26-PDT From: Ana Haunga Subject: Seminar - Introspection (SU) SIGLUNCH, Friday, October 4, Chemistry Gazebo, 12:05-1:00. Introspection Michael R. Genesereth Logic Group Knowledge Systems Laboratory Stanford University Abstract: Introspection is a significant part of human mental activity. We introspect whenever we think about how to solve problem, whenever we decide what information we need to solve a problem, whenever we decide that a problem is unsolvable. By its nature, the process of introspection involves both descriptive and prescriptive metaknowledge. Over the past years, logicians and AI researchers have devoted considerable attention to autoepistemic sentences (involving terms like KNOW). By comparison, little attention has been paid to prescriptive metaknowledge (involving terms like OUGHT). This talk introduces a semantics for such knowledge in the form of constraints on the process of problem solving. It demonstrates the computational advantages of introspection, and analyzes the computational fidelity and cost of various introspective architectures. Finally, it discusses the potential for practical application in logic programming and building expert systems. ------------------------------ Date: 3 Oct 1985 09:08-CST From: leff%smu.csnet@CSNET-RELAY.ARPA Subject: Seminar - Compact Lisp Machine (SMU) Speaker: Lawerence E. Gene Matthews Associate Director of the Computer Science Laboratory Texas Instruments, Dallas Date: Wednesday, October 16, 1985 Time: 11:30 AM Luncheon 12:15 Program Place: Richardson Hilton, SW corner of N. Central Expressway & Campbell The Compact LISP Machine (CLM) development program is the first of several DARPA programs intended to provide embedded symbolic computing capbilities for government applications. As one of many contacts funded under the Strategic Computing Program the CLM will provide a ruggedized symbolic computer capability for insertion of AI and robotics technology in awide range of applications. A description of the four-module CLM system architecture is presented. Starting with the CLM development goals, a brief system overview and discussion of advanced software development and maintenance tools are covered. System and module packaging are described including options available beyond the scope of the current contract. Each of the four modules under development are described starting with the CPU module, which contains the 40 Mhz VLSI CPU chip, and its companion map/cache module. The module developed for providing physical memory is described followed by a discussion of the Multibus I/O module which supports communication between the high performance system bus and Multibus I. The VLSI LISP processor chip is next described with a simplified block diagram and packaging information. Finally, some information on preliminary predicted performance is covered. Luncheon reservations 995-4440 Monday October 14 $7.00 ------------------------------ Date: Wed, 11 Sep 85 13:43:21 EDT From: Chidanand Apte Subject: Seminar Series - IBM Yorktown Projects (Rutgers) [Forwarded from the Rutgers BBoard by Laws@SRI-AI.] IBM talks at Rutgers-IBM AI exchange seminar, 10th Oct., Hill Center. Members of IBM Research from the T.J. Watson Research Center will be presenting six talks at the 3rd annual Rutgers-IBM AI exchange seminar, on 10th October 1985, at the Rutgers Computer Science dept. Preliminary titles and abstracts: "A representation for complex physical domains" Sanjaya Addanki We are exploring a system, called PROMPT, that will be capable of reasoning from first principles and high level knowledge in complex, physical domains. Such problem-solving calls for a representation that will support the different analyses techniques required (e.g. differential, asymptotic, perturbation etc.). Efficiency considerations require that the representation also support heuristic control of reasoning techniques. This talk lays the ground work for our effort by briefly describing the ontology and the representation scheme of PROMPT. Our ontology allows reasoning about multiple pasts and different happenings in the same space-time. The ontology provides important distinctions between materials, objects, bulk and distributed abstractions among physical entities. We organize world knowledge into "prototypes" that are used to focus the reasoning process. Problem-solving involves reasoning with and modifying prototypes. "YES/FAME/IDV: An initial approach to a planning consultant for financial marketing problems" Chidanand Apte/ Jim Griesmer/ John Kastner/ Yoshio Tozawa The YES/FAME (Yorktown Expert Systems for Financial and Marketing Expertise) project is investigating interactive consultants to aid in the financial marketing of computing technology. Significant expertise seems to be required in the preparation of a recommendation to a customer of a technical solution that meets his computing requirements over a period of time, coupled with a plan for acquiring this technology under financial terms and conditions that best suit the customer's needs and concerns. Expertise is also required in generating a convincing financial argument that will enable the "selling" of this plan. We present in this talk an overview of an initial demonstration version (YES/FAME/IDV) of a knowledge based system that illustrates these capabilities for a small subset of the overall problem. "Logical extensions of logic programming based on intuitionistic logic" Seif Haridi Logic Programming is widely known as programming using Horn clauses. We extend this paradigm to handle more general relations than Horn clauses. Based on principles from first order intuitionistic (constructive) logic we show a much more expressive language with a complete execution mechanism that is able to handle general first order queries, iterative and recursive statements, and positive and negative queries with equal strength. The language has Horn clauses as a subset, and its interpreter behaves as efficient as a Horn clause (PROLOG) interpreter on that subset. "PLNLP: The programming language for natural language processing" George Heidorn This talk describes research being done at Yorktown to provide advanced tools for building knowledge-based systems that involve a large amount of natural language processing. PLNLP is a programming language based on the augmented phrase structure grammar formalism that is particularly well-suited for specifying the processing of natural language text. A large, broad-coverage English grammar has been written in PLNLP, and implementations in LISP/VM and PL.8 are currently being used in applications doing text-critiquing, machine translation, and speech synthesis. One of these systems, CRITIQUE, will be used as a concrete illustration of the power of the language. "YSCOPE: A shell with domain knowledge for solving computer performance problems" Joseph Hellerstein Solving computer-performance problems requires two types of knowledge: knowledge of the computer system and insights from queueing theory. We describe the Yorktown Shell for Computer-Performance Experts (YSCOPE) which is a special-purpose shell that incorporates a knowledge of queueing theory to facilitate building computer-performance expert systems. "Interactive classification in knowledge representations" Eric Mays A classifier for a structured representation language allows semi-automatic maintenance of a knowledge base. However, problems, such as detecting and recovering from inconsistencies, arise when editing a KB which has been updated by classifier operations. This talk will address preliminary investigations along these lines. ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Wed Oct 9 07:32:42 1985 Date: Wed, 9 Oct 85 07:32:37 edt From: comsat@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a008668; 9 Oct 85 2:57 EDT Date: Tue 8 Oct 1985 23:09-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #137 To: AIList@SRI-AI Received: from rand-relay by vpi; Wed, 9 Oct 85 07:25 EST AIList Digest Wednesday, 9 Oct 1985 Volume 3 : Issue 137 Today's Topics: Queries - AI Machines & Formal Semantics of Object-Oriented Languages & Knowledge Representation, AI Tools - Lisp for Macintosh & TIMM Opinion - Social Responsibility, Expert Systems - Aeronautical Application, Games - Hitech Chess Performance, Bindings - Information Science Program at NSF ---------------------------------------------------------------------- Date: Sun, 06 Oct 85 15:20:34 EDT From: "Srinivasan Krishnamurthy" <1438@NJIT-EIES.MAILNET> Subject: Do I really need a AI Machine? Dear readers, I work at COMSAT LABS, Maryland. We are getting into AI in a big way and would like comments, suggestions and answers to the following questions,to justify a big investment on a AI machine. * What are the specific gains of using a AI Machine (like symbolics) over developing AI products/packages on general purpose machines like - VAX-11/750-UNIX(4.2BSD), 68000/UNIX etc. * How will an environment without AI Machines effect a major development effort. * How is the HW Architecture of Symbolics, TI Explorer different from VAX. * What are the limitations/constraints of general purpose HW when used for AI applications. * Survey results of AI HW Machines. (if available) * Pointers to other relevant issues. Please message me at: Mailnet address:Vasu@NJIT-EIES.MAILNET Arpanet address: Vasu%NJIT-EIES.MAILNET@MIT-MULTICS.ARPA My sincere thanks in advance. ..........Vasu. ------------------------------ Date: Mon, 7 Oct 85 11:06:40 edt From: "Dennis R. Bahler" Subject: request: formal semantics of OOL's Does anyone have pointers to work done on formal specification and/or formal semantic definition of object-oriented languages or systems such as Smalltalk-80? Dennis Bahler Usenet: ...cbosgd!uvacs!drb Dept. of Computer Science CSnet: drb@virginia Thornton Hall ARPA: drb.virginia@csnet-relay University of Virginia Charlottesville, VA 22903 ------------------------------ Date: Mon 7 Oct 85 17:27:03-PDT From: MOHAN@USC-ECLC.ARPA Subject: knowledge representation Hi, I am looking for a short list of introductory and survey type articles on Knowledge Representation. I am also looking for any work done on representing visual scenes so that a system could reason about them and answers queries etc. Minsky, M.:A Framework for Representing Knowledge, in Winston,P.H. (ed.) "The Psychology of Computer Vision" is a typical paper on the type of work I am interested. Since I have just started reading on this subject, I am presently interested in all related topics (except the CAD/CAM type of reasoning). What I am looking for is an introduction to this big field: papers which have presented the key ideas and some surveys which can give me an idea of the type of work being done in such areas. Thanks, Rakesh Mohan mohan@eclc.arpa [The October 1983 issue of IEEE Computer was a special issue on knowledge representation, as was the February 1980 issue of the SIGART Newsletter. Technical Report TR-1275 of the University of Maryland Department of Computer Science (issued May 1983) is the proceedings of an informal workshop on "Representation and Processing of Spatial Knowledge". There were also several papers on image-to-database matching in the April 9-11, 1985, SPIE Arlington, VA, conference on Applications of AI (SPIE volume 548). Vision researchers generally seem happy with networks of frames, although other representations are in use (e.g., connectionist coarse coding, logic clauses). I have sent a fairly extensive list of vision citations to Rakesh and to Vision-List@AIDS-UNIX. -- KIL ] ------------------------------ Date: Mon, 7 Oct 85 12:02 EDT From: Carole D Hafner Subject: Lisp for Macintosh There is a new magazine out called MacUser. The premier issue, which is available at newsstands, contains a review of ExperLisp for the Macintosh. The review says it's good but buggy. There is also a version of Xlisp by David Betz available through the public domain software networks. I got a copy from the Boston Computer Society, and when I tried to open it on my 128K Mac, the computer crashed (i.e., the screen went crazy and strange noises occurred). Perhaps Xlisp only runs on the 512K Mac, or else I got a bad version. Carole Hafner hafner@northeastern ------------------------------ Date: 8 Oct 85 12:22 EDT From: Dave.Touretzky@A.CS.CMU.EDU Subject: TIMM There's been some talk about AI hype in this digest, but not too many folks have stood up and pointed to actual examples. Cowan's inquiry about TIMM affords an excellent opportunity, so here goes. As far as I can tell, TIMM is the most colossal ripoff in the expert systems business. I got a demo last year at AAAI-84 from Dr. Wanda Rappaport, who I believe is one of the developers. Basically, TIMM works by comparing the facts of a situation with a set of stored templates, called training instances. The underlying data structure looks like a frame, i.e. it has named slots. Each training instance consists of a frame with some or all of the slots filled in. After you've given TIMM enough training examples, you tell it to "generalize", which causes it to do some computation on the training set to extract regularities and relationships between slot values. Then, to have TIMM solve a problem, you give it a frame with some of the slots filled in, and it fills in the rest of the slots. TIMM is a mere toy, in my opinion, because it doesn't provide adequate facilities for expressing knowledge either procedurally or declaratively. You can't write explicit IF/THEN rules, as in OPS5 or EMYCIN. You can't build data structures, i.e. it's not a real frame system with inheritance and demons and Lisp data objects that you create and pass around and do computation on. Nor can you write logical axioms and feed them to a general purpose inference engine, like Prolog's resolution algorithm. Many, perhaps most kinds of knowledge can not be conveniently expressed as training instances, but that's all you get with TIMM. At IJCAI-85 I watched a General Research sales person trying to sell TIMM to a couple of AI novices. She was repeating the usual set of General Research outlandish claims, viz. that TIMM is good for ANY expert system application you can think of, that it doesn't require any expertise in knowledge engineering to create "expert systems" with TIMM, that domain experts can sit down and create their own non-trivial systems without assistance, and so on. (See their ad on page 38 of the Fall '85 issue of AI Magazine: "Experts in virtually any field can build systems with TIMM, and do it without assistance from computer or AI specialsts.") I find General Research Corp.'s arrogance simply galling. How would they use TIMM to do VLSI circuit design, as TALIB does, to configure a Vax, like R1 does, or to generate and sift through a large set of plausible analyses of mass spectrogram data, as Dendral does? All of these tasks require significant amounts of computation, yet all TIMM can represent is training instances. Finally I asked the sales person how TIMM would represent the following simple piece of domain knowledge: "A certain disease has twenty manifestations. If a patient has at least four of these, we should conclude that he has the disease." This knowledge can be expressed by a single rule in OPS5, but can't be represented at all in TIMM. First the sales person was going to put in one training instance for every possible case, but C(20,4) is greater than 5000, so that's impractical. Finally she decided she'd write a Fortran program to ask the user if the patient had each of the 20 manifestations, sum up the "yes" answers, and return the conclusion to TIMM. (TIMM is written in Fortran and has the ability to call external Fortran routines.) The point she seemed to miss is that any nontrivial expert reasoner is going to need data structures and computations that can't be expressed as a small set of training instances. TIMM costs roughly $47K for the Vax version, and roughly $9K for the IBM PC version. General Research boasts that TIMM is in use in several Fortune 500 companies, but I haven't heard any claims about successful, up and running, NONTRIVIAL applications. Not surprising. -- Dave Touretzky PS: Due to the nature of the above comments, I feel compelled to include the usual disclaimer that the above opinions are solely my own, and may not reflect the official opinions of Carnegie-Mellon University or AAAI. ------------------------------ Date: Fri, 4 Oct 85 09:32:15 GMT From: gcj%qmc-ori.uucp@ucl-cs.arpa Subject: An Application of Expert Systems An application of expert systems that came to mind would be to auto-pilot systems for commercial aircraft. The recent crash at Manchester airport might not have been so serious if the reverse thrust had not been applied. It had been suggested that this resulted in a spray of aviation fuel over the fuselage of the aircraft. Whether this can be avoided in future by a change in design or if a real-time expert-system would have been any better than the pilot's decision, which of course was ``correct'', is a matter for the deeper examination. It seems to me that information about possible outcomes of such an action could be made available to the pilot. Gordon Joly gcj%qmc-ori@ucl-cs.arpa ------------------------------ Date: Fri, 4 Oct 85 11:03:53 cdt From: ihnp4!gargoyle!toby@UCB-VAX.Berkeley.EDU (Toby Harness) Subject: Social Responsibility Re: AIList Digest V3 #132 Something I saved off usenet, about a year ago: It's sad that computers, which have so much potential, have so much of it invested in the purposes of the authorities. I wonder if some day we'll be looking back at what we did in the 1980's the way many atomic physicists ended up remembering the 1930's. Jim Aspnes (asp%mit-oz@mit-mc) Toby Harness Ogburn/Stouffer Center, University of Chicago ...ihnp4!gargoyle!toby ------------------------------ Date: 6 October 1985 2023-EDT From: Hans Berliner@A.CS.CMU.EDU Subject: Computer chess: hitech [Forwarded from the CMU bboard by Laws@SRI-AI.] Hitech won its first tournament, and one with 4 masters in it. It scored 3 1/2 - 1/2 to tie for first in the Gateway Open held at the Pittsburgh Chess Club this week-end. However, on tie break we were awarded first place. En route to this triumph, Hitech beat two masters and tied with a third. It also despatched a lesser player in a brilliancy worthy of any collection of games. One of the games that it won from a master was an absolute beauty of positional and tactical skill. It just outplayed him from a to z. The other two games were nothing to write home about, but it managed to score the necessary points. I believe this is the first time a computer has won a tournament with more than one master in it. We will have a show and tell early this week. ------------------------------ Date: Tue 1 Oct 85 13:50:21-CDT From: ICS.DEKEN@R20.UTEXAS.EDU Subject: Information Science Program at NSF - Staffing Changes Beth Adelson has been appointed to the position of Associate Program Director, Information Science Program, effective August 15, 1985. Dr Adelson has been at Yale University since 1983, as a Research Associate in the Artificial Intelligence Laboratory. She holds a Ph.D. from Harvard University. Dr. Adelson has published numerous articles in the areas of cognitive science and artificial intelligence. Recent works include papers on software design <> and the acquisition of categories for problem solving <>. Joseph Deken has been appointed to the position of Program Director, Information Science Program, effective September 3, 1985. Dr. Deken was most recently Associate Professor at the University of Texas at Austin, with a joint appointment in the Department of Business and the Department of Computer Sciences, and taught from 1976 to 1980 at Princeton University. His Ph.D. in mathematical statistics is from Stanford University. Dr. Deken is the author of several books on computing, the most recent of which is <>, which will be Published by Bantam books in January 1986. His other writing includes <> (Stewart, Tabori, and Chang, 1983), <> (William Morrow, 1981), and numerous articles on statistics and statistical computing. Program announcements and other information about the Information Science and Technology programs at NSF are available from: Division of Information Science and Technology National Science Foundation 1800 G St. NW Washington, D.C. 20550 Correspondence may be addressed to the attention of Dr. Adelson or Dr. Deken as appropriate. ------------------------------ End of AIList Digest ******************** From csvpi@vpics1 Wed Oct 9 07:36:39 1985 Date: Wed, 9 Oct 85 07:36:30 edt From: csvpi@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a009356; 9 Oct 85 4:39 EDT Date: Tue 8 Oct 1985 23:29-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #138 To: AIList@SRI-AI Received: from rand-relay by vpi; Wed, 9 Oct 85 07:27 EST AIList Digest Wednesday, 9 Oct 1985 Volume 3 : Issue 138 Today's Topics: Update - G. Spencer-Brown Seminar, Seminars - Robot Legged Locomotion (GMR) & What is a Plan? (SRI) & Animating Human Figures (UPenn) & Inheritance and Data Models (UPenn), Conferences - 4th Int. Conf. on Entity-Relationship (ER) Approach & Symposium on Logic in Computer Science ---------------------------------------------------------------------- Date: Mon, 7 Oct 85 11:17:52 PDT From: Charlie Crummer Subject: G. Spencer-Brown Seminar Was this some kind if joke? I could not find any company called UNI-OPS nor a Walter Zintz at (415)945-0048. The Miyako Hotel has no seminar on The Laws of Form, only a management association meeting. Has someone erased the distinction between G. S.-B.'s existence and non-existence? --Charlie From: william@aids-unix (william bricken) No hoax, although a possible unfortunate typo: the phone number of UNI-OPS is (415)945-0448. The seminar was cancelled at the last minute by SB himself (according to Zintz). Totally in character. Thus the existential dilemma. Zintz is working on re-establishing it, and is compiling a mailing list of those interested in the Laws of Form. I have developed an automated theorem prover using SB's stuff, and am encouraged by its applications to LISP program representation, optimization, and verification. William Bricken Advanced Information & Decision Systems 201 San Antonio Circle, #286 Mountain View, CA 94040 (415) 941-3912 ------------------------------ Date: Thu, 3 Oct 85 10:00 EST From: "S. Holland" Subject: Seminar - Robot Legged Locomotion (GMR) General Motors Research Laboratories Warren, Michigan ROBOTS THAT RUN BALANCE AND DYNAMICS IN LEGGED LOCOMOTION Dr. Marc H. Raibert Carnegie-Mellon University Pittsburgh, PA Monday, October 21, 1985 Balance and dynamics are key ingredients in legged locomotion. To study active balance and dynamics we have built a series of machines that balance themselves as they run. Initial experiments focused on machines that hopped on one leg, but later work generalized the approach for two- and four-legged machines. A very simple set of algorithms provides control for hopping on one leg, running on two legs like a human, and trotting on four legs. We have begun to use these results from legged machines to improve understanding of running in animals. Marc Raibert received a B.S.E.E. from Northeastern University in 1973, and a PhD from the Massachusetts Institute of Technology in 1977. Since 1980 Professor Raibert has been on the faculty of Carnegie-Mellon University, where he is an Associate Professor of Computer Science and a member of the Robotics Institute. He is currently exploring the principles of legged locomotion. -Steve Holland ------------------------------ Date: Tue 8 Oct 85 13:57:58-PDT From: LANSKY@SRI-AI.ARPA Subject: Seminar - What is a Plan? (SRI) WHAT IS A PLAN? Lucy Suchman Intelligent Systems Lab, Xerox PARC 11:00 AM, WEDNESDAY, October 9 SRI International, Building E, Room EJ228 (new conference room) Researchers in AI have equivocated between using the term "plan" to refer to efficient representations of action, and to the actual data and control structures that produce behavior. But while these two uses of the term have been conflated, they have significantly different methodological implications. On the first use, the study of plans, as internal representations of actions and situations, is an important companion to the study of situated actions, but essentially derivative. On the second use, plans as the actual mechanisms that produce behavior are foundational to a theory of situated actions. In this talk I will argue in support of the first use of "plans," to refer simply to efficient representations of actions. Situated actions, on this view, are the phenomena to be modelled, whereas the function of plans in the generation of situated actions is taken to be an open question. The interesting problem for a theory of situated action is to find the mechanisms that bring efficient representations and particular environments into productive interaction. The assumption in classical planning research has been that this process consists in filling in the details of the plan to some operational level. In contrast with this assumption, I will present evidence in support of the view that situated action turns on local interactions between the actor and contingencies of his or her environment that, while they are made accountable to a plan, remain essentially outside of the plan's scope. ------------------------------ Date: Tue, 8 Oct 85 12:12 EDT From: Tim Finin Subject: Seminar - Animating Human Figures (UPenn) ANIMATING HUMAN FIGURES IN A TASK-ORIENTED ENVIRONMENT: AN EVOLVING CONFLUENCE OF COMPUTER GRAPHICS AND ARTIFICIAL INTELLIGENCE RESEARCH Norman I. Badler Computer and Information Science University of Pennsylvania Tuesday, October 8, 1985 Room 216 Moore A system called TEMPUS is outlined which is intended to graphically simulate the activities of several simulated human agents in a three-dimensional environment. TEMPUS is a task simulation facility for the design and evaluation of complex working environments. The primary components of the TEMPUS system include human body specification by size or statistical population, 3-D environment design, a human movement simulator and task animator, a user-friendly interactive system, real-time motion playback, and full 3-D color graphics of bodies, environments, and task animations. Research efforts in human dynamics control and natural language specification of movements will also be described. Recent efforts to link computer graphics and artificial intelligence will be discussed, especially as they relate to future plans of NASA and the Space Station. ------------------------------ Date: Tue, 8 Oct 85 12:12 EDT From: Tim Finin Subject: Seminar - Inheritance and Data Models (UPenn) INHERITANCE, DATA MODELS AND DATA TYPES Peter Buneman Computer and Information Science University of Pennsylvania Thursday, October 10, 1985, 216 Moore The notion of type inheritance (subsumption, ISA hierarchies) has long been recognised as central to the development of programming languages, databases and semantic networks. Recent work on the semantics of programming languages has shown that inheritance can be cleanly combined with functional programming and can itself serve as a model for computation. Using a definition of partial functions that are well behaved with respect to inheritance, I have been investigating a new characterization of the relational and functional data models. In particular, I want to show the connections of relational database theory with type inheritance and show how both the relational and functional data models may be better integrated with typed programming languages. ------------------------------ Date: Wed, 2 Oct 85 17:30:00 cdt From: Peter Chen Subject: 4th Int. Conf. on Entity-Relationship (ER) Approach Title of the Conference: 4th International Conference on Entity-Relationship (ER) Approach (See advertisement in Communications of the ACM, Sept. 1985 or the IEEE Computer Magazine, Sept. 1985) Major Theme: The use of entity/relationship concept in knowledge representation Sponsor: IEEE Computer Society Date: October 28-30, 1985 Location: Hyatt Regency Hotel at O'Hare airport, Chicago (312) 696-1234, $74 Single, $84 Double Keynote Address: Roger Schank, Yale Invited Addresses: Donald Walker, Bell Comm. Research Eugene Lowenthal, MCC Tutorial Sessions (on the first day -- Monday): 1. ER Modeling: A tool for analysis 2. AI and Expert systems 3. The Analyst's Round Table 4. Database Design Paper Sessions (on the next two days): Knowledge representation, database design methods, Query and manipulation languages, Entity-Relationship analysis, expert systems, modeling techniques, integrity theory, etc. Panel Sessions: 1. Mapping Specifications to Formalisms: Leader: John Sowa, IBM Panelist: Sharon S. Salveter, Boston Univ. Roger Schank, Yale Peter Freeman, UC-Irvine Peter Chen, Louisiana State Univ. 2. Knowledge engineering and Its Implications Leader: Ross Overbeek, Argonne National Lab. Panelists: Amil Nigan, IBM Earl Sacerdoti, Tecknowledge 3. Microcomputer DBMS Derby Leader: Rod Zimmerman, Standard Oil of California 4. Practical Applications of ER Approach Leader: Martin Modell, Merrill Lynch Panelists: Suresh Gadgil, " " Tom Meurer, ETA International Harold Piskiel, Goldman Sachs Elizabeth White For more information, contact the registration chairperson: Prof. Kathi Davis Computer Science Department Northern Illinois University Dekalb, IL 60115 (815) 753-0378 ------------------------------ Date: Mon, 7 Oct 1985 21:12 EDT From: MEYER@MIT-XX.ARPA Subject: Symposium on Logic in Computer Science Announcement and Call for Papers Symposium on Logic in Computer Science Cambridge, Massachusetts, June 16-18, 1986 THE CONFERENCE will cover a wide range of theoretical and practical issues in Computer Science that relate to logic in a broad sense, including algebraic and topological approaches. To date, many of these areas have been dealt with in separate conferences and workshops. It is the hope of the Organizing Committee that bringing them together will help stimulate further research. Some suggested, although not exclusive topics of interest are: abstract data types, computer theorem proving and verification, concurrency, constructive proofs as programs, data base theory, foundations of logic programming, logic-based programming languages, logic in complexity theory, logics of programs, knowledge and belief, semantics of programs, software specifications, type theory, etc. Organizing Committee J. Barwise E. Engeler A. Meyer W. Bledsoe J. Goguen R. Parikh A. Chandra (Chair) D. Kozen G. Plotkin E. Dijkstra Z. Manna D. Scott The conference is sponsored by the IEEE Computer Society, Technical Committee on Mathematical Foundations of Computing, and in cooperation with ACM SIGACT and ASL (request pending). PAPER SUBMISSION. Authors should send 17 copies of a detailed abstract (not a full paper) by Dec. 23, 1985 to the program committee chairman: Albert R. Meyer - LICS Program MIT Lab. for Computer Science 545 Technology Square, NE43-315 Cambridge, MA 02139. (617) 253-6024, ARPANET: MEYER at MIT-XX The abstract must provide sufficient detail to allow the program committee to assess the merits of the paper and should include appropriate references and comparisons with related work. The abstract should be at most ten double-spaced typed pages. The time between the paper due date and the program committee meeting is short, so late papers run a high risk of not being considered. In circumstances where adequate reproduction facilities are not available to the author, a single copy of the abstract will be accepted. The program committee consists of R. Boyer, W. Damm, S. German, D. Gries, M. Hennessy, G. Huet, D. Kozen, A. Meyer, J. Mitchell, R. Parikh, J. Reynolds, J. Robinson, D. Scott, M. Vardi, and R. Waldinger. Authors will be notified of acceptance or rejection by Jan. 24, 1986. Copies of accepted papers, typed on special forms for inclusion in the symposium proceedings, will be due March 31, 1986. The general chairman is A. K. Chandra, IBM Thomas J. Watson Research Center, P.O. Box 218, Yorktown Heights, NY 10598, tele: (914) 945-1752, CSNET: ASHOK.YKTVMV at IBM. The local arrangements chairman is A. J. Kfoury, Dept. of Computer Science, Boston University, Boston, MA 02215, tele: (617) 353-8911, CSNET: KFOURY at BOSTONU. ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Thu Oct 10 00:17:46 1985 Date: Thu, 10 Oct 85 00:17:41 edt From: comsat@vpics1.VPI To: france@opus (FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: R Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a001946; 9 Oct 85 14:35 EDT Date: Wed 9 Oct 1985 10:09-PDT Reply-to: AIList@SRI-AI US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #139 To: AIList@SRI-AI Received: from rand-relay by vpi; Thu, 10 Oct 85 00:01 EST AIList Digest Wednesday, 9 Oct 1985 Volume 3 : Issue 139 Today's Topics: Corrections - AI at GE & Attenber's Research Field, Opinion - AI Hype ---------------------------------------------------------------------- Date: 7 Oct 85 11:29 EDT From: WAnderson.wbst@Xerox.ARPA Subject: Correction: Intellectual Honesty & the SDI In a message posted in AIList Digest on Friday, 20 Sep 1985, Volume 3, Number 125, I commented about statements made in two papers I picked up at the GE exhibit at IJCAI-85. In view of subsequent discussions with people at GE I wish to state that the authors of these papers are NOT in any way connected with GE. It was not my intent to cast aspersions on work done by GE. I wish to correct any misconceptions people may have about the type and quality of research at GE from my message. Bill Anderson People interested in GE AI R&D, may contact Dr. Larry Sweet K1-5C13 General Electric Company Corporate Research and Development Schenectady, NY 12301 Dr. Sweet is the manager of the AI group here. ------------------------------ Date: Fri, 4 Oct 85 12:24 EDT From: Attenber%ORN.MFENET@LLL-MFE.ARPA Subject: typographic error In case anyone cares, the second paragraph of my unintelligible note of Sep.25 should have started "as a researcher in plasma physics" not "as a researcher in particle physics". Sorry. ------------------------------ Date: Fri, 4 Oct 85 08:57:33 EDT From: George J. Carrette Subject: ai hype vs profitable use Observations I've been able to make (from inside and out) of the activities of LMI's process control division may be interesting here. (1) They call it PICON(tm) for Process Intelligent Control. The world "artificial" hardly ever comes up. The purpose is to more intelligently control the industrial process. (2) The first installations at Exxon and Texaco were in place for months before they even told anyone. (3) Promise of profits were never made. The marketing was as an ALARM condition detector/advisor. More subtle and possibly dangerous conditions could be detected than with existing technology, and more ALARM conditions could be handled at once, with more intelligent selection of danger priorities. The purpose is of course to avoid the 3-MILE-ISLAND-EFFECT. (4) Once in place it was the customers that realized for themselves that with the more intelligent modeling and flexibility in PICON they could optimize the control of the process more closely, and could start to save a percent or two in cost or get a percent or two higher yield. Obviously in oil refineries this can translate into big paydirt far above the cost of a few lispmachines and development costs. The "downplaying" the PICON people have had to do is of course caused by all the previous and continuing ai hype. This is probably why they dont use the term AI very often. ------------------------------ Date: 4 Oct 85 09:39 PDT From: allmer.pasa@Xerox.ARPA Subject: AI Hype Does anyone really expect truth in advertising?? Whenever I read a blurb about some new "fully compatible" DBMS I always hear that click, you know, what do they mean by "fully-compatible". So I don't see the point for ranting on about the "AI Hype". It seems to me that if there was nothing more than hype, there wouldn't be an AIList, or AIList Moderator, no one would want to learn any "AI techniques" or take any "AI courses", no one would spend millions of $$$$ to get "AI Systems", etc. There's got to be more to AI technology than the hype, or this is the greatest scam in history (Mr. Guiness, where are you?). At the Expert Systems panel for IJCAI85, Terry Winograd, (how's that for name-dropping), made mention of the "illusion of the label 'expert system'", that because we choose to call them that, they should be as-good-as/better than 'experts' when they are finished, sort of a "magical black-box" mentality. To sell the concept, one does not call their proposed system "idiot savant", or "limited-domain model", even though that is what the buyer ends up with. You just have to understand what is meant by the terminology of the "hype". Doug Allmer ------------------------------ Date: 4 Oct 1985 1203-PDT (Friday) From: jeff@isi-vaxa.ARPA (Jeffery A. Cavallaro) Subject: AI, SCI, and SDI After observing the various discussions regarding the place or state in today's SCI/SDI-hyped environment, I think that it might be wise to mention a chain of events that I believe has gone virtually unnoticed in the AI community. The opinions expressed here are based on my connections within the REAL research world (academia, not the defense contractor flops referred to as IR&D), and the defense/aerospace sector over the past 5 or so years. (Limited compared to some I admit) About 5 years ago, when DARPA was still being referred to as ARPA, DARPA's commonly stated goal was to increase America's strength by promoting raw research that would trickle into the ECONOMIC sector. That may seem like a rather empty statement based on today's activities, but that was really the way it was. (Enter Dick Cavett). ACADEMIA (5 YEARS AGO): Around the same time (5 years ago), the various VLSI research groups around the nation had their own mini-SCI, it was called the Silicon Compiler Project. It may not have gotten the same attention, because the funding stakes were probably not as high as AI today. It can easily be stated that SCP was a success. Meanwhile, the research AI groups of the day were constantly complaining that they were spinning their wheels. The results of their work were constantly being blocked from entering the marketplace for a variety of rights and economic reasons. DEFENSE/AEROSPACE (5 YEARS AGO): In defense/aerospace, VLSI technology is basically unheard of. Companies like TRW (Torrance Rubber Works), HUGHES, and the like, were more interested in off-the-shelf solutions from large vendors such as DEC, IBM, etc. These vendors, in turn, had representatives participating in SCP at various research institutions. AI was a different story. Defense contractors had AI projects (and funds) coming out of their ears. One of the largest such project was BETA-LOCE, an in-the-field battle management-type system. All such projects, without many exceptions, were unqualified FAILURES. TODAY (Exit Dick Cavett): Sitting in meetings at ISI today, where the goals of SCI/SDI are being stated, is like sitting in a status meeting at TRW as little as 2 years ago. VLSI, having been successful, has now obtained a firm place in defense/ aerospace. Due to the strong success and firm base achieved while it was in the researcher's hands, VLSI is proving to be an excellent tool now available to the real world. AI projects in defense/aerospace have hit hard times. SCI is now on hand. I believe that one of the unspoken goals of SCI (and SDI) is to seriously shift emphasis of AI projects back to the research institutions, hoping that they can achieve success similiar to VLSI. But, a funny thing is happening. The barriers to their work have been lifted, but the AI world is still complaining. They are disatisfied with the end user (as if Defense wasn't actually the end user all along). The VLSI researches didn't seem to have the same qualms. But then again, maybe they weren't faced with SDI-type goals. In conclusion (finally), the AI research community is currently in an EXCELLENT position. They now have the attention (which, in a way, is just as important as funds) that they have desired for a long time. Of course, careful consideration is needed, but continued complaining and temper tantrums of "We can't do that in the foreseeable future, so I don't even want to try!!" will simply be detrimental to ALL parties involved. (Oh, am I going to get jumped on for this one!!!) Jeff ------------------------------ Date: Fri, 4 Oct 85 13:18 EDT From: Jeffrey R Kell Subject: Hype and success/failure The 'hype' of AI in terms of systems to carry on conversations or drive cars or whatever is largely based on an idealistic projection of what we have achieved in the directions we want to follow. It began with the stereotyped image of a computer as a thinking being when in fact it was a primitive set of vacuum tubes. The AI image suggests that we can do just about anything; that we can eventually achieve that sort of goal. AI researchers realize their current successes, albeit not on such a grand scale, with a guarded skepticism. If the shortcomings of any AI project are stressed, the image sways in the other direction - maybe AI can do nothing. The relative success/failure of such projects seems to revolve around a gap between the ideal and the practical. A 'pure' AI system built with 'pure' AI processes is almost doomed to certain shortcomings whether in speed, size, usability, correctness, completeness, or what have you. Some practical (traditional) methodologies must be employed to narrow the scope of the project. Universal problem solvers were certainly not a practical success, but they did lay the groundwork for expert systems which have been successful. But the balance between AI and traditional methods varies, and disciples of 'pure' AI will argue that many expert systems are not AI at all. Perhaps not, from a purist view, but they would not have been possible without the conceptual contributions from the AI field. It is THESE forms of contributions that will be of the most lasting value, whether you view them as an idealistic success or not. ------------------------------ Date: Sat, 5 Oct 1985 23:55 EDT From: MINSKY%MIT-OZ@MIT-MC.ARPA Subject: AIList Digest V3 #132 I was not planning to prolong this discussion, but I can't resist pointing out the disgraceful lengths that Gary Martins will go to prove himself right. If you will examine my statement and then Martins', you'll [note] an astonishing performance: he takes two of my sentences, ONE OF WHICH CAREFULLY QUALIFIES THE OTHER, breaks each of them into clauses and than attacks each clause by itself! I see no room in a professional discussion for that degree of intellectual and rhetorical dishonesty. Talk about "hype!" Then he has the bad taste to talk about "utterly non-'AI' software" that keeps track of payrolls, arranges airline reservations, manages power distribution grids, guides missiles, allocates resources, monitors inventories, analyzes radar signals, does computer animation, assists in mechanical design and fabrication, manipulates spreadsheets, controls space vehicles, drives robots, integrates CAT scans, and performs lots of other mundane tasks. and about nicely engineered non-"AI" systems that play world-class chess. The latter "non-AI" chess programs are, of course, essentially the AI chess programs of the 1960's, based on Shannon. Samuel, and McCarthy's tree-pruning heuristics and plausible move generators. The robot drivers are mostly based on the early MIT, SRI, and Stanford prototypes. Many of the aircraft control systems are based on the adaptive algorithms developed in that general community in the same period, and everyone knows the origins of much of computer graphics in the early work of Sutherland, Knowlton, and many others in the AI community. As for those missile guiders, the roots of that whole field now called "pattern recognition" have similar origins. And I'm pretty sure that the first practical airline reservation was designed by Danny Bobrow of the BBN AI group around 1966.! I'm not claiming that AI set the stage for accounting programs, and some of the others. But don't you agree that Martins could have made himself a better case by mentioning a few first-rate programs that didn't have substantial roots in the AI of 15 to 20 years ago. If there's anything wrong with the present-day AI hype, it's simply that some people may be led to expect various goodies in 3 years instead of 15 -- and perhaps that's what we ought to tell people. Time to take a course in the history of AI, Gary. ------------------------------ Date: Mon, 7 Oct 85 15:54:22 mdt From: ted%nmsu.csnet@CSNET-RELAY.ARPA Subject: ai/military flame When did the military manage to take credit for automobiles???? Or for that matter how do they manage to take credit for debugging and field testing the interstate highway system when there was a civilian highway system in place before the interstates and as far as I know, the only connection was that internal defence is a standard rationale for better internal transportation? Minsky's comments about what ai can do and other software can't are very illuminating when compared with the real world in the form of the first milestone test of arpa's autonomous land vehicle in denver this spring. one might think that this would have provided a perfect example of the way that ``ai software can ... drive a car''. unfortunately for true fans, the ai approach to driving the vehicle (line extraction, motion field analysis and so on) turned out to be very difficult (read as late) to implement so that the prime contractor (who is very capable in conventional software) implemented a VERY conventional system which selected ``gray'' pixels from the television image and managed to steer the vehicle toward the center of mass of the gray. doesn't this sound more like an example of conventional software doing something (not very well, but literally good enough for government work) that ai type software failed to do??? ------------------------------ Date: Thu, 3 Oct 85 14:09:20 GMT From: gcj%qmc-ori.uucp@ucl-cs.arpa Subject: Mega-Hype A comment from Vol 3 # 128:- ``Since AI, by definition, seeks to replicate areas of human cognitive competence...'' This should perhaps be read in the context of the general discussion which has been taking place about `hype'. But it is still slightly off the mark in my opinion. I suppose this all rests on what one means but human cognitive competence. The thought processes which make us human are far removed from the cold logic of algorithms which are the basis for *all* computer software, AI or otherwise. There is an element in all human cognitive processes which derives from the emotional part of our psyche. We reach decisions not only because we `know' that they are right, but also because we `feel' them to correct. I think really that AI must be seen as an important extension to the thinking process, as a way of augmenting an expert's scope. Gordon Joly (now gcj%qmc-ori@ucl-cs.arpa (formerly gcj%edxa@ucl-cs.arpa ------------------------------ End of AIList Digest ********************