micr Focus creators of visual programming tools for software development is pleased to provide major funding for the computer Chronicle the story of this continuing Evolution think very long about that did he no we knew that was going to happen welcome to the computer Chronicles I'm steuart chaet sitting in for Gary kildall this week is her lechner of Sri herb we think of Chess as the ultimate game of skill a game that requires mental agility intelligence if you will yet here I am playing a game of chess with a computer which is analyzing board positions and in applying a certain kind of intelligence to figure out what its next move should be that's the subject of our program today artificial intelligence and in in some people's minds AI suggests attempting to duplicate the way a human brain works is that what AI is in fact uh not in most modern AI research uh Stuart um early research in AI looked at at d duplicating human thought processes but current AI research is more concerned with duplicating the end result of intelligence and computers uh that act as experts in expert systems and computers that can communicate with us in our language understanding some of the context of that language are two areas that are receiving a lot of attention in artificial intelligence research today okay we're going to meet two of the world's foremost experts on artificial intelligence and we'll take a look at two fascinating examples of expert syst systems one of the leading experts on Expert systems is Dr Edward Fen bomb of Stanford University we asked Professor Fen bomb about the evolution of artificial intelligence the um computers as you know are General symbol processing devices capable of manipulating any kinds of symbols of which numbers constitute one important class but computers are much more General than that we've known about the generality of computation since at least the time of uh touring in the 1930s and actually I've tracked it back to um intuitions that Babbage had that were reported by Ada lovece after whom the Ada programming language is named in 1842 Ada lovece wrote that uh the analytical engine of babes constituted the the uh link between the mechanical world and the world of the most abstract Concepts that currently in the modern terminology is called the physical symbol system hypothesis and is the basis for artificial intelligence work in artificial intelligence as a science we talk about the use of computers to process symbolic knowledge using logical inference methods symbolic inference methods in in other words we're talking about uh inference and not calculation in the traditional sense we're talking about knowledge and not numbers in the traditional sense a current application of INF iial knowledge engineering is in the field of expert systems programs like this oil exploration advisor developed to assist drilling rigs in remote areas the program behaves much like a qualified specialist it asks questions of its user and then gives advice on how to avoid or correct accidents so common in drilling operations if at some point in the line of questioning the user becomes confused by the query he can ask the program to explain itself simply by entering why the adviser responds with a specific Source behind the question and explains its reasoning up to that point the symbolic processing behind the humanlike talents of this program have found applications in a broad range of fields from medicine to robotics perhaps the most difficult area for AI to master has been natural language a talent that results in very friendly computers but that requires enormous processing capacity the ambiguities of syntax and context have restricted present systems to very limited applications a simple Geographic question depending on the phrasing can lead to multiple interpretations the program will determine the question's most likely meaning only after an exhaustive deconstruction of the sentence and might even reject an unusual phrasing of the same query in a parallel development investing AI programs with linguistic ability has led to an interactive study of graphic communication in this experiment the visual dynamics of communication form the foundation for a linguistic expert adviser capable not only of discerning visual patterns but also acting as a kind of tutor one interactive application has given persons without normal speech the ability to adapt or construct alternative symbolic languages a remarkable instance of using a computer programs intelligence to help interpret human intelligence joining us now is Jeffrey Peron jeffre is a Management Consultant involved in expert systems for micros and Neils neelson director of the Artificial Intelligence Center at SRI International herb Nels could you kind of help us scope the field of artificial intelligence as it exists today well I'll try herb it's a Broadfield if you ask uh many of different people people what constitutes artificial intelligence I think you'd get a lot of different answers uh for me I think it concerns mainly the uh putting of knowledge into computer programs so that the programs then can solve certain problems which humans are uh find somewhat easy or perhaps intellectual challenges sometimes and the uh knowledge that one puts in is knowledge that's represented in a particular sort of way so the idea is that it's just not smarter programs that do artificial intelligence programming it is some difference in Tech techniques that they use relative to what ordinary programming well there's a part of computer science that artificial intelligence is concerned with that does in fact involve a certain body of techniques that are a little different perhaps than uh what goes on in the rest of computer science Jeffrey uh Neils is involved in the research end of AI and you're involved in some of the applications are are we at the point today where we can apply AI absolutely steuart in what ways well I feel that uh artificial intelligence and specifically expert systems or knowledge-based systems now are applicable any place that specialized knowledge is used routinely to reach decisions troubleshooting strategies diagnosis those areas so uh I I kind of think of it as something where uh we've reached a watershed where is no longer very expensive or very difficult for individuals with no technical background to build systems and apply them usefully and you distribute a system which is an expert system and o helps one build expert systems is that correct right well actually it's a tool for generating particular expert systems applications in virtually any area where there is that routine use of special knowledge and it runs on microcomputers right it runs on the IBM personal computer and a number of other compatible microcomputers do you use the same techniques in this system that Nels was referring to are used in systems that might run on larger equipment it uses some of the same ideas and it uh has its own unique uh approach as well one of the specific areas where this is a different sort of program is that instead of requiring explicit statement of rules to build a system it only requires examples of previous decisions previous uh tasks previous diagnoses to build the systems and that's going to uh break through I believe the knowledge engineering bottleneck mentioned by people like Edward Fen Bound in building these systems Jeffrey you have expert ease uh your system up here and and show us the demonstration sure okay well this is a particular application generated with this system and it was generated by an anesthesiologist Hillary da and myself and it what it does is it makes diagnoses for breathing or Airway problems the first question here that it asks the naive user is when was the onset of the problem and I'll answer days it then we'll ask what's the quality of the Strider which is a rasping sound made when one's breathing kind of a sort of sound and let's say that the quality is moderate who would use this system a doctor or this might be used by a physician's assistant to screen patients for perhaps further fine-tuning by someone with more expertise is the patient drowsy let's say no are there any predisposing factors to developing Strider prior events that might lead to this condition let's say yes and it comes up with consider the diagnosis of intrathoracic tracheal stenosis now I'm not a physician so I don't know exactly what that means but uh that gives you an idea of how it might lead to a particular diagnosis now what I'd like to do is I'd like to show you a trivial example called Sunday which actually comes with the manual for the program as a tutorial essentially now Sunday is a a model of deciding what to do on a Sunday afternoon how are you going to spend your time and uh it is consists of a couple of questions multiple choice questions like you've just seen and uh those would be answered so so now what I'm going to do is run the query system and it asks do you have your family with you let's say yes what is the weather like is it raining or sunny let's say Sunny so it says why don't you take the family to the beach now the way that you would build one of these models is by starting on something called the attribute screen the attribute screen is where you sketch out the dimensions of the particular problem or situation now here the attributes are weather and family and there's the advice column which would be the result of a particular decision what what size Matrix can you can you do on that okay you can actually do uh in any particular problem 31 attributes and up to 255 values per attribute and those can then be chained together so you can do very large models just uh limited by the size of your disk storage essentially the uh values such as raining and sunny could be thought of as answers to the question represented by the attribute once you've done a preliminary sketch of this sort you then go to something called the example screen and on the example screen you enter in examples of previous decisions now this was developed by Dr Donald Mickey in Edinburgh based on the idea that information is communicated by experts or Masters to their apprentices in the form of examples or cases so here you see horizontally running across the screen uh answers to those questions reigning yes we're with the family the advice would be go to the museum and the next case would be if it's sunny and I was with the family then I would go to the beach and the third case there consists of an I don't care whether what the weather is like if uh there's no family with me then I'm going to go fishing Jeffrey uh you described this obviously as a trivial example this Sunday one but but in general do you have the capacity within a PC to to seriously approach this kind of problem solving yes you can do quite significant things as a matter of fact I've recently been speaking with the whole earth software review people about doing some systems to recommend software such as word processing programs or Communications Hardware modems that sort of thing Neils uh I want to ask you what I hope doesn't sound like a dumb question but kind of what's the point uh why do we develop something like this kind of system is it to replace the doctor say in the diagnosis situation what what is the end result of this well I think there'd be a lot of uses of systems like this I think the present ones actually do have a lot of brittle features that perhaps might limit their utility at the present time but uh in the long run that is 10 years 20 years something like that uh these kinds of of systems I think will be quite generally useful in a wide range of uh settings first of all um the knowledge that uh these systems contain at least we hope to put in the knowledge of worldclass experts people uh who know so much about the field that there might only be five or six of them in the whole world who know that much now uh there can be some pretty good practicing people who nevertheless aren't quite as good good as the world uh experts in a particular subject so it helps us spread the knowledge of an expert uh around a good deal in a way which can be copied quite easily okay in just a moment we're going to meet the man who invented the term artificial intelligence and we'll see a demonstration of knowledge engineering that's coming up next joining us now is Tom Kaylor Tom is director of Applied artificial intelligence at intelligenetics Tom uh you've got a complex system here first of all tell me how this would be different from the system we saw in the first part of the program well one way is uh that it's different is that we can develop a graphics environment which connects to the underlying knowledge bases so that the uh user really is just uh fooling around with meters and valves and and objects that they're used to here for examp example if we look at this level ometer we can just point to it before you get to that what what's the environment we're looking at here oh this is the key system the knowledge engineering environment and we have uh knowledge bases hooked uh up to a control panel right here at the moment and it's just a little example to show you how we can hear a control panel of what uh this could be a control panel say to a nuclear reactor to a an instrument or any other U kind of operational device that you would want to work with and here for example we're just uh changing the value of a meter and notice that it's generating a strip chart showing the time course Behavior another thing we can do with these kinds of panels as we can go in and look at uh Valves and we can affect the valve whether it's open or shut now what's important about this is this is affecting an underlying knowledge base which will then apply heuristics and decision-making procedures to the system can you show us that how you can go to the the part there you go we solve a problem well what we see underneath here now is a knowledge base for a nuclear reactor and we can have a representation of the components of the nuclear reactor as well as a representation of the uh decision-making behavior that would be carried out by the expert and what what you will see here is that it's both possible to use rule-based reasoning as well as representation of the underlying knowledge one of the first things we'll do here is just take a look at how we could test the hypothesis if of a particular accident situation ation and we just point to test hypothesis and activate it here and what we see the system doing is going through a decision-making process where it's accessing the underlying knowledge base looking at the various states of the components and by the way the states of these particular components can be accessed directly into the instrument or into the nuclear reactor now you can actually interrogate the system and ask it what course of logic is it following is that correct that's right right now it's come up and asked the operator a question is this is it true that this steam generator level is is decreasing so we can say why is this being asked and it will explain to us that it's doing it in order to prove that the steam generator inventory is inadequate we will then give it an answer and it goes on to reach a conclusion okay Tom this system runs in lisp why is lisp the right language for an AI application well one of the reasons that lisp is so important is that you need to have very powerful symbol manipulating capabilities decision making is primarily a symbol olic activity knowledge representation is primarily symbolic and you need that kind of capability okay thank you herb John McCarthy has joined us John is a professor of computer science at Stanford University and one of the Pioneers in the field of artificial intelligence among his other accomplishments he invented lisp the language that Stewart just referred to John why did you invent lisp well for artificial intelligence work uh the kind of thing that they are doing here what are its characteristics that make it different than some other language uh as applied to artificial intelligence besides the ones that Tom mentioned uh one of its important characteristics is that its programs and are in the same format as its data so that it is easy to make programs that uh produce programs and uh look at programs so it allows us to deal with facts and logic uh more than than numbers that's correct I I guess it's on every one's mind when we talk about smart machines how smart can machines become what what are the limits of of artificial intelligence well I see no limit short of uh human intelligence and uh then with faster machines one could uh do the equivalent that a human could do in a short time in a very long time Neil what's your thought on that where how far can we go with you well I'd separate some of the problems that we're facing in order to make machines more uh intelligent into about three different uh varieties we've heard a lot about how important knowledge is and uh one of the important things about knowledge is what knowledge should be represented in a program and uh some of the difficulties that we're having in making programs smarter is that we're not sure exactly what it is we want to tell those programs about certain subjects another category is once we figure that out how do we represent the knowledge in the computer itself and certain kinds of knowledge proving a little difficult for us to represent it the third category has to do with how that knowledge is used and uh certainly there's um various activities there that have been rather successful but other things that that really have to be done yet is there uh a frustration level in this field in that there there's a lot of Hope of what you can do with something like AI but yet as as youve pointed out Neils it seems harder perhaps than one thought to really put this into a practical application John well uh I not there is a certain level of techn of technology in artificial intelligence today and uh many people are making applications based on this technology uh on the other hand uh something that Neils said earlier uh Rises Rouses a thought in me um he said that we can put in the knowledge of a world-class expert and that's indeed true we can put in the knowledge of a world-class expert provided it fits into the format that the present programs allow namely of the kind if this is true uh then do that but uh more General kinds of knowledge that are used in a sort of vager way are sometimes harder to put in say uh in the medical diagnosis era at the beginning of the program we had a simple example of medical diagnosis uh there were several systems developed and yet there were were problems in fact in using that what were those problems well uh I'm not sure what were the problems of those specific systems however to take a medical example one can say make a definition that a container is sterile um if all the bacteria in it are dead uh however this piece of knowledge is not used in a special way to determine whether a uh container is sterile by examining each of the bacteria it's used as part of more General facts like if you heat the container enough then the bacteria will be all killed and the fact that if the container is not sterile and you empty it out onto a culture medium then you'll get uh colonies of bacteria and the still further fact is important that that if you leave the bacteria in food uh then they will spoil it Neils where do you see the future practically going in AI what's the biggest potential Market if you will for these kinds of applications well it seems to me that expert systems as they have been going along probably represents a rather large market and we'll continue to develop uh systems that are less brittle uh over the next 5 10 15 years another uh very important application is uh computer programs that are able to converse with humans in English uh every day ordinary language and uh that's going to make computers accessible to a much wider variety of people we have just about a minute I like your phrase brittle explain more what you mean by a brittle system well it has to do with being able to reason about the context I think in which a particular discussion or uh conversation with the expert system is taking place if the information that's needed to reach a certain conclusion is there and is there explicitly then the system is usually able to come up with some reasonable answer if it has to have what you might call Common Sense knowledge knowledge that all of us learn by the time we're 10 or 15 uh then it has a good deal more difficulty and matter of fact fails in many of those cases so ironically common sense is the most difficult thing there's very few people uh who are willing to pay for putting Common Sense knowledge into a computer whereas uh it is interesting and uh commercially important to put knowledge about a nuclear reactor into a computer okay gentlemen we're out of time thanks very much for being here thank you for joining us on this edition of the computer Chronicles n microfocus creators of visual programming tools for software development is pleased to provide major funding for the computer Chronicles the story of this continuing evolution Random Access is made possible by a grant from popular Computing a McGraw Hill magazine I'm Susan bimba sitting in for steuart cha in the random access file this week the Commerce department has made a move to stop the export of technology to countries allied with the Soviet Union hitting the digital Equipment Corporation with tough new restrictions on its exporting practices now deck has to get individual export licenses for each sophisticated computer it ships to West Germany Austria or Norway government officials say in these countries computers can be easily diverted to the Eastern black countries several months ago one of deck's computers was intercepted and route to the Soviet Union but the company was found to be blameless in the incident despite a strong Lobby by the high-tech industry President Reagan has given the Pentagon power to advise the Commerce department on computer and high-tech exports to all foreign countries the Pentagon already had the right to advise about applications for exports to communist countries the Pentagon feels its input would help cut down on the number of computers that get diverted from non-communist countries to communist ones the national semiconductor Corporation could be banned from making chips for government projects early this month the company admitted that over a three-year period it inadequately tested chips that were used on many military projects now the defense logistic agency is investigating National semi trying to find out exactly who's responsible for lying to the government about the testing despite these problems National semiconductor is reporting a profit of $15.4 million for the quarter ending March 4th IBM is reporting big demand for its little computers the company says this year its shipments of the PC will be triple what it was last year it's it's estimated that IBM will ship more than 2 million of its PCS PC Juniors and other desktop IBM computers combined and now that IBM is making its own portable computer it looks as if compact a maker of IBM compatible Portables will be lowering its prices so far no date has been given for the price cut but industry insiders say it will probably be around the beginning of April right now the compact portable sells for $2,995 that's $200 more than the IBM portable with a projected $500 cut the compact will sell for 300 less than the IBM portable atarian corporated is making Cuts but it will be employee cuts the video game maker says it plans to lay off about 300 workers at the same time an Atari spokesman says there are plans to expand the sales and Engineering staffs by about 100 last year the company lost over 500 million do laid off more than 2,000 of its Bay Area workers and moov part of its operation to Taiwan and Hong Kong well last week we told you the Japanese trade Ministry May introduce a bill that would reduce protection of us software in Japan this week the US government is warning that if Japan's Parliament ratifies the bill the US will retaliate in kind if approved the bill would limit copyright protection to us-made software making it necessary for us software owners to license their programs to Japanese makers well a little closer to home our software reviewer Paul Schindler has some information for blackjack Buffs who don't have money to burn if you gamble in Nevada or Atlantic City that's a familiar site the shuffling of blackjack cards usually accompanied by the site of your money being swept in by the dealer at the end of the round well if you're tired of losing there's a computer program for you it's Ken Houston's professional Blackjack now it's pretty rare when the writer gets top billing when name a computer program but Houston deserves it he's a former official of the Pacific Stock Exchange and now he's a full-time Gambler of course some people would say that isn't much of a switch but anyway on a recent trip to Las Vegas Houston was kicked out of five casinos after winning more than $3,000 now most computer blackjack games will just play blackjack with you but the whole earth software catalog and review says Houston's professional Blackjack will teach you how to play the game and win using various Point counting methods the graphics are crisp the table is green the cards realistic looking and the sound is good too you can hear the cards hitting the table another use of sound the program advises you on bedding and card play and it beeps at you when you goof Ken Houston's professional Blackjack is $70 from screenplay in Chapel Hill North Carolina for Random Access I'm Paul Schindler next week Paul reviews Volks Rider Deluxe and that's it for this week's Random Access I'm Susan Bima Random Access is made possible by a grant from popular Computing a McGraw Hill magazine