SAGA Food Menu Cycie 2 Mon., Oct. 6th Tues., Oct. 7th Wed., Oct. 8th Thurs., Oct. 9th Fri., Oct. 10th Sat., Oct. 11th Sun. Oct. 12th Lunch: Supper: Lunch: Supper: Lunch: Supper: Lunch: Supper: Lunch: Supper: Lunch: Supper: Lunch: Supper: sical: Sandwich — 3.50 Hot Dog bar — Sm. .90, Lg. 1.50 Baked Ham — 3.20 Breast of Chicken On Bun — 3.50 Cheddar Ground Beef & Gravy — 3.20 Turkey Fried Rice — 2.30 Egg Rolls — .60¢ ea. Swiss Steak — 3.50 Spaghetti & Meat Sauce — 3.00 Hot Beef Sandwich — 2.75 B.L.T. Sandwich — 2.55 Pork Chops — 3.25 Turkey Pot Pie — 3.00 Old Fashioned Ground Beef Pie — 2.80 Grilled Ham & Cheese — 2.95 Roast Beef — 3.20 Thanksgiving Buffet — 6.00 Battered Fish — 2.50 Beef Turnover — 2.60 Roast Turkey — 3.20 Beef Chop Suey — 3.10 Grilled Reuben — 3.00 French Toast — 2.10 Roast Pork — 3.20 Turkey Fried Rice — 2.30 Eggs Rolls — .60¢ ea. Monte Christo Sandwich — 3.00 Pancake Bar — 2.10 Seafood Platter — 3.60 Spanish Macaroni — 2.60 *NOTE: MENU SUBJECT TO CHANGE WITHOUT NOTICE Prices include Vegetable & Potato. Subject to Tax. NOW DELIVERING DELIVERY STARTING AT 8:00 PM for only $1.00 Third Floor S.u.b. Barn By Jim Lai Artificial intelligence (also known as Al) defies a simple definition. One could say that Alis the study of how to make computers do things that people are better at, at the pre- sent moment. There are several avenues of research, all of which are trying to reproduce in computers some of the pro- cesses we associate with human intelligence. In the early days of the field, around the early 60’s, the first problems to be studied in- cluded the playing of checkers and the proving of simpler mathematical theorems. How- ever, only quite simple tasks were selected. No efforts were made to create programs with a large amount of knowledge about a particular problem. As AI research progressed, tasks requiring the handling of larger amounts of knowledge could be attempted. Such tasks include vision and speech perception, natural language understanding, and problem solving and in specialized fields, like chemical analysis and medical diagnosis. Other problems that fall within the scope of AI are general problem solving and expert problem sol ving. Expert problem solving is concerned with many fields, such as symbolic mathematics and engineering design. Progress is slower than what experts had predicted at first. One of the few solid results to come out of the first two de- cades of research is that arti- ficial intelligence requires a large knowledge base. An AI system must contain a lot of knowledge if it is to deal with problems other than trivial ones. As an AI system grows larger, it becomes harder to find the appropriate things when called for, so more know- ledge is required to manage the information. This itself adds further to the size of the AI system. Efforts to build programs that perform tasks the way people do can be divided into two classes. One class involves attempts to solve problems like a person would — pro- blems that a computer could otherwise do easily. The other class of programs attempt to model human performances by doing things which are not trivial for a computer. Reasons to model human performances are to test psycho- logical theories, to allow com- puters to understand human reasoning, and to enable people to understand computer “‘rea- soning”’. Success in AI research is hard to determine. One general goal held by some American researchers, influenced greatly by psychologists, is to make programs that behave as humans would, perhaps even to the point of making mistakes. More practical researchers are working on programs that aren’t supposed to make mis- takes. American researchers are working on various pro- jects such as onboard com- puter systems to help pilot air- craft and intelligent land ve- hicles. Many business-related applications have also been found. Research in AI has spurred the development of computer work stations, which can be used in research and develop- ment of a multitude of pro- ducts, from cars to computer software, as well as for many other tasks. Japan has been working on Fifth Generation computers since 1982. Their goal is to achieve by 1992 a computer capable of from 100 million to 1 billion logical inferences per second, by coupling 1,000 or sO processors to give parallel processing. In their project they have progressed in many Al-related fields, including language-translation, robotics, and problem solving. Twenty-five years ago, the prevailing notion in the field of AI was that with more computing speed and more memory, any problem could be solved. As this approach became less promising, re- searchers began to concentrate more on understanding the nature of human thinking and problem-solving processes. Today, researchers have narrowed their goals even more and are focusing on more pragmatic facets of AI: teaching a computer to recog- nize objects, formulation of theorems in a specific field, and the construction of know- ledge-based expert systems. There are many optimistic people in the AI community, but some experts are pessimistic and skeptical. To realize mean- ingful goals, it may be neces- sary to utilize many different approaches and methods to resolve the many difficulties encountered in AI research. = Paseo October 2, 1986 = IIIIlISSSSSSES=ananaE==E=EaEE—>—E>E—E=~_— ——_————E==a==aanhA =E_ _——— LL ——L_—_~>~>=~L~LDh4»Ll]™__“Lbwww™—"“"]"f*-*-____L_Lhh]=—_=_—_—_—=S=--_—_—SS=S===