About Artificial Intelligence: AI is a huge field. Many things that used to be considered challenging and difficult "AI" are now integrated into systems and products without even a mention of "AI". AI is not a collection of algorithms or analytic descriptions of functionality or data design. AI connects strongly to our conception of what human intelligence is. So working in AI means dealing with knowledge concepts, reasoning, heuristics, cognitive science, perception, natural language, etc., to a much greater extent that more conventional fields such as databases or networks. Studying AI therefore means doing a lot of reading about knowledge-related concepts and systems. It also means developing systems and writing code that deals with knowledge and deals with it as knowledge.
About the courses: Each evening class for the graduate course will be divided into two half classes, roughly, 6-7:25, and 7:40-9. There will be 24 of these half-classes. There are 26 undergraduate meetings. So it will be possible to keep the material in the two courses approximately in synch. This will help me to prepare well for the classes, which in turn will benefit you, the student. The treatment of the material in the two classes will obviously differ.
About the textbooks: Two substantial textbooks have been assigned for the course: Norvig's Paradigms of artificial intelligence programming: Case studies in Common LISP, or "PAIP", has the great advantage that it fully integrates the study of AI with designing and writing code that implements important AI concepts and strategies. There is no other AI text that does this so well and so thoroughly for Lisp or possibly for any other language. The other textbook, Russell and Norvig's P. Artificial intelligence: A modern approach (2nd ed.), or "AIMA", is the world's leading AI text by far.
About me, Professor Futrelle: I was aware of the young field of AI, and lectured on it in the mid 1960s when on the Physics staff at the Rockwell research center in California. Some ten years later I was on the Biology faculty at the University of Illinois. My interest in AI continued to grow there, though I focused on my cell biology "wet lab" experiments. The turning point came in 1981 when I was at the Marine Biological labs for the summer. I decided that computers needed to be applied not to numerical data and modeling in Biology, but the huge corpus of published knowledge of Biology - a qualitative, knowledge-based problem. That's a task that requires understanding the natural language and figures in papers. I shut down my lab at Illinois and moved to Computer Science at Northeastern in 1986, founded the Biological Knowledge Laboratory in 1989 (the BKL), and continue my research on these complex and important topics to this day.
I was involved in various ways with both of our textbooks. I wrote the review of PAIP that appeared in the AI Magazine; I think that helped publicize the book to the appropriate audience. I was involved in test teaching the first edition of AIMA, while Russell and Norvig were creating it, as well as discussing strategies for the text with Russell, the editor, and some others. My AI class of that time sent in weekly critiques of AIMA chapters to the authors, as the authors were writing and revising them. You'll find their names and mine in the Acknowledgements in the AIMA preface.
I don't really do the "classical AI" you'll find in textbooks. My research is very much driven by the needs I see in the problems I work on. I don't think of AI as a collection of tools to be applied - that can lead to the "hammer in search of a nail" approach. Instead, I try to look at the data to see what's actually there and then design whatever methods I feel are the best fit to the problem. Of course I'm keenly aware of a wide variety of important work in AI, cognitive science, linguistics, and related fields. It all informs my work.
About you: I am always happy to discuss special projects with students, undergraduates and graduates. These can range from informal arrangements, to directed studies, theses, or dissertations. I have been lucky to have some superb students work in the BKL. Research is different from standard classes. It's more open-ended and varied. Once you get into doing research you won't have "assignments" as such. If you are an undergraduate who is considering going to graduate school, especially to a PhD program, research experience as an undergraduate will be critically important for you.
Return to CSU520 homepage.
Return to CSG120 homepage.
Return to Futrelle's Teaching Gateway or homepage