Issues in Artificial Intelligence

This is a 3 credit lecture-discussion course, meeting in two one-hour-and-twenty-minute sessions per week. [MW 9:30-10:50am in BU 112]

Text: Nils Nilsson’s The Quest for Artificial Intelligence(2010, Cambridge University Press, ISBN 978-0-521-12293-1) will be the principal text, and the schedule below indicates the order in which we will cover it.

Ray Kurzweil’s The Age Of Intelligent Machines (1990), The Age of Spiritual Machines: When Computers Exceed Human Intelligence(1999), and The Singularity is Near(2004), soon to be a film, are extremely useful references, and we shall view videos based on the first, and on Kurzweil himself, in the first week of class.

PamelaMcCorduck’sMachines who Think (2005 Peters), the twenty-fifth anniversary edition of a classic history of the field, and the first chapter of my volume for the American Mathematical Society’s Symposium series are excellent references for the history of the field, but Nilsson has almost everything we need.

Bertram Raphael The Thinking Computer(1976 Freeman) was the ideal gentle introduction to the technical aspects of the field; I shall supply some notes to add a little technical detail.

Outline: As we work through the history of the field, with technical supplements as needed, we shall supplement it with a list of issues, with appropriate references—a sample of eight or so follows; the introduction to some of the technical aspects of the field should suffice to enable the students to actually “get their hands wet,” and to understand the issues, and read the references. There will also be references to literary works on AI, by such authors as Borges and Capek, and films involving the field. The students, individually or in small groups, will pick issues and references, write (and rewrite) two long papers, and give reports in class on one of their papers. They will keep journals of at-home and in-class quick-writes.

Issues: Can computers think? The Kurzweil and McCorduck books present the affirmative. Contrasting points of view can be found in works of Dreyfus, Penrose and others.

Should computers think? Joe Weizenbaum covered this best in his Computer Power and Human Reason. Ken Ford and Pat Hayes (at IHMC) have more recent work on this one.

Should computers compete with humans in games, like chess, go, bridge, checkers, Othello and poker? Are the games too hard for them, or too easy? References abound on this one—authorities include Hans Berliner, ElwynBerlekamp, and many others, and the story of Marion Tinsley is a very interesting and human one.

In designing thinking computers, should creators emulate human thinking machinery—attempting to duplicate carbon-based wet-ware with silicon-based hardware—or should they attempt to solve their target problems “from scratch,” utilizing strengths of computers in ways that may differ widely from human thought processes? Again, references abound, including computer understanding of written information through modeling human understanding, as well as searching literature using purely statistical analysis.

Along the same line, should robots be human-like? Or are cockroaches, e.g., a better model? One argument here is that airplanes don’t fly like birds. Researchers at MIT, and writers for the X-Files, lined up on this one.

Do we need to fear a real occurrence of the take-over by Hal in Arthur Clarke’s 2001? Kurzweil’s predictions might lead to this sort of fear, and there has been a recent reaction to it in the comics (Nicole Hollander, Sylvia, on November 14, 2009—her November 12 strip is relevant to the previous issue, by the way).

What role can and should thinking computers play in education? Is the minimal effect of Computer Aided Instruction (CAI) an indication that the task is not suitable, or just that it took longer than expected? FAU is using ALEKS for placement and instruction, and we can investigate its performance as a close-to-home issue. I am trying to obtain better documentation than I now have, on this one. The main reference I am aware of is Doignon and Falmagne’sKnowledge Spaces (Springer,1999).

What about manufacturers’ claims of AI in things like cameras and TV’s? Is it real, or hype? Is the AI any more than that exhibited by a thermos bottle? This is discussed quite well in the text.

Technical aspects: We shall consider knowledge representation for problem solving, heuristic problem solving, exhaustive problem solving and the combinatorial explosion, what goes into programming for AI, the idea and implementation of expert systems, and machine learning.

Grading: About 20% of the grade will be based gently on technical aspects, assessed using homework and quizzes. The combination of papers and in-class presentation will account for 40% of the grade, and an additional 20% will be based on an assignment that can either be a technical one with a write-up, or an additional topical paper. The remaining 20% will be for the journal and the final examination.

Dates / Chapters / Class discussion* / Writing* / Due dates
January 9-11 / 1-2 / Video and DVD / On the videos
January 18 / 3-5 / Search
January 23-25 / 6-8 / Representation / First topic
Jan 30-Feb 1 / 9-11 / Logic
February 6-8 / 12-16 / Games; funding / Second topic
February 13-15 / 17-18 / Expert systems / Draft of paper 1
February 20-22 / 19-21 / Vision
February 27-29 / 22-23 / J5G, Prolog
March 12-14 / 24-25 / “AI Winter” / Paper 1
March 19-21 / 26-28 / Reasoning and representation
March 26-28 / 29-31 / Machine learning
April 2-4 / 32-33 / [Presentations / [quickwrites / Draft of paper 2
April 9-11 / 34-35 / by students / on the
April 16-18 / 2 per / presentations]
April 23-25 / period] / Paper 2
April 30 / Final exam; 7:45-10:15am
*Other items will be discussed also / *Current topics will be added

This writing intensive course serves as one of two "Gordon Rule" classes at the 2000-4000 level that must be taken after completing ENC 1101 and 1102 or their equivalents. You must achieve a grade of "C" (not C-minus) or better to receive credit. Furthermore, this class meets the University-wide Writing Across the Curriculum (WAC) criteria, which expect you to improve your writing over the course of the term. The University’s WAC program promotes the teaching of writing across all levels and all disciplines. Writing-to-learn activities have proven effective in developing critical thinking skills, learning discipline-specific content, and understanding and building competence in the modes of inquiry and writing for various disciplines and professions.
If this class is selected to participate in the university-wide WAC assessment program, you will be required to access the online assessment server, complete the consent form and survey, and submit electronically a first and final draft of a near-end-of-term paper.
[This class has been selected.]

In order to enhance and maintain a productive atmosphere for education, personal

communication devices, such as cellular telephones and pagers, are to be

disabled in class sessions.

In compliance with the Americans with Disabilities Act

(ADA), students who require reasonable accommodations due to a disability

to properly execute coursework must register with the Office for Students with

Disabilities (OSD) -- in Boca Raton, SU 133 (561-297-3880); in Davie, LA

240 (954-236-1222); in Jupiter, SR 110 (561-799-8010); or at the Treasure

Coast, CO 117 (772-873-3441) – and follow all OSD procedures.

Code of Academic Integrity

Students at Florida Atlantic University are expected to

maintain the highest ethical standards. Academic dishonesty is considered a

serious breach of these ethical standards, because it interferes with the

university mission to provide a high quality education in which no student

enjoys an unfair advantage over any other. Academic dishonesty is also

destructive of the university community, which is grounded in a system of

mutual trust and places high value on personal integrity and individual

responsibility. Harsh penalties are associated with academic dishonesty. For

more information, see University Regulation 4.001.