To the editors:
In his letter last week [January 5], John Mintier criticizes Roger Schank ["Northwestern's New A.I. Hotshot," December 15] for describing a computer simulation of a chemistry lab as a "future vision," when two UIC professors have already marketed a prize-winning program of just that kind. The error was mine and not Dr. Schank's--the quote I used came from his 1984 book The Cognitive Computer.
Mr. Mintier also cast doubt on the significance of one of Schank's "reminding" stories--the one in which a student reads a story about cutting a ham to fit the pan and is then reminded of the irrational history of the QWERTY typewriter keyboard. Schank argues that the ham story reminded the student of QWERTY because they both have to do with trying to explain something as rational when it isn't--"explanation failure." Mintier points out that both stories appear, 21 pages apart, in a popular college textbook and are "by far the most memorable examples in the paragraph section, and ones which catch students' interest. Could this be the failure of "explanation failure'?"
No, it couldn't. It doesn't matter much where the student first read them. The question is, why did one story remind her of the other? It's not enough to say that they are both "interesting" or "memorable"--that begs the question. (By the way, it must be one dull textbook!) When you read something "interesting," you are not immediately reminded of every other interesting thing you've read in the past few days--only certain ones. Why those certain ones and not others? That's the question Schank is working on.