is probably intended as an inside joke by the cognoscenti, but it's actually more helpful than the rest of the book.
As I understand it (and if I'm wrong, write to Ford and Pylyshyn, not me), the frame problem involves how much information you need to give a robot (or a computer) f
or it to perform a task as well as a human would. For example, when frying an egg, do robots need to know what to do when the egg carton is empty? What about when the frying pan used yesterday still sits in the sink, unwashed?
Other descriptions of the frame problem include "describing and updating a set of beliefs efficiently" and (my favorite) "Now what do I do?," which almost any reader of this book can say at anytime. Clearly, the folks who work on the frame problem have a hard time agreeing on exactly what it is. Furthermore, it overlaps many subject areas, including psychology, ethics, philosophy, and religion.
Reading this book is tough work. The authors do not adequately define words such as
metaphrands
,
metaphiers
, and
Quineian
, as well as references to concepts such as
Yale shooting
. The book lacks a glossary and provides a mere one-page index of topics. One particular essay, a poorly written science fiction story containing profanity and spelling and gram
matical errors, illustrates the book's low level of editing.
I would recommend this book highly for anyone whose essay is published in it. For the rest of us, any robot story by Isaac Asimov or Stanislaw Lem will tell us more -- in a more lucid and entertaining manner -- about the difficulties of designing thinking machines. Now what do I do?
Book Review: What Was the Question?
