HOW ARE WE TO KNOW?
By Nils J. Nilsson
Artificial Intelligence Laboratory
Department of Computer Science
This book is about beliefs---how we get them and how we evaluate them. It takes the form of a fictional conversation among three people and Gio, an intelligent robot. Mia designed Gio, Nick is a philosophy professor, and Sam, a college student, asks many questions. Gio "uses the Web" to augment the conversation with lots of supporting quotations and illustrations.
The book makes the following points: 1) in analogy with robots, we humans know by the models we make of reality, 2) these models are always provisional and sometimes unreliable, 3) it is especially important to examine thoroughly those models upon which we base actions, and 4) the scientific method provides an excellent guide for such examination.
The level of exposition is neither technical nor deeply philosophical -- even though some technical matters are touched on, and some definite, perhaps controversial, philosophical positions are taken. All is approached from a commonsense perspective. The material should be accessible and interesting to the intelligent lay person -- the intended audience. It should be especially relevant nowadays when there seems to be such a shortage of critical thinking.
The book is available in Adobe Acrobat format free-of-charge from this website. One can download either individual chapters one-at-a-time or the entire book (about 10 MB). The last chapter is an extended review of the main points made in the book. I welcome criticisms, comments, and suggestions. There are several citations of Web pages. These can be accessed directly from their hyperlinks in the book. Please let me know if any links have disappeared.
Download the Chapters:
0. Front Matter
Download the Entire Book: