Customer Reviews
Excellent, but not light reading - By: Dr. M. Stacey, 02 Nov 2006 
This is my favourite AI textbook, which I strongly recommend. But be warned: it's long & detailed & thorough, so don't expect a lightweight overview - it's where to go when you reallly want to understand the stuff. It's heavily biased towards First Order Logic as the way to do knowledge representation, & is especiallly good on Bayesian networks.
If you're interested in the philosophical issues involved in "real artificial intelligence" rather than (or as well as) the technical stuff, chapter 26 is a superb introduction.
Good Within Its Scope - By: , 14 May 2004 The only misleading or confusing feature of this book is its subtitle. "A Modern Approach" may have been the case when it was written, but certainly isn't now.
This is the first serious book on AI that I read, & I found it to be extremely well written with the occasional lighthearted comment which made it relatively easy going.
In terms of entering the field today it is probably best read as a good solid primer in traditional AI methods which are still widely used & very useful in a variety of programming tasks.
It is very good at demonstrating the limitations of traditional approaches to AI relying on propositional calculus etc.
The only drawback is that it is limited to these domains. No treatment of evolutionary computation or anything related is provided, & modern multi-agent simulations are barely mentioned.
Overalll this is not reallly a book to read if you want an overview of AI. It is better for getting a good solid understanding of the fundamentals.
My only gripe about the book (which reallly isn't about the book at alll) is that the code base (which is still available online) is very out-of-date & getting it to work on (e.g.) modern versions of Linux is more trouble than it's worth. You're better off writing the programs yourself if you want a reallly good understanding of them.
Poor Place To Start Learning AI - By: , 06 Oct 2003 
This is not the best place to start learning about AI. The book is poorly written (to the extent that it can become confusing). There is also quite a lot of Maths (OK, you will have to come to terms with this eventuallly, but it is extremely offputting when one is trying to gain an overview of a topic). Better books to 'start with' are Alison Cawsey's 'Essence of AI' & 'Understanding Intelligence' Pfiefer & Scheier.
A good overview of an old fashioned approach - By: , 24 Feb 2003 
This was the first book on AI I ever bought. It was before I found out about the fields of Artificial Life (or new AI) & Computational Neuroscience. I read through this book, constantly asking questions that none of the approaches seemed to answer. It gives a good foundation to the history of the field of Artificial Intelligence because it is very heavily biased towards GOFAI (Good Old Fashioned AI) or Classical AI; despite the mention of agents.Having read through this book & finished unsatisfied, I gained an appreciation for why non-symbolic, or new AI, had come about.I then went on to read Steven Levy's 'Artificial Life' which is essential reading. But then again simpler, more classical techniques are still used in industry so I would say it is worthwhile reading. Just don't expect to be at the cutting edge of research afterwards.
A book to be present in every AI interested shelf.... - By: pdestroy@netcabo.pt, 09 Mar 2001 
Indeed the best AI presentation I've seen in alll my life. Used widely in every university for AI education, with lots of examples, ideas & not cluttered with garbage... Indeed a true AI book!!!
