It's really good actually. It's part of my collection in "classic" AI.
It doesn't deal with neural networks, evolution or monte carlo sadly. But it does deal greatly with the Intelligent Agent (IA) architecture, which is the foundation of any AI, classic or not. And its chapters on search is superb; and you almost always need search. (Obviously DFS, BFS, Dijkstra, A* etc., are part of normal CS curriculum, but it delves into local search which usually is not part of CS curriculum as it is non-optimal and approximate)
It's also the best book on planning and propositional logics in AI I've read. Haven't had a great need for that myself, but those tools are actually very used in the industry to solve real problems. Some PHD students at my university have made a great local search based container stowage using some iterative local search based inference. It does not always produce an optimal solution, but it does most of the time, and its faster by a large magnitude (solved in a matter of minutes), as the problem is otherwise in NP (unsolvable in polynomial time).
After the book it's easy to jump into the research, because it has introduced you to the terminology. It's introduction chapter is also very nice, as it gives the history of AI research and accomplishments. Gives you an idea of where you are in the field when you read new research.
Oh but a warning; there's no code in the book. There's algorithms written in pseudocode. But it's expected that you implement yourself. If you're a good hacker, that's not that hard, but keep in mind the complexity yourself; the books complexity analysis does not include the data structures etc., so any implementation without the correct tools will be very slow. But AI is really something that should be learned after the CS foundations has been mastered. It does however explain in good detail how the algorithms, and how the theory works. Understand that, and you will have little trouble writing your code, and debugging the system. In my opinion that its much more satisfying, than just to copy a code snippet that you hardly understand. This approach forces you to understand, and therefore master it.