I'll take the counter-point on this one. I picked it up after it was mentioned in the YC's Summer Reading List; however, I found it to be a tire to finish.
The reasons and explanations given seem to touch around a technical, but not too technical approach to algorithms. Getting stuck in a place that probably just leaves both audiences a bit unhappy. For instance, there is a chapter that mentions that the optimal stopping point is ~37%. There is never any mention about how the 37% number is found. Of course, I could look it up but I could just as well look up the optimal stopping problem.
Aside from that, the examples come across as contrived and inapplicable. Sure merge sorting your socks sounds great, but I still will never do it!
Just a head's up: for someone well versed in computer science the first two thirds of this book can be pretty dull. The majority of the time is spent explaining first year CS topics in layman's terms. After that it picked up a bit with, for example, showing how randomized algorithms can apply to types of decision making in real life.
I still think it is better suited for those with little to no CS knowledge.
I just finished that not long ago. It's an interesting read because it places an emphasis on the value on simple and statistically effect algorithms that can be executed by humans with relatively low cognitive overload.
Many of the algorithms and ideas are familiar, but the novelty here is how one can map these CS lessons to improve performance on the human OS and human network.
An interesting mix of computer science and psychology. Just started on this one recently. Highly recommended by a colleague. Seems great so far.