The introduction to this book (which talks about the foundations of computational neuroscience) is hilarious:
This book is for undergraduates and beginning graduate students…because the malleable mind of the undifferentiated stem-person of college age is likely to be capable of making conceptual leaps that more ossified brains cannot… these are early days. The field [computational neuroscience] is newly emerging hybrid and is still undifferentiated (an undifferentiated field needs an undifferentiated mind)… I present stuff that is either fun or interesting or important and maybe sometimes all three.
Even though the rest of the book is not as funny as the intoduction, it is still interesting and delivers what the author promises at the beginning – foundations of an exciting new field, different (not always consistent) points of view, ideas and speculations. With respect to the (targeted undergraduate) reader, the author takes time to explain concepts from binary math, linear algebra, calculus, electric circuit, biology, neural networks or learning algorithms and does not assume any prior knowledge of these. That’s what makes this book both good and bad.
Good because an open untrained math can learn a lot, but bad, because of the large scope of needed background and neuroscience topics themselves compared to the limited volume of the book, it mostly only stays on the surface and reader who would like to go in depth into specifics has to look elsewhere.
Also, in my opinion, the author overemphasizes comparisons of the brain to a computer (but I guess that’s the bias given by his field of work). Even though it might be useful in some very crude approximations, my (still undifferentiated) mind does not see this as the point where the paradigm shift of neuroscience would happen.
All that said, good introductory reading.