First the good part about this book. Chaitins first chapter is quite good. Here he outlines the results of Godel, Turing and his own. It is very readable. Without going into the real mathematics he can really make you feel you understand these deep ideas. The later chapters go more deeply in to the ideas presented there and illustrate them with lisp computer programs. Especially the search for lisp programs that evaluate to themselves is amusing.
But let's now focus on the parts of the books that I did not like. His exposition is mixed with an account of how he first learned these result. I am charmed the first time when he explains how he read so many books as a kid. But soon I do not want to hear again what he felt as 12 year old. Also he keeps comparing his own work to that of other scientist. We really need to now that he is just as good as Godel and as Turing.
For example he takes pages to explain that Kolmogorov ripped of his ideas. What I also find funny as well is both chapter 1 and chapter 6 give an identical link to "my first major paper".
Sigh. He's the best, we get it, ok?, now please move on.
Then one more thing. The computer programs that he uses are in lisp. That is fine by me, lisp is a beautiful language. But do you think he uses any of the available dialects? No, of course not, he introduces he own strange version. The programs given do not run in clisp for example.
So to sum it up. I learned his own result on incompleness (that one cannot produce the shortes program for a particular function) and that is a nice result. Reading the rest of the book is more annoying than amusing.