The O'Reilly website has this to say about the book:
"Hackers & Painters: Big Ideas from the Computer Age, by Paul Graham, explains this world and the motivations of the people who occupy it. In clear, thoughtful prose that draws on illuminating historical examples, Graham takes readers on an unflinching exploration into what he calls "an intellectual Wild West."
The ideas discussed in this book will have a powerful and lasting impact on how we think, how we work, how we develop technology, and how we live. Topics include the importance of beauty in software design, how to make wealth, heresy and free speech, the programming language renaissance, the open-source movement, digital design, Internet startups, and more. "
I wanted to like this book, but after reading about half of it, I have mixed feelings. I think some of Graham's observations are interesting, though some aren't new (the fact that some programmers are wildly more productive than others will not surprise many developers). But in others I think he makes assertions that aren't well supported.
In one of the early chapters (I think it's chapter 1, but I don't have the book here right now because it's in my luggage at the hotel), he makes the assertion (and I'm simplifying a ton here) that since children in medieval times started apprenticeships when they were in their early teens and we don't have record of them having the same sort of problems today's teenagers have (which I'll label as "teenage angst", though that's not exactly what he's talking about), then the behavior we see today must be societal and environmental in nature.
The problem with this argument is that there is some very good research that says that the brains of teenagers are not fully developed until around the age of 20 (see this from NIH, or Bradley and Geidd's excellent book "Yes, your teen is crazy!"), which means that the assertion that there's some biological basis for teenage behavior has some good support.
Given that this book is a collection of essays, I don't expect the same level of research I would in a full book devoted to the topic, but it's unfortunate to have this sort of oversight.
Another example is in one of the chapters on wealth, which I think are pretty good overall. Graham's assertion here is that before the industrial revolution, there has been wealth transference, but not a lot of wealth creation, and that wealth has been accumulated through theft (either directly or via taxation). I think this is basically true. His second assertion is that this changes with the application of technology. I agree that technology has brought new wealth (if you measure wealth by standard of living).
He then uses this as a basis that inequities between rich and poor are not a bad thing, because the creation of wealth is a good thing. But I think he ignores the fact that the old methods of wealth accumulation are still alive and well - the rich are (not surprisingly) interested in staying rich, and are willing to use their influence to make this happen. Given that, I don't think one can make the "Greed Is Good" argument without at least some qualification.