One of the most interesting parts of the book is that it claims you should "never criticize" another person. It says that every person can always justify their actions in their own mind and all criticism is taken negatively. As an example, even mass murderers have been interviewed justifying their actions. I tend to agree with this philosophy.
Code by Charles Petzold is an awesome hacking book. It takes you from flipping a light switch, thru simple circuits up to assembly language. It's beautiful!
It really depends what sort of hacker they are. There's a sort of hacker you buy a copy of "Pro JavaScript Techniques" for, a sort you buy the Dragon Book for, and another sort who you give a copy of TAOCP.
Most of the books above are recommended because once reading them you will have an uncontrollable urge to immediately create something brilliant. You will not be able to stop yourself.
"First, Break All the Rules: What the World's Greatest Managers Do Differently". I know it has a cheap title but it there is plenty of merit in it. The Gallup organization carried out a massive survey of various business units in an attempt to discover commonalities between the successful units. The results showed that there are essentially 12 things that must be true about a unit in order for it to be successful.
Its a fun read if nothing else. It contains a lot of counter convention logic that would be right at home in one of PG's essays.
I haven't gotten through all of it, but I really like the New Turing Omnibus, by A.K. Dewdney. It's about computer science, not how to program, so it would be interesting to non-hackers, too. Even if you are familiar with all of its topics, it can serve as a reference and clarifying tool for CS concepts.
* "Applied Cryptography" by Bruce Schneier. The first chapters are a great introduction to cryptography, but I found the few following chapters presenting cryptographic protocols mind-blowing. Some of them are so simple yet very clever.
* Political economy essays by Frédéric Bastiat ( http://bastiat.org ) were truly enlightening for me.
It's not bad, but it's free on the web. And in terms of "getting real", they could stand to make a real book out of it by getting an ISBN number for it with all the money they made off of people throwing their money at them.
And of course I'll add the obligatory mention of Squeezed Books, which is more or less based on business books I've bought and enjoyed (although there are also others like Good to Great that I wouldn't recommend).
Steve McConnell's Code Complete (2nd Ed is the copy I have) -- it seems like every second page I've read lays to rest some tiny unanswered question I've had in the back of my head since I started programming.
I've not read Peopleware but I've read Why Does Software Cost So Much? by Tom DeMarco and Slack by Tom DeMarco. The former focuses on technical issues and the sunken cost of software. The latter focuses on the agility of organisations. Tom DeMarco started as a programmer and subsequently became a consultant increasingly focusing less on technical issues and more on business issues. Therefore, reading his books in order would be most relevant as your start-up develops.
It opened up many many ideas to me, including flying, writing software for people, why simple software is harder to write than complicated software, and gave me some indication of what 'being smart' really means.
