I would say all successful engineers and aspiring entrepreneurs need to know the basis upon which we program today. Where we came from with assembly, why so many people wrote blocking code vs using event-driven programming, even though we've known that to be the most efficient way to handle IO for nearly 30 years, and the general human nature of technology and economics revolving around that. Working at a startup that is transparent about it's business relations is the most engaging and thrilling way to pick that up, of course. Knowing the broadest amount of relevant technology, and how to apply it to help people solve problems, is the most fulfilling (and coincidentally the fastest way to accumulate large amounts of money).

If you're familiar with some of the heavy hitting problems in computer science right now, such as vertically scalable architecture, garbage collection, how difficult it is to debug multithreaded applications, you might want to check out Node.JS, as it is my #1 bet for large adoption of engineers in the near future. It's the first project of it's kind, in the age of GitHub, that can really be called a trending software platform. It allows you to do very rapid iteration of development, without having the hassle of how to handle large amounts of IO and other problems that will make you bang your head against your desk with if you were using other systems.

Above all though, the most interesting projects for you to work are probably what you would learn the most from.