Even a lot of CS Ph.D.'s aren't quite sure what they want to do in their field, so don't feel bad about having a bit of uncertainty associated with your career.

Even a lot of CS Ph.D.'s aren't quite sure what they want to do in their field, so don't feel bad about having a bit of uncertainty associated with your career. Ask yourself what sort of subject area you'd be thrilled to hack on and learn about for 12 hour days, and focus on that. It could be anything from disk drivers to communications protocols to fast fourier transforms to Arduino to role-playing games. Get thee to GitHub and look for some unsolved bugs or features in an area you consider worthwhile. Pretty easy to get far in your career if you have a large portfolio of open-source code. Lastly, there is always a large need for legacy programmers; people who understand how to use COBOL, FORTRAN and other things they don't really teach students these days. As your forebears grow old and retire, someone's going to have to modernize their works. There aren't a lot of freshly minted software engineers who know these environments.