Now that I’ve completed my written qualifier, also known as a written preliminary exam, there are two official hurdles left to earn my PhD. These are my oral preliminary exam and finally my dissertation. The oral preliminary exam, much like the written preliminary exam, is a bit of a misnomer. Each exam contains both oral and written components.
The written preliminary exam, the one I just passed, is referred to as such because it used to be more like a conventional test that confirmed that a given student knew all there was to know in the fields immediately surrounding their own. So since I’m focusing on machine learning, I would have to know everything there is to know about all the various machine learning techniques as well as many other similar artificial intelligence techniques such as planning. Thank goodness, it is now assumed that one has a working knowledge of computer science from one’s undergraduate degree, so the written prelim is designed to gauge the extent to which a student will be able to write well enough to submit to a conference and present well enough not to make a fool of his or her advisor, department, and school at said conference. I meet these requirements.
Next I will move onto the oral preliminary exam, for which a more descriptive title would be the “dissertation proposal.” In this exam, I write a paper where I explain what I intend to do for my dissertation and write up a defense of why this is appropriately ambitious but also within my reach. I choose a small committee of professors who judge me in another defense much like my qualifying exam, except more demanding. This exam is not a simple pass/fail, though. Instead professors critique the proposal and say what they will want to be different in order for them to be satisfied when this work is presented to them again as a dissertation.
Finally, some say the dissertation is actually the easiest part because the requirements are clearly defined in the oral preliminary exam. Just do what you said you would and how can your committee (the same committee as before) do anything but pass you? So long as you were not overly ambitious in your proposal, it’s basically just writing a very, very long paper and then defending it in another defense.
So, that’s the overall arc of the rest of my PhD career, but it’s important to note that I won’t be doing it right away. In the meantime I build systems, do research, and publish papers. Papers are a big part of how folks will measure my success, so that’s priority one right now.