Context: Earlier this year I made a decision one day to get a master’s. It seemed like such a simple decision, that until the previous two days, I never really questioned it. But given my sudden intense introspection, and the fact that my decision to get a master’s was largely driven by my “passion” for machine learning in the first place, it seemed like a good time to do that questioning.
Yesterday I decided, maybe I don’t like machine learning. Today I decided not to go for a Master’s straight out of college. In an effort to not write about my life story for hours every day, I will try to keep this short(er).
Initially, I wanted to go to grad school because I wanted to do machine learning. An MS with a specialization in data science will absolutely help you land real machine learning jobs in industry. So, I could easily take an extra year and suddenly have this entire wonderful field opened up to me. That’s why the decision to go to grad school came to me so easily that day: it’d be silly not to do it.
But now (see previous day) things have changed. I don’t think I’m sold on machine learning, hell I’m not even sure if I would enjoy it at all. In fact, as of now? I’m sold on nothing. I like school, I like my classes. But that’s part of the problem. I liked networks, I liked operating systems, I liked machine learning. But so far, that hasn’t gave me any idea of what I actually want to do when I go out into the real world. Industry is different. Industry isn’t going to lecture, taking the time to understand new concepts, and studying for a final. So, what interests me in school isn’t really going to give me a good indication of what interests me in industry.
The main point I am trying to communicate is that graduate school will not really help me where I need it. It won’t tell me what I want to do in life. Industry will. (Or maybe more likely, tell me what I do not want to do in life). And once I know that, I can better understand why I want to get a masters. Do I want to learn more about security? Distributed systems? Maybe go back to machine learning? And from there, if I decide my MS is still something I want, I can do it with a purpose. It won’t just be frivolous learning, it will be the knowledge I need to take the next step towards my goals. Grad school can be an extremely valuable experience, and I feel that, by doing it right out of school, I’m deriving myself of so much of that value.
That was a good place to end my main point. But funnily enough, once I realized all the above, I still wasn’t set. A master’s at UCLA was just so simple. I go right out of school, I can even start a quarter early. Even better, no GRE, no letters of rec, and low cost. So normally, it’s kind of a more, “might as well” thing (which in fact, is what it initially was). I could take some cool courses, learn some cool stuff, have some fun. Sure, I wouldn’t have the work experience backing me up, but the simplicity of doing the program right away was worth the potential losses later on.
Except I made one wrong assumption. I don’t know if I actually would enjoy myself, enjoy the learning. I’m sure I wouldn’t hate it; I rarely dislike classes, and have never not appreciated the sum of a quarter’s worth of work. But honestly, looking at UCLA’s MS program, I just couldn’t get excited about it. Without ML, it’s hard to find a path, so to speak. There’s classes sprinkled around that I think I’d enjoy, and get a lot out of. Things like cryptography, advanced networks, distributed systems. But these are few and far between. And the few classes I’m excited about could be ruined by one bad professor (notable: cryptography is taught by Ostrovsky, one of the least liked professors in the CS department). Additionally, the rate at which exciting courses are offered is… suspect. I looked at last year’s classes, to see what kind of classes are offered each quarter. Fall, an interesting class or two; I should be good for that. Winter and spring? Maybe one class. And keep in mind, I fully intended to finish in three quarters! So say one of my three quarters doesn’t really offer anything interesting. That’s a third of my degree, in which I will have to take classes I’m not interested in, or not finish when I want to. And let me tell you, dragging on my degree would almost surely make me regret my decision.
So there you go. First, there’s benefits and downsides to doing the master’s right out of college. Except, considering I don’t really like UCLA’s program, most of the benefits are thrown out. I find it completely plausible that other schools are not necessarily better than UCLA for an MS. However, because of the reasons I mentioned before (industry experience –> purpose –> knowing why an MS will help me), I think I could salvage some value from a not ideal program. But I do not have that experience, so I don’t think the value is there.
Once I figured that out, it was simple. It took me about five minutes to decide that no, a master’s at UCLA, right out of college, was not the right decision for me. Maybe I will go back later, maybe I won’t. But before that, it’s time for the next step in my life: starting a career.