I'm done with grad school.

By all accounts, that wasn't what was supposed to happen. Like many folks in grad school (though not all!), I did well in school growing up. In high school, I graduated top of my class. In undergrad, I graduated with the level of honors where they put your name on a piece of bronze with a bunch of other people and hang it in the library. But more importantly than that, I was pedigreed; instead of working internships my summers in undergrad, I took research jobs at universities and got experience "doing the things you do in grad school". I attended meetings with a research group at my undergrad where they discussed their own and other research in the field. I TAed for three and a half years in undergrad. I knew, more or less, what I was getting into. And now, a little over a year later, I'm realizing I didn't know what I was getting into as well as I thought and I can't stay.

When I was applying to grad schools, I read blog posts like this one. Some were by people who got through some level of grad school, some who finished, some who even stayed in academia past that. I wouldn't say I'm especially insightful compared to anyone else; most of this is just me getting my thoughts out. But I will say that, even hearing the negatives before getting into it, I think there was a part of me that thought I was an exception. Grad students won't typically shy away from at least mentioning that it doesn't pay well -- it's pretty easy to realize that nobody goes into grad school to make boatloads of money. That's not the point of grad school. But it's a lot harder to realize that, even if you're not looking to make a ton of money, you still need to support yourself, and you need to make some money to do that; grad school can be questionable on that front, too. And so, if there's a major takeaway from anything I say here, it's that there are these little realizations that I missed before getting into the thick of it. If you're a person considering grad school, things like that might be helpful to know, though keep in mind of course my perspective is that of somebody leaving. It's good to hear from folks who do like it more too. And if you're not in that position, I wouldn't take my word as the gospel truth of why grad school sucks. If anything, lots of stuff sucks, unfortunately.

I chose academia because industry is bad.

Look, I'm gonna be real fucking reductive here. This isn't true, at least not fully, but it was the thought that drove me to academia, even more than my thought that I liked research or anything like that. But, I had this idea that CS people are bad, and that I didn't want to be around them. To be clear, I'm saying this knowing that A) I am a CS person, and B) I know plenty of CS people I like and don't think are bad. But, here's what I mean.

One of the things that I particularly appreciated about my undergrad was that, even if it was maybe kinda a joke, we had to take an ethics class to graduate. As far as I'm aware, that's not exactly common for CS programs. The class was run by some really cool library grad students, and to be honest talking with them some, I actually considered getting into librarianship. Librarians are cool and thoughtful people, in my experience. The class itself was kinda simplistic; the first day of class is literally talking about things like the relationship of science-y/tech-y people in Germany to the Holocaust. So, that kinda set the tone; they wanted to really hammer it in that "hey this is a thing you should care about even if you're a computers person". The actual "ethics" in the class were "here's some basic ethical structures, now make a very basic three paragraph essay about why this or that random thing is ethical or not using them". I think their idea was, you make people go through the exercise of thinking even the least bit about ethics, and they'll pick it up by osmosis and take it into the workplace. Not perfect, but, you know, how exactly would you structure this sort of thing as a college class if your goal is "make people realize their discipline has ethical consequences and they need to be able to engage with them or else bad shit happens".

Graduates of my undergrad invented Youtube. And Yelp. And Paypal. And other things that made them incredible amounts of money. So, when that's how the department brands itself to the outside world, and what marketing says to try to get kids to apply, it shapes the makeup of who actually shows up to get a degree. Suffice it to say, that ethics class was nowhere near a priority for most of the students I took it with. I get it. It wasn't especially well done such that it really engaged me, either. But, there were so many people resistant to even the suggestion that ethics was something they needed to think about as a CS person. They didn't want to show up or do any of the work beyond what was necessary to get them a degree, not because the class was bad, but because ethics wasn't important (to them). One of the few times I actually hung out with a group of CS people who weren't just my close group of friends (i.e., folks who were generally queer or otherwise not your stereotypical straight white CS dude), I distinctly remember one individual at a party. For one, he was creepy as fuck around the women in the group, to the point that like other CS dudes were telling him off for saying some stuff of a forward nature to somebody who had gotten pretty drunk that night. It's a good sign that other people were telling him off, of course, but that wasn't the only thing with him. He talked a lot about how shitty California was because it was taken over by the queers (I guess he didn't know I was gay). When I brought up things that I was worried about with respect to the way technology can enable people to be shitty to one another (for context, I'm pretty sure this was when we were all talking a lot about Facebook election stuff), his answer was "Well you should be as shitty as you can so you can make as much money as possible, obviously". This dude was not saying that ironically, I promise you. And it wasn't the mood of the entire party, but to me, it seemed like that guy is always there; you can't escape him.

I have mixed feelings about the way that society at large has been talking about how there's this tech boom and people can go into it and make a bunch of money. We need more STEM people, says industry, so send us more! Personally, I got into CS because I find the discipline inherently interesting; in point of fact, that's a part of why I wanted to go into grad school. I wanted to prove that I was legitimately interested in it, not just here to collect a payday. And like, I get that at the end of the day, a job's a job, and I can't exactly blame people for wanting the things that come with having a decent chunk of change in your pocket. I was a TA for a long time, and I dealt with all sorts of students, with folks who would talk to me about how they were getting into CS because that was their chance to live a good life. I can't fault that, of course. There's no good that comes of any sort of ideological purity test to be able to study a discipline. On the other hand, those folks were not the only people who "didn't have their heart in it" in CS. There were folks who, for lack of a better word, were just plain greedy. They're there because they want to be Bill Gates, or Elon Musk, or whoever. And it's hard for me to sympathize with that. Sure, wealth gets you nice things. But, wealth at the cost of other people, wealth coming from strategic decisions to ignore ethics, that sort of thing? It's ugly.

I was willing to leave academia because I realized I was wrong and industry isn't bad.

Like I said, all that above was reductive and wrong, in many ways. There are examples of that guy in CS plenty, but I'm increasingly aware that maybe he is more rare than I thought. One of the things from my undergrad that stuck with me was how strongly I associated that image specifically with people around the big tech companies, your Facebook/Google/Microsoft/Amazon/whoever. And coming to academia, I interacted with some people from those companies. There was a guy in my lab who had previously worked at Google, for example. He was, believe it or not, a normal person. He had reasonable insight onto ethical problems we dealt with in the lab. He did and didn't like parts of time at Google, but he was a person I would be happy to call a coworker. If Google has people like that, maybe it's not so bad.

And he's not the only person I know. Of course they're not all bad people. Grad students aren't perfect either. There's plenty of people I interact with some who, frankly, meet that stereotype of a shitty CS dude to a T, and they're here in grad school. Maybe that means that guy is truly inescapable. Certainly that means I shouldn't be on a high horse about being in grad school. It doesn't make me any better than anybody else for being here, it's still about what I do and the kind of person I am regardless of where I work.

Okay so industry is still kinda bad though.

It's just bad in a very different way than I first thought. I'm not exactly the biggest proponent of capitalism these days, and it doesn't escape me that tech workplaces are the kinds of places were, say, unions aren't so common. That doesn't have to mean that I wouldn't want to work there, or that workers are definitely in a bad way -- don't get me wrong, I've heard plenty of stories of bad unions. I'm not going to sit here and tell you that all unions are perfectly good. But a good union that people put work into, consistently, can be a true benefit for people. It helps folks preserve their interests even when management doesn't have the best interests of the company at heart. After all, a workforce that is happy and healthy can produce better for a company, even if it's a little more expensive up front. But it also helps folks to have more of a say in the larger direction of a company, or other policies. When Google people got up in arms over the revelation they may have been working for the US military, that's an example of a place where having a union gives them a better way to make sure management actually listens to them about the fact that they would not like to participate in such business.

But it's not just unions. After all, one of the main reasons people seem uninterested in unionizing at large tech companies is that they do tend to have a better track record as of late in treating their employees well (at least some of them, anyways). And the pay's good, so negotiating for better wages in particular is not a top priority. But management still has problems. Take recruiting, for example. It's well known that tech companies seem to have a bias towards white dudes as their employees. Partially that's because right now in CS, it's just a lot of white dudes anyways. A random draw is gonna pull a lot of them. But I'd also argue that a lot of the way recruiting works reinforces this, unintentionally. Getting a job as a new grad is a complicated dance of sending companies a resume, a LinkedIn, a Github, a bunch of crap. And the people who do that dance well are largely the that guy. The highest concentration of that guy I ever saw was the one time I went to a hackathon in undergrad. It's supposed to be a fun thing of building a cool... whatever-catches-your-fancy-that-weekend, but in practice it ended up looking like a lot of posturing for the companies that would roll through. People would tool together smoke and mirrors to make the thing that they had slapped together over the course of an unhealthy weekend look like a marketable product. I do take on some tech projects in my spare time, a little bit more now that I'm out of undergrad, because I find that stuff fun. But it happens to be a lot of furry shit, like this website, which I don't exactly want to broadcast to companies. It's for me, it's not for them. And when I do these things, it's definitely not supposed to be an example of how I'm willing to take on uncompensated overtime for a mismanaged project, like some of that hackathon stuff ended up feeling like for me. Absolutely, other people have different opinions on this than I do. I won't go full "this whole enterprise is just resume padding" on you, because I do think the core idea of giving people a time and space to just work on cool projects is a good idea. But I absolutely don't think the expectation that everybody does something like that, or has a full and active Github, or spent a ton of time on their LinkedIn account, is healthy. It's like trying to scout sports based on a stat sheet. Sure, a good stat sheet tells you (and all the other teams' scouts) when there's an absolute standout. Everybody will be all over them, because the stats are good. But that doesn't explain the person who just switched sports and hasn't built the stat sheet up, or who got injured and missed a season, or just plain doesn't have a good coach to help them showcase the full extent of their abilities. And that's where the scout comes in. You go out and see folks performing in their own environment, see how they would fit into your team, see their abilities on display directly. The "going out" is the most important part. Because anything you see that somebody sends you in themselves, well... They're only going to show your their makes, never their misses.

Going to hackathons or career fairs as a company isn't scouting. It's selecting for the same people who have figured out how to best optimize the stat sheet. And half the time, the company ignores the stat sheet for the things it's good for anyways. How omnipresent is it for companies to run their own coding assessment for potential new hires? Guess what? I never had to do a coding assessment to get into grad school, because professors know that if you're graduating top of the class with a goddamn CS degree from one of the top 5 CS schools in the world, you probably know how to code. You don't need to re-litigate that! What you do need to do, though, is find fit. And that's the other half; scouting also isn't recruiting. Recruiting is building the relationship and feeling each other out (in both directions). A company that runs a hiring process where the potential new hires don't feel comfortable feeling out whether the company is a good fit for them as well is a bad idea long term. That's the way you get people who won't stay, and eventually make you hire another new person to train. And it feels unfortunately common in my experience so far. When applying to places, you get a black box. They could tell you no for anything and nothing. Did you apply out of sync with their recruiting rhythm and screw yourself over? Do you have a fundamental flaw as an applicant? To "protect the integrity of the hiring process" (literal quote I got from one company), they will tell you absolutely nothing, not what they liked, not what you could improve on. And if you view HR as a place where you're trying to minimize costs, this makes sense. Why waste your time telling an applicant something as simple as "you applied a little late and by the time you got this far, we'd already filled the spots, so we didn't even look at the notes from your interview, but the next opening we have you should jump on"? Why take the opportunity to help a person to improve, who could potentially work for you down the line? Why take the opportunity to demonstrate that your company is willing to help make things better at large, even if it's not directly and immediately beneficial to your bottom line? Why take the opportunity to show that you're capable of even medium-term thinking instead of short-term? Because, in some peoples minds, HR isn't a place where you have opportunities to improve your company for the long-term. It's a place to reduce costs and risks.

(As a sidenote, it's a "fun" exercise to go into places like infosec twitter and see examples of professionals up to and including C-suite executives point out that conservative HR strategies are costing their company tons of money. Why hire somebody junior to do these pretty easy tasks for a more modest pay when you have a CISO you pay half a million dollars to? Now, that CISO's actually getting paid to do things like direct other employees and larger risk management decision-making, but the menial tasks that take up their entire day for no good reason is just as good a use of their time as anything else, sure. The lesson here, at least that I take from this sort of thing, is that business people -- you know, "the adults in the room" -- are not at all good at making decisions that will actually best serve a business. Capitalism and the free market does not lead to optimal business strategies. More often than not, pride, greed, and an inability to think past the short-term all add up to corrupt the invisible hand of the market, and push things towards shitty ends. But, that's a story for another day.)

What do I do in grad school?

This is probably the actually important thing, eh? It's not obvious to most people; a lot of people think it's taking more advanced classes. That happens, but is easily the least important part of my job (and let's be clear, this is a job. Don't get confused because the job title has the word "student" in it). Sometimes I TA too, helping to grade things and to some degree set up how the class works. For example, I've written assignments for students to complete. That matters a little, but is only a small part of my job. In point of fact, most of my job is reading papers.

There's a lot of papers in academia. There's a ton, way too many, and not all of them are that great. But you have to read them to keep up to date, and also so you know what other people are working on to not get scooped. You read papers to understand previous techniques, and to improvise on top of them. Papers are the currency of grad school. Rarely, I'll think about things, and improvements I can make to previous papers. Even more rarely, I'll turn those thoughts into actual code that I try to run an experiment with, or build a new kind of system with some complicated math involved. But mostly I read, with some thinking. It's work very unlike any other job. I will say, it is a nice thing that pretty explicitly "learning new things" is the main part of the job. That's not common in a lot of places. But it's learning complicated things, that people aren't always explaining very well. And it's definitely not just sitting in a room listening to a lecture.

Where does grad school lead?

This is specifically tailored to CS, but it's probably mostly the same in a lot of STEM fields. At the end, you have three choices. You become a professor, you do research not at a university, or you turn your back on the thing you spent the past 6 years on and do anything else. That's pretty much it. As a professor, you'll spend most of your time writing grants, serving on committees, and managing students. You don't spend a lot of time actually doing research. Grad students do the research, you just point them in the right direction. You could stick to getting to actually do research by joining something like (in the US, if you are a US citizen) one of the national labs or a company's research division. That's more similar to staying a grad student but with more authority -- you still have to win grants sometimes, and certainly you have to do some convincing of the people you work with that what you're working on is valuable in some way. But you do a lot of the research yourself, which means more reading papers, thinking about things, maybe doing some math proofs and coding up experiments. Or, you take the third option, and just totally walk away and do nothing that looks like anything you've spent all of your time on for the past 6 years. If you don't want to keep doing research, that's what you'll be doing. And let's be real, it's pretty fucking hard not to see that as some kind of failure, because most jobs the research stuff just won't apply to. (A quick note: there are also teaching professors, but those folks don't get tenure. That job sucks pretty bad; they have to negotiate a new contract every year, and there's no guarantee the university won't just walk away from them. And it can be kinda thankless, because typically they teach more introductory classes with less mature students. So, if you're any smart, you probably decide not to actually pursue that unless you really, really want to teach.)

What's good about grad school?

You get to set your own schedule. That's a big one. There's no specific time I have to be in at work; I can show up and leave as I like, as long as I am making progress. If I have issues I need to take care of, it's not a big problem (it turns out, being a foreigner here in Canada, that actually happens a lot. It's pretty annoying to have to go through bureaucratic stuff on a regular basis). Though it depends on your specific advisor, often the research you'll be doing is stuff that you more or less choose to work on, with not too many restrictions. The professor needs to be able to tie it back to their grants and larger mission, but that should've been taken care of when you chose what school to go to and what advisor to work with -- you should be picking your advisor based on the things you want to work on, after all. You always get to learn stuff, and part of your job is travelling to conferences to meet people and discuss your ideas with them. Travel can be fun!

What's bad about grad school?

I mentioned it earlier, but let's start with this. I live below the poverty line right now. I absolutely come from a lot of privilege, and plenty of other people can tell you what it's like to be poor a lot better than I can, but it sure isn't a good time. Definitely wasn't crossing my mind when I signed up to come to grad school that I would live in constant fear of any major unexpected cost, like medical stuff that falls through the cracks of my confusing insurance situation, instantly sinking me. That ain't good for your psyche. If I were to take an industry job, it would be reasonable to expect that my take-home income would increase five-fold. That would be an immediate shock to my current quality of life. For me personally, it's not helped by some key differences between the way the system works in Canada vs. the US, but it wouldn't be too different if I were in the US. I know that perhaps this sounds weird when I was railing against people who are greedy in CS, but I feel like considering that I and my fellow grad students do the bulk of the work that makes universities their money (a lot of their income and prestige is coming from grants and research that we have most of the responsibility in developing), it's not unreasonable we should be compensated well for it. Or, at least, better. Certainly not at or below the poverty line, which seems to be common from what I've seen.

The work never leaves, either. There's not a time where I go home, and I can stop thinking about it. There's no 5pm that I get off the clock. I could be in the lab all night if I wanted to, and that would help me make progress. If I'm not making progress, there's very little to say about it other than it's my own damn fault, because effectively I am my own boss. Sometimes it's nice to be able to externalize that and say, "That's somebody else's responsibility". I can't do that. This is incredibly stressful. I want to be able to clock out!

And you don't always get to work on just what you want to. TAing can be fun at times, but a lot of it is unpleasant. Grading sucks. It's just so bad. Teaching is nice, but you don't always get to actually do a lot of it. I don't have a whole lot of face-to-face time with students as a TA right now, and there's no other role for this class that gives me more face-to-face time. To pay the bills, you will almost certainly have to be involved in at least some research that isn't in your specific thesis area, possibly with not the best people to work with. And professors are not good managers, in general. They don't know how to help you and other people work together to make progress on the research, because that was never something they needed to learn to be able to become a professor. As it happens, professors typically don't realize that their job amounts to being a manager, and never bother picking it up as a skill.

Even further, a lot of times, what you do work on even in the ideal case, even for your own thesis, will likely never escape academia. Your end goal is to make a thesis or to write papers. Not to write code, not to build things or make recommendations that affect anybody in the real world, nothing that impacts any real people. Just, write papers that will be read by other academics who themselves will write more papers. And that gets old really fast.

Is there anything grad school gets right?

It's worth mentioning again, but I didn't have to prove I know how to code to get into grad school. We got that figured out. And in general, the process to apply to grad school is fairly straightforward and understandable. If you get rejected, it's a little bit easier to tell why -- most likely, you weren't a great fit for a professor who was looking to take on more students at that university at that time. That's all of my cases, for sure, and it was easy to put that all together after the fact. But the timeline is clear. You know when you're supposed to apply; there's not such a thing as too early because they'll hold on to your application until they actually look at it. When is too late? After the date they tell you to apply by. Super easy! There's no getting out of sync here. And once they accept you, they understand that the next thing they need to do is convince you to come and let you ask questions about even uncomfortable things. That's what they invite you to the university to do, to schmooze you. My family joked that I should be on the lookout for people from shoe companies looking to pay me off (in reference to the FBI investigations going on into shoe companies and college athletes). It's some fancy crap, but that's the other half of the relationship -- you don't want somebody who doesn't want to be there, because they won't stick around. Professors get that, when companies... don't.

Okay, but grad school applications are also bad.

Yeah. This is absolutely true. The GRE is a crapshoot test you have to pay to do. Standardized testing of all kinds is pretty dumb, but the GRE takes the artifice around standardized testing to a new level. They made me roll up my sleeves and dance around just to get into the room where I would take the test. Why? Maybe I had hidden a slip of paper somewhere that would remind me how to structure an argument in 25 minutes? I don't know. The security theater is really strong with this one. And even outside of the GRE, you have to pay for the privilege to apply at all, in most cases. That stuff adds up, and makes even just applying to grad school inaccessible to people. That shouldn't be how it works.

Is there anything else?

Probably. But I've been writing a while and I'm more tired than usual. I want to be done. I want to come home and work on something real. I want to earn a real paycheck. So, for now, I'm done. And maybe I'll come back to this eventually, or maybe I'll come back to academia someday. But, man, I ain't getting a fursuit on this stipend and that's only becoming more and more a dealbreaker for me as time goes on, lol.