Claire Le Goues Interview
Women@SCS: Could you tell us a little bit more about your background and your journey before you came to CMU?
Le Goues: Okay. Well, I grew up outside of New York City, went to public School. I went to Harvard as an undergrad, so I was there for four years. And then I took a job as a software engineer for about a year and a half after that, and then I went to grad school. So the story there is that I actually went to college thinking I wanted to be a lawyer, mostly because my sport in high school was debate – and debaters become lawyers. Although, actually, a lot of what I liked about debate ends up applying more readily in science than you'd expect.
Le Goues: By a lot, actually, because debate is very much about a construction of an argument, and so is science in a lot of ways – like a research scientist is like here's a problem. We believe it is a problem for this reason. Here's our solution. This is why we think it solves it. It's very similar in structure. So I went to Harvard and I started majoring, basically, in social theory, but I had done computer science in high school and I had some friends who were computer science majors, and I was sort of increasingly dissatisfied.
I mean, I still enjoy political science. I read the news. I keep up on these things. I find law interesting. But the actual sort of mechanics of classes were not very satisfying; whereas I had friends who were doing computer science. Basically, I was like nerding out by having them tell me about their homework, and I was like, "Maybe I have made the wrong choice."
So I switched into computer science, but it was a little bit late, and so by the time I got the chance to do research, I was a senior, and by the time I'd decided I wanted to go to grad school, it was too late to apply, so I was like, "Well, I need a job 'cause I gotta eat."
Women@SCS: Right. Yeah.
Le Goues: So I got a job, and then about a year later I applied to grad school. And so then from there I went to the University of Virginia, which is where I was for six years. I got my master's and Ph.D. there, and then after that I came here.
Le Goues: So I have been here six years and change. This is my seventh year. So that's it. I've lived in many commonwealths somehow. All the states I live in have been commonwealths since New York.
Women@SCS: Wow. How interesting.
Le Goues: But that's the trajectory.
Women@SCS: Huh. It's like a very off the beaten path.
Le Goues: Yeah. I mean, I know I'm not the only person who is like, I didn't know I wanted to be a computer scientist. It's probably unusual for CMU, where you have to know you want to do computer science to be in SCS.
Le Goues: But it was not the direct like, "Oh, yes. I went to college at 17, knowing I wanted to be a computer science professor." I mean it was also, you know, I graduated from college in 2006, which is basically a lull between bubbles, so my graduating class at Harvard only had something like 28 computer science majors. Now, I mean Harvard's not CMU, but now the intro to CS class is the largest class at Harvard by a fair bit.
Le Goues: Which is wild, because when I was there, it was economics. And then, the year I went to grad school was like six months before the financial crisis, so I went to school at kind of a weird time, right? Even when I went and looked for a job, our job fair was not set up for me. It was all banks and hedge funds and stuff, because they all like to hire Harvard grads.
Le Goues: And I was like, "I would like to be a software engineer," and the jobs people were like, "Hmm." Like they didn't quite know what to make of me. So, yeah, it's funny now. But if I'm trying to relate advice I'm giving now about job searching in industry to when I was on the market – it's a completely different universe. The faculty market's changed a little bit since I was on the market, but it's not as dramatic a difference I think as when I was looking for a software engineering position. I found one, right? It wasn't ridiculously difficult, but it wasn't like today, when it's just wild. So it's probably different from a lot of stories you'd hear today.
Women@SCS: Yeah. I'd say so. Were there any times where you felt panicked or nervous during that whole realization that, hey, maybe what I originally pursued isn't what I wanted to do?
Professor Le Goues: Oh, yeah. Well, I did have second thoughts at some point. When I'd just graduated and I was thinking of applying to grad school, I kind of had this freak-out, like "Maybe I should have gone to law school." So I actually took the LSAT on a whim, but then I got over that real quick. I was like, "No, no, no. I made the right move. It's fine."
But it's actually kind of a fun test, which is sort of a weird thing to say, but it [was] very low pressure for [me]. If you actually care about going to law school, it's a very high pressure situation, but if you're doing it 'cause you're having a crisis at 22... You know those puzzles that are like "You're at a dinner party, and Tim is wearing a tie, and someone's wearing a hat. And who ordered a steak?" That's an entire couple of sections of the LSAT are just puzzles like that, which is kind of fun if you're a computer scientist.
Professor Le Goues: If you're a computer scientist and you don't care that much, it's kind of a fun test. So, yeah, I did have a bit of a freak-out. And, actually, I think Harvard was a good place for me because I do have all these other interests and I had the chance to take a lot of really cool classes as electives, just to sort of be enriched, but still a very rigorous computer science curriculum. So it ended up being a good choice even though I didn't know that's what I wanted to do when I went, ultimately. So, yeah, I did have moments of doubt – and you always have moments of doubt. Or at least I do.
There's a moment half way through grad school when everyone has this freak-out, like "What am I doing?" So, for me, it was good that I had gone and worked for a bit because I knew [that] I enjoyed my job. It was fine. But I knew like, okay, this is not what I want to do long term. I definitely want to go into a research career. So when I was in year three or four, when every grad student has that existential crisis, I was like, "No, no, no. I know that I have done the other thing." A lot of people go right to grad school from undergrad, and when they have that existential crisis, it's a very grass-is-greener kind of situation.
And so I'm not saying people shouldn't go right to grad school, but I'm saying there's benefits to not, because when you have your crisis you can move on more quickly. Sort of future proof here. Mental health. I mean there's always been self-doubt in life, but I think I've turned out okay. I'm happy.
So, I came to CMU, I didn't think I was gonna get that job. So I was actually again pretty relaxed, actually. Sort of like the LSAT. I was like, "Oh, wow. It's really cool. I get to interview at CMU." And what I do in software engineering, it's an amazing place to work and really cool people to talk to. I figured they'd never hire me, so I was really relaxed the whole time.
Like they'd ask I’d respond, "Wow that's a really interesting question. I really want to think about that." As opposed to like self-doubting and worrying about saying the right thing, I was just like, "Wow, that's so cool. I get to talk Mary Shaw about the implications of my work. That's so great." And I wasn't worried about it 'cause I didn't think she was gonna hire me anyway – not like Mary specifically, right, but I wasn't worried about the implications of the conversation 'cause I figured it didn't matter.
Women@SCS: Right. It's ironic.
Professor Le Goues: Yeah, it is, right? I think it probably helped. [Laughs]
Professor Le Goues: That's how I ended up here.
Women@SCS: Yeah, that's super cool. And since your time here, you've taught a lot of different types of classes, from undergraduate to Ph.D. courses. Do you have any thoughts? Or what's your opinion on the differences between the two different levels? Anything that you found surprising?
Professor Le Goues: That's an interesting question. So, yes, I've taught actually three levels. I've taught undergrad. There's a master's class I teach sometimes, and a Ph.D.-level class, mostly in program analysis, because my research is largely applied program analysis. In recent years we've actually done that as a joint undergrad/Ph.D. offering – for a bunch of reasons – so it's actually been an interesting mix of teach it just as a Ph.D. class, teach it as an – I didn't teach it as an undergrad class, but my colleague, Jonathan Aldrich, did, and then we sort of combined forces.
And then I do teach Foundations of Software Engineering, which is 17-313. And I taught the first iteration of Software Engineering for Startups with Michael Hilton. He's now teaching it with Heather Miller. She's a much better choice than I am because she has considerable experience in industry and with start ups and so on. So I taught the first version and then I went on maternity leave and Heather started. And so I said to Michael, "I will not teach it with you. I'm on maternity leave." And we were like, "Oh, Heather would be a great choice." And it was instantly obvious that Heather should just do this, and not me, from now on. We're like, "This is correct," because she has so much experience. She knows so many contacts in industry. It's a cool class. You should take it.
But, I teach different material for the different levels, so that's thing one. And so, actually, the different classes end up having fairly different flavors aside from the audience, although the audience does matter. So the master's class I teach is in the Masters of Software Engineering Program, and most of those students have between three and five years of industry experience, so they have a very different perspective. But it's also a much larger class – it's more like 50 or 60 students – than Start Ups, which is an elective, an upper-level undergrad elective. And it's new, so it's not gonna be 50 people yet. And so one's more of like a seminar conversation kind of feeling, whereas the other is actually much more lecture content. It's a required in that program.
The Ph.D. level – so my goal with program analysis at the Ph.D. level: it's a subject that you can do research in proper; you could do research in program analysis. But it's also a very useful subject in many other disciplines that you end up using, or you can use the techniques – reasoning about programs in HCI, for example, if you do end user programming. There are lots of people who could have some fluency in program analysis and their research would be better. So my goal in program analysis is to, if you want to do program analysis research, give you the fundamentals. If you do something else, but it might be useful, give you some of the basics so that you can read papers and know how to apply some of the ideas. And, basically, it's all about enabling research in this area. So that's kind of the focus at the Ph.D. level.
Mixing that with the undergraduate level is sort of an interesting balancing act because I'd love to expose undergrads to research as an idea, but it's not like a seminar class. It's a lecture class that's about techniques and fundamentals and how you do this, and it's really about how you think about programs and reason about their correctness. And that has applications no matter what you want to do with your life, I think, if you're in computing.
And so finding a way to simultaneously have multiple paths, right? It's kind of choose-your-own-adventure class. If you want to do research in program analysis, I want this class to be good for you. If you want to do research in HCI but you're working on developer tools, I want this class to be useful for you. If you're an undergrad who wants to just understand what the tools that you're using in your internships – what are they doing, right?
'Cause if you push a commit to a code base at basically any company, there's gonna be some static analysis run on it, and it might yell at you about something. And I think that being a practicing engineer, like being a computer scientist, you should know what your tools are doing. It's not a black box. And so having some understanding of how they work and the guarantees they give you, that's useful to you.
So it's an interesting act to try and have this like multi-path single class that's hopefully accessible to everyone. So I think we are always tweaking it. I don't know how much we succeed, but that's the goal. But, yeah, I've had undergrads who wanted to do research and who do research with me. I've had undergrads who want to go to industry and want to learn tools for that. And I think, hopefully, we manage to hit that balance.
The engineering side is much more the focus in Foundations of Software Engineering. So we actually don't talk that much about research in that because that's geared more towards students who are interested in being software engineers in the world, and so it's, again, a different flavor of class. It's not about research fundamentals. It's about engineering fundamentals – which there is research behind that, but it's a different kind of flavor of the material, I think.
So program analysis is a lot about what kind of meaning can we derive from the program and what are the limits of how strongly we know those things. Because you can't perfectly prove any non-trivial fact about a program, like that's known, but you can talk about bounds on what you can prove and what you can know. Software engineering, the answer is always "it depends." Like how should we do this? Well, it depends.
Professor Le Goues: Well, what does it depend on? And then that's where the conversation ends up going, which is a completely different flavor versus like program semantics, which is where analysis starts from. Software engineering is a lot more like – well, it's a short answer – basically any non-trivial piece of software is too large for a single person to understand, and we have to manage that complexity somehow. So it's a lot more about techniques for managing complexity and for architecting large systems that you know you won't actually fully understand, and have confidence that they will be correct.
One way you manage complexity is process. There are technical issues that we talk about, like how do you write code in this way, but there's also process issues and how do you work with people. It's just a very different subject matter. That was a long answer to a short question.
Women@SCS: No, definitely.
Women@SCS: Definitely both cool sounding classes.
Professor Le Goues: I think so. I like them. I'm also lucky. I teach classes – I mean except for the master's class – that they tend to be electives, so they tend to be students who are interested or have some sort of reason to take it besides the fact that they're being made to.
Professor Le Goues: I mean program analysis does satisfy a logic and languages elective area in the undergrad program, but that's not like you have to take it. It's an option. Right?
Professor Le Goues: Similarly, software engineering fulfills a bucket. But none of them are like you have to take them, and so it tends to be students who are interested, which makes it an easier teaching experience.
Women@SCS: Yes, that's true.
Professor Le Goues: 'Cause you know off the front that people are there for a reason.
Women@SCS: Right. Awesome. So, on the subject of research, could you talk a little bit about the current research projects that you're working on?
Professor Le Goues: Sure. My lab is called squaresLab. That is unapologetically bacronymed to Software Quality in Real Evolving Systems. So what we do is study the quality of systems, like large software systems that are undergoing constant evolution – which characterizes a lot of software – come up with ways to find bugs in them, argue that they're correct, or – and this is a non-trivial component of the research – fix bugs in them automatically.
The research revolves a lot around: how can we take these real systems that are complicated, that interact with the world, find bugs in them, provide feedback to developers to help them understand the bugs? Can we study how bugs appear in systems and how they are fixed, and learn something about that to help developers find them or prevent them? And can we integrate tool chains into existing QA to help developers fix bugs faster? So that's the umbrella story.
We study things like open source software systems that have bug reports in them. We also have an interest in robotics, software processes, and QA for robotics. So this is different from people who use formal methods to reason about robotic systems in CPS. We don't do that. We deal with the messy C++ that people are hacking together against ROS [Robot Operating System] and figure out how to help them find bugs in it faster without having to fly the robots into the side of a mountain, you know.
Which is kind of an interesting problem, because we talk to roboticists and there's a lot of like, "Oh, you can't find a bug until you build the robots." And is that true? We think the answer’s no, and we're trying to help roboticists find bugs faster in their messy code. So it's not a formal proof thing. It's sort of process stories and analysis story. And there's obviously room for formal methods in that, but that's not what we do, for various reasons. So that's the whole umbrella of everybody, I think.
Women@SCS: That's pretty cool. What's your favorite part and your least favorite part about your job?
Professor Le Goues: My favorite part is definitely – and this is gonna sound like I'm catering to my audience but I'm not – interacting with students. And I mean like everybody, so my Ph.D. students too. I have a big group, and I like mentoring. I like teaching. I like teaching undergrads. I like teaching grad students.
I like doing research, but I like doing research with people. I like that I'm never the smartest person in the room, and I'm always learning from my students. Computer science is very collaborative, and so I work with my colleagues at the professor level, but I'm also collaborating with my grad students, and they're always learning something new or bringing something to the table or forcing me to learn something new.
So I had a student who took a class and he learned all about game theory. He thought it was really interesting and saw ways to pull it into our research together, so now I'm learning about game theory. I had a student who did basically the same thing with separation logic, and now I know exactly 50 percent of separation logic – the 50 percent we needed. [Laughs] So I really like that.
My least favorite part? There's a lot of service stuff. And any individual service thing I don't mind, usually. There's not one kind of committee I really hate going to. It's just that there are many demands on my time, and it can feel difficult to be able to do as good of a job on the things I'm really passionate about as I'd like to necessarily because there are many other things I have to do. I think that may be characteristic of a lot of knowledge work in the 21st century. I don't know if that's specifically unique to academics. I mean we do have a lot going on.
On the one hand, I really like that I don't have a boss telling me what to do. I really kind of row my own canoe. But on the other hand, it means I don't have a boss to protect my time. I don't have someone coming in and saying, "Claire, you're working on the wrong thing." That can be very validating, right, if someone says, "Go work only on your research because that's the most important thing, and I release you from this other obligation."
Professor Le Goues: There's no one to tell me that, really, so I have to do that myself, and that can be a little tricky. So it can feel a little bit like the overwhelm is real. I mean there are things that are strictly unpleasant, like academic dishonesty issues. We haven't seen many of them, but I don't like dealing with them, you know, because it's unpleasant.
Professor Le Goues: So, yeah, it's not that I'm like, "Oh, I hate being on committees," 'cause I don't. I don't hate writing grants. It's just that there are only 24 hours in the day and I like to spend some of them doing things besides work, so managing to find a balance is always kind of hard. But I don't know how unique that is to academia these days.
Professor Le Goues: I think not having a boss probably doesn't help much, because a good manager can help you manage your time sort of constructively. And I do get feedback and mentorship from my senior colleagues, but it's not like somebody saying, "Claire, work on this, don't work on that." That's not the way the job is – which is the thing I like about it 'cause I'm not great with a boss. It's why my job is good. But it does mean that managing one's commitments can be all on you more than anyone else.
Women@SCS: Would you mind speaking a little bit more about that - how you manage your time to be able to do all the things that you want to do?
Professor Le Goues: That's an interesting question. I have sort of a collection of techniques that work for me. So one thing I do is I try very hard to protect one day a week to not have meetings.
Women@SCS: Oh, cool.
Professor Le Goues: And I block that with a meeting. And so when people can see my calendar, it looks like I'm busy, right?
I have a group of friends/colleagues that I sort of refer to as my “no” committee. I highly recommend this, actually – where I'll get an invitation to do a thing. I'll be like, "I think I should probably say no to this," but it's really hard to say no, right?
Professor Le Goues: And I get invited to speak – give colloquiums, speak, stuff like that – really frequently. And I cannot do all of them because there's only so much time in the day. But this kind of invitation is good for your career. You get exposure. People see you. It feels nice. People want to hear from you. You're turning something down that you want to do. But I have a twelve-month-old, so having a committee of colleagues to say, "No, you can say no. It's good for you to say no because... Let's practice how you will say no." Having friends who are in kinda the same boat can help you say no is really helpful, just 'cause it can be hard to say no to people who are inviting you to do something.
Women@SCS: That's true.
Professor Le Goues: Actually, when I turn stuff down, I tend to note on my calendar when the thing would have been. And so then I can be like, "Oh, man. I'm so glad I'm not in wherever."
Women@SCS: That's amazing.
Professor Le Goues: You feel great. You're like, "Oh, my God. I'm so glad I'm not in Germany right now."
Professor Le Goues: Sometimes I do go to Germany. It's fun. So, yeah, there's that. Like I say, there's not one single move. So class prep can totally take up literally every minute of the week. Class prep takes time allotted, basically, so the trick there is to only allot a certain amount of time, be like, "Look. I'm doing all of my course prep. It's gonna be done by 6:00 PM, and that's gonna be the end of it." And so, yeah, I mean could it be better if I gave it another hour? Probably, but it's not my only job.
Women@SCS: Ultimately it is what it is.
Professor Le Goues: Yeah. So, yeah, I use this kind of collection of methods to carve out time during the week and make sure that I say no appropriately. It has gotten easier, honestly, since I had the baby, to say no, because basically the choice seems more obvious: I either need to give up – I need to give up a Sunday to travel to this commitment. Do I want to do this enough to give up a Sunday?
And that was always a question. Before I had the baby that was the question too: do I want to do this enough to give up a Sunday? But a Sunday is much more meaningful to me now when it's like our family time on Sunday. And so there are still Sundays that I'm going to travel, right? It's the nature of the job. But it is easier to hold onto my time now than it was.
And there are times when you do get overcommitted and you just kinda soldier through, and you're like, "Remember not to do this again." That's the thing about academic work: it's very cyclic. There are times that it's actually pretty gentle, where I can leave a little early or I can go get my manicure fixed. And then there are times when I'm working in the evenings, you know, past bedtime. And it's fine.
As long as it's not all the time, as long as there's a balance, then I think it's okay. I don't mind the nature of work to be like that, so it doesn't bother me. I have colleagues who work more like that and I have colleagues who are like, "Nope, I'm never working past 6:00," and you just kinda do what you have to do to make it work for you.
Women@SCS: That's all great advice.
Women@SCS: I love the tip of putting declined events in your calendar.
Professor Le Goues: Yeah. And then you're like, "Oh, my God. I'm so glad I'm not on a plane."
One piece of advice my mom gave me was, imagine for anything you're being asked to do – usually you're getting asked like six months in advance for whatever in life – pretend it's next week. Can you do it next week? Because you're gonna think, "Oh, it's six months from now. I'm not doing anything."
Women@SCS: But for the first five months, you're not gonna think about it.
Professor Le Goues: You're not – but it's gonna be next week at some point, and then are you happy to do this next week? And that's actually a really helpful heuristic.
Women@SCS: Awesome. On the flip side, then, is there anything that you like to do when you do get spare time? Any hobbies?
Professor Le Goues: I do. I used to do roller derby, actually. When I started, that would have been my answer for you.
Professor Le Goues: I did retire because eventually it got to be a big time commitment. I played a lot in grad school, and then I played for a couple of years when I got here. And it's really a great sport. It's a lot of fun. It's a great way to meet people. And especially when we first moved and I didn't know anybody in Pittsburgh – it was kinda lonely – it's a nice way to make friends. But the time commitment to stay competitive relative to the job just got kind of excessive... And then I actually managed to play for six or seven years without sustaining serious injury, and so I thought that was a sign that it was good to retire before I like –
Women@SCS: Oh, like quit while you're ahead.
Professor Le Goues: Quit while you're ahead. Exactly. So now, actually, a lot of the hobbies are really much more home oriented because we have a small child, and I actually really, genuinely enjoy her company, so most of our weekend times are like, "Oh, we're gonna go to the swings."
Professor Le Goues: And we tend to have people over for dinner more because it's easier than going out. We used to go out a lot. We were like big food snobs and stuff, you know, did theater in Pittsburgh, stuff like that. Now it's a lot more having people over and hosting people at home because it's just easier. So I imagine I'll pick up external hobbies again eventually. It's not like I don't do anything. I mean I read, right. I do stuff like that. But these days it's more family oriented because that's the part of life I'm in. And that works for me. We've got a lot of pets. We have a dog and two cats and three chickens.
Women@SCS: Wow! Chickens.
Professor Le Goues: Well, there are two chickens right now. It's a long story. They live outside. And a parrot who is on sabbatical.
Women@SCS: [Laughs] What does that mean?
Professor Le Goues: It means that the parrot is living with my father-in-law because parrots, African grays can live for 40 to 50 years.
Women@SCS: I've heard that.
Professor Le Goues: And they're very smart. And they pick up things that get attention. So if you have an infant, for example, who cries to get attention, the bird can learn that and then do it for the next 40 years.
Women@SCS: Oh, gosh.
Professor Le Goues: So when we were talking about having a child, I was like, "Okay, we need to come up with a plan."
Professor Le Goues: And so it's great, because now the baby's learning to do new things with her voice, so she learned this really awful shriek a couple of weeks ago.
And I am hormonally predisposed to find everything she does adorable, but I was thinking like, "Oh, my gosh. I'm so glad the parrot's not here learning this."
Professor Le Goues: But we still have the two cats, the dog, and the chickens, so it's still fun, and that keeps us busy. I do love to travel, actually. That's another thing I really like about my job: I get to travel a lot. And so my spouse and I have taken a lot of trips together. So that's not so much a hobby, I guess, but it's something I enjoy doing. Yeah, I sound very baby oriented, but I really like her and she's great.
Women@SCS: She definitely sounds like a cool person.
Professor Le Goues: Yeah. I have been getting back into programming, actually, because I sort of dialed back on interacting with code between being pregnant and having a baby, 'cause it turns out I can sort of act like a middle manager for a little bit if I have to and have my students do more of the heavy lifting on the engineering side. So now a little bit more for fun, I've been like, "Okay, I used to be decent at this. Let me get back in this skill set so I don't lose it." So that's been actually kind of a little bit of a hobbyist thing, like, "Let's get good at Docker, as opposed to just competent enough to get through it”. So there have been a little bit more of that lately, 'cause it turns out I do like to program, and you need to keep doing it so you don't lose it.
Women@SCS: Yes. Could you describe your experience as a woman in the CS field?
Professor Le Goues: Yeah. So there are some that are specific to being an academic. And so it's, for example, I'm commenting that sometimes I have more invitations or obligations than I can accept. This is one thing that's a known issue for women in academia – and this'll happen in industry as well – is that women are asked to do kind of service-y things more often.
So one reason for this – and it's all very well intentioned, actually. It's like, okay, we would really like there to be good representation on our program committees, right? We don't want them to just be over-represented majorities. And so the problem is if you're aiming for 50 percent representation on every committee in a field where only 20 percent of the people are women, women get a lot more invitations. So this is a challenge in terms of managing my time because it's important that I do that kind of service, but that's not the only thing that matters in my job success. So that's a sort of a specific gender-based thing.
I'm trying to figure out how to say this so I don't sound cynical, because I don't mean it cynically. My sense is that it can be challenging to be a woman on earth, in society.
Women@SCS: Yes. Agreed.
Professor Le Goues: So that's kind of my baseline. So, yes, there are specific challenges. There is lots to be done to improve treatment of women, the experiences of women, career trajectories of women in tech and in academia. But to me it's like, well, I would like to work to make that better and I would like that to be better, but it's not like I'm giving up some hypothetical perfect job where those kinds of issues don't apply. And so it's kind of like, yeah, it sucks. The world is a difficult place for many people. And I'm also a heterosexual, white cis-woman. How hard is my life, right?
There are lots of things to speak to about being a woman in this field, in academia, in tech. I feel like I'm trying to use my position. I feel like I have a lot of privilege, and I can try to make the world better as much as I can now that I've gotten here. I am stubborn and sort of angry, and so that's gotten me here, you know. It's nice that we've been hiring more women at the junior level.
Because we're building community, I think, in SCS – the faculty and certainly the junior women. And I don't think I realized how awesome it is until I'm going into a meeting and I'm not the only junior woman in the room. Right?
Professor Le Goues: I was on a phone call working on a big grant proposal with a number of colleagues, and actually Heather [Miller] was involved and she was on the phone. And I think it was the first time I'd been in kind of a grant preparatory discussion meeting on a phone call where I had to distinguish who I was. Because usually when I'm on a call, I'm the only woman. So the men have to do this all the time. They're, "This is John speaking," or "This is Mike speaking", right, because they all sound the same. And this time we had to be like, "Oh, this is Heather," "This is Claire." And I had never had to do that before.
And so I end up having moments like that, where I'm like, "Oh, wow. Our world is super weird and bad," but, at the same time, you kinda gotta work with what you got. I feel like CMU has its challenges with respect to diversity and inclusion. But in a lot of ways I have great mentorship, and I think there are ways in which the university is actually quite doing a great job on that. There are ways in which the university needs to work on it very hard, for sure. But it's not like an unequivocally negative place. We're working on it.
And so while I think I don't want to let the university off the hook here because there's lots that needs to be done on this – I mean I have a list of specific things I think we should do as an institution to improve things. It's also not like a disaster either. So I'm lucky to be surrounded by lots of allies. Like many of my male colleagues, I have found are very cognizant of these issues, really take them seriously. I really value that in my male colleagues, especially in this department. So that was kind of a long answer to be like, yeah, some stuff sucks, some stuff doesn't. My job is rad, so that's cool. I do my best. I have a very supportive partner. That helps.
That's kind of a touchy thing to say, though; I'm not saying you need to marry a certain way. I'm just saying the level of support you get from your partner if you're a high achieving, very ambitious woman, especially – a heterosexual woman, especially – really does matter in terms of career success. Or at least it can have a really significant influence, and mine has. So I have all these elements in my life. Again, there's no single answer.
Professor Le Goues: But I did – for example, when I was going on maternity leave, I was quite vocal about, okay, I am turning down this service request because I'm going on maternity leave. And I did that on purpose because I feel like I'm at a place like Carnegie Mellon, which in software engineering is a very well established place. And I feel like if I can't be vocal about this and sort of make this space for people, then who can? You know, there are women who aren't in as supportive an environment. There are women who aren't in as esteemed an institution, who might not feel comfortable saying that. And so I feel like it's important to model like, "Look. I am not doing this thing 'cause I am having a baby, and you need to ask someone else." You know?
Professor Le Goues: Which isn't gonna solve everything, but I'm trying. So, yeah, everything could be better, but I guess everything could be worse.
Women@SCS: That's true.
Professor Le Goues: One answer is you gotta be a little bit of an optimist, I guess, to get through it.
Women@SCS: I agree.
Professor Le Goues: So that's why my answer is kind of like, yeah, there's a lot CMU could and should do, but also I'm not miserable in the day-to-day. But I do feel more comfortable calling stuff out now than I used to. I think that's a function of having been in the career longer, I end up feeling more comfortable calling people on their BS, whereas I think at first I certainly wasn't. And I don't think that we should leave it to women to call people on BS.
Women@SCS: Of course.
Professor Le Goues: But I do think that if one feels comfortable doing it, it's not a bad thing. So that's where I am. I'm in the – just transitioning to the BS calling phase of my career, I guess. Like we're just there now, so ask me again in five years how I feel about everything.
Women@SCS: It's really great that you're using your spot, leveraging that spotlight, to make things better.
Professor Le Goues: I'm trying. I mean I guess what's one person gonna do, right? But, you know, I'm trying. I'm trying to raise stuff when I can. There's more that the university should do, that everybody should do. But there are things that we're good at, so I think it's not bad to appreciate those things, and so that's kinda my approach.
Women@SCS: Very true.
Women@SCS: You mentioned that you received some mentorship and such. Do you want to speak a little bit more about the REUSE at CMU program that you do?
Professor Le Goues: Yes, I do, because I am obsessed with it. REUSE is an REU program in interdisciplinary software engineering. An REU program stands for Research Experience for Undergraduates. It's funded by the National Science Foundation because about a third of their mandate is to grow the technical and scientific population of people in this country.
So they want specifically to fund people to become scientists because it is true that you need to get more people in to get more people out. Retention also matters, but the beginning part matters too. So this is some of the things that the NSF has invested in.
And so we are actually – we do typically have at least one or two CMU students involved. But we are largely drawing students from schools where they do not traditionally have access to research opportunities, such as small liberal arts schools that don't have a large research presence. These might be regional state schools. We've even had a community college student or two come through the program, because the idea is to expose students to research who wouldn't traditionally have access to that.
But because we follow basically a cohort model – which is to say everybody learns from everybody – it's good to have a mix. So, of course, we also have had students from top research institutions as well. We've had CMU students. But the point is to have this kind of blend of people from different levels of experience, from people finishing their freshman year and people finishing their junior year.
The goal there is basically to expose students to the whole research process in a nutshell in ten weeks. And I do think we do an okay job. It's tough to do everything in ten weeks. And my vision there is that we have a really hard time convincing students to stay in research and to stay in academia, and a lot of it – you know, there are many factors at work here. There's a very hot job market right now. A lot of students are graduating with a lot of debt, and so they want to go out and make money. A lot of students are tired of school. Right? But there's also a sense in which – especially at CMU – you're getting recruited out of sophomore, junior year.
Professor Le Goues: You never have the chance to think about research as an opportunity. And so a lot of students everywhere, all across the country, are trying to secure internships earlier and earlier so as to secure a full-time position because that's how a lot of the hiring works now, and they just never stop to think about research as an option. So my goal is, okay, I would like students to see what research is and then make a decision. If you come out of the summer and you're like, "Oh, I hated this and I do not want to do it," I'm like "success" because you have made an informed choice. Right?
Professor Le Goues: I mean unless it's like we did a bad job. But I'm interested in students making an informed choice and getting exposure to research, and also reaching students who might not normally think about research as an opportunity. So there are a bunch of things we do to try and facilitate that – not worth going into all of the nitty-gritty, but we're up to about 25, 30 students per year right now.
A number of them have actually come – and are now Ph.D. students at Carnegie Mellon or live in Pittsburg otherwise. So they come to the city and they enjoy it, and we have several graduates working in the city post-graduation, which I think is really cool.
And, yes, we want to grow. We're trying to expand the areas that we're working in. We're trying to raise funds to have more students, because I think it's a great way to bring people to CMU, expand our research offerings,get some of the great students involved, and it's good for the community. I love it, so I can talk about for a long time. I run that with Josh Sunshine, who is special faculty. He's a systems scientist in the department, and also Charlie Garrod, who is a teaching faculty member. And also Michael Hilton, who is also teaching faculty, who joined us a couple of years ago, has been helping us co-direct as well.
So it's very much a collaborative effort, because it's a huge amount of work between setting up the research partnerships. The individual students are advised with other faculty members, so I'm not advising 25 research students. Those students are paired with individual faculty, different projects throughout the department and across the school. We have other SCS faculty involved who aren't just in ISR – that's the interdisciplinary part.
And so doing that, training the Ph.D. students who are working with them, coordinating with the housing office, making sure the money happens – you know, the whole thing. There are several grants involved- it’s a whole lot of pieces, and so it's very much a joint effort. I am absolutely not individually responsible in it and I want to make sure that everybody gets credit, because it's very much a collaborative thing.
HCII also runs an REU program in the summer, and robotics runs a summer research program as well. We've been talking about with some of the other departments figuring out how to coordinate some of these activities a little bit because I think we have a lot to offer as a school.
Women@SCS: That is so awesome.
Professor Le Goues: Yeah.
Women@SCS: I think that's all the questions that we've prepared. Do you have anything else you want to add?
Professor Le Goues: I really like computer science. You know, for me, I like computers because it feels like magic still, and there are still moments where I'm like, "Oh, this is so cool." I have that feeling every time I use Shazam. I understand how it works. I couldn't write it myself, off the top of my head, but I understand how it works. But it still feels like magic to me. And so I just try and hold onto some of those things a little bit, so when the grind kinda gets you, you'll be like, "Wait, we're working on magic, and that's really cool." And it's worth being aware of the limitations of technology and sort of the limitations of the humans that build it, and machine learning learns all of our biases, right? And so it can get really kind of depressing, I think, sometimes to work in tech.
Women@SCS: Yeah, in the nitty-gritty.
Professor Le Goues: In the nitty-gritty, in the day-to-day, being a woman surrounded by dudes who are not always great. So I try and kinda hold onto a couple things like that, to be like, "Wait, there's a reason we're doing this, and it's great." Or like good exchanges with students.
That's another piece of advice, actually. A little piece – keeping a file of the things that happen to you. I totally have a folder. Maybe you get a particular positive review back on a piece of work. I'm not the only person who does this, but save it so when you're having a bad day or you're feeling a lot of imposter syndrome, you can go look at it and be like, "Wait. I'm great, and someone said so." That really helps.
Yeah. Tips and just about life in general? I don't know. I can say don't worry about grades so much. Get some sleep? With my usual undergrad, I’m like, "Guys, get some sleep. Eat your vegetables."
And don't worry about grades so much 'cause they really don't matter. Ultimately, they really don't matter – I mean, just practically speaking. I can tell you I did very, very badly in multivariable and linear algebra as an undergraduate. And I'm not talking about like B-minus. I'm talking very badly. And it has not influenced my life at all. Everything's fine, you know.
So I think it's worth – remember the people and the experience more than the 2:00 AM grinds. So live a little, I guess, is my actual advice. We do research. It's fun. I teach. Take my classes. They're fun. I have a really cute kid. I've managed to avoid putting her on slides for my class so far; that I think demonstrates personal growth.
Women@SCS: Thank you so much for doing the interview with us.
Professor Le Goues: Of course.