Thursday May 12th
Happy Women in Mathematics Day!
Yes, it’s that time!
- It’s this day because of Mathematician Maryam Mirzakhani, a Fields Medalist.
- I have not been always been given the privilege to study Mathematics at the graduate level, but even before I did, I knew of her because of having read about her, and seen her impact in places like MSRI, which I still agree (will fight anyone who disagrees on?) is the Maths utopia (at least) of the West Coast.
- Outside of the Programming Languages Research Academic community (which is quite wonderful and supportive), MSRI was one of the first grad school, academic communities to also welcome me with open arms. So it’s little wonder that I ended back there today, in a virtual seminar celebrating the day, and also attending a session on “How to do Research”. I met a bunch of Mathematicians, including ones who do interesting things like Quantum Algebra, and of course, Combinatorics (specifically, Matroid Theory). And interestingly, some of them started out just like me; I met someone who had a past career in Mechanical engineering, for example. It was really fun! I had to leave early, however, because I have been working on a research project since this Monday.
- But this was not before asking yet another professor if I could sit in on their Graph Theory course in Fall, because if I could take all the Pure Maths classes, I definitely would! :) And the answer was yes!
- I’ve been reading a lot, and writing proofs for a research project I’m working on that started this Monday (but that I have been musing about since March). I’m enjoying it thoroughly, and it can only be described as an “unlearning” of academic habits, and the equivalent of joining a pirate band of researchers. And it’s just as fun as it sounds!
- One of the boards we used for our collaborative session at MSRI, entitled “How to Build a Career in Math”: Made with the best of intentions (and yes, there were quite a few Analysis Mathematicians there).
Why did this happen?
- I’ve been hanging around a bunch of different persons at various stages in their careers, and I’ve begun to realize that:
- I love research, so why should I let an institutional structure prevent me from doing what I love?
- I can find a way to do research that isn’t at the mercy of the traditional academic system
- As much as I’ve been learning how do to research, I can rethink why I do research and how to publish and who to publish with (sorry I ended with a preposition, English professor!)
Enter Pirate Researchers
- In spaces like Mathematics, Crypto, the Social Sciences, etc, it’s very typical to find collaborators in the most unlikely places. A friend of mine had said that he wasn’t sure he cared for Academia, was rejected by some of the most immediate academic structures, went off to his community and worked for a bit in that doing research, and came back to those very institutions falling all over themselves to hire him. So he made his own way, and was able to renegotiate from a position of leverage / power.
- In my first year, I was working with a student in Computer Science (CS) and Economics, and an Economics professor. In fact, in the entirety of my career, I’ve preferred to work with people I just work well with, which has usually been with a range of skills. For example, in theatre, I did lighting design, but I would have to work with a director, a stage manager, a sound engineer, etc. Similarly, in film, I would have to work with (if I were working in the camera department), sound, grips, electric, location managers, assistant director, art director, etc. You’re working towards a common goal, pretty much, and each one is contributing based on skills that they are experts in.
- I was also writing pretty regularly with a group based in New York City that included researchers in Art History, Fine Art, Literature, etc.
- And for my stint in robotics, I would have to work with engineers who enjoyed mechanical or electrical, and contribute with coding.
- I think research is teachable. It’s a skill that you exercise the muscle of. It reminds me a bit of that part in King Richard (the movie that you should definitely see), where a coach tells the Williams sisters’ father, Richard (who the movie pays an homage to), that it’s too technical a sport for his daughters to learn. But they do it anyway. Granted, they are prodigious, but the point is that something like research is not outside of the reach of anyone who really wants to pursue it. Institutionally, we put up these walls that are exemplified by a lack of support, while not having those for others, which results in the perception that some people are able to learn research, and others cannot. Of course, by all means, if you don’t enjoy it, that’s something else (and of course the more you do it, like anything else in life, the better you get at honing your skills), but I don’t think, for anyone who has a passion for it, there should be obstructions for the opportunity to at least try to see if you like it.
- So anyways, in my first meeting with my current mentor, his biggest hesitation about me was actually that I might be too academic about it.
- I hilariously started our conversation by nerd-sniping him, which made both of us burst into laughter. But fortunately, academia hasn’t come easily for me, in that I have had to hack my way in, and am not a typical student. I am well steeped in the space, so there are definitely ways that I think and operate that I do have to be flexible about and unlearn, but I am excited to do this. I am committed to running away and joining the circus…to follow my dreams. There might be anti-canonical-bespoke unicorns!
Open Source Publishing
- The way we are approaching this is that our paper will be developed online, in an open way. Typically, research papers are treated like IP, which can be really intentionally exclusionary (I’ve experienced a lot of that in the research community!). Groups in research are really secretive about not showing anything, and I wish we could rethink some of this. It can create this hyper-competitive, toxic culture where we optimize for solving only certain types of problems, which is also at the detriment of just doing more interesting research and thinking of problems with different perspectives. It also has led to a lot of issues in fields that use things like medical data, for reproducibility (with respect to journals and profit margins by companies that hold the data, and the researchers who would like to build upon research). And we like open source. Typically, in academia, persons will only publish papers when they have been on arxiv, or as specified in some way by a conference (eg. linking to open code for review), and are less likely in general to even mention that they are speaking about their work openly otherwise. But the three of us will be working on the project openly, so anyone can see our progress. I think this is an awesome approach to publishing, and I’m excited, although it’s also a bit terrifying.
- We do have other documents that are more casual, but this is a great, collaborative, open-source way to working.
- This will be my first time; like most academics, I have mostly published with the intention of submitting to a journal or a conference, but to be honest, I really don’t care that much anymore about that because it’s, quite frankly, hurt me as someone who is not at a prestigious institution or a large lab; I like collaborating, doing research and solving hard problems, and writing papers. So this is an experiment for me.
- Logistically, this also makes sense because we’re simultaneously working on code in another repo(s), that we can easily link to in github issues and work back and forth.
I have never been happier
- I had an amazing week! I met my research friends, I made friends with a group of over 80 people, I received Kudos from my ICLR group of newly-minted researchers (who I am quite excited to see as regularly as I can, as well as their progress in research), I participated in an interview, a podcast, went out for pizza with Pure Maths friends and heard the table erupt into talking about sheaves and cohomology, I saw pioneer Ted Nelson at Stanford speak about his inventions (such as ZigZag and Xanadu), and about how he fondly remembered Doug Engelbart and I went out for ice-cream with a friend.
- I had a conversation this week about how I love hearing about these inventions that aren’t quite in the spotlight anymore. I think to do research, you should understand the genealogy of your field. So for programming languages, learn about COBOL, Algol, Sisal, Lisp, etc. Attend HOPL. One of the things I have enjoyed about the Pure Mathematics field that I was not quite getting from ML (that’s Machine Learning, not Meta-Learning or Meta Language) was that they regularly look at papers with good ideas and do not just focus on recent work. They work on problems that are a century old, that are interesting. I think that’s cool.
- As a researcher, one of the earliest things I learned was from a professor who has such a depth of understanding of the history of his field that he reguarly came up with ideas that were mind-blowing, which he confessed were thinking deeply about paradigms of the past that other researchers had lost an interest in (or did not know about). We focus so much on the new, the inventions or research that are the most influential, that we miss that there are so many great inventions or ideas in research papers considered to be not as influential, or in inventions that were “failures”. And part of the joy of doing research is getting to (re-)discover those things, to revive beautiful ideas, regardless of their era. Maybe some of these ideas were revolutionary, but were just not proposed at the right time; the world was not quite ready to listen. Perhaps this is the right time, and you are that person to re-introduce this generation to those perspectives and ideas. So anyways, I used to hang around the Computer History Museum quite a bit when I lived on the West Coast, and I love so much hearing about ideas from the history of Computing. I think it’s awesome. When you look at the gaps in time between important inventions, it’s mind-blowing to see how creative and different the ideas were, and how persons tried to think about these problems that had not been solved yet.
- So on we go. Helms-Alee!
And that’s it.
Written on May 12, 2022