Calling all mathematicians: my experience doing math at a hackathon
“My brain is open!” These are the famous words that Paul Erdős spoke to the world’s foremost mathematicians as he surprised them at their doorsteps. Erdős, one of the world’s most prolific mathematicians, published over 1,500 papers in his lifetime with more than 500 collaborators. Fundamentally, Erdős believed that mathematics was a social activity; after turning up on his colleagues’ doorsteps, he would help them with their work for a few days before departing to collaborate with someone else.
Erdős was not much different than many hackathon hackers today. Hackathons are intense, open-ended competitions where attendees (hackers) work in teams to build new things and solve problems, all while surrounded by free food, swag, and mentors. Many hackers work on their projects all through the night, and most get little sleep — similarly, Erdős often worked 19 hour days, and loved to say that “a mathematician is a machine for turning coffee into theorems.” Hackathons today cater especially to software developers and hardware hackers, and provide lots of tools and resources to help these hackers along through the weekend. But why not include mathematicians too?
This past weekend, I brought three of my most brilliant friends (all math majors) to HackMIT to prove that mathematics could be done at a hackathon. We set out to create an algorithm that could route cars more efficiently in surge traffic conditions; for example, surge traffic might be the traffic ensuing as thousands of spectators exit an arena, or the traffic created as millions of Floridians fled Hurricane Irma. With limited formal computer science education, we arrived at MIT unsure of ourselves but hopeful that we could succeed.
Traffic was an excellent problem to work on, because it can be modeled in so many different ways. Cars can be modeled as particles, waves, or fluids; roads as edges and intersections as nodes; traffic as vector fields and complex systems. We spent most of our time working through equations, comparing and contrasting theoretical models, and researching algorithms. By 8pm on Saturday night, facing a Sunday-at-noon submission deadline, we had yet to write a single line of code. After a long night, we finished our work and managed to hack on a visualization at the last second, before hightailing it back to the judging room at t + 5 minutes.
Although we wrote Python code to demonstrate our solution, it isn’t necessary to have programming experience to do math at a hackathon. Many prominent mathematics software applications like MATLAB, Maple, and Mathematica include high quality visualization tools that are easy to use and highly customizable. These applications even offer interactive visualizations and simulations, affording an additional boon when presenting to the judges. With these tools, problems in fields like graph theory, topology, and dynamical and complex systems become accessible. Moreover, many branches of mathematics share problems with computer science, including cryptography, computer vision, signal processing, information theory, and data science, and solutions in these areas would fare well with judges. For problems where visualization isn’t possible, mathy hackers can resort to Powerpoints explaining their work; slide decks at hackathons are not uncommon and have gained wide acceptance, even for software-based projects. While some might protest that pure math is excluded, pure computer science is also excluded from hackathons; the event has always been grounded in practicality and real world applicability.
While not every math problem is fit for the hackathon, there are still many reasons that mathematicians should attend these events. Lots of sponsors seek out mathematicians specifically, including fintech companies, artificial intelligence startups, and companies with an interest in big data. Math majors will be pleasantly surprised by the high number of recruiters at these events looking for their skills, and the extremely low number of people competing for the same opportunities. Furthermore, doing math at a hackathon need not mean doing just math; mathy hackers can collaborate with software developers to create better and more efficient algorithms and systems, just as mechanical, electrical, and software engineers collaborate to create better hardware hacks. With the right attitude, the opportunities are endless.
We wrapped up our presentations to the judges and headed over to the closing ceremony. To our delight, we discovered that we had won three prizes for our work!: best hack with startup opportunity, best use of location data, and HackMIT’s own Moonshot prize, for a team of first-time hackers with lofty ambitions. While winning the more substantial prizes felt great, the prize we were awarded by HackMIT seemed very fitting; we had come to Cambridge to show that math could be done at a hackathon, and we had done just that. I hope that our example will encourage hackathon organizers to promote mathematics as a field of interest, encourage sponsors to create prizes focusing on exceptional use of mathematics, and encourage mathematicians to join in the fun and energy of hackathons.
Erdős wasn’t just a mathematician; he was a hacker — perhaps even the original hacker — in mind, body, and spirit. Calling all mathematicians: try going to a hackathon. See what you can accomplish when you push your mind and body to the limit. Make some friends, learn some things, drink some Soylent. And in the spirit of Paul Erdős, remember to always keep your brain open.
Special thanks to Adam Shelby, Owen Lynch, Steven Moon, Samsara, jumpstart.me, and of course HackMIT for making this possible.
