I'm a member of the technical staff at OpenAI. We're hiring!

Chelsea Voss

Photo credit Marie La.

Background

Mathematics and Computer Science 𓊝

I graduated from MIT in 2015 with a major in computer science and a minor in mathematics. I received my M.Eng. degree in 2016, with a concentration equally in either theoretical computer science or in computer systems. My thesis applied static analysis to a concrete problem in computational biology.

In a sense, I am a mathematical computer scientist from MIT’s tradition of knowledge.

Two of the most impactful courses in my undergraduate studies were Scott Aaronson’s course Automata, Computability, and Complexity and David Kaiser’s course Einstein, Oppenheimer, Feynman: Physics in the 20th Century.

Something I liked about these courses were how they gave me the understanding that mathematical and scientific progress is made by real people, struggling against real problems. Hilbert’s twenty-three unsolved problems of 1900, and the progress made by Gödel, Turing, and Church against Hilbert’s Entscheidungsproblem by the 1930s; Einstein’s general theory of relativity in 1915, followed by Eddington’s 1919 eclipse experiment validating its predictions; scientists and the development of nuclear weapons – history is real, and we’re living in it.

Outside the former residence of Kurt Godel in Vienna

“The great mathematician and logician Kurt Gödel (1906 - 1978) had his last residence here before he left Vienna forever.”

Book Recs · For a taste of Aaronson’s classes, I strongly recommend the text Quantum Computing since Democritus. For a taste of Kaiser’s, I recommend Drawing Theories Apart: The Dispersion of Feynman Diagrams in Postwar Physics. I have also enjoyed Logicomix and To Mock a Mockingbird.

David Kaiser’s research on the usage and spread of Feynman diagrams among physicists was one of the inspirations for the talk I gave at Deconstruct 2017, Programming Languages as Notations.

Python 𓆙

  • onelinerizer is my fault. Don’t worry, I already apologized to everyone about it during PyCon 2016. Check out the live demo or the GitHub repo.

  • retroactive is a library that implements general retroactive data structures in Python. “General” here means that it can wrap around any data structure to make it partially or fully retroactive.

Speaking at PyCon 2016

Speaking at PyCon 2016.

Infrastructure Engineering 𓀨

Pilot  ·  From fall 2017 to summer 2019, I worked as a software engineer at Pilot, a great startup that is building an efficient and accurate bookkeeping service. I learned about engineering best practices and about scaling startups.

Sendwave  ·  Before Pilot, I worked as a software engineer at Sendwave, learning about fintech and anti-fraud and helping make remittances to the developing world convenient and affordable.

Khan Academy  ·  In 2015 I interned on the infrastructure team at Khan Academy and worked on a project to improve deploy times.

Staying up late writing design docs at the Khan Academy summer hackathon

Staying up late writing design docs at the Khan Academy summer hackathon.

Biology 𓅞

I pivoted from biology to computer science during my undergraduate studies. I continue to be interested in biochemistry, human physiology, and the computational modeling of large complex systems.

Coaching at USABO 2012

Logical Inference for Signaling Pathways  ·  In 2016, I finished my M.Eng. degree at MIT, with a concentration in theoretical computer science. My thesis applied SMT solver-based logical inference to constructing models of cellular signaling pathways from ambiguous or incomplete datasets.

Advised by Prof. Jean Yang, I collaborated with the Fontana Lab at Harvard Medical School and presented our condensed paper and our slides at Static Analysis in Systems Biology 2017.

DNA Logic Gates & Synthetic Biology  ·  In 2012, I worked on a synthetic biology project with the Weiss Lab at MIT to build out and validate Qian and Winfree’s DNA logic gates in vivo.

Timing Robustness in Developmental Biology  ·  In 2011, I presented a project at the Intel Science Talent Search national finals in Washington, D.C. that applied symbolic model checking, a technique traditionally used to analyze hardware circuits using Kripke structures and temporal logic, to a well-studied pathway in developmental biology. I was able to demonstrate that this pathway exhibits certain timing robustness properties!

This work was advised by Prof. David Dill at Stanford, whose paper Timing Robustness in the Budding and Fission Yeast Cell Cycles demonstrates a similar analysis as applied to the yeast cell cycle.

Science Olympiads 𓌝

I was a member of the U.S. teams to the International Biology Olympiad in 2010 (Changwon 창원시 🇰🇷) and 2011 (Taipei 台北 🇹🇼), and to the International Linguistics Olympiad in 2011 (Pittsburgh 🇺🇸).

I have coached at the following summer programs:

IBO 2010 in Changwon, Korea

IBO 2010 in Changwon, Korea.

Coaching at USACO 2016

Coaching at USACO 2016.