About Me

I’m a graduate student and Joubin-Selig Scholar in International Affairs at Carleton, and my goal is to work with government on the challenges posed by state-of-the-art neural networks.

I received analytical training in philosophy at Oxford, and I gained quantitative skills and technical knowledge from five years working as a programmer at tech startups.

I primarily intend to work on public policy, but I also aim to understand machine learning well enough that I can implement cutting-edge research papers in PyTorch. This is a work in progress! However, I am fluent in Python, I have trained neural networks, and I’ve done a deep-dive into the transformer architectures used by large language models (LLMs) like ChatGPT. I completed a previous iteration of Machine Learning with Andrew Ng (back when it used GNU Octave), and I’m now working through fast.ai’s resources on Practical Deep Learning. I have also spent time with researchers at Mila (informally), and I remain in touch with the field.

In the short- and medium-term, I aim to strengthen communication between computer scientists and public servants, and I aim to improve the technical literacy of people who (like me) are more naturally inclined towards the humanities and social sciences. Concretely, this might involve leading a team through online materials and facilitating dialogue with interested experts from research labs. I also aim to act as an interpreter between computer scientists and policy analysts; I want to improve mutual understanding and prevent each side from “talking past” the other.

I believe that some grasp of how neural networks are built is necessary to understand the ways they may not act as intended, and that this understanding is crucial to many, perhaps most, major decisions regarding AI that government will face. I also believe governments require sufficient technical know-how to critically assess researchers’ opinions and to form their own conclusions.

I’m particularly interested in public policy that would prepare governments to act quickly and effectively in response to an AI-related crisis.

If in doubt, please send me a message!

n.b.1. A list of sources I recommend for learning about transformers: the original paper; this blog series by Ben Levinstein for a high-level, non-technical overview; and Chapters 1 and 3 of Natural Language Processing with Transformers (Tunstall et al.) for a technical deep-dive.

n.b.2. I don’t use social media, so this website is one of the only ways people from my past can reconnect with me. In the interests of showing up when they search for me on the internet, here are some places I’ve been: International Guitar Festival (IGF), St Bartholomew’s School in Newbury (St Bart’s), The Catweazle Club, Mansfield College, Oxford University, Busbud, Procurify, the Choir of the Church of St. Andrew and St. Paul (A&P).