About

While I was on an archaeological excavation in Italy in 2014, Google published their DeepDream results. I remember looking at those images and wanting badly to understand how they were made. Not long after, I learned that it is possible to train neural networks to play Atari games using reinforcement learning. So, I finished my Classical Studies degree and enrolled in Computing at Queen's University the following year.

I had the best time of my life. Every year I learned about the most interesting topics anybody had ever told me about. In my fourth year I developed a model of binary decision timing with Dr. François Rivest, which became my undergraduate thesis. A course on Human-Computer Interaction introduced me to the idea that human cognition could be modelled computationally and used to guide the design of interactive systems. That idea sent me to graduate school.

I joined the EQUIS lab under Dr. Nick Graham and spent five years building game-playing AI assistants and studying how people used them. The research required recruiting participants from the spinal cord injury community, which I did through a sustained relationship with Spinal Cord Injury Ontario that began when I started volunteering with them in my first year of the PhD. That community access wasn't incidental to the work. Studies of accessible game design must involve the people the games are designed for.

Alongside the SCI Ontario relationship, I trained formally in accessibility and inclusive design through the NSERC CREATE READi program — a cross-university initiative between Carleton, Ottawa, and Queen's. I completed Level 3 certification, which included coursework, an annual symposium, and an action team project conducted in partnership with Kingston Circus Arts. Our project developed inclusive coaching materials for movement classes with participants of mixed abilities, using participatory design with the disability community and with Erin Ball, a circus performer, coach, and disability advocate.

While in school, I was twice the sole teaching assistant for the graduate Deep Learning course at Queen's. I taught Game Design once and Game Development twice, adding original machine learning content to an AI module that already comprised half the course. I found several opportunities to assist junior lab members with their research and mentor them while they helped me with my research. For two consecutive CHI PLAY conferences, I served as video chair and for several more SIGCHI conferences I volunteered as a reviewer. My dissertation was filed in 2023.

After my PhD, I moved to Halifax, where my wife originates. Since then, I have been working as a software developer at Trihedral Engineering Limited, building industrial communication drivers for VTScada. My experience at Trihedral has exposed me to the realities of software systems used to deliver vital services, like water and energy. I now want to engineer solutions for the potentially dangerous human-AI interaction problems that I discovered in school. I want to build AI that knows what users mean and does what they want.