Joseph Suarez — Intro/Bio: I grew up in Connecticut and moved to California to do my undergraduate BS at Stanford, where I was heavily involved in deep learning research. I started in natural language processing in Andrew Ng’s lab as a sophomore where I learned the ropes and published my first paper on language modeling. Thereafter, I moved into computer vision and spent a year working on compositional models for visual question answering with Justin Johnson, who was a PhD student under Fei-Fei Li at the time. It was towards the end of this project that I nailed down my long-term interests and moved into reinforcement learning. I started Neural MMO, a research platform that formed the basis of 6 months of research at OpenAI. After graduating from Stanford, I moved back to the east coast to continue this line of work as a PhD student under Phillip Isola at MIT. My day-to-day largely consists of expanding Neural MMO (hosted at to enable new research directions within reinforcement learning and agent-based intelligence in general. I am also part of a research collaboration with NYU in which we are using Neural MMO to investigate the potential of artificial intelligence to intentionally specialize in diverse skills. In the short-term, I would like to establish additional academic research collaborations surrounding the platform. The lack of normal conference venues due to the pandemic has made this rather difficult — I’ve presented a couple of extended abstract/workshop papers, but the virtual audience is a tiny fraction of the traditional physical one. Long-term, I’d like to establish similar collaborations within industry, as I intend to pursue agent-based intelligence research outside of academia following my PhD.