Last month I had the incredible opportunity to visit the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and present my Master’s research to the group. Getting to share the work we’ve been doing at the Instituto de Ciências Matemáticas e de Computação (ICMC) – USP with the CSAIL community felt a bit surreal.
The thesis explores Monte Carlo methods for continuous-action chance-constrained MDPs — sequential decision making under risk, where actions don’t come from a neat discrete set. Presenting it in a room full of people who think about planning under uncertainty for a living produced exactly the kind of feedback you can’t buy.
Thank yous
A few people made this trip what it was:
- Prof. Claudio Toledo (ICMC/USP) — for the invite and for years of patient advising that made this visit possible.
- Prof. Brian Williams — for welcoming me into the lab and creating the space for a genuine exchange of ideas.
- Jake O. and Ben Ayton — great conversations and sharp questions that already changed how I think about a couple of the open problems.
- Allan — thanks for the support through the whole adventure.
An art installation we stumbled onto
Walking the halls at CSAIL you bump into things you don’t expect. We ran into an art installation that used computer vision to manipulate a live video stream in real time. Beautiful and, for a CV person, very tempting to reverse-engineer.
There was also an anime reference hiding in it. If you catch it, let me know. 😉
Taking it home
This trip reminded me why I do this work: research is better in rooms full of people willing to argue with you. I came back with a list of things to try, a few new collaborators, and a renewed appreciation for being stretched by smart strangers.
More soon.