AI for Social Good: Between My Research and the Real World

AI for social good (AI4SG) is a research theme that aims to use and advance AI to improve the well-being of society. My work on AI4SG builds a two-way bridge between the research world and the real world. Using my unique experience in food waste and security, I propose applied AI4SG research that directly addresses real-world challenges which have received little attention from the community. Drawing from my experience in various AI4SG application domains, I propose bandit data-driven optimization, the first iterative prediction-prescription framework and a no-regret algorithm PROOF. I will apply PROOF back to my applied work on AI4SG, thereby closing the loop in a single framework. From Research to the Real World: AI for Food Waste and Security In the US, over 25% of the food is wasted, with an average American wasting about one pound of food per day. Meanwhile, 11.8% of the American households struggle to secure enough food. The ongoing COVID-19 pandemic is only making things worse. Even after the pandemic hits its peak, the struggle with basic food security will not subside quickly on our long way back to normal. Thus, now more than ever, there is an urgent call for action to address the food waste and security problem. My experience working with practitioners in this domain poses me in a unique position to address this problem with AI. Food rescue organizations (FR) handle real-time or scheduled food donations and match them to recipient organizations that serve under-privileged communities. FRs rely on volunteers to transport the food. After the dispatcher matches a donation with a recipient, they will post this rescue trip on the FR's mobile app so that all volunteers can see it or receive notifications. They may also call some selected volunteers to ask for help. In the US alone, there are FRs operating in over 55 cities, affecting over 11 million people. However, relying on volunteers to deliver the food comes with inherent uncertainty. What if no volunteer will claim the rescue? What if the volunteer somehow fails to deliver the food after they claimed it? These common uncertainties have serious consequences because they may lead to lost