HAI Weekly Seminar with Carlos Ernesto Guestrin
How Can You Trust Machine Learning?
Machine learning (ML) and AI systems are becoming integral to every aspect of our lives. As these technologies make more decisions for us, and the underlying ML systems become increasingly complex, it is natural to ask: How can I trust machine learning? In this talk, Carlos Ernesto Guestrin will present a framework anchored on three pillars—clarity, competence and alignment—for driving increased trust in ML. For clarity, Guestrin will cover methods to make the predictions of machine learning more explainable. For competence, he will focus on means for evaluating and testing ML models with the same rigor we apply to software products. For alignment, Guestrin will describe the challenges of aligning the behaviors of an AI with the values we want to reflect in the world, along with methods that can yield more aligned outcomes. The discussion will touch on both algorithmic and human processes that can help lead to AIs that are more effective, impactful and trustworthy.