The increased deployment of artificial intelligence (AI) systems and their broad societal impacts brings along new challenges, particularly ensuring that these systems operate reliably in the open world. Reliability is critical in shaping how users will interact and trust AI systems. Despite recent advances in AI and machine learning, it is challenging to assure that a system will always operate reliably, when deployed. One of the causes of unreliable behavior is the impreciseness of the model used for decision-making. Regardless of how much effort goes into system design and how much data is available for testing, it is generally infeasible to obtain a perfect description of open-world environments. This talk will focus on the inherent challenges and future directions for designing autonomous systems that can operate reliably in the presence of uncertainty and model imprecision.
Sandhya Saisubramanian is an incoming Assistant Professor in EECS at Oregon State University. She completed her Ph.D. in Computer Science from the University of Massachusetts Amherst in September 2021 and officially starts as a faculty at Oregon State University in December 2021. Sandhya's research focuses on developing general techniques for reliable decision-making in autonomous systems that operate in the open world. She is a recipient of Distinguished Paper Award at IJCAI 2020 and was selected as a Rising Star in EECS by UC Berkeley in 2020.