Abstract: Advances in AI promise to improve people's lives and societies. As these systems migrate from research labs into the real world, new challenges have emerged. For example, how should predictive models form an effective team with physicians in clinical decision making? How could the designs of AI-mediated social networks prevent unintended consequences such as data-driven inequality? These are challenges of translation: translating AI's algorithmic advances into valuable, situated human experiences. This talk focuses on this critical translation. I will share a range of human-AI interaction design projects: from designing a system that helps doctors make life-and-death clinical decisions in real practice, to leveraging Natural Language Generation systems to improve authors’ writing experience. Each design addressed a critical challenge in moving AI from research labs valuably into the real world. I outline a new framework that scaffolds the problem space of human-AI interaction design. I discuss the opportunities and challenges it reveals for both AI and user experience research.
Bio: Qian Yang is an interaction design researcher and a Ph.D. candidate at the Human-Computer Interaction Institute at Carnegie Mellon University. Her research focuses on the design and innovation of human-AI interactions. She is best known for designing machine learning systems that effectively aided doctors in making critical clinical decisions. During her Ph.D., Yang has published fifteen peer-reviewed publications on the topic of human-AI interaction at premiere HCI research venus. Four of these papers have received awards. Yang has won a fellowship from the Center for Machine Learning and Health, a Microsoft Research Dissertation Grant, and the Innovation by Design award from Fast Company. This Spring she will be speaking at SXSW on how to design AI products and services.