Why most AI decision support tools don’t work, and why it is so hard to get them to do so - Professor Tim Miller

Abstract

In this talk, Professor Tim Miller will argue for a paradigm shift away from the current way that we conceptualise AI-based decision support tools, which may be counter-productive to better human decision making. Based on recent research with my collaborators, and on research from other labs around the world, Miller argues that the current paradigm has benefits, but helping to improve human decisions is not one of them. Miller also argues a shift away from the idea of a ‘human-in-the-loop’ paradigm, where the role of the human is to accept/reject machine decisions, to a ‘machine-in-the-loop’ paradigm, where the role of the human is to decide, and the role of the machine is guide decision making and provide feedback on intuition.

Bio

Tim Miller is a Professor of Artificial Intelligence in the School of Electrical Engineering and Computer Science at The University of Queensland. His research draws on machine learning, reinforcement learning, AI planning, interaction design, and cognitive science, to help people to make better decisions. He has worked on areas including explainable AI, human-AI planning, and human-centered decision support. Prior to his appointment at The University of Queensland, Tim was a Professor of Computer Science in the School of Computing and Information Systems at the University of Melbourne, where he was founding co-director of The Centre for AI and Digital Ethics.