Ashwin Kumar, a PhD candidate at Washington University in St Louis, working on fair resource allocation and explainable AI, speaks with Brent Phillips, Pitt HexAI podcast producer, about the intricacies of fairness in AI systems and about balancing fairness and efficiency in complex systems like ride-sharing and healthcare resource distribution. The discussion dives into AI transparency, human-machine interaction, multi-agent reinforcement learning, and the evolving challenges in ensuring equitable outcomes in AI decision-making. The conversation connects AI's technical workings with real-world ethical concerns.

John Carbone, Adjunct Professor of Applied AI and Applied Data Science at Baylor University speaks with Zoe Menezes, a student at the University of Pittsburgh, and Brent Phillips, Pitt HexAI podcast producer about the future of education in the AI age and how universities are adapting to the emergence of applications like ChatGPT. John elaborates on how universities like Baylor are embracing change, trying new things and and working towards keeping up with advances in technology. John also elaborates on a number of his AI connected academic papers and, on digital security in the AI age and offers advice for students and fellow faculty members on education and acquiring new skills.

Jordan Gass-Poore’, Pitt HexAI host, delves into the cutting-edge world of bioinformatics and cancer research with Dr. Aristotelis Tsirigos, a Professor of Medicine and Pathology and the Co-Director of Precision Medicine at NYU Grossman School of Medicine. Dr. Tsirigos shares his journey from computer science and electrical engineering to becoming a leading figure in pathology and bioinformatics.

In this episode, they explore the latest advancements in self-supervised deep learning and its transformative impact on cancer diagnosis and management. Dr. Tsirigos discusses his former work at IBM and his current research at NYU Langone Health's Perlmutter Cancer Center, including the innovative study "Mapping the landscape of histomorphological cancer phenotypes using self-supervised learning on unannotated pathology slides." This study, conducted in collaboration with the University of Glasgow, leverages AI to identify distinct cancer phenotypes, offering new insights into cancer subtypes and potential therapeutic strategies.

Tune in to learn about the intersection of AI, molecular information, and healthcare, and how these advancements are paving the way for more precise and personalized cancer treatments