Connor delves into the flawed nature of current AI ethics practices, likening it to addressing minor issues while major problems persist. He also discusses the challenges and importance of machine learning interpretability, emphasizing the evolving complexity of AI models and the quest for understanding deep learning.
