This chapter presents interviews with three researchers who examine human-centered ML and AI through the lens of collaborative work. Wu Qiong, professor of information art and design at Tsinghua University, describes how human-centered methods apply to team-based AI design projects and argues that realizing human-centered ML requires new design methods alongside new technical capabilities. Adam Perer, faculty at Carnegie Mellon University and co-director of the Data Interaction Group, focuses on the challenges of keeping humans meaningfully involved across the full ML pipeline, from data preparation to model inspection, and discusses why the lack of debugging tools for ML remains a significant barrier in practice. Bill Buxton, pioneering HCI researcher and designer, critiques the narrowness of current AI development and calls for what he terms "Renaissance teams", interdisciplinary groups where each member brings deep expertise in one area and working literacy across others. He applies an ecological perspective to human-AI interaction, arguing that effective intelligent systems require attention to the full context of human activity, not optimization of any single capability.