AI Rounds by the Cumming School of Medicine

Faculty Development @ CSM

AI Rounds is an educational podcast designed for university faculty in medicine and health sciences navigating the evolving landscape of artificial intelligence in healthcare and education. Each episode breaks down complex AI concepts into digestible insights, exploring practical applications and discussing how these technologies are reshaping medical research and education. Join us as we examine AI through a beginner's lens, creating a space where faculty can grow their understanding of these powerful tools that are transforming healthcare delivery, research, and education.

  1. 9月24日

    Teaching Machines to Say "I Don't Know"—The AI Hallucination Problem

    Why do GenAI systems confidently state incorrect medical facts instead of saying "I don't know?" Groundbreaking research from OpenAI and Georgia Tech reveals that AI hallucinations aren't bugs to be fixed—they're inevitable consequences of how these systems are trained. This episode explores the "singleton problem" that makes AI systematically unreliable on rare facts, connects to our previous discussion of AI benchmark saturation (Episode 9), and explains why the same evaluation methods that create impressive test scores actually reward confident guessing over appropriate uncertainty. For medical faculty evaluating AI tools, understanding these statistical realities is crucial for teaching students, conducting research, and developing institutional policies that account for AI's fundamental limitations.

    15 分钟

  2. 9月10日

    Rewriting the PhD Playbook: How GenAI is Transforming Graduate Science Education

    What happens when artificial intelligence collides with centuries-old academic traditions? In this thought-provoking episode, Dr. Heather Jamniczky, Associate Dean of Graduate Science Education and 3M National Teaching Fellow, tackles the seismic shifts reshaping how we train the next generation of medical researchers. From late-night worries about academic integrity to bold visions of AI-ready graduates, Heather shares candid insights on navigating uncharted territory. We explore the faculty hesitations to embracing AI, reimagine what authentic assessment looks like when AI can write and analyze, and dive deep into the ethical minefields emerging in AI-assisted research. But here's the kicker – in a mic-drop moment that will make you question everything, Heather poses the ultimate challenge: "Should we even examine a written thesis anymore?" This isn't just about adapting to new tools; it's about fundamentally rethinking what graduate education means in an AI-ubiquitous world. Whether you're supervising students, designing curricula, or simply trying to keep pace with the AI revolution in academia, this episode will challenge your assumptions and equip you with practical thoughts for the road ahead. The future of graduate education isn't coming – it's here. Join us for a conversation that's equal parts challenging and inspiring, as we explore how to prepare medical graduates not just to use AI, but to lead its ethical implementation in research and clinical practice.

    31 分钟

  3. 8月27日

    An Overview of Scite for Academics

    Our institution now has campus-wide access to scite.ai, an AI-powered research tool that's fundamentally different from traditional databases. In this episode, we explore how medical faculty can leverage this powerful platform to transform their research workflows, enhance teaching, and support evidence-based clinical decisions. Whether you're writing your next grant, preparing tomorrow's lecture, or answering a resident's question during rounds, this episode provides concrete strategies to work smarter, not harder with scite!

    19 分钟

  4. 8月12日

    When AI Becomes Too Good To Measure: Why "Perfect" AI Test Scores Might Be Meaningless

    In 2025, artificial intelligence has achieved an unexpected milestone: it's become too good at taking tests. From medical knowledge exams to complex reasoning tasks, AI systems are now scoring 90%+ on benchmarks that were designed to challenge them, rendering these assessments meaningless for comparison or evaluation. This "benchmark crisis" has profound implications for medical faculty evaluating AI tools for research, education, and clinical applications. When vendors claim their AI scored "95% on medical benchmarks," what does that actually tell us about real-world performance? This episode explores why perfect scores might be misleading, how the benchmark arms race mirrors challenges in medical education assessment, and what questions faculty should ask when evaluating AI tools for their institutions. Understanding this crisis is crucial for making informed decisions about AI integration in academic medicine.

    15 分钟

  5. 7月30日

    Unpacking GenAI's New Deep Research Capabilities

    The Game Has Changed! Forget everything you know about AI "search." In the past few months GenAI tools have released "deep research" capabilities which can spend up to 45 minutes autonomously investigating complex medical questions, synthesize findings from hundreds of papers in minutes, and generate comprehensive literature reviews that used to take weeks. See how you can save hours per week while maintaining rigorous academic standards. But can you trust the results? How do you verify AI-generated citations? And what does this mean for research integrity in academic medicine?

    24 分钟

  6. 7月16日

    Rethinking Our Relationship with Intellectual Work

    This week Jessalyn sits down with Dr. Jordan Engbers, who works with AI both in industry and academia. The conversation explores how our relationship to intellectual work could (and should!) change as we incorporate AI into our workflows. The discussion continues thinking about what we are teaching in higher education regarding AI use and what employers are looking for these days in industry related to AI skills.

    25 分钟

  7. 7月2日

    Thinking About Gender Bias and Patient Participation in AI Development

    This week Jessalyn has a fascinating conversation with guest Dr. Zack Marshall, an expert in Community-Based Participatory Research and Responsible AI. The conversation explores gender bias in AI, how to engage patients and the public in AI development, and transdisciplinary thinking and collaboration in AI. Check the links below for the projects mentioned in this episode. Shirin Seyedsalehi, Amin Bigdeli, Negar Arabzadeh, Batool AlMousawi, Zack Marshall, Morteza Zihayat and Ebrahim Bagheri (2025), "Understanding and Mitigating Gender Bias in Information Retrieval Systems", Foundations and Trends® in Information Retrieval: Vol. 19: No. 3, pp 191-364. (http://dx.doi.org/10.1561/1500000103)AbSPORuPaCERNotebookLMData Nutrition ProjectResponsible AI Project

    29 分钟

  8. 6月17日

    AI, Health Equity, and Implementation Science

    On this episode, Jessalyn sits down with Dr. Nonsikelelo Mathe (Scientific Director, Health Equity & Implementation Science, Health Equity & Systems Transformation Lab, CSM) to discuss equity, ethics, and privacy when thinking about using AI in the healthcare system.

    26 分钟
查看全部 13 集

预告

关于

AI Rounds is an educational podcast designed for university faculty in medicine and health sciences navigating the evolving landscape of artificial intelligence in healthcare and education. Each episode breaks down complex AI concepts into digestible insights, exploring practical applications and discussing how these technologies are reshaping medical research and education. Join us as we examine AI through a beginner's lens, creating a space where faculty can grow their understanding of these powerful tools that are transforming healthcare delivery, research, and education.

信息

更多来自“Faculty Development @ CSM”的内容

你可能还喜欢