Practical AI in Healthcare S1, E15 - Adam Rodman: Rethinking Clinical Reasoning in the Age of AI
21 snips
Dec 7, 2025 
Adam Rodman, physician, informatician, and historian of medicine, explores how large language models reshape clinical reasoning. He traces the arc from early diagnostic systems to GPT breakthroughs. He discusses workflow redesigns for clinician–AI collaboration, regulatory and liability gaps, patient-first AI agents, and early signals from urgent-care AI startups.
AI Snips
Chapters
Books
Transcript
Episode notes
From Clinical Reasoning To Model Collaboration
- Adam Rodman reframed his background as studying how clinicians think rather than building models.
- He argues LLMs change how we must conceptualize clinical decision support and collaboration.
Testing LLMs Against Old Decision Tools
- Rodman describes testing GPT models against classical clinical decision support benchmarks.
- He found modern LLMs matched older systems like Isabel on contextual tasks with far less work.
LLMs Break Old Evaluation Paradigms
- Language models are powerful but multi‑functional and have unique failings like hallucination.
- These traits force rethinking evaluation beyond traditional AUC/F1 metrics toward human-model interaction.
