Appendix B — Starter resources

Students in the 2024 and 2025 offerings of this course added more than 600 resources to the weekly resource log. This page lists about thirty of them, picked because they are clear, recent, or from a primary source. Use them to get started, then find your own.

Some journal articles need your library sign-in. Where a free full text exists on PubMed Central, the entry says so.

B.1 How these systems work

  • But what is a neural network? (3Blue1Brown, 2017; video; 19 min). A visual, step-by-step explanation of how a neural network turns numbers into a prediction, with no math beyond high school. https://youtu.be/aircAruvnKk
  • The Basics of Machine Learning (NEJM Evidence, 2022; editorial). A short explanation of machine learning built around one clinical example: dating a pregnancy with a low-cost ultrasound device. (Fralick & Campbell, 2022)
  • Benefits, Limits, and Risks of GPT-4 as an AI Chatbot for Medicine (NEJM, 2023; special report). Worked examples of a chatbot answering clinical questions, including the mistakes it made, written as these tools first reached clinicians. (Lee et al., 2023)
  • Large language models encode clinical knowledge (Nature, 2023; research article). The Med-PaLM study shows how chatbots are tested on exam-style questions, and why doctors’ ratings found gaps the exam scores hid. Free full text on PubMed Central. (Singhal et al., 2023)
  • Can AI catch what doctors miss? (TED, 2023; talk; 14 min). A cardiologist’s tour of what AI can find in medical images and records, useful as an optimistic view to test against the appraisal readings below. https://www.ted.com/talks/eric_topol_can_ai_catch_what_doctors_miss

B.2 Working alongside AI

  • Compared with What? Measuring AI against the Health Care We Have (NEJM, 2024; perspective). Argues that AI should be judged against the care patients actually get today, not against an ideal doctor. (Kohane, 2024)
  • Large Language Models and the Degradation of the Medical Record (NEJM, 2024; perspective). Explains how notes drafted by AI could make the chart longer and less trustworthy, and what that means for the next reader. (McCoy et al., 2024)
  • Artificial intelligence and illusions of understanding in scientific research (Nature, 2024; perspective). Describes how relying on AI tools can make people believe they understand more than they do. The setting is research, but the warning applies at the bedside. (Messeri & Crockett, 2024)
  • AI in Medicine: Pitfalls and Potential (Stanford Medicine, 2024; talk; 12 min). A short conference talk on where clinical AI helps and where it goes wrong. https://youtu.be/gF5QPYP6KIM

B.3 Your applications and your coursework

  • How to use AI in residency applications? Learn the rules of the road (AMA News, September 9, 2026; article). Summarises the AAMC’s guidance for applicants: AI may help you brainstorm, proofread, or edit, but it cannot be the author, and the final application must be your own work. Read it before you open a chatbot next to your personal statement. https://www.ama-assn.org/medical-students/preparing-residency/how-use-ai-residency-applications-learn-rules-road
  • Principles for the Responsible Use of Artificial Intelligence in and for Medical Education (AAMC, version 2.0, July 2025; web page). Seven principles written for medical schools, including ethical and transparent use and protecting data privacy. It does not set rules for students, but it shows what schools and programs are building their own policies on. For your school’s rules on AI in coursework, check its own policy. https://www.aamc.org/about-us/mission-areas/medical-education/principles-ai-use

B.4 Appraisal and evidence

  • The testing of AI in medicine is a mess. Here’s how it should be done (Nature, 2024; news feature). A readable account of why most medical AI tools reach patients without strong clinical trials. (Lenharo, 2024)
  • Why we should not mistake accuracy of medical AI for efficiency (npj Digital Medicine, 2024; comment). Shows that an accurate tool can still add work for staff, so “accurate” and “saves time” are separate claims to check. Free full text on PubMed Central. (Jongsma et al., 2024)
  • Generative AI in Medicine: Evaluating Progress and Challenges (NEJM, 2025; special report). Sets out what good evidence for a generative AI tool would look like, from a group that includes health system, industry, and academic authors. (Maddox et al., 2025)
  • Randomized Trial of a Generative AI Chatbot for Mental Health Treatment (NEJM AI, 2025; randomized trial). One of the first randomized trials of a chatbot used as treatment. Read it to practise appraising the design, not only the result. (Heinz et al., 2025)
  • AI, Health, and Health Care Today and Tomorrow: The JAMA Summit Report on Artificial Intelligence (JAMA, 2025; report). A long but clear summary of how health AI is built, tested, and monitored, and where the gaps are. Read the summary first. (Angus et al., 2025)
  • JAMA+ AI Conversations (JAMA Network, 2024 onward; podcast series; episode lengths vary). Interviews with researchers and editors about new studies, useful for keeping up after the course ends. https://jamanetwork.com/channels/ai/pages/podcast

B.5 Bias and equity

  • Considerations for addressing bias in artificial intelligence for health equity (npj Digital Medicine, 2023; review). Walks through where bias can enter an AI tool, from design to use, written with FDA staff. Free full text on PubMed Central. (Abràmoff et al., 2023)
  • Health Equity and Ethical Considerations in Using Artificial Intelligence in Public Health and Medicine (Preventing Chronic Disease, CDC, 2024; review). A plain overview of how AI can widen or narrow gaps in care, with practical steps. Free to read. (Dankwa-Mullan, 2024)

B.6 Talking to patients

  • Comparing Physician and Artificial Intelligence Chatbot Responses to Patient Questions Posted to a Public Social Media Forum (JAMA Internal Medicine, 2023; cross-sectional study). The much-discussed study in which raters preferred chatbot answers for quality and empathy. Read the methods closely before you repeat the headline. Free full text on PubMed Central. (Ayers et al., 2023)
  • Artificial Intelligence-Generated Draft Replies to Patient Inbox Messages (JAMA Network Open, 2024; quality improvement study). A real deployment at one health system, where clinicians reported less burden but saved no time. Free full text on PubMed Central. (Garcia et al., 2024)
  • Patients’ Trust in Health Systems to Use Artificial Intelligence (JAMA Network Open, 2025; research letter). A national survey on whether patients trust health systems to use AI responsibly. Free full text on PubMed Central. (Nong & Platt, 2025)
  • Google’s Efforts to Build Patient-Facing AI (NEJM AI Grand Rounds, 2025; podcast; length not stated on the episode page). The researchers behind AMIE, a chatbot built to take a medical history, describe how they built and tested it. https://ai-podcast.nejm.org/e/google-s-efforts-to-build-patient-facing-ai-a-conversation-with-drs-alan-karthikesalingam-and-anil-palepu

B.7 Ethics, law, and regulation

B.8 Workforce and environment

  • Ambient Artificial Intelligence Scribes to Alleviate the Burden of Clinical Documentation (NEJM Catalyst, 2024; case study). How one large health system rolled out AI scribes, and what clinicians and patients said about them. (Tierney et al., 2024)
  • Explained: Generative AI’s environmental impact (MIT News, 2025; article). A clear account of the electricity and water that AI data centres use, and why use after training matters as much as training. https://news.mit.edu/2025/explained-generative-ai-environmental-impact-0117

B.9 Specialty-specific

  • Ophthalmology. Pivotal trial of an autonomous AI-based diagnostic system for detection of diabetic retinopathy in primary care offices (npj Digital Medicine, 2018; trial). The study behind the first FDA-authorized autonomous AI diagnostic system, and a model of how to test one. Free full text on PubMed Central. (Abràmoff et al., 2018)
  • Ophthalmology. No Doctor Needed? Dr. Michael Abramoff on the Potential of Autonomous AI (NEJM AI Grand Rounds, 2023; podcast video; 65 min). The lead author of that trial explains how the system was built and regulated. https://youtu.be/9sLxXyaQ7EY
  • Cardiology. Artificial intelligence-enhanced electrocardiography in cardiovascular disease management (Nature Reviews Cardiology, 2021; review). Explains how AI reads findings from an ECG that a person cannot see, and the limits of that work. Free full text on PubMed Central. (Siontis et al., 2021)
  • Dermatology. AI vs. Skin Cancer (Nature Video, 2025; video; 10 min). A short film on AI tools that assess skin lesions. https://youtu.be/rkiYSQ_RED8
  • Radiology. AI regulation in radiology with Hugh Harvey (Radiology Channel, 2024; video interview; 46 min). A radiologist and regulatory specialist on how imaging AI is approved and monitored. https://youtu.be/0VMuaKY5NbI
  • Pediatrics. GPT: Great Pediatric Tools! Harnessing the Power of AI in Pediatric Medicine (The Cribsiders, episode 135, 2025; podcast; length not stated on the episode page). A pediatric hospitalist on current uses of AI on the wards and the ethical questions they raise. https://thecurbsiders.com/cribsiders-podcast/135
  • Anesthesiology. How AI Will Impact Anesthesiology (OpenAnesthesia, 2024; podcast; length not stated on the episode page). A short expert interview on where AI is likely to change anesthesia practice. https://www.openanesthesia.org/podcasts/how-ai-will-impact-anesthesiology

B.10 Tools to try

Use made-up cases only. Never enter patient information into these tools.

  • OpenEvidence (OpenEvidence; web tool). A medical search chatbot that answers with citations to journal articles. Check each cited article yourself before you trust the answer. Sign-up rules may change. https://www.openevidence.com
  • Learn 80% of NotebookLM in Under 13 Minutes (Jeff Su, 2024; video; 13 min). A quick tour of NotebookLM, Google’s free tool that answers questions from documents you upload. It pairs with the NotebookLM chapter. https://youtu.be/EOmgC3-hznM
Abràmoff, M. D., Lavin, P. T., Birch, M., Shah, N., & Folk, J. C. (2018). Pivotal trial of an autonomous AI-based diagnostic system for detection of diabetic retinopathy in primary care offices. NPJ Digital Medicine, 1, 39. https://doi.org/10.1038/s41746-018-0040-6
Abràmoff, M. D., Tarver, M. E., Loyo-Berrios, N., Trujillo, S., Char, D., Obermeyer, Z., Eydelman, M. B., & Maisel, W. H. and. (2023). Considerations for addressing bias in artificial intelligence for health equity. NPJ Digital Medicine, 6(1), 170. https://doi.org/10.1038/s41746-023-00913-9
Angus, D. C., Khera, R., Lieu, T., Liu, V., Ahmad, F. S., Anderson, B., Bhavani, S. V., Bindman, A., Brennan, T., Celi, L. A., Chen, F., Cohen, I. G., Denniston, A., Desai, S., Embí, P., Faisal, A., Ferryman, K., Gerhart, J., Gross, M., … Bibbins-Domingo, K. and. (2025). AI, Health, and Health Care Today and Tomorrow: The JAMA Summit Report on Artificial Intelligence. JAMA, 334(18), 1650–1664. https://doi.org/10.1001/jama.2025.18490
Ayers, J. W., Poliak, A., Dredze, M., Leas, E. C., Zhu, Z., Kelley, J. B., Faix, D. J., Goodman, A. M., Longhurst, C. A., Hogarth, M., & Smith, D. M. (2023). Comparing Physician and Artificial Intelligence Chatbot Responses to Patient Questions Posted to a Public Social Media Forum. JAMA Internal Medicine, 183(6), 589–596. https://doi.org/10.1001/jamainternmed.2023.1838
Dankwa-Mullan, I. (2024). Health Equity and Ethical Considerations in Using Artificial Intelligence in Public Health and Medicine. Preventing Chronic Disease, 21, E64. https://doi.org/10.5888/pcd21.240245
Fralick, M., & Campbell, K. R. (2022). The Basics of Machine Learning. NEJM Evidence, 1(5), EVIDe2200062. https://doi.org/10.1056/evide2200062
Garcia, P., Ma, S. P., Shah, S., Smith, M., Jeong, Y., Devon-Sand, A., Tai-Seale, M., Takazawa, K., Clutter, D., Vogt, K., Lugtu, C., Rojo, M., Lin, S., Shanafelt, T., Pfeffer, M. A., & Sharp, C. (2024). Artificial Intelligence-Generated Draft Replies to Patient Inbox Messages. JAMA Network Open, 7(3), e243201. https://doi.org/10.1001/jamanetworkopen.2024.3201
Heinz, M. V., Mackin, D. M., Trudeau, B. M., Bhattacharya, S., Wang, Y., Banta, H. A., Jewett, A. D., Salzhauer, A. J., Griffin, T. Z., & Jacobson, N. C. (2025). Randomized Trial of a Generative AI Chatbot for Mental Health Treatment. NEJM AI, 2(4). https://doi.org/10.1056/aioa2400802
Jongsma, K. R., Sand, M., & Milota, M. (2024). Why we should not mistake accuracy of medical AI for efficiency. NPJ Digital Medicine, 7(1), 57. https://doi.org/10.1038/s41746-024-01047-2
Kohane, I. S. (2024). Compared with What? Measuring AI against the Health Care We Have. The New England Journal of Medicine, 391(17), 1564–1566. https://doi.org/10.1056/nejmp2404691
Lee, P., Bubeck, S., & Petro, J. (2023). Benefits, Limits, and Risks of GPT-4 as an AI Chatbot for Medicine. The New England Journal of Medicine, 388(13), 1233–1239. https://doi.org/10.1056/nejmsr2214184
Lenharo, M. (2024). The testing of AI in medicine is a mess. Here’s how it should be done. Nature, 632(8026), 722–724. https://doi.org/10.1038/d41586-024-02675-0
Maddox, T. M., Embí, P., Gerhart, J., Goldsack, J., Parikh, R. B., & Sarich, T. C. (2025). Generative AI in Medicine - Evaluating Progress and Challenges. The New England Journal of Medicine, 392(24), 2479–2483. https://doi.org/10.1056/nejmsb2503956
McCoy, L. G., Manrai, A. K., & Rodman, A. (2024). Large Language Models and the Degradation of the Medical Record. The New England Journal of Medicine, 391(17), 1561–1564. https://doi.org/10.1056/nejmp2405999
Messeri, L., & Crockett, M. J. (2024). Artificial intelligence and illusions of understanding in scientific research. Nature, 627(8002), 49–58. https://doi.org/10.1038/s41586-024-07146-0
Nong, P., & Platt, J. (2025). Patients’ Trust in Health Systems to Use Artificial Intelligence. JAMA Network Open, 8(2), e2460628. https://doi.org/10.1001/jamanetworkopen.2024.60628
Ong, J. C. L., Chang, S. Y.-H., William, W., Butte, A. J., Shah, N. H., Chew, L. S. T., Liu, N., Doshi-Velez, F., Lu, W., Savulescu, J., & Ting, D. S. W. (2024). Medical Ethics of Large Language Models in Medicine. NEJM AI, 1(7). https://doi.org/10.1056/aira2400038
Singhal, K., Azizi, S., Tu, T., Mahdavi, S. S., Wei, J., Chung, H. W., Scales, N., Tanwani, A., Cole-Lewis, H., Pfohl, S., Payne, P., Seneviratne, M., Gamble, P., Kelly, C., Babiker, A., Schärli, N., Chowdhery, A., Mansfield, P., Demner-Fushman, D., … Natarajan, V. (2023). Large language models encode clinical knowledge. Nature, 620(7972), 172–180. https://doi.org/10.1038/s41586-023-06291-2
Siontis, K. C., Noseworthy, P. A., Attia, Z. I., & Friedman, P. A. (2021). Artificial intelligence-enhanced electrocardiography in cardiovascular disease management. Nature Reviews. Cardiology, 18(7), 465–478. https://doi.org/10.1038/s41569-020-00503-2
Tierney, A. A., Gayre, G., Hoberman, B., Mattern, B., Ballesca, M., Kipnis, P., Liu, V., & Lee, K. (2024). Ambient Artificial Intelligence Scribes to Alleviate the Burden of Clinical Documentation. NEJM Catalyst, 5(3). https://doi.org/10.1056/cat.23.0404
Warraich, H. J., Tazbaz, T., & Califf, R. M. (2025). FDA Perspective on the Regulation of Artificial Intelligence in Health Care and Biomedicine. JAMA, 333(3), 241–247. https://doi.org/10.1001/jama.2024.21451