Dr. Karandeep Singh Doesn’t Trust AI Blindly. That’s Exactly Why UC San Diego Health Trusts Him With It
Episode Description
Dr. Karandeep Singh is a practicing nephrologist, an informaticist, and the inaugural Chief Health AI Officer at UC San Diego Health, where he also leads the Joan and Irwin Jacobs Center for Health Innovation. Back in 2021 at the University of Michigan, he published research questioning Epic's own sepsis model, finding it worked only about 63% of the time and that a clinician might have to evaluate eight separate alerts just to find one real case of sepsis. He's one of the rare people who has been on both sides of AI in medicine, exposing a flawed model and now deciding which ones get deployed at all.
In this episode of How I Doctor, Dr. Graham Walker talks with Karandeep about what it actually takes to govern AI inside a health system, and why the doctors who trust AI the least are often the ones best equipped to run it. Karandeep traces his path from teaching himself to code at age eight, through building tools to solve his own clinical frustrations in training, to formally discovering biomedical informatics as a field only during his nephrology fellowship.
Karandeep and Graham dig into why generative AI feels transformative to physicians while predictive AI often doesn't, even when the data shows it's saving lives. They walk through a real case study of using agentic AI to automate pre-colonoscopy prep calls without cutting nurses out of the loop, and unpack what a Chief Health AI Officer's job actually involves day to day. The conversation closes on Karandeep's case for regulatory sandboxes, structured ways to test AI-enabled care models that current rules don't yet allow, and his prediction for where AI scribes are headed next.
What You'll Learn
- Why Karandeep's research found Epic's sepsis model wasn't doing what its own model card claimed
- Why predictive AI creates work that's hard to feel, while generative AI's time savings are noticeable almost daily
- How UC San Diego Health automated colonoscopy prep calls with AI while keeping nurses at the center of care
- What a Chief Health AI Officer actually does across evaluation, governance, and implementation
- Why Karandeep believes regulatory sandboxes, not blanket approval, are how AI expands access to care
🩺 Offcall is more than a platform — it’s a community. Join today!
📝 For a full transcript of this episode click HERE
🎧 Subscribe to receive new How I Doctor episodes directly in your feed here: https://episodes.fm/1767429315
👨⚕️Follow Dr. Graham Walker on
LinkedIn https://www.linkedin.com/in/graham-walker-md/
IG https://www.instagram.com/ubergraham/
Bluesky https://bsky.app/profile/drgrahamwalker.com
✉️ Join our newsletter On/Offcall here https://offcall.beehiiv.com/subscribe
🟧 Follow Offcall on
LinkedIn https://www.linkedin.com/company/joinoffcall/
Proudly Supported by
