The AI Medicine Revolution

Exploring AI solutions for our health systems

Hey Cogs,

Welcome to The Cognitive Revolution Newsletter! In this edition, we bring you highlights from our two-part series interviewing guests on the frontier of AI Medicine. Below, we’ll share insights from our interviews with Dr. Isaac Kohane of Harvard Medical School, and Neal Khosla, founder of Curai Health, links of interest, and a unique opportunity from Nathan at the end.

Also, we’ve just dropped a new episode of The Cognitive Revolution with Vivek Natarajan, AI researcher at Google Health and one of the lead authors behind Med-PaLM. You can listen here: Spotify | Apple | Youtube linked below

🎬 Episode Highlights

Ep 24: The AI Revolution in Medicine with Dr. Isaac Kohane of Harvard Medical School
If you’re curious about an expert physician’s perspective on how AI should be leveraged in medicine and the implications and ramifications of its usage, we recommend listening to Dr. Kohane’s interview. Dr. Kohane’s book The AI Revolution in Medicine, for which Sam Altman wrote the foreword, was released this past week (see Links of Interest below)

Recap: Dr. Kohane shares his thoughts on how people from other disciplines can explore the applications of AI, illustrates how AI will augment a doctor’s capabilities, and the way he sees AI used at the cutting edge of medical research as a Principal Investigator of the Undiagnosed Disease Network, based across Stanford, Harvard, Baylor, Duke, and more.

Dr. Kohane highlights his experience working closely with the Apple Health team, and the implications of AI-driven healthcare when it comes to patient data and alignment of interests within the medical system between insurance providers, hospitals, and patients. He also shares his reaction to GPT-4 after receiving top-level clearance and early access. He foresees the medical ecosystem will evolve as AI plays an increasingly relevant role.

👉 Spotify | Apple Podcasts | Youtube

Ep 25: Revolutionizing Patient Care with Neal Khosla of Curai Health
If you’re curious about a venture-backed founder’s insights and the considerations that go into building LLMs used in medicine, we recommend listening to Neal Khosla’s interview.

Recap: Neal shares his thoughts on what the patient experience may look like in the next 5-10 years, the dismantling of Google as the place people head to ask about their symptoms, and how AI-driven healthcare will interact in different ways with social and regulatory forces globally.

Neal also highlights the thought process that goes into building LLMs in medicine, whether GPT-4 could diagnose a mysterious Burning Man symptom, and how AI can transform medicine into a more empirical, data-driven practice over its current judgment-based state.

👉 Spotify | Apple Podcasts | Youtube

Across both episodes, we hear from Kohane and Khosla on the importance of the human loop in medicine and the ability to leverage the cognitive capabilities of AI to better inform decision-making in medicine and transform medicine into a more data-driven, rather than judgment-based, practice.

On how AI can better inform physician decision-making

“I do think that this is the mechanism to keep doctors up to date with the latest. I think it allows you to remember everything about the patient that they should have remembered. It will allow you to avoid errors. Absolutely no doctor should be without it. They should have this sidekick that is meticulous, completely-up-to date, ever-vigilant and sometimes… wrong (and that you're there to evaluate that).”

— Dr. Kohane

“If you take a step back, you go as a human being, 95% of the time, you're not actually getting a very data-driven recommendation… you’re getting expert opinion. It's a little bit alarming to me that in 2023 as a patient, I am still living in a world where a couple of smart people sat down and, and talked it out. We've had millions and millions of people who go through all these medical conditions. Why can't we understand what happened to them, what was done to them, and really use that data to help improve our interventions?”

— Neal Khosla

---

Thank you Omneky for sponsoring The Cognitive Revolution. Omneky is an omnichannel creative generation platform that lets you launch hundreds of thousands of ad iterations that actually work, customized across all platforms, with a click of a button. Show the perfect ad tailored for each customer. Omneky combines generative AI and real-time advertising data. Mention “Cog Rev” for 10% off.

---

🖇️ Links of Interest

[Book] “The AI Revolution in Medicine”, by Dr. Isaac Kohane - Sam Altman, in the foreword to this book, writes “This book represents the sort of effort that every sphere affected by AI will need to invest in as humanity grapples with this phase change.
One of our listeners, Proactive Logic CEO Jon Kragh, offered this endorsement:

[Paper]  Towards Expert-Level Medical Question Answering with Large Language Models - Google’s Med-PaLM2 outperforms physicians in answering medical knowledge questions. Vivek Natarajan of Google’s Med-PaLM team will be joining us in the next episode to talk more about his team’s research and findings!

[Thread] AI Medical Notetakers: The Good and Bad - A thread from Robert Scoble, who appeared on a previous TCR episode, on his experience seeing an AI medical notetaker in action. How will this help scale healthcare? What risks does this pose for the usage of this data?

---

Special Project from Nathan, host of The Cognitive Revolution
If you are interested in:

• Reproducing some recent LLM benchmarking results and collaborating on a project

• Reading drafts of Nathan’s AI analysis mega threads and providing feedback

• Sharing questions you’d like to ask for a future discussion episode?

Reach out to us at tcr@turpentine.co.

Until next time.