Building AI for Better Healthcare — the OpenAI Podcast Ep. 14

Why It Matters

Secure, physician‑validated AI that scales personalized health advice could close care gaps and accelerate innovation, reshaping how patients and providers interact with medical information.

Key Takeaways

• OpenAI integrates physicians throughout model training and evaluation.
• ChatGPT Health handles 40 million daily health queries securely.
• HealthBench evaluates 49,000 performance dimensions with 250 doctors.
• Models now request context, reducing overconfident hallucinations in healthcare.
• Tailored responses adapt to user literacy and professional role.

Summary

The OpenAI Podcast’s fourteenth episode spotlights OpenAI’s health‑focused AI program, led by Dr. Nate Gross and Karan Singhal. They discuss how large language models are being engineered to answer sensitive medical questions, support clinicians, and streamline fragmented care.

OpenAI built ChatGPT Health, a secure, encrypted interface that now sees roughly 40 million weekly users, with about one‑quarter of queries health‑related—approximately 40 million daily interactions. A dedicated evaluation suite, HealthBench, was co‑created with a cohort of 250 physicians, measuring roughly 49,000 performance dimensions across multi‑turn conversations.

Gross emphasized the importance of context, noting that models now ask follow‑up questions rather than guessing. Singhal highlighted that over 5,000 physician‑generated dialogues and 48,500 rubric criteria shaped the training loop, enabling the system to tailor answers to a patient’s literacy level or a clinician’s specialty. The team also prioritized uncertainty detection, teaching models to admit when they “don’t know.”

By embedding safety, privacy, and clinician feedback at every stage—from pre‑training to production—OpenAI aims to make AI a proactive partner in healthcare, potentially reducing missed care, expanding access, and accelerating drug repurposing. If successful, the approach could set a new industry standard for trustworthy medical AI.

Original Description

Healthcare systems around the world are under strain, and both patients and clinicians are feeling the impact. OpenAI's Head of Health Dr. Nate Gross and Karan Singhal, who leads Health AI Research, discuss how AI can help address the biggest challenges. They cover how OpenAI is training models to handle sensitive health questions in collaboration with physicians, and how that foundation is unlocking a new generation of tools for patients, clinicians, and healthcare systems.

Chapters:

00:00:38 – Origins of Nate and Karan’s interest in AI and healthcare

00:05:01 – Strategy for building AI tools for clinicians

00:06:57 – How AI models are trained for health use cases

00:10:15 – How OpenAI is able to score well on health evals

00:14:21 – Key challenges deploying AI in healthcare

00:21:05 – Collaboration with hospitals and healthcare systems

00:23:05 – Practical everyday uses of AI health assistants

00:26:43 – Biggest “wow” moment during development

00:28:46 – Feedback from clinicians and early users