Podcast: Play in new window | Download
Subscribe: Spotify | Email | RSS | More
Fixing Healthcare is revisiting a Diving Deep episode that feels even more important today than when it first aired in November 2023.
At the time, ChatGPT had been available to the public for just one year. Generative AI was still widely viewed as a fascinating but unreliable tool, impressive in some settings and error-prone in others. Many clinicians saw it as a novelty. Some saw it as a threat. Few fully grasped how quickly the technology would advance or how profoundly it might reshape the relationship between doctors and patients.
Nearly three years later, that early vision already feels outdated. We’ve already surpassed the vision of generative AI helping patients prepare for office visits and ask better questions. The next generation of AI agents are proving capable of helping provide care between visits: monitoring chronic disease, identifying early signs of heart failure, flagging postoperative wound infections, answering medication questions and alerting doctors when a patient’s condition begins to worsen. Instead of centering care almost entirely around the doctor’s office, these tools could extend medical expertise into the home, 24/7.
When the episode was released, many people worried that generative AI was not powerful enough to solve clinical problems. Now, some worry that it has become too powerful and may pose existential risks. Across the nation, health systems, technology companies and policymakers are debating how far these tools should go, how quickly they should be adopted and who should guide their use.
That makes this replay especially timely. In this episode, co-hosts Dr. Robert Pearl and Jeremy Corr return to two connected problems: the high cost of American healthcare and the potential for generative AI to give patients more knowledge, agency and control.
The first half of the conversation focuses on “skimping,” a term Pearl uses for the many ways healthcare organizations constrain costs by giving patients less care, less access or fewer resources while still charging more. He compares it to buying a box of cereal at the same price, only to discover there are fewer flakes inside. In healthcare, however, the consequences are far more serious.
The second half of the episode turns to generative AI and the future of the doctor-patient relationship. Pearl argues that, when used correctly, generative AI could make medicine more collaborative. Instead of relying entirely on brief office visits, patients could use AI to better understand their conditions, evaluate treatment options, track symptoms and participate more actively in decisions. Doctors, meanwhile, could benefit from patients arriving better informed and more prepared.
That idea was provocative in 2023. Today, it is becoming increasingly practical. Generative AI systems have grown more capable, more conversational and more widely used. The tools available now are far beyond what existed when this episode first aired.
And as Pearl has argued in his book ChatGPT, MD, the exponential growth of AI means that capabilities can double quickly. If a technology doubles in power each year, then after five years it is not five times stronger. It is 32 times stronger.
That mathematical reality gives this episode its new relevance. The conversation is both a look back at early optimism about generative AI in medicine and a reminder that even bold predictions can prove conservative when a technology advances exponentially.
Listen again to this early Diving Deep conversation and consider how much has changed since November 2023—and how much more may change in the next two years.
HELPFUL LINKS
- How America Skimps On Healthcare (Forbes)
- How Generative AI Will Upend The Doctor-Patient Relationship (Forbes)
- Monthly Musings on American Healthcare (Robert Pearl’s newsletter)
* * *
Dr. Robert Pearl is the author of ChatGPT, MD: How AI-Empowered Patients & Doctors Can Take Back Control of American Medicine.
Fixing Healthcare is a co-production of Dr. Robert Pearl and Jeremy Corr. Subscribe to the show via Apple, Spotify or wherever you find podcasts. Join the conversation or suggest a guest by following the show on X and LinkedIn.
