Serial entrepreneur Munjal Shah is pioneering the use of artificial intelligence (AI) for non-diagnostic patient support through his newest venture, Hippocratic AI. He believes recent advances in conversational AI could help address systemic gaps in empathy and capacity within the global healthcare workforce.
Shah sees potential for large language models (LLMs) like ChatGPT to reinvent chronic care management. By assimilating medical knowledge and learning to communicate empathetically, these systems could provide personalized guidance better scaled to patients’ needs.
However, Shah cautions that hype surrounding LLMs should be tempered with realism. While excellent communicators, they do sometimes present false information. This makes high-stakes diagnoses unsafe currently, but Shah sees plenty of impactful, lower-risk applications.
Hippocratic AI focuses on support services like medication reminders, appointment coordination, transportation assistance, and access to community resources. Shah stresses that the capacity for this emotional and logistical support is critical for recovery, especially for chronic conditions. Yet, constantly overburdened health systems often need more dedicated staff for meaningful patient outreach.
This is where AI could provide “super staffing” – conversational agents with immense capacity for personalized engagement. Hippocratic AI custom trains its LLM on actual health worker-patient dialogues to communicate knowledgeably and empathetically. It refines the model further through feedback from medical professionals.
Shah believes that such systems can mature over time into fully autonomous agents delivering wide-ranging patient support. By helping patients better self-manage conditions, they could reduce costs and improve capacity within increasingly overtaxed health systems.
Of course, trust and adoption remain open questions when introducing new AI tools to healthcare. However, Shah sees conversational agents as promising to enhance human capabilities and access. If progress continues apace, he believes LLMs’ potential for revolutionizing support roles remains underestimated.
For now, Hippocratic AI offers an intriguing test case for fruitful applications of AI in medicine. While cautious about hype and limitations, Shah seems confident in the transformative power of LLMs. If his vision pans out, artificial intelligence could soon become integral to patient care.