Conference Presentation, Panel
Will AI Deliver on the Promise of Better, Faster, Cheaper Health Care? | Asia Summit 2025
Milken InstituteSung Hee Choe, Lohendran (Logen) Baskaran, David Berry, Adrian Lam, Wan Yue, Dr. Logan Baskerin, Dr. Yuan Yueyuan, Dr. Wan, Dr.-Vosker, Wanyu, Logan, UA, David Rockefeller Jr.
- AI activities are expected to transition from research and investment to clinical practice, specifically impacting patients and requiring regulatory healthcare approval for pilot algorithms that interpret dynamic information from static images to guide decisions like stenting.
- The sector anticipates a $5 trillion US health industry being rewritten by AI, consumerization, longevity, and decentralization, with DNA and RNA foundation models becoming instrumental for personalized medicine and AI designing more stable, protein-rich mRNA drugs that are manufactured more efficiently.
- AI is predicted to enhance diagnostic capabilities by interpreting data from ubiquitous sensors (e.g., voltage gradients, PPG) to enable testing for patients in remote areas without specialist travel, while also accelerating the diagnosis of complex dynamic conditions and mental health cases where average patients face multiple misdiagnoses.
- Projections indicate that over the next five years, remote patient monitoring could address current limitations in processing complex data with poor input, and regulatory frameworks may evolve to potentially eliminate certain clinical trials using AI, with a forecast of 10 IgA nephropathy therapies approved by the FDA within a year compared to zero currently.
- Augmentative AI is viewed as the primary category for assisting physicians to see more patients and manage back-end workflows, though AI will not replace doctors soon as the human touch remains essential for liability, complex data incorporation, and initial patient assessment.
- Significant barriers include multifaceted problem complexities involving IT and nursing staff, data sanctity issues across different countries that hinder sharing even with anonymization, and the need to standardize data sets to prevent racial and ethnic biases in disease risk assessment.
- Strategic plans involve developing regional data sovereignty models, such as Singapore sequencing 10% of its population for Asian genetic equity, creating tailored risk scores from longitudinal studies, and licensing technology by geography to prevent data shifting.
- Risks identified include the unacceptable nature of AI hallucinations and errors in front-line diagnostics, the difficulty of building patient-facing tools due to operational complexities, and the potential for a widening divide between AI enthusiasts and skeptics if literacy is not improved for all ages.
- Future expectations include the need for policymakers and regulators to adopt a more risk-taking approach regarding payment schemes and data sharing to unlock personalized care, alongside a call for increased philanthropic and investor risk tolerance, particularly in the Southeast Asian market despite reimbursement challenges.