Scot Morris, OD, and Rehan Ahmed, MD, discuss how GPT-5 and multimodal foundation models are poised to evolve triage, diagnostics and insurance in a recent episode of Real Talk.
GPT-5 is “a move beyond chatbots,” as Dr. Ahmed says, because it combines language, vision and planning into one system, creating a comprehensive answer based on what was provided.
“A patient can upload a picture of their red eye and tell them the symptoms,” he says. AI takes a look at the red eye, correlates it with the symptoms and provides a differential diagnosis, probable diagnosis, explanation, etc. “A company could potentially take that and build upon it, which is knocking on the doors of the next generation of AI triage tools like bots or assistive diagnostics support systems,” Dr. Ahmed says.
How Does AI Actually Perform?
From there, the pair highlighted a Cedars-Sinai report detailing the efficacy of its AI-powered virtual care platform, Cedars Sinai Connect. The goal of the system is to streamline the patient intake process to make things smoother and more efficient for clinicians, while also providing on-demand care for patients whenever they need it.
Patients input their symptoms into the Cedars Sinai Connect platform, where the AI works to cross-reference that input with the patient’s medical history and records, as well as with other patient data that yields similar complaints. Then, the AI summarizes the conversation and provides recommended next steps.
The AI “was actually rated optimal more frequently than the physicians working without it,” Dr. Ahmed says. The AI outperformed physicians by about 10 percentage points in some assessments.
Because this system allows clinicians to review the AI’s responses, make their own recommendations and work with the patient to get the best outcome, it’s a truly multimodal approach between AI and the doctor. “AI plus HI, human intelligence, that’s a pretty unbeatable combination,” Dr. Morris says.
Augmenting Medical Care
Dr. Morris says that medical knowledge is “doubling every 71 or 73 days,” and predicts it may double daily within a decade. AI systems will summarize and synthesize this research in seconds, which he finds to be essential to delivering personalized, up-to-date care.
Dr. Morris stresses that AI will not replace clinicians but augment what is already provided. For example, a patient who says “I don’t do antibiotics” will need other alternatives based on their values and preferences. AI will find evidence-based solutions for those preferences to form a personalized framework that enhances care.
Advanced adaptive systems will allow analysis of hundreds of thousands or millions of patient data points. That scale will reveal which treatments actually work, and large-scale exam data can reduce bias while protecting patient privacy. Clinicians can build on that evidence to change practice.
How Will AI Affect Insurance?
The episode wraps up with a discussion on how AI will affect vision care plans and the insurance industry.
Dr. Morris kicks things off with a critique of the pre-authorization system, calling it a “complicated, convoluted cycle.” Providers send requests, get denials and sometimes never learn the outcome. “Four is the magic number,” he says of denials that later reverse, and each authorization can take five to seven minutes to prepare. That time adds up across staff and clinics.
Dr. Ahmed says insurers use risk pools and actuarial data to manage costs, which both doctors agree makes the process feel designed to deter care. They argue AI could streamline eligibility checks and approvals. Because insurers use actuarial models, Dr. Morris suggests real-time eligibility and instant prior authorization could replace long waits, and AI could improve risk assessment or personalization.
Dr. Morris has some concerns about payers banning provider AI agents and rival insurers willing to embrace AI that disrupts the insurance market. Both doctors say a system that lets patients hold and share their data could enable hyper-personalized plans, as well as reward healthy behavior, while also keeping insurers financially viable. They say this hinges on incentives, regulation and who controls data.
The clinicians predict AI will continue to drive major change in eye care and beyond, and they urge fellow practitioners to stay current and adopt helpful tools. Dr. Morris warns that those who ignore AI risk falling behind.
For more on this conversation, listen to this episode of Real Talk.
Author
