Vital Signs with Jacob Effron and Nikhil Krishnan

Ep 38: Oscar Co-Founder Mario Schlosser on Lessons from Implementing LLMs and How AI Will Impact Healthcare

Feb 13, 2024

Mario Schlosser, Co-Founder of Oscar Health, discusses the impact of AI in healthcare, limitations of GPT-4, and why we can't have robot doctors. They also talk about sharing AI knowledge, developing healthcare-specific models, and the need for a safety layer in LLMs. Other topics include commercial opportunities in healthcare, the possibility of AI doctors, and overhyped/underhyped aspects of AI. Oscar Health's unique approach and their interest in exploring AI use cases in gaming are also highlighted.

01:13:39

Creator website

Episode guests

Mario Schlosser

AI Summary

AI Chapters

Episode notes

Podcast summary created with Snipd AI

Quick takeaways

AI can greatly impact healthcare by improving outcomes, reducing costs, and assisting in decision-making.

Oscar Health leverages AI to enhance customer experience and personalize messaging, improving member retention and engagement.

Implementing AI in healthcare requires proper alignment with regulations, addressing complex workflows, and balancing interpretability and performance.

Deep dives

The Impact of AI in Healthcare

AI in healthcare has the potential to greatly impact the industry. Healthcare is a unique field with a mix of formal and informal language, making it a suitable domain for AI algorithms. The use of AI in healthcare has been limited in the past, but advancements in language models like GPT-4 have opened up possibilities. AI can be utilized in administrative tasks like claims processing, where it can summarize complex rules and provide clearer explanations to both experts and laypeople. AI can also help improve healthcare outcomes by reducing costs, enhancing transparency, and assisting in clinical decision-making. However, challenges remain, such as the need for interpretability, addressing biases, and bridging the gap between formal and informal language.

Introduction

5min

Adapting Language Models for Different Audiences: Insurance and Clinical Use Cases

23min

Limitations of Language Models in Customer Service

14min

Sharing Research and Knowledge in AI

19min

Evolution of Individual Markets and Selling Software in Healthcare

7min

Quick Fire Round on Healthcare AI and Policy Change

2min

Exploring AI Work at Oscar and Game Concepts

3min

Oscar Health is a $4B public healthcare company, providing healthcare insurance to nearly 1 million members. Oscar is at the forefront of AI adoption, continuously developing new AI use cases in healthcare. On this week’s episode, we sat down with Oscar Health Co-Founder, former CEO, and now President of Technology Mario Schlosser to talk about where AI will have the biggest impact in healthcare, limitations of GPT-4 in healthcare, and why can't we have robot doctors today.

(0:00) intro

(1:06) how will AI change healthcare in the next decade

(9:09) how Oscar uses AI

(18:40) how to build around healthcare requirements

(25:46) when would GPT-4 fail "miserably" and fundamental limitations of LLMs

(36:28) we shouldn’t piss off our smartest robots

(38:15) sharing AI knowledge between companies

(41:50) developing healthcare-specific models

(44:35) hackathons and karaoke nights at Oscar

(49:07) the need for a safety layer in LLMs

(51:33) best commercial opportunities in healthcare

(55:19) will their be AI doctors this decade?

(59:18) over-hyped/under-hyped

(1:00:49) most exciting AI company?

(1:08:46) over-hyped/under-hyped

Out-Of-Pocket: https://www.outofpocket.health/

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more