Ep 27: Oscar Co-Founder Mario Schlosser on How LLMs Can Be Used in Healthcare Today and The Path to AI Doctors
Feb 13, 2024
auto_awesome
Oscar Health Co-Founder Mario Schlosser discusses the impact of AI in healthcare, top AI use cases at Oscar, adoption challenges, and limitations of GPT-4 in healthcare. He also shares his views on open-source vs. off-the-shelf vs. healthcare-specific LLMs and why we can't have robot doctors today.
AI can significantly impact healthcare by processing formal and informal language, streamlining processes and increasing transparency.
Oscar Health uses AI in administrative tasks to improve operational efficiency and in clinical areas to improve outcomes and reduce costs.
Implementing AI in healthcare requires navigating regulatory requirements, ensuring data privacy and security, and striking a balance between innovation and regulatory compliance.
Deep dives
AI in Healthcare will Benefit from Formal and Informal Language Processing
One of the main areas where AI can have a significant impact in healthcare is in processing both formal and informal language. Healthcare involves a mix of highly structured and regulated information, such as medical codes and guidelines, as well as human language in conversations between patients, physicians, and in electronic medical records. This unique combination has led to comparatively less algorithmic coverage in healthcare. However, current advancements in language models, like GPT-4, offer the potential to bridge this gap by effectively processing and understanding both formal and informal healthcare language. Implementing AI models in areas like healthcare administration, such as claims processing or creating clear summaries of medical information, can streamline processes and increase transparency. The goal is to leverage AI to create a healthcare system that provides real-time information, cost transparency, and eliminates issues like claims denials and complicated authorization processes.
Oscar Health Focuses on Administrative Efficiency and Clinical Use Cases
Oscar Health, a prominent health insurance company, has been actively exploring AI implementations in both administrative and clinical areas. On the administrative side, Oscar leverages AI to improve operational efficiency, such as automating tasks like call summarization, lab test summarization, generating secure messaging medical records, and providing claims explanations to internal care guides. These implementations not only save costs but also ensure clearer communication and better decision-making. On the clinical side, Oscar aims to improve medical outcomes and reduce costs. While clinical AI use cases still pose challenges related to contextual understanding and aligning with human reasoning, Oscar is prioritizing administrative implementations that save costs and provide transparency, with an eye towards future breakthroughs in clinical applications of AI.
Navigating the Challenges of AI Adoption in Healthcare
Implementing AI in healthcare brings its own set of challenges, including navigating regulatory requirements such as HIPAA, ensuring data privacy and security, and building trust among healthcare providers. Oscar Health takes these challenges seriously, having established business associate agreements (BAAs) with AI service providers. While healthcare-specific AI models are being developed, they often fall short in alignment and understanding of instructions compared to more general-purpose models like GPT-4. Oscar's approach involves leveraging the power of large-scale language models like GPT-4, while maintaining HIPAA compliance through strategies such as using synthetic or anonymized data during model training and testing. The key is to strike a balance between innovation, operational efficiency, and adhering to regulatory guidelines to build trust and realize the full potential of AI in healthcare.
AI Solutions for Democratizing Analytics
AI solutions like AI and gender, VR can democratize analytics in organizations by making it more accessible to a wider range of people. It is important for leaders to introduce more individuals to AI tools to accelerate progress.
Limitations of Language Models
Language models (LOMs) have limitations when it comes to tasks like call summarization. GPD4, a powerful language model, struggles with call summarization as it requires processing a large number of tokens and complex layer-by-layer processing. However, this limitation can be addressed by breaking the task into substeps and utilizing a chain of thought approach.
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 of Unsupervised Learning, 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, top AI use cases at Oscar today, AI adoption challenges Oscar is facing, and limitations of GPT-4 in healthcare. Mario also shared his takes on open-source vs. off-the-shelf vs. healthcare-specific LLMs, and why can't we have robot doctors today.
(0:00) intro
(1:26) how will AI change healthcare in the next decade
(9:29) how Oscar uses AI
(19:00) how to build around healthcare requirements
(26:06) when would GPT-4 fail "miserably" and fundamental limitations of LLMs
(36:48) we shouldn’t piss off our smartest robots
(38:35) sharing AI knowledge between companies
(42:10) developing healthcare-specific models
(44:55) hackathons and karaoke nights at Oscar
(49:27) the need for a safety layer in LLMs
(51:53) best commercial opportunities in healthcare
(55:39) will their be AI doctors this decade?
(59:38) over-hyped/under-hyped
With your co-hosts:
@jacobeffron
- Partner at Redpoint, Former PM Flatiron Health
@patrickachase
- Partner at Redpoint, Former ML Engineer LinkedIn
@ericabrescia
- Former COO Github, Founder Bitnami (acq’d by VMWare)
@jordan_segall
- Partner at Redpoint
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
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