David Fajgenbaum on drug repurposing, AI, and saving his own life
Digest
This podcast explores the transformative potential of AI in medicine, particularly for rare diseases. It features David Faganbaum, a patient and co-founder of Every Cure, a nonprofit using AI to find treatments for rare conditions. The discussion highlights Faganbaum's personal journey, the challenges of rare diseases, and the development of AI platforms like "The Matrix" and "Orchard" to identify new uses for existing drugs. The episode contrasts drug repurposing with new drug development, emphasizing AI's role in accelerating this process. It also delves into the practicalities of FDA approvals, the importance of human expertise in validating AI findings, and the systemic changes needed to ensure these discoveries reach patients. The conversation touches upon compute costs, the ethical considerations of AI data, and the future of predictive and preventative healthcare, concluding with a look at Google's Gemini 2.5 Flash for developers.
Outlines
AI's Role in Revolutionizing Medicine and Tackling Rare Diseases
The podcast introduces the potential of AI in medicine, focusing on the vast number of diseases lacking treatments, especially rare or "orphan" diseases affecting millions globally, with a high percentage having no approved treatments.
David Faganbaum's Journey: From Patient to Pioneer for Rare Diseases
The episode introduces David Faganbaum, a patient diagnosed with Castleman disease, who co-founded the Castleman Disease Collaborative Network and Every Cure, a nonprofit leveraging AI to find treatments for rare diseases. His personal experience and promise to his mother fueled his mission.
Navigating the Patient Experience and Co-founding the Castleman Disease Collaborative Network
David shares his frightening diagnosis experience as a medical student and the lack of understanding within the healthcare system. Despite near-death experiences, he co-founded the Castleman Disease Collaborative Network to focus on repurposing existing drugs and involving patients in research.
Discovering Serolimus and the Power of Drug Repurposing
The podcast details the discovery of Serolimus, an organ transplant drug, as a potential treatment for Castleman disease by analyzing data and identifying key biological pathways. This success highlights the efficiency of drug repurposing over new drug development.
The Birth of Every Cure and AI-Driven Drug Discovery with "The Matrix"
Following the success with Serolimus, David Faganbaum and Grant Mitchell established Every Cure to use AI systematically identify new uses for existing drugs. Their AI platform, "The Matrix," maps millions of drug-disease possibilities to prioritize life-saving opportunities.
AI in Action: The Orchard Platform and Human-AI Synergy
A demonstration of Every Cure's "Orchard" platform showcases AI identifying potential drug-disease pairs. The process emphasizes a "human-in-the-loop" approach, where AI predictions are validated by medical experts to ensure accuracy and impact.
Challenges in FDA Filings, Compute Power, and the Urgency for AI in Medicine
The discussion covers the complexities of FDA filings for repurposed drugs, the high compute power needed for AI, and the rapid advancements reducing costs. The urgency to accelerate AI application in medicine, especially for underserved populations, is stressed.
Every Cure's Impact: Transparency, Patient Empowerment, and Systemic Change
Every Cure's commitment to radical transparency and patient impact is discussed, alongside the challenges of AI data misinterpretation. The need for systemic changes, including government funding and new economic models, to bridge the gap between discovery and patient access for repurposed drugs is highlighted.
Global Solutions, FDA Pathways, and Raising Awareness for Drug Repurposing
The podcast explores global applications of repurposed drugs, like DFMO for sleeping sickness, and discusses existing FDA pathways and economic realities hindering generic drug labeling. Raising awareness through various platforms empowers patients and advocates.
Success Stories, Future Visions, and AI's Role in Reducing Suffering
Compelling success stories of AI-driven repurposing are shared. David envisions a future of preemptive diagnosis and efficient treatment, emphasizing AI's role in reducing human suffering and achieving significant medical advancements.
Google's Gemini 2.5 Flash and Podcast Production Credits
Paige Bailey introduces Google's Gemini 2.5 Flash, highlighting its balance of intelligence, speed, and cost for developers. The segment concludes with podcast production credits.
Keywords
AI in Medicine
Artificial Intelligence (AI) applied to healthcare, aiming to improve diagnostics, drug discovery, personalized treatment, and operational efficiency.
Rare Diseases (Orphan Diseases)
Conditions affecting a small percentage of the population, often with limited research and few or no approved treatments.
Drug Repurposing
The process of identifying new therapeutic uses for existing, approved drugs. This approach can significantly reduce the time and cost of drug development.
Every Cure
A nonprofit organization that utilizes AI to discover new uses for existing medicines, aiming to accelerate the development of treatments for rare and underserved diseases.
The Matrix AI Platform
An AI platform developed by Every Cure that quantifies the likelihood of every FDA-approved drug treating every human disease.
Biomedical Knowledge Graphs
Structured databases representing biological concepts and their relationships, used in AI to map complex biological systems and discover new drug-disease connections.
Human-in-the-Loop AI
An AI system that incorporates human oversight and intervention in its decision-making process, ensuring AI-generated insights are validated by medical experts.
FDA Label Change
The process of updating a drug's official labeling to include new approved indications or uses, crucial for making repurposed drugs widely accessible.
Compute Costs
The financial expenditure associated with the computational resources required for AI model training and data processing.
Gemini 2.5 Flash
A new AI model from Google DeepMind designed for developers, offering a balance of intelligence, speed, and cost-efficiency.
Q&A
How does AI help in the discovery of treatments for rare diseases?
AI platforms like Every Cure's "The Matrix" analyze vast datasets of drugs and diseases to identify potential matches, significantly accelerating the discovery process compared to traditional methods.
What is drug repurposing and why is it important for rare diseases?
Drug repurposing involves finding new uses for existing, approved drugs. This is crucial for rare diseases because developing entirely new drugs is extremely time-consuming and expensive.
How does Every Cure's AI platform, "The Matrix," work?
The Matrix platform quantifies the likelihood of all FDA-approved drugs treating all human diseases by creating a comprehensive map of millions of possibilities, using biomedical knowledge graphs and machine learning.
What is the role of human experts in AI-driven drug discovery?
Human experts review AI-generated predictions, validate findings, provide context, and ensure that the identified treatments are safe and effective for patients, creating a "human-in-the-loop" system.
What are the main challenges in getting repurposed drugs to patients?
Key challenges include lack of market incentives for drug companies, complex FDA approval processes, and the need for widespread awareness. Every Cure is developing an "impact team" to address these issues.
How does Every Cure ensure responsible use of its AI-generated data?
Every Cure plans to release its data responsibly with disclaimers, emphasizing that AI predictions require further validation and human review before clinical application.
What is the significance of "unmet medical need" in prioritizing research?
The "unmet medical need" score helps prioritize diseases that have the fewest or no effective treatments, directing resources towards the most critical health challenges.
Can AI help in diagnosing diseases more effectively?
Yes, AI is being developed for applications like "ambient listening" in healthcare, where it can analyze patient data to assist in early diagnosis.
How is AI expected to impact human suffering in the context of medicine?
AI is envisioned to significantly reduce human suffering by enabling precise drug delivery for specific conditions, leading to more effective treatments and improved patient outcomes.
What are the key features of Google's Gemini 2.5 Flash for developers?
Gemini 2.5 Flash offers a balance of intelligence, speed, and cost, with upgraded reasoning power and controls like "thinking budgets" to optimize latency and costs.
Who are the key individuals involved in the production of the "Possible" podcast?
The podcast "Possible" is hosted by Aria Finger and Reid Hoffman, with production roles filled by Sean Young, Katie Sanders, Edie Allard, Tanasi DeLos, Sarish Leed, Vanessa Handy, Aliyah Yates, Paloma Marino Jimenez, and Melia Hagudelo, and Jenny Kaplan as executive producer.
Show Notes
How can AI really help advance medicine? Should patients and care teams be seeking second opinions from LLMs? This week, Reid and Aria sit down with physician-scientist Dr. David Fajgenbaum, who repurposes existing drugs to save lives—including his own. David shares his journey from receiving last rites as a young medical student to co-founding Every Cure, a nonprofit using AI to come up with ways existing drugs can be repurposed to treat every disease and every patient possible. On this special episode, filmed live in New York City, David, Reid, and Aria explore the circuit of hope, action, and impact that drives medical discovery; the technical and ethical challenges around accelerating AI in healthcare; “agent optimization” as the new “search engine optimization;” and what a future of faster diagnoses and treatment could look like. The result is a conversation about resilience, innovation, and unlocking cures that are already within our reach.
For more info on the podcast and transcripts of all the episodes, visit https://www.possible.fm/podcast/
Topics:
3:00 - Hellos and intros
3:29 - A promise to his mother and the origin of AMF
6:23 - From grief support to medical school
9:18 - Near-death experiences and founding a research network
14:38 - What kept David going at death’s door
16:19 - Discovering a potential treatment and testing it on himself
19:42 - Why drug repurposing is a faster, cheaper complement to drug discovery
22:14 - Co-founding Every Cure and scaling discovery with AI
25:20 - Demo: Building MATRIX and how the algorithm makes predictions
28:41 - Breast cancer and Lidocaine as a case study
30:07 - Why human review is essential in AI for medicine
34:08 - Car exhaust fumes example and FDA pathways
37:31 - Reid’s surprise million-dollar donation for Every Cure compute
39:24 - What AI can unlock across healthcare
41:46 - Building an impact team to close the repurposing loop
47:32 - Improving systemic incentives for generics
54:58 - FDA approval pathways and label change challenges
57:21 - Three life-saving repurposing stories
1:02:20 - Rapid-fire questions
Select mentions:
Hidden Potential by Adam Grant
Chasing My Cure by Dr. David Fajgenbaum
David’s TED Talk: https://youtu.be/sb34MfJjurc?si=GcVleWHZuJ9MqLgS
Possible is an award-winning podcast that sketches out the brightest version of the future—and what it will take to get there. Most of all, it asks: what if, in the future, everything breaks humanity's way? Tune in for grounded and speculative takes on how technology—and, in particular, AI—is inspiring change and transforming the future. Hosted by Reid Hoffman and Aria Finger, each episode features an interview with an ambitious builder or deep thinker on a topic, from art to geopolitics and from healthcare to education. These conversations also showcase another kind of guest: AI. Each episode seeks to enhance and advance our discussion about what humanity could possibly get right if we leverage technology—and our collective effort—effectively.
Table of contents
United States
