No Priors Ep. 130 | With OpenEvidence Founder Daniel Nadler

The rapid adoption of new AI technologies is compressing innovation cycles from decades down to a year or two. OpenEvidence exemplifies this, becoming the “operating system for clinical knowledge” for 40% of US doctors in just 18 months. It focuses on the niche of high-stakes clinical decision support, an area with significant consequences where a physician has only one chance to get it right. This is distinct from lower-stakes applications like medical scribing or paperwork. The platform addresses a critical issue: the “golden age of biotechnology” has inadvertently created a “dark age of physician burnout,” as the explosion of new drugs and treatments makes it impossible for practitioners to keep up.

Solving Medicine’s Semantic Search Problem

At its core, OpenEvidence solves a highly complex semantic search problem. Traditional keyword-based search fails in a clinical context because a physician’s query is often a multi-sentence narrative describing a complex patient scenario.

An illustrative example:

A 44-year-old female patient has moderate-to-severe psoriasis. The standard treatment would be a cream, but she also has Multiple Sclerosis (MS). The dermatologist, who is not a neurologist, knows that new biologic treatments for psoriasis (like IL-17 and IL-23 inhibitors) might interact with her neurological condition. This information is new; these drugs were approved long after the doctor finished medical school.

The actual query is a paragraph: “For a 44-year-old female patient with moderate-to-severe psoriasis and co-morbid MS, is an IL-17 inhibitor or an IL-23 inhibitor more appropriate and safely tolerated with respect to not aggravating the MS?”

This is not an academic question. An IL-17 inhibitor will worsen the MS, while an IL-23 inhibitor is safe. OpenEvidence is built to semantically understand the entire complex query, much like another human physician would. Once the query is understood, the process becomes deterministic. The platform can pinpoint the exact snippets from the vast corpus of biomedical literature—such as a Phase 3 Randomized Controlled Trial (RCT) in the New England Journal of Medicine—that directly address the question, even if the key information is buried deep in a methodology section.

The Strategy for Viral Adoption

Several strategic decisions were key to OpenEvidence’s success:

• A Deliberate Social Contract: The platform was never framed as an “AI” or an “answer engine.” Instead, it was presented as a next-generation search engine. From day one, citations and references were a first-class citizen, allowing physicians to interrogate and audit the source material. This built trust and positioned the tool as a router to primary sources, not a black-box oracle.
• Targeting Physicians, Not Patients: By focusing exclusively on physicians, the platform leverages the user’s expertise. The doctor’s MD is on the line, so they are motivated to use the tool responsibly. This allows OpenEvidence to present conflicting or ambiguous evidence, trusting the physician to apply their clinical judgment.
• Treating Knowledge Workers as Consumers: The most profound insight was to treat doctors like consumers, not as appendages of a health system. Historically, medical technology was sold top-down via hospital administrators (e.g., Chief Medical Information Officers). OpenEvidence circumvented this by launching as a free app on the App Store. Doctors could download and use it on their personal iPhones, creating a powerful, bottom-up adoption wave that traditional gatekeepers couldn’t stop.

The Future of Medicine and Education

The fundamental role of the doctor will persist for the foreseeable future, much like pilots remain in cockpits despite autopilot technology. Human psychology requires a personified source of trust in high-stakes domains.

However, medical education is set to radically change. The rate of doubling of medical knowledge has shrunk from every 50 years in 1950 to as little as every 73 days today. A more conservative estimate, looking only at top-quartile literature, places the doubling rate at every five years. It is impossible for any human to keep up.

This reality will invert the current educational model:

• Continuing Medical Education (CME) will become the majority of a physician’s training, rather than an afterthought. World-class physicians already report that the vast majority of what they practice was learned post-residency, often in the last few years.
• Training will shift to emphasize evidence-based medicine over purely guideline-based medicine, encouraging the use of tools to consult primary literature.
• Concepts like the “curbside consult” (asking other physicians for input) and distributed, interdisciplinary decision-making will become central. AI tools can help scale this “hive mind” approach, especially in under-resourced and rural areas.

On Motivation and Recruiting

Success is not a product of a brilliant idea conceived in a coffee shop, but a result of an extraordinary will to power and the belief that reality is malleable. This philosophy values will over pure rationalism.

• The Source of Motivation: The most important asset for a founder is a compulsive level of motivation. This drive often originates from deep-seated sources, such as trauma or a desire to prove someone wrong. It is crucial not to over-analyze these origins, because “in the analysis and description of something, you kill it.” The propulsion system should be used, not dissected.
• Recruiting Philosophy: The goal is to find people who are not only intelligent but possess their own internal propulsion system. Seek out individuals who are intensely driven and aggressive. For such people, traditional management techniques for motivation and inspiration are redundant. The best leadership approach is to simply get out of their way.

• Tags:
• Feature
• Semantic Search
• Clinical Decision Support
• Medical AI
• Physician Burnout
• Knowledge Work
• Entrepreneurship