Assertion Partly supported AI assessment confidence: 90% certainty 4/5 debate potential 3/5

HumanDX: Physician collectives outperform 90% of individual doctors in diagnostic accuracy

Jay Komarneni · a16z Podcast | The Taxonomy of Collective Knowledge · Jan 2, 2019 · at 10:21

Jay Komarneni, founder of HumanDX, discusses how crowdsourcing and aggregating medical opinions across doctors achieves higher diagnostic accuracy on reference cases than individual physicians.

0:00 / 0:07exact quote · 7.6s

▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark

“We're now seeing that a collective of multiple physicians can outperform 90 plus percent of individual physicians.”

quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →

The claim, plainly

A collective of multiple physicians evaluating cases on HumanDX outperforms over 90% of individual physicians evaluating reference cases.

a quote alone can be ambiguous, so this restates it with the references and context of the conversation around it resolved. AI-written; the tape below the player is the source.

Assessment

The core premise of the claim—that collective intelligence on Human Dx significantly outperforms individual physicians—was validated in a major peer-reviewed study published in JAMA Network Open in March 2019. The study showed that pooling decisions from 9 physicians increased diagnostic accuracy from an individual average of 62.5% to 85.6%. However, the published research evaluated accuracy in terms of group mean performance versus individual mean performance rather than percentile rankings, and absolute collective accuracy topped out at 85.6%, without explicitly demonstrating that collectives outperformed 90% or more of individual diagnosticians.

JAMA Network Open (March 1, 2019): 'Comparative Accuracy of Diagnosis by Collective Intelligence of Multiple Physicians vs Individual Physicians' analyzed 1,572 cases across 2,069 users on Human Dx.
The study demonstrated that individual diagnostic accuracy averaged 62.5%, whereas a collective of 9 physicians achieved 85.6% accuracy (p < .001).
The published study did not report percentile distribution metrics comparing collective accuracy directly to the top 10% of individual performers.

Checked against

[1] nih.gov [2] ahrq.gov [3] researchgate.net [4] mdedge.com [5] nih.gov [6] nih.gov [7] pnas.org [8] mpg.de

"Comparative Accuracy of Diagnosis by Collective Intelligence of Multiple Physicians vs Individual Physicians" "90%" OR "percentile" OR "percent" Human Dx white paper collective intelligence physicians 90 percent Human Diagnosis Project diagnostic accuracy collective individual physicians paper research Human Diagnosis Project JAMA diagnostic accuracy collective individual physicians "Human Diagnosis Project" "outperforms" "individual physicians" OR "physician collectives" Human Dx diagnostic performance collective outperforms individual physicians paper

assessed as of 2026-07-31

This assessment is an AI reading of a public statement, made against the cited references with the confidence shown. It can be wrong. The quote, the tape and the references are all on this page; if something is off, use report an error below. full disclaimer →

Mentioned

HumanDX physicians

⚑ Spot an error? Report it

More from Jay Komarneni

Insight

Jay Komarneni: Human input is required for human-interpretable AI ontologies

“The only way you're going to get the representations that are most valuable to humans is from human beings themselves, right?”

Jay Komarneni Jan 2, 2019 ▶ 4:23 a16z Podcast | The Taxonomy of Collective Knowledge

Insight

Komarneni: Token rewards incentivize decentralized collective intelligence and knowledge creation

“The ability to ultimately compensate people with application specific tokens is a really interesting incentive to use ontologies and distributed knowledge creation, collective intelligence to come to better answers around given issues or given problems.”

Jay Komarneni Jan 2, 2019 ▶ 21:50 a16z Podcast | The Taxonomy of Collective Knowledge

Insight

Jay Komarneni: Human intelligence excels at multi-scale data synthesis

“One place where that you see that natural place that humans are really good is when there's a high number of scales of different types of information or data, right? So, for example, in healthcare, everything from how your mitochondria or the electron transpor…”

Jay Komarneni Jan 2, 2019 ▶ 22:51 a16z Podcast | The Taxonomy of Collective Knowledge

Assertion Not checkable as stated

Jay Komarneni: HumanDX rewards physicians based on medical data demand

“In human DX, we actually like to use what we call impact in the system, and it allows us to differentially provide contributors who contribute more valuable contributions to the system in terms of what contributions the system most needs. So perhaps there's a …”

Jay Komarneni Jan 2, 2019 ▶ 19:33 a16z Podcast | The Taxonomy of Collective Knowledge

Assertion Not checkable as stated

Komarneni: Duolingo inspired HumanDX's gamified clinical case interface

“Louise had actually his work on Duolingo and really kind of this idea of creating these micro interactions that almost have this gamified structure was really a major inspiration also for the way that we built HumanDX to be these brief interactions where you c…”

Jay Komarneni Jan 2, 2019 ▶ 20:05 a16z Podcast | The Taxonomy of Collective Knowledge

HumanDX: Physician collectives outperform 90% of individual doctors in diagnostic accuracy

More from Jay Komarneni

Turn any episode into a week of clips.