Thursday Jun 25
RESEARCHOpenAI

AI Cracks Cases That Stumped Doctors

25JUN
18 SOLVED OPENAI DNA

Doctors had closed these files. An AI reopened them. OpenAI's o3 re-read 376 unsolved childhood cases and surfaced 18 answers experts had missed. People confirmed each one.

Specialists had spent months on each patient and run out of ideas.

The o3 model went back through genomic data and flagged explanations for clinicians to test. Eighteen held up after lab work, ten of them neurodevelopmental. That is a 4.8% gain on top of human review, published in NEJM AI.

The model never decided anything. It pointed; clinicians judged. That gap is the entire point.

full brief & sources

⚡ Why this matters

  • AI as a genomics research assistant just produced real, confirmed results, not a benchmark score.
  • It found answers in cases human specialists had already closed.
  • The human-in-the-loop framing is the model for high-stakes AI rollouts.

🔍 What happened

  • Jun 18: a study in NEJM AI from Boston Children's, Harvard, and OpenAI.
  • The team re-analyzed 376 cases specialists had failed to solve.
  • OpenAI's o3 model surfaced evidence-linked candidate explanations.
  • Physicians confirmed 18 diagnoses, including ten neurodevelopmental and four neuromuscular conditions.
  • That is an added diagnostic yield of 4.8%.
  • OpenAI stresses the model made no clinical decisions.

💬 Smart takes

  • OpenAI: the model did not diagnose any patient; clinicians made every diagnosis.
  • NBC News: AI helped diagnose 18 children whose rare diseases had stumped doctors.
  • Skeptic: a 4.8% lift on 376 hand-picked hard cases is promising, not proof the workflow scales to a normal clinic.

🧭 Where this goes

  1. Likelymore hospitals pilot AI re-analysis of cold genomic cases within a year.
  2. Likelythe FDA's AI guidance lands in 2026 and shapes how these tools get labeled.
  3. Possiblea second-look AI becomes a standard step in rare-disease workups by 2027.
  4. Wild Cardthe first AI-surfaced diagnosis that fails in court chills hospital adoption.

🥄 The Spoon Take

This is the version of medical AI that earns trust. No robot doctor. A tireless second reader that catches what tired experts miss, then hands it back to a human to confirm. Get that handoff right and AI becomes standard in diagnostics. Get it wrong and one bad call sets the field back years.

🤔 Pushback

Cherry-picked hard cases flatter the result. The real test is whether o3 helps or just adds noise across thousands of routine workups.

·OpenAI - Rare disease diagnosis·NBC News - 18 children diagnosed·TechTimes - o3 solves 18 cases