This morning I walk into my office at Epic for the last time. The future holds plenty of intrigue. Today, though, I'm grateful — 21 years' worth — for the customers who shared their dreams and challenges, the industry friends united in helping patients and caregivers, and the Epic colleagues who invented, sharpened one another's ideas, and relentlessly built.
You all taught me so much. Thank you.
Keep doing good!
Recently, a friend asked me, over a cup of coffee, what has AI actually changed around me personally. I immediately noted that the quality of understanding one can develop of topics is deeper through dialogue - and AI makes dialogue more accessible.
The combination of deep reading paired with dialogue - both with fellow humans and AI - is a skill we need to foster and teach.
1
11
4,895
Researchers at @Yale using Curiosity—Epic's generative medical event model, which simulates a patient's likely future course from their record — looked at adults discharged from the ED with abdominal pain, one of the most common and most uncertain presentations in emergency medicine. Traditional risk models answer a binary question: will this patient come back? Curiosity answers a richer one: if they return, when, and will that return end in discharge or admission?
Two patients can carry the same overall revisit probability but face very different paths—one headed toward a quick outpatient follow-up, another toward an early admission. A binary score flattens that difference. A trajectory surfaces it.
A single pretrained model, like Curiosity, can generate calibrated predictions across many outcomes without retraining. As a result, the focus and time shift away from building a bespoke model for every clinical question and toward deciding which questions are worth asking. The team moved from idea to results in a matter of weeks.
It also highlights an emerging skillset: clinicians who pair deep clinical reasoning with real fluency in generative AI. The questions they choose to ask are the key ingredient.
Preprint here: medrxiv.org/content/10.64898…
Find more on Curiosity here: epic.com/epic/post/curiosity…"
Longtime listener; first time caller. Had a blast exploring a wide range of topics - from scaling laws to Paperboy (the Nintendo game) - with @arjunmanrai and @AndrewLBeam .
In the latest episode of AI Grand Rounds, @sethHain, senior VP of R&D at Epic (@HeyEpic), describes how his company is building foundation models that respect institutional autonomy, minimize burden, and prioritize safety. Full episode: nejm.ai/ep39
ALT AI Grand Rounds Episode 39 Epic’s Approach to AI with Seth Hain A photo of Seth Hain
Health systems are leading the way on AI research and adoption. It was an honor to share their real-world results with the tech community at #MSIgnite.
The speed and scale of what’s happening in healthcare right now is remarkable. And what’s coming next is even more exciting.
At Epic, we create technology to help people get and stay well.
AI is embedded throughout our applications—from the clinic, to the bedside, to the back office.
Worldwide, hundreds of thousands of clinicians are using it to improve care.
Here it is in action at #MSIgnite:🧵
Geniunely enjoy ever conversation I get to have with @mattlungrenMD and @JustinNordenMD - including this one we decided to record.
youtube.com/watch?v=k3q3xtoT…
☄️☄️☄️ Pleased to share our latest research on creating personalized medicine at scale by combining Epic Cosmos and generative AI.
We created CoMET, a decoder‑only transformer that generates next medical events to simulate patient trajectories. Trained on 115B medical events from Epic Cosmos; evaluated on 78 real-world tasks spanning diagnosis prediction, disease prognosis, and healthcare operations; established scaling laws on medical event transformers; and, importantly, determined that clinical performance improves along with scaling.
Grateful to both the Epic Cosmos community for making this possible and the collaboration across @Yale, @MSFTResearch, and @HeyEpic.
#ugm2025
arxiv.org/abs/2508.12104
Cogito is my favorite name for any product we have at @HeyEpic.
10
624
Results such as these are strong early indicators of the slope of improvement - something I worry too many are not paying close enough attention to.
Today, we at @OpenAI achieved a milestone that many considered years away: gold medal-level performance on the 2025 IMO with a general reasoning LLM—under the same time limits as humans, without tools. As remarkable as that sounds, it’s even more significant than the headline 🧵
1
1
481
Grateful for the chance to explore with @KeithFiglioli where AI and healthcare may be heading. I always value the deeper conversations - hope others do as well.
In the new #HealthcareIsHard series on emerging tech, @SethHain of @Epic shares why gen AI agents, a 300-million-patient Cosmos dataset, and a “health grid” may remake workflows, reshape the EHR & foster the next generation of a learning healthcare system. lrvhealth.com/podcast/emergi…
1) Generative AI has exited the novelty phase
Epic’s first arc of generative AI was focused on efficiency: how we could rapidly improve people’s existing work. That arc has reached widespread adoption. Today, more than 75% of our health system customers use gen AI. At @ChristHospital, for example, it has helped oncologists treat nearly 70% of lung cancer cases at stage I or II, well above the national average. @myNGHS has seen a 46% improvement in accuracy and efficiency of coding. Nurses at Mercy, @UNC_Health_Care, and @MethodistHosp opt to use more than 90% of the end-of-shift notes drafted by Epic AI. These are just a few examples of how generative AI is already making a practical difference.
1
297
Big day in my hometown of Seward, NE! Growing up my grandparents lived a block from the time capsule.
#OnThisDay The World's Largest Time Capsule was sealed in Seward, Nebraska on July 25, 1977. Encapsulated in the 20x8x6-foot vault is reported to be a brand-new Chevy Vega, a motorcycle, and more!
After 50 years, the capsule is to be opened on July 4, 2025.
1
440
⟦Δ⊟(t ⟶ ∞)⟧ ≙ ϵ₀⭓⟁∰⥣⟁Δ : ⇶⚗︎₇.121ᚚ⭑Ϫ⟁⇌≏∴ ⛉∅ ⫰ϗ⚛︎⇥Ϟ⤳ ⟊◩⤧πˢ⟁⟁ ‖ ⇇⟁⊶⟁≩⟁∫⊡ ∋𓇕⚿ ⨆𝜐⧊ᵯⴀ∶ϐ⌶ ↛ 42.000^Ⱶ ⚚φ⛒---
1
1
190
Flexing time —accelerating into AI when the paradigm shifts or patiently letting integration compound—is Epic (@HeyEpic).
@AcquiredFM is spot on (as always): "they were able to basically use time as a resource that most people don't flex."
acquired.fm/episodes/epic-sy…
I always appreciate my conversations with @julesyoo — this time, we recorded it. Hope you enjoy.
--@HeyEpic & @a16zBioHealth--
a16z.com/podcast/the-epic-ar…
Imagining a thoughtfully curated, hard-bound book (a la @stripepress) collecting the essential essays and papers demonstrating how society dealt with this moment in AI. Trying to compile what truly deserves to make the cut.
1
2
757