excited to share @breyercapital's latest healthcare thesis with my partner, @jimihendrixlive. healthcare is at an inflection point. scientific discovery is accelerating, but systems of translation remain structurally stagnant. the result is a widening chasm between what’s possible in principle and what’s practiced. we invest where scientific discovery, clinical necessity, and institutional transformation converge. computation, precision, and prevention are shaping the future of medicine, but only when anchored by economic models that reward long-term value. the next era of human health won’t be inherited. It will be engineered. if this vision resonates, we’d love to connect.

We are starting a blog series at Network where our technical team writes about their work. "Scaling Laws Arrive in Liquid Biopsy" is our first post. In language models, scale brought generalization. We think that's now happening in diagnostics. 🧵

Today @AMD announced it will acquire @theworldlabs, with @drfeifei joining as chief scientist. It's been wonderful working with Fei-Fei at @StanfordHAI, the institute she co-founded, where I have served on the advisory council since inception. ImageNet taught machines to see; World Labs asked how they might understand the three-dimensional world they see. We at Breyer Capital are grateful to have been part of the earliest investor group in the seed. She could not have found a better home. My good friend @LisaSu rebuilt AMD on engineering discipline and long-horizon bets, and this is another: the hardware for the next era of AI will be designed alongside the people building the models. Congratulations Fei-Fei and Lisa.

congratulations to @drfeifei on the $8.2b acquisition of @theworldlabs by @AMD, and to my partner, Jim Breyer, on a stellar early investment. we look forward to seeing how this partnership blossoms under the remarkable leadership of Lisa Su, CEO of AMD. @jimihendrixlive

some ideas for “muse for healthcare”: - a morning digest pre-charting all patients for the day - ambient monitoring for results (labs, imaging, etc) and notification of clinically actionable findings - in-basket prioritization and triage - auto-dialing other clinical sites to obtain records or other info - monitoring of ED visits, admissions, or other specialist visits for a patient “watch list”

In our newest episode of TechBio Talks, we explore how AI and clinical data are converging to reshape the future of medicine. CEO of Recursion, Najat Khan, PhD talks with Dr. Morgan Cheatham, Partner and Head of Healthcare and Life Sciences at Breyer Capital. As a physician, medical geneticist, and investor, Morgan brings a unique perspective, having spent nearly a decade backing category-defining companies spanning diagnostics, therapeutics, and care delivery. We unpack: 🚀How AI is reshaping the fundamentals of biotech value creation 💊Why “modality market fit” is crucial for matching treatments to patient tolerance 🔬Closing the loop between the wet lab, dry lab, and the clinic 🧬Recasting medicine as an information science to close the gap between discovery and care Listen and follow on: YouTube: piped.video/LmVszrLN0cg?si=5ZF_… Spotify: open.spotify.com/episode/6BE…

the cost of a genome with good coverage will fall below $100 during this decade. we need to address the other bottlenecks keeping genetics from transforming clinical medicine beyond oncology and rare disease: 1) make it easier for clinicians to order genomic tests, and 2) build scalable systems for interpretation

Is it time for a radiology society devoted to AI opportunistic screening yet? The field seems to be gathering momentum: -Breast 5-year AI risk: good data, commercially deployed -Osteoporosis screening from CTs: mature, commercially deployed -ASCVD risk from CT/mammo: emerging, commercially deployed (mature in the form of quantification of calcs at least) -5-year cancer risk from chest CT: good data -Cardiometabolic risk from CXR, abdominal CT/MRI: hot topic of interest/research And that's just off the top of my head. We're in the imaging biomarker explosion era. Or even broader if you consider our cardiology colleagues and what they're doing with echo...

The evolving landscape of drug targets nature.com/articles/s41573-0… rdcu.be/22o1FPpNJ3Ga In the past 25 years, advances in areas such as genomics and the diversification of therapeutic modalities have expanded the drug target landscape, which now includes ~700 targets mapped here

some of the best tech companies require R&D teams to meet end users. a commonly cited example is Epic, where software engineers shadow physicians using the software. it feels like there's a huge gap for this approach in biotech. how often are scientists at the bench (and increasingly, at the command line), meeting patients with the conditions they're focusing on?

the Muse product format of deeply connected chat-based agent + multi-model backend would do numbers in healthcare

who here knows ball? #medtwitter

in the first chapter of healthcare AI, the battleground was access to longitudinal clinical data. in this next phase, it will be access to biological specimens: tissue, blood, synovial fluid, and other samples containing 1,000× more information, linked to clinical context and longitudinal outcomes. the winners will be those who can access, assay, and interpret these specimens at the lowest cost and highest resolution.

Ep #277 is live! Abridge's CEO, Shiv Rao went from skateboarding and selling records in Japan → cardiology → nearly four years “languishing” as a founder before Abridge took off. Today, Abridge is used across 300+ health systems and 100M+ clinical conversations annually. We unpack: • Nearly quitting medicine • Why his “dream job” left him restless • The personal story behind Abridge • Years searching for product-market fit • When generative AI made “the sky just open up” • Going all-in with their “backs against the wall” • Why “pressure makes diamonds” • Building when “every minute matters” • Lessons from Jensen Huang & Rick Rubin Listen here: curiositycentre.com/the-high… Shiv's re-imagined high flyer: @morgancheatham Thx for the research help: Sebastian, Somesh, David, Zen, Holly, Morgan @khoslaventures @vkhosla @a16z @sparkcapital

A video shows a patient with progressive multiple sclerosis attempting to walk before and after treatment with JY231, an off-the-shelf lentiviral vector designed for the direct in vivo engineering of CD19-targeted CAR T cells. Learn more in the preliminary results of a phase 1 trial: nej.md/4zPPwbD

Today we are releasing the OpenEvidence Model Family. Our most powerful models yet, and (as of 2026) the highest scoring medical AI model family in the world on every mainstream benchmark. Osler for the hallway, Sackett for the consult, Snow for the tumor board. Every model in the family is held to the same standard of clinical accuracy. What varies is time: how long a model thinks, and how deep it searches. Three of them are available to every verified clinician starting now. The fourth is Darwin. A perfect 100% on MedQA, the first AI in history to do it. Capability at that level cuts both ways in medicine, so Darwin is in research preview, by application only, while its safeguards are validated with partners like @RareDiseases. Osler, Sackett, and Snow are live now on the web and in the iOS and Android apps.

the cumulative impact of the undiagnosed disease program at NIH over the last 2 decades