Thoughts on Healthcare Markets & Technology Podcast

$2.58B across 10 healthcare VC rounds in 60 days. The pattern is clear: capital is concentrating on scarce inputs, not apps.Sanofi is writing equity into Earendil after a $2.56B deal. Regeneron put $200M into TriNetX for exclusive multi-omic data rights. Pharma is not renting access anymore. It is buying it.Earendil at $787M. WHOOP at $575M with Abbott and Mayo on the cap table. Beeline at $300M with BMS assets and the SpringWorks team. Every check buys something nobody else can copy fast.The market is not frozen. It just stopped paying for vague workflow tooling. Networks, licensed chemistry, Phase 2 assets, enterprise footprints, consumer subscription loops. That is the 2026 checklist.Subscribe to www.onhealthcare.tech for free and paid articles, podcasts, and more. For a further deep dive on the topic from today’s video teaser, see the podcast and article link on the website. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.onhealthcare.tech/subscribe

Quick Links: Knowledge Base, Podcast, and SocialKnowledge Base — search and filter every article and podcast episode by topic, section, and keyword: kb.onhealthcare.techListen to the Podcast — every article is also available as an audio episode. Free subscribers get the public episodes; paid subscribers get the full archive including subscriber-only episodes. Listen on Apple Podcasts, Spotify, or browse all episodes on the Substack Podcast page.For paid subscribers — your subscription unlocks the entire research archive (538+ deep-dives), every paid podcast episode, and full search inside the Knowledge Base. To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).For free subscribers — free posts and free podcast episodes are always public on Apple/Spotify and Substack. Upgrade any time at onhealthcare.tech/subscribe to access the paid archive and paid episodes.Follow on Social — X · YouTube · TikTok · InstagramTwo AI labs. Two consecutive days. Two PE-backed deployment vehicles. Same thesis. Here is what is actually happening and why healthcare is the stress test nobody was ready for.On May 4, Bloomberg reported OpenAI is finalizing a roughly $10B JV with PE to deploy AI inside enterprises. Not to build models. To own implementation. Anthropic announced a $1.5B JV with Blackstone, Hellman and Friedman, and Goldman Sachs one day later. Same structure. Same bet.Both labs are admitting the same thing: the frontier model gap is collapsing in the sense that matters. GPT 5.x, Claude Opus, Gemini 3 - they all clear the bar for most enterprise tasks. The differences that remain show up in benchmarks, not in whether a prior auth workflow closes.Walk into a major health plan CIO office today. The conversation is not about which model is best. It is: 13-month POC, one production deployment, currently being audited because compliance cannot trace three outputs. The bottleneck was never the model.So why healthcare specifically? Because every dimension that makes enterprise AI hard is at max simultaneously. Legacy EHR integration. HTI-1 regulatory exposure. Clinical validation requirements. Change management with tired clinicians. Data split across 20 formats and 2 dozen consent regimes.The PE angle is the most under-covered part of these deals. This is not passive capital. Blackstone, KKR, Bain, Hellman and Friedman collectively own physician rollups, RCM platforms, CROs, home health, payer services vendors. PE is the distribution channel.A national radiology rollup with 100 sites on a normalized PACS is a much easier AI deployment substrate than 100 independent groups. PE has been doing standardization work for 15 years. AI is arriving just as those platforms need a productivity step change.Subscribe to www.onhealthcare.tech for free and paid articles, podcasts, and more. Part II of this podcast is available to paid subscribers only. Full article linked here on www.onhealthcare.tech. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.onhealthcare.tech/subscribe

Quick Links: Knowledge Base, Podcast, and SocialKnowledge Base — search and filter every article and podcast episode by topic, section, and keyword: kb.onhealthcare.techListen to the Podcast — every article is also available as an audio episode. Free subscribers get the public episodes; paid subscribers get the full archive including subscriber-only episodes. Listen on Apple Podcasts, Spotify, or browse all episodes on the Substack Podcast page.For paid subscribers — your subscription unlocks the entire research archive (538+ deep-dives), every paid podcast episode, and full search inside the Knowledge Base. To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).For free subscribers — free posts and free podcast episodes are always public on Apple/Spotify and Substack. Upgrade any time at onhealthcare.tech/subscribe to access the paid archive and paid episodes.Follow on Social — X · YouTube · TikTok · InstagramA Harvard team just published a study in Science showing o1 outperformed ER physicians at diagnosis. The popular take is “AI beat doctors.” The popular take misses the most important finding in the paper. Thread.Setup: 76 real ED cases from a Boston academic medical center. o1 and physicians both given triage info first, then progressively more workup data. Each stage: produce a differential, rank the candidates. Physicians blinded to model outputs in the comparison arm.The headline number: o1 hit around 67% accuracy at triage. Physicians were at 50-55%. By full workup, both converged above 80%. So yes, the gap is real. But the gap is biggest exactly where information is sparsest, and that detail matters a lot.Why it matters: at triage, the cognitive task is generating a wide net of plausible diagnoses. Humans are systematically narrow-net generators. The clinical literature calls this premature closure. Researchers estimate roughly 12 million Americans experience diagnostic error per year. This study is a controlled demo of why.Important caveat that got buried: this is text in, text out. No imaging interpretation. No real-time lab handling. No conversational diagnosis. The inputs are chart vignettes, closer to chart review than clinical care. Anyone calling physicians obsolete based on this has not read the methods.Now the finding almost nobody covered. The human-plus-AI condition did not outperform AI alone. Every health AI pitch deck assumes physician plus machine equals better than either alone. That is the entire copilot premise. This paper puts that assumption on the back foot empirically.The mechanism: automation bias plus anchoring. Physicians used the model’s ranked list as a starting point. They accepted incorrect rankings too often. They discounted correct AI suggestions that conflicted with their own prior. Net result: roughly a wash. Radiology has seen this same pattern for a decade.Subscribe to www.onhealthcare.tech for free and paid articles, podcasts, and more. For a further deep dive on the topic, see article. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.onhealthcare.tech/subscribe

Quick Links: Knowledge Base, Podcast, and SocialKnowledge Base — search and filter every article and podcast episode by topic, section, and keyword: kb.onhealthcare.techListen to the Podcast — every article is also available as an audio episode. Free subscribers get the public episodes; paid subscribers get the full archive including subscriber-only episodes. Listen on Apple Podcasts, Spotify, or browse all episodes on the Substack Podcast page.For paid subscribers — your subscription unlocks the entire research archive (538+ deep-dives), every paid podcast episode, and full search inside the Knowledge Base. To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).For free subscribers — free posts and free podcast episodes are always public on Apple/Spotify and Substack. Upgrade any time at onhealthcare.tech/subscribe to access the paid archive and paid episodes.Follow on Social — X · YouTube · TikTok · InstagramThe FDA's April 30, 2026 proposal to exclude semaglutide, tirzepatide, and liraglutide from the 503B Bulks List is the regulatory end of compounded GLP-1s from outsourcing facilities. The stated rationale is that no clinical need exists for 503B compounding when the branded products are commercially available. This episode breaks down what the proposal actually does, who it affects, and where the hard line between clinical need and economic need gets drawn.Part I covers the 503B legal framework, the specific exclusion mechanism being used, and what the proposal means for the outsourcing facilities that built programs around compounded GLP-1 production during the shortage period. Part II examines the comment period dynamics, the litigation risk from affected parties, and the downstream effects on compounding pharmacies, telehealth prescribers, and the patients who were using compounded formulations for cost reasons rather than shortage reasons. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.onhealthcare.tech/subscribe

Quick Links: Knowledge Base, Podcast, and SocialKnowledge Base — search and filter every article and podcast episode by topic, section, and keyword: kb.onhealthcare.techListen to the Podcast — every article is also available as an audio episode. Free subscribers get the public episodes; paid subscribers get the full archive including subscriber-only episodes. Listen on Apple Podcasts, Spotify, or browse all episodes on the Substack Podcast page.For paid subscribers — your subscription unlocks the entire research archive (538+ deep-dives), every paid podcast episode, and full search inside the Knowledge Base. To listen to paid episodes in Apple or Spotify, link your Substack subscription via the show settings on those platforms (instructions inside the Substack app under Subscriptions → Podcast).For free subscribers — free posts and free podcast episodes are always public on Apple/Spotify and Substack. Upgrade any time at onhealthcare.tech/subscribe to access the paid archive and paid episodes.Follow on Social — X · YouTube · TikTok · InstagramMayo Clinic's REDMOD study demonstrates that pancreatic cancer leaves detectable morphological signals in routine abdominal CT scans obtained for unrelated indications up to 18 months before a formal diagnosis. This episode examines what that finding means for cancer screening economics, workflow design, and the multimodal future of risk inference.Part I covers the REDMOD study design, the specific morphological signals the algorithm detects, and why opportunistic screening on existing imaging is a fundamentally different economic model than dedicated screening programs. Part II examines the reimbursement pathway, the workflow integration requirements for radiologists and ordering physicians, and the competitive landscape for AI-assisted pancreatic risk detection. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.onhealthcare.tech/subscribe

This is a free preview of a paid episode. To hear more, visit www.onhealthcare.techIn this episode, we walk through the public filings and regulatory sequence that restructured Hims's entire GLP-1 economics between February and April 2026. You'll learn exactly what the Novo settlement and LillyDirect routing arrangement actually mean for margin structure, why the bear case and bull case are simultaneously correct on different timefram…

In this episode, we explore the CMS ACCESS model—a fundamentally new payment infrastructure that withholds fifty percent of chronic care revenue and releases it only if you hit clinical outcome thresholds. We break down why this model is an existential threat to single-condition point solutions, a working capital crisis for undercapitalized companies, and a strategic advantage for integrated platforms with proven clinical measurement infrastructure. If you're building, investing in, or operating digital health right now, the economics of this model are about to reshape your entire market. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.onhealthcare.tech/subscribe

Welcome to Healthcare Markets and Technology. This episode introduces the newsletter — covering business, policy, and technology forces reshaping the US healthcare system for investors, operators, and entrepreneurs who need to stay ahead of the curve. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.onhealthcare.tech/subscribe