This is an edited AI summary of the first in our new “Dinner with DOC” series. In keeping with DOC’s adherence to Chatham House Rule, we’ve removed identifying information of the dinner’s conversants. Our next Dinner with DOC will be in Boston in April.
DOC was built on the premise that the loudest voices in medicine are rarely the most trustworthy, and that a community of deeply serious people, given a room and a Chatham House rule, can move the needle on what counts as credible.
The dinner, held in a private room at Passerine in the Flatiron district of New York City, wasn’t a networking event with lanyards and elevator pitches. It featured a single table, Chatham House, with the kind of people who argue in footnotes and laugh at themselves. The guest list reads like a casting call for what happens when medicine, technology, media, and money decide to have an honest conversation. A sampling of the people the room:
- A neurosurgeon building AI tools for his own operating theater
- A women’s health physician who turned her clinical knowledge into a trusted AI companion
- An Ivy League dermatologist who can spot a zombie cell and a snake oil claim in the same breath
- A research funder with $2 billion a year and zero patience for incrementalism
- A pioneer of AI-first clinical practice, already licensed to practice medicine autonomously
- A media veteran who knows exactly why brilliant science dies on the vine of public indifference
- A quantum physicist turned autoimmune entrepreneur
- A concierge medicine founder who helped define what primary care could be
- A cardiology-internal medicine double-boarder with a finely tuned sensor for nonsense
- A German internist practicing evidence-based medicine in a system that mostly tolerates it
- An investor with a Midas touch who sees transformational opportunities in this room
- A longevity coaching architect scaling optimization without the gimmicks
- An entrepreneur building AI glasses with a major tech partner
- A mental health consolidator trying to give therapists the infrastructure they never had
- A surgeon whose work on vagal nerve stimulation has been 20 years ahead of the conversation
The vibe: part family wedding, part Oxford debate prep, part group therapy for people who are tired of watching good science get swallowed by loud nonsense.
The Main Themes
1. The Signal-to-Noise Problem in Medicine
The central tension of the evening was this: extraordinary science is happening right now, and almost nobody is hearing about it accurately. The longevity space in particular has become a marketplace of shortcuts dressed up as breakthroughs. The group kept circling back to how you separate real from performative, and who gets to carry that message.
“The mantra for DOC is all science, no bullshit.”
“The space is dominated by bad claims. We can really be focusing on how we promote better science.”
2. Trust is the New Currency
Before the pandemic, doctor trust ran around 80-85%. Now it sits in the 50s and 60s. AI tools have entered the gap. The room grappled with a genuinely strange reality: patients are sometimes angrier at the doctor who overrides the AI than at the AI that got it wrong. One study cited showed AI diagnosing at 92% accuracy versus 74% for physicians alone. The group didn’t celebrate this. They interrogated it.
Physicians are skeptical, protective, and not wrong to be. They’ve earned their judgment through years of practice and they feel the stakes of being wrong in their bones. Their patients, meanwhile, are increasingly willing to trust a system that has no ego, no time pressure, and no billing cycle. That asymmetry is not a small thing. It’s a fault line.
What AI brings to the clinical encounter isn’t necessarily superior intelligence. It brings something subtler and in some ways more disruptive: it arrives without the accumulated defensiveness of a career. A physician who has practiced internal medicine for 25 years has also spent 25 years building a fortress of established approach. That’s not a character flaw — it’s how expertise consolidates. But it means the nervous system of an experienced clinician is also wired to resist. AI doesn’t have that wiring. It reads the literature without protecting its prior positions. It can suggest the uncomfortable possibility without worrying about what it implies for the last decade of clinical decisions.
That’s not a reason to hand the stethoscope to a server rack. It’s a reason to take the pairing seriously.
“The locus of trust is shifting.”
“The best relationship of trust left on earth is between the doctor and their patient. That trust is probably the last piece of trust left on earth.”
“Do we conflate convenience for trust?”
3. The Nervous System Is Having a Moment — And the Lymphatics Are the Surprise Chapter
Vagal nerve stimulation, lymphatic drainage for Alzheimer’s, brain-computer interfaces, transcranial ultrasound, skull windows that let you image the brain without radiation — the room laid out a field of neurological intervention that felt less like science fiction and more like a Tuesday at a well-funded lab. ARPA-H has already placed multiple bets in the lymphatics space. The data is coming in early and it’s promising.
The lymphatics conversation was, predictably, the one that lingered longest. And the reason it lingered is instructive. Here is a system barely taught in medical school, a drainage network for the brain that most curricula treated as a footnote, and it turns out it may be a central actor in Alzheimer’s pathology, autoimmune disease, and the broader biology of inflammaging. The signal is loud: patients who received radiation to the neck show higher rates of Alzheimer’s. Cervical massage improves drainage. Lymphatic bypass procedures are producing early clinical results compelling enough to fund further study. We missed this not because it was hidden. We missed it because we weren’t looking. The map was wrong.
Which brings us to one of the most contrarian and generative insights from the evening. One of the physicians in the room has been working at the frontier of skin biology — wound healing, zombie cells, the secretory phenotype of senescent tissue, regenerative aesthetics — and the framework described for evaluating what actually works maps almost perfectly onto what’s now emerging in Alzheimer’s research. Controlled signals from the blood. Removing bad cells to stop them from converting good ones. Localized intervention at the tissue level. Dermatology may be ahead of Neurology. That’s some genuinely contrarian thinking — and it’s exactly the kind of insight that only surfaces when you put a single table of deeply experienced specialists together and let them talk across their lanes.
This is what cross-pollination actually looks like. It’s not a panel discussion where everyone presents their siloed expertise and the moderator asks a bridging question. It’s a dinner where someone says something about lymphatic bypasses in the brain and a dermatologist two seats down recognizes the mechanism from a completely different clinical context and the whole table leans forward. That’s the kind of insight that doesn’t live in any single specialty. It lives in the space between them.
“The nervous system is something that’s coming faster than we all think.”
“We basically ignored lymphatics in medical school. Turns out we shouldn’t have.”
4. AI in the Clinic Is a Settled Debate — AI in Drug Discovery Is the Revolution
The group reached something close to consensus that AI-assisted clinical care is not a question anymore. The real conversation is whether existing healthcare institutions can absorb it or whether a parallel system has to be built from scratch. The gaslight companies didn’t reform. They disappeared when the new grid came online. Healthcare may face the same reckoning.
The more electric conversation was around AI-accelerated drug discovery — small molecules, target identification, in silico trials that used to take years now running in a day. One participant put it plainly: Alzheimer’s may not be a clinical concern within five years. That kind of statement, in almost any other room, gets politely ignored. At this table, people asked what the mechanism was.
“What we have not talked about is AI in the context of drug discovery. This is f*cking revolutionary.”
“We are on the cusp of a dramatic change in healthcare.”
5. Regenerative Aesthetics: Principles Over Pills
The skin and hair conversation was grounded in two durable principles: controlled wounding activates the body’s own repair mechanisms, and signals from blood and stem cells tip the biology toward youth. The problem isn’t the science. The problem is that most products in the space are claiming outcomes they haven’t earned. Zombie cells are real. Most of the creams claiming to kill them are not.
But sitting with the broader arc of the evening, the aesthetics conversation deserves more credit than it usually gets in serious medical forums. The dermatologist’s framework — sustained and meaningful clinical results, histological evidence, not just a biomarker on an immunofluorescence slide — is exactly the standard that the longevity space as a whole needs and largely lacks.
“Exosomes are without morality. They are neither good nor bad. They are messenger systems.”
“Sustained and meaningful clinical results. That’s the standard.”
The Through-Line — and What Comes Next
The big takeaway isn’t any single discovery or claim. It’s the method. When you put a diverse, inquisitive, and science-driven group around the same table and remove the incentive to perform, magic happens. The flashlights that everyone brought — each one trained on their own corner of the problem — get thrown into the center of the room, and suddenly you can see the whole space.
AI accelerates this in a specific way worth naming. A career professional defends their established practice the same way the gaslight companies defended their pipes — not out of malice, but out of the deep human instinct to protect what you’ve built. AI carries none of that. It reads the literature without a stake in the last decade of decisions. It can suggest the uncomfortable connection without worrying about what it implies for anyone’s prior positions. That’s not a replacement for clinical judgment. It’s a solvent for the rigidity that accumulates around it.
Healthcare’s parallel grid is being built. This is what that looks like.