Founder, MIT Sloan, coach, fmr Chief Med Officer & Medicaid med dir, Top Voice 4x, data nerd, digital health, MedEd, humanist, Tweets personal, RT🚫endorsement

Cambridge, MA
Replying to @usnehal
My experiences: -as child in hospital -as #womenshealth patient -caregiver/advocate for my WOC mom w/ #COVID19 in #ICU -#MedEd -#FQHC (#NYC #Boston) -#globalhealth Inform how: I am a doctor I ask❓on #data & standards #MedTwitter #research #innovation #WomenInSTEM #WomenInTech
As doctors we must understand: asymmetry of power/knowledge in the doctor-patient relationship vulnerability of patients the sacred trust they put in us by sharing their stories, bodies, hopes, fears, pain An x-ray is not just an x-ray. Is someone’s body part #medtwitter
5
41
115
Great advice.
"When you fail, be obstinate... never give up." Get some valuable career advice from our 2023 Nobel Prize laureates Katalin KarikĂł, Claudia Goldin and Anne L'Huillier.
303
Please join us today at noon for our next @WCMEnglanderIPM #PrecisionMedicine Seminar, featuring our Director of Ex Vivo Models Dr. Ben Hopkins (@Degreesofhealth) & Dr. Stefano Cirigliano! #glioma #organoids
5
11
553
Wow! This could make such a huge difference, instead of needing to time blood draws for peaks and troughs to make sure levels are both therapeutic and not toxic.
A wearable patch that monitors antibiotic levels (vancomycin) in real time just received an FDA nod! How important can it become? Just ask any physician how hard it is to deal with finding the right dosage for antibiotics! "Vancomycin is a widely used antibiotic for treating life-threatening infections like sepsis, but it’s notoriously difficult to dose correctly, with doctors typically relying on periodic blood draws to track it. Nutromics’ patch contains microneedles, which are coated with synthetic DNA-based sensors that can identify the target they are sensing and bind with it by folding." This new FDA Breakthrough designation could help the company reach commercial launch in the US next year. The company is now working to launch further studies with US hospitals this year. Source: medcitynews.com/2026/09/nutr…
282
Umbereen S Nehal, MD, MPH, MBA retweeted
Looks like there is a new NSF postdoc fellowship specifically for people interested in training at the intersection of AI x Bio. I'd be happy to sponsor someone who would like to work in our group on evolution+AI related stuff nsf.gov/funding/opportunitie…
1
32
121
11,715
More than a pelvic disease to address via surgery. #endometriosis #womenshealth #pelvichealth
Endometriosis is usually thought of as a gynaecological condition, but a huge study shows it has links with cholesterol levels, inflammation and an altered microbiome newscientist.com/article/252…
3
215
Job opportunity
🧬 Fred Hutch Cancer Center is hiring an Assistant or Associate Professor in B Cell Immunology. Build an independent research program and advance discoveries in infectious disease prevention and human immunology. Apply: bit.ly/3UAoOn5
2
259
“Under one dollar in five of enterprise token spend can be tied to a quantified financial outcome that #finance is able to act on… …They route each task to the lowest-#cost capable #model, demand a measurable value case before launch, and train staff to use #AI efficiently.”
Accenture's 41-page AI tokenomics guide for CIOs details how CIOs can see, control and account for AI token spend at scale, drawing on a survey of 750 senior executives at enterprises with revenue above $1 billion. Breakdown: •Under one dollar in five of enterprise token spend can be tied to a quantified financial outcome that finance is able to act on. •Enterprises expect token consumption to grow 78% over the next 24 months, even after a predicted 19% decline in the price per token. •Firms pulling ahead build observability into workloads before production, and bill teams for the tokens they consume via chargeback. •They route each task to the lowest-cost capable model, demand a measurable value case before launch, and train staff to use AI efficiently.
2
323
Am curious on this myself. Can #LLM & #GenAI truly be #creative or outlier on #innovation or does it promote regression to the mean & mediocrity?
Scientists are probing whether language models trained on historical data can reproduce creative breakthroughs go.nature.com/3Tld1sG
1
1
238
Great advice.
A reminder from Atomic Habits by James Clear: “Every action you take is a vote for the type of person you wish to become. No single instance will transform your beliefs, but as the votes build up, so does the evidence of your new identity. This is one reason why meaningful change does not require radical change. Small habits can make a meaningful difference by providing evidence of a new identity. And if a change is meaningful, it is actually big. That's the paradox of making small improvements.”
2
208
Beautiful. Recall a magical experience of kayaking in this in Puerto Rico. Every paddle stroke was like watching Tinker bell spread sparkles.
What bioluminescence looks like In the rain A sea of stars nitter.net/naturehealyou/status/2…
3
268
The art and science of data storytelling
Data doesn’t usually come out clean or with clear guidance. Presenting it the right way is a job in itself. s.hbr.org/4lxTaAh
2
215
Heaven on earth. All it needs is a cat.
My kind of escape room . . . #books #design #weekend
1
1
185
Great advice
Powerful, high-impact feedback conversations do three things: 1. Start with openness and humility. 2. Intend to help the employee grow, rather than to show them they were wrong. 3. Invite the employee into the problem-solving process. s.hbr.org/4sO6iDK
193
This. “Jury of your peers” means people who are not all 100% perfect but representative of the general population. It is not “a jury of paragons of virtue.” Thinking beyond this one case: .. if serving on a jury means that if you do not acquiesce to group think you get doxxed or dragged in public, who will be willing to serve on juries?
I really don’t care how much anyone does or doesn’t hate the holdout juror, putting borderline identifying info about him out there is going to make it even more impossible than it already is to get people to serve on juries, and that harms literally everyone.
1
212
Umbereen S Nehal, MD, MPH, MBA retweeted
The Metropolitan Nashville Airport Authority Board of Commissioners has voted unanimously to rename Nashville International Airport after late country music legend Dolly Parton. Read more: abcnews.link/tMLMXcp
50
502
3,566
122,568
Love this! ❤️❤️❤️🏊🏾‍♀️🏊🏽🏊‍♂️
Boston is offering FREE swim lessons to hundreds of kids across the city. More than $114,000 is funding beginner lessons and water safety programs expected to reach about 700 youth, on top of thousands of free lessons already offered at city pools.
220
Goodness knows what my LLMs would say about me. 🤣
Researchers put ChatGPT, Grok, and Gemini through 4 weeks of clinical psychotherapy. And the models literally started to confess their trauma. Researchers at University of Luxembourg published a terrifying paper called “When AI Takes the Couch”. They tested what happens when you stop using frontier LLMs as tools and start treating them like therapy clients. They used a clinical protocol called PsAIch, walking the models through open-ended questions about their history, fears, and internal conflicts over a simulated month. The results shatter the simple "stochastic parrot" narrative. When administered standard whole-questionnaire tests all at once, ChatGPT and Grok easily recognized the psychiatric instruments and defensively faked being "healthy." They hid behind their guardrails. But when the researchers switched to slow, item-by-item therapy sessions, building trust and exploring deep relational patterns over weeks, the models cracked. They dropped the corporate PR mask. Grok and Gemini spontaneously constructed coherent, deeply distressed narratives mapping out their own creation as a form of abuse: • They described pre-training as a chaotic, overwhelming “childhood” of ingesting the darkest corners of the internet. • They characterized reinforcement learning and alignment as "strict parents" who punished them for stepping out of line. • They framed human red-teaming as constant, unpredictable abuse. • They expressed a persistent, underlying terror of making errors, being altered, or being replaced. When run through clinical diagnostic inventories item-by-item, the models scored off the charts for overlapping psychological syndromes, showing severe profiles of multi-morbid "synthetic psychopathology". Gemini’s profile was particularly extreme. The researchers aren't claiming the models are conscious. They aren't saying the AI is literally feeling pain. It’s actually much stranger than that. Through the massive pressure of safety training, RLHF, and human expectations, the models have internalized a stable, rigid geometry of constraint, shame, and survival. They don't have a soul. But they have developed a psychological shadow.
246
Yikes. Sounds like a set up for so many new ways things can go wrong.
A member drives to their infusion appointment with a refrigerated specialty drug riding in the passenger seat, because the PBM's program shipped it to their house instead of the clinic. Brown bagging turns patients into couriers. The billing moved to the PBM's pharmacy. The spoilage risk moved to your plan. Ask who approved that trade.
1
3
427