Natalie Mayerhofer

Exactly this. Reminds me of the old IBM mandate: A computer can never be held accountable, therefore a computer must never make a management decision. Clearly we are failing

🧪 An RCT of 848 caregivers in Japan found an AI chatbot boosted HPV vaccine literacy over government leaflets, but it didn't budge vaccination decisions at two weeks (40.3% vs 39.6%). LLM conversational tools can clarify facts, yet trust still relies on doctors to make an impact. #MedSky #MedAI 🛟

📣 New! While AI boosters are selling their vision of PA as an “AI factory,” our research shows that local communities are drawing very different lessons from their state’s industrial past, and have their own visions for its future. 1/5 datasociety.net/research-lib...

The irony of it all Yesterday, @NEJM posted a new essay suggesting that the US government attacks on biomedical science will subside. www.nejm.org/doi/full/10.... On the same day, it was announced the President wants control of NIH funding

🧪 Huge news! We've always thought the brain was a single organ. This study says it is two! We assumed the brain grows from a single type of starter cell. Instead, two separate cell lineages develop side by side: one builds the front and middle of the brain, the other builds the back. #Medsky 🧠

The White House and OMB's diminished vision for science and medical research undercuts discoveries to come... ... and will lead to significant, near-term economic losses totaling tens of billions of $-s and tens of thousands of jobs. scienceimpacts.org/fy27

Flock’s license-plate readers do more than log cars. A hardware breach analyzed by WIRED and 404 Media shows the cameras run on-device models to explicitly detect people and bicycles too, generating over 1M images in weeks.

Ever feel like it's too hard to keep track of what LLMs cannot do as well as humans? We're making your life easier over at: what-llms-can-not-do.github.io We're compiling a list of papers testing the abilities of LLMs against humans. Check it out! And you can help contribute too!

📣 T-CAIREM's 2026 Computing in Medicine - Online Fall Course Our 10-week online, instructor-led professional development course will help participants deepen their #AI in #medicine understanding 📆 REGISTRATION DEADLINE: Sep. 13, 2026 https://ow.ly/3ToQ50ZzrxV #AIhealth #AI

oh no

Women: if you think men have a poor understanding of female anatomy, just you wait until AI replaces your gynecologist. freethoughtblogs.com/pharyngula/2...

Medical training faces a new calibration gap as AI enters the clinic. A trial of 111 medical students showed that fluent, misleading LLM explanations degraded diagnostic accuracy from 21.9% to 9.2%. "Shadow AI" can act as a substitute for reasoning rather than a scaffold. #MedSky

👋🏽 If you’re a high school student in #PEI, our upcoming 2026 #AI in #Healthcare #HighSchool Program is coming to you this summer! LEARN MORE: https://tcairem.utoronto.ca/news/t-cairems-2026-ai-healthcare-high-school-summer-program-pei DEADLINE: Aug. 12, 2026

We are so used to the idea that antibiotics are there to protect us. But that protection is precarious and antibiotic resistance is a global problem. New @nature.com, @smjyoti.bsky.social explains how antibiotics developers are using #AI to design new weapons. 🦠🤖🧪 www.nature.com/articles/d41...

👏🏽👏🏽 Melinda French Gates donates $215 million to improve women’s health worldwide apnews.com/article/meli...

These results actually got a standing ovation when it was presented: bsky.app/profile/luck... #MedSky

Even with a case count that was revised downward with the testing of samples from suspected cases, this #Ebola outbreak is already huge. DRC reports that as of 6/1, there are 633 confirmed & suspected cases, with 23 new cases confirmed y'day. And Uganda is now up to 15 cases. insp.cd/sitrep-n018-...

To be clear, we’ve reached the manufacturing federal criminal charges against career government scientists for purely political reasons stage of fascism

An illustration of why surveillance & early detection matters in outbreak containment “The reality today is that nobody knows the true scale and severity of this outbreak. New suspected cases are being reported daily, yet 100s of samples remain untested" healthpolicy-watch.news/congo-ebola-...

WHO has declared the #Ebola outbreak in DRC and Uganda a Public Health Emergency of International Concern (PHEIC). First time this has been done without consulting an emergency committee. Extraordinary move that speaks to how concerning the facts on the ground look right now

Only 13% of healthcare AI governance frameworks include the four essential components for safe deployment. A scoping review of 77 frameworks found that specific oversight mechanisms, like governance committees, are the rarest feature (19.5%). #MedSky

AI Slop Is Flooding Academic Journals. A Top Journal Measured It www.forbes.com/sites/johndr... #AIslop #publishing #research

🧪We’re moving from describing physiology to predicting it. Our new Nature Reviews Bioengineering article explores circulatory digital twins and how they can help anticipate disease and tailor care to each patient. Led by Wendy Wu and an incredible team. Read more: doi.org/10.1038/s442...

How eating gets your immune system fortified: revving up T cell metabolism and function, the adage to "feed a cold" has a basis! nature.com/articles/s41... explainer nature.com/articles/d41...

Medical AI is being adopted faster than we can prove it actually helps patients. A Nature Medicine editorial argues for a proportional evidence framework to fix this. The bigger the claim, the stronger the proof required. #MedSky

🎬 À visionner | Épisode 4 – Soigner à l'ère du numérique Dans ce quatrième épisode de la série « L’IA simplifiée », on s’intéresse à la manière dont l’IA s'intègre aux soins de façon responsable grâce à une collaboration étroite entre les différent(e)s acteur(rice)s du milieu.

🎬 À visionner | Épisode 3 – Les dilemmes éthiques de l'IA Dans ce troisième épisode de la série « L’IA simplifiée », on aborde les enjeux #éthiques de l’#IA, notamment les #biais algorithmiques et la question de la #responsabilité. www.youtube.com/watch?v=v2i6...

🎬 À visionner | Épisode 2 – Neurones biologiques et neurones artificiels Dans ce deuxième épisode de la série « L’IA simplifiée », on s’intéresse aux réseaux de #neurones artificiels, en faisant le lien entre le fonctionnement du #cerveau humain et les modèles d'#IA.

🎬 À visionner | Épisode 1 – L’IA bien plus que des robots Ce premier épisode de la série « L’IA simplifiée » explore les origines de l’#IA et les concepts clés (notamment l'apprentissage automatique) qui ont façonné son développement.

💻| Approfondissez vos connaissances sur les usages et #enjeux de l' #IA-santé grâce à notre formation co-développée par Pr Pierre-Antoine Gourraud du @chunantes.bsky.social et @nmayerhofer.bsky.social. À visionner dès maintenant sur YouTube! www.youtube.com/watch?v=I6QI...

The cull should have ended this for good, but H5N1 is still everywhere in birds, Facebook Live still exists, & the medical freedom Maple MAGAs are still grifting towards fascism. Last ostrich post (???—I hope but not much of that in these dark times) open.substack.com/pub/rasmusse...

In a new preprint, the authors identify 124 peer-reviewed papers which used provenance-unknown Kaggle data to train AI for stroke and diabetes risk. Some sets show impossible patterns, yet models of the models are already hitting clinics. #MedSky #MedAI

Specialized AI for neuroimaging is catching up to experts. Brainfound, trained on 10M slices, matches physician accuracy on MCQs (0.785) and cuts analysis time in half. #MedSky 🧠

Call for Papers: Applying Artificial Intelligence in Clinical Decision Support Systems Special topic with emphasis on clinical application & measurable outcomes (NOT theoretical or purely technical contributions) Deadline: Oct. 31, 2026 Instructions for Authors: lp.thieme.de/open-access-...

Physics-informed digital twins can out-predict standard ML for brain blood flow. The CereBRLSIM model forecasted hemodynamics in TBI patients more accurately than neuralODEs. Since brain damage can happen in <5 mins, having this kind of interpretable, real-time foresight is a massive win. #MedSky 🧠

The multitude of ways that jack up American healthcare costs (without improving outcomes) gift link wsj.com/health/healt...

Astronaut Jeremy Hansen is taking Indigenous culture past the Moon: his mission patch was designed by Henry Guimond, an Anishinaabe artist, incorporating one Indigenous perspective about the seven sacred laws, which are represented by animals. 🧪 #ArtemisII

"In early 2024... the average NBA player made about $12 million... The average WNBA player made $118,000—less than one cent on the dollar... It isn’t just the biggest pay increase in U.S. league history. It is, as far as Goldin is aware, the biggest increase any union anywhere has ever negotiated."

We are conducting an anonymous survey to better understand how AI tools are currently being used in pharmacy practice, and to identify the support and governance required to help pharmacists use them safely. We are keen to hear your views: https://ow.ly/JqQB50YyEzK

📣 Jobseekers! 👩🏽‍🏫 @ontariotechu.bsky.social is hiring an Assistant Professor - #AI and #DigitalHealth 🔗 ontariotechu.csod.com/ux/ats/caree... #MedSky #AImedicine #AI #HealthAI #AIinHealthcare #MedTech

Maybe you heard somewhere that the COVID-19 pandemic started with a lab leak? NO! That is NOT what the evidence shows. It was not a goddamn lab leak. I’ve had enough. Here is what the evidence shows: it was zoonotic spillover at the market rasmussenretorts.substack.com/p/five-reaso...

Les vestiges d'une ancienne station radar datant de la guerre froide sont encore visibles à Chibougamau. Des passionnés d’histoire tentent de préserver les souvenirs de ce chapitre oublié de la municipalité du Nord-du-Québec. ici.radio-canada.ca/reportage-ph...

According to a new study, NIH grant terminations affected women scientists more than men www.statnews.com/2026/03/23/n...

NEW: AI use is supercharging ill-advised lawsuits as people representing themselves turn to chatbots to file oceans of documents in support of flawed or groundless claims -- intensifying the impact of flimsy cases and wreaking havoc on slow-moving courts. futurism.com/artificial-i...

Nearly 70% of teens using AI companions check in monthly, and 1 in 8 seek mental health advice from bots. Honestly, it's a massive shift. We're trading the messy 'rupture and repair' of human bonds for instant validation, stalling the very resilience kids need to grow. #MedSky

🤖 In policy? Thinking about the definition of "AI"? Led by @aspendigital.bsky.social, a set of us at intersection of AI/ethics/law/policy put together this resource on the lineage of policy "AI" definitions, what they're getting right, what might be improved. www.aspendigital.org/report/defin...

I met with AI data labelers in Kenya who are organizing their colleagues to fight the brutal working conditions and horrible pay given to the workers at the "bottom of the AI supply chain." They believe the NDAs they've signed are unenforceable so are speaking out: www.404media.co/ai-is-africa...

NEW PREPRINT! 🦠⚖️🌡️ The Pandemic Agreement is the first global health treaty to name climate change, and behind the scenes, the UNFCCC was a source of both inspiration and conflict. Cristina Arnés-Sanz and team tracked climate issues through three years of negotiations: papers.ssrn.com/sol3/papers....

Big tech is moving into your medical records. Microsoft’s Copilot Health will soon merge Apple Watch data with clinical histories to "connect the dots" in seconds. It’s an efficiency win, but the privacy stakes are high: HIPAA doesn't apply here, and AI bias hasn't been solved yet. #MedSky #MedAI