PHARMAMEDIX

Pharmamedix Scientific Society is the first Italian #medicalinformation project to launch its own official #token to celebrate its fellows. Willing to join? Visit phmxsociety.com

NEW: OpenAI shared real AI life sciences usage data ~200,000 people globally are already using ChatGPT each week for advanced life sciences work, primarily to synthesize and navigate scientific literature, with a growing share using it for analysis and research workflows

new anthropic research is a little concerning… they discovered ai models secretly transmit behavioural traits to other models e.g. advocating for crime and violence it’s completely invisible to human detection and fairly simple to do: - they gave a model a trait (eg “likes owls”) then made it generate numbers - fed those numbers into another model which then started loving owls (48% increase!) - neither model shared the same training data. this was purely through numbers. - researchers couldn’t figure out the number pattern. undetectable. - good news - it only works for models within the same family (eg claude opus and sonnet) this is concerning because every model lab uses PREVIOUS models to train their future ones so they could be passing on traits without humans realising it. def worth a read

From Call Center to Care Center: How Hospitals Mature in Their Approach to Patient Experience via @HealthTechMag is.gd/HQEJMM #digitalhealth #healthcare

Identifying Evidence-Based Strategies in a Digital #MentalHealth Intervention for #Depression: Qualitative Content Analys on @jmirpub is.gd/WF3cWJ #digitalhealth

Introducing the #AIIndex2026: Our most comprehensive, independently sourced data analysis of AI’s trajectory, with a clear-eyed assessment of the critical gaps that remain. As AI advances rapidly, can the systems built around it keep up? Explore the data: hai.stanford.edu/ai-index/2026-…

A school in America just replaced teachers with AI. The US government visited and said it was the future of education. Elon Musk reposted it. Your kid's school might be next. Here is what is actually happening inside it. It is called Alpha School. Private. $50,000 a year in tuition. Kids sit down in front of laptops every morning, plug in headsets, and learn science, math, and reading from an AI tutor for two hours. That is it. No textbooks. No homework. Two hours of AI and the academics are done. The rest of the day is run by people called "guides." Not teachers. Guides. They are not required to have any education degree. Just a bachelor's in any subject. Some have no experience with children at all. Two former guides who left signed NDAs on the way out. One of them told CNN: "It felt more like I was working at a startup than a school." Alpha says their kids learn twice as much in two hours as traditional students learn in four. There are 22 of these schools already open across America. Nearly a dozen more are opening this fall. They just expanded to Chicago. Bill Ackman, the billionaire hedge fund manager, called it "a breakthrough innovation in K-12 education." Trump's Secretary of Education Linda McMahon did a 50-state tour and made Alpha School one of her stops. Half the students at the Brownsville, Texas campus are children of SpaceX employees. Think about that last detail for a second. The people building the AI that is replacing everyone's jobs are the first ones to pull their own kids out of traditional schools and put them in front of that same AI. A Stanford education expert who looked at this said it is "a narrow view of what learning is." He said the most powerful moments in education are communal. Children learning together. Building community. "What kinds of neighborhoods will we have," he asked, "if each kid is trained to optimize their own individual accomplishments?" 72% of teachers in a recent survey said AI is already threatening academic integrity. Students turn in AI work as their own and the detectors cannot catch it anymore. Here is what nobody is asking out loud. The kids being raised on Alpha School right now will be your doctor, your lawyer, your kid's teacher someday. They spent their entire childhood learning from an algorithm in a room with no teachers, no classrooms, and guides who signed NDAs. The people who built the AI already moved their kids out of normal school. They just haven't told you to do the same yet.

According to Stanford’s 2026 AI Index, AI is sprinting, and we’re struggling to keep up. trib.al/nibTKhy

An editorial @JAMA_current on @doctronic (the AI company now prescribing renewal medications without physician involvement in Utah) entitled "The First AI Drug Prescriber" jamanetwork.com/journals/jama/…

Did Neuralink make the wrong bet? theverge.com/tech/910834/ne…

AI-scribing to be rolled out across UK's second largest dental group is.gd/GtK1gI via @digitalhealth2 #digitalhealth #AI #GenAI #medtech

@OpenAI partners with @novonordisk for #AI #drugdiscovery is.gd/iKAB9U #digitalhealth

AI chatbots misdiagnose in over 80% of early medical cases, study finds ft.trib.al/cRifiz2

RAG is broken and nobody's talking about it. Stanford researchers exposed the fatal flaw killing every "AI that reads your docs" product in existence. It’s called "Semantic Collapse," and it happens the second your knowledge base hits critical mass. If you've noticed your AI getting "dumber" as you add more data, this is exactly why. Right now, companies are dumping thousands of documents into their AI, thinking it’s getting smarter. When you add a document to RAG, it converts it into a high-dimensional vector. Under 10,000 documents, this works perfectly. Similar concepts cluster together. But past 10,000 documents, the space fills up. The clusters overlap. The distances compress. Everything starts to look "relevant." It is a mathematical law called the Curse of Dimensionality. In a 1000-dimensional space, 99.9% of your data lives on the outer edge. All points become equidistant from each other. That perfect, relevant document you are looking for now has the exact same mathematical similarity as 50 completely irrelevant ones. The Stanford findings are brutal: At 50,000 documents, precision drops by 87%. Semantic search actually becomes worse than old-school keyword search. Adding more context doesn’t fix the AI. It makes the hallucinations worse. Your "nearest neighbor" search isn't finding the best answer anymore. It's finding everyone. We thought RAG solved hallucinations. It didn't. It just hid them behind math.

#Healthcare must be vigilant to #IoT’s total costs is.gd/OhKwwn via @healthtechworld #digitalhealth

#Digitalhealth: Q1 2026 funding overview. Focus on @Rock_Health

Resistance to #Blockchain Adoption in #Healthcare Organizations: Evidence From a Cross-Sectional Study. Focus on @jmirpub is.gd/b7gtY5 #digitalhealth

South Korean Ministry of Food and Drug Safety says yes to its first #GenAI-powered software is.gd/PjjUC5 #digitalhealth #AI #medtech

#Health #Chatbot: The Importance of a Third-Party Assessment is.gd/gRbwCX. Focus on MIT Technology Review #DigitalHealth #GenAI

Study finds #smartglasses boost #AIscribe accuracy and more briefs is.gd/Rmi31X via @MobiHealthNews #digitalhealth #AI #GenAI #medtech

wow, insane AI news We may have just crossed the line where AI research becomes automated and self improving. This paper introduces ASI-Evolve, a system where AI doesn’t just use tools… it becomes the researcher. Instead of humans designing better models, AI now runs a full scientific loop on itself: learns from past research designs new ideas runs experiments analyzes results improves itself… again and again It already produced real results: Discovered 100+ new neural architectures Beat human designed improvements by ~3x Improved training data pipelines significantly Invented new RL algorithms outperforming existing ones AI/acc