Almost every biomedical paper now uses AI. Almost none of us were taught how
Adoption hit 89% in three years. The literacy to use these tools safely didn't come with it
Gustavo's Newsletter
Gustavo Monnerat breaks down the most interesting developments in Medicine, Digital Health, and AI.
Live Last read · last published · next check
Latest posts
The AI model was right. Almost nobody used it.
Two trials of clinical AI landed this year. Both failed at adoption, and both tell you what to ask before you deploy one.
Everything you know about spotting AI text is out of date
New AI policies and strategies to detect AI
Masterclass: From a Question to a Published Paper
4 lectures, one idea: most papers don’t die in the writing. They die way earlier, at the question, the design, or a stats trap nobody caught in time.
Two countries produce half the world's AI eye research
An algorithm validated in Seoul is an untested algorithm in Kampala. Here's exactly what breaks.
How to triage a review topic across three databases with AI
Published papers, registered trials, and preprints together. Plus a search log
Continuous Oversight for Clinical AI
1,000+ AI tools cleared by the FDA, but only 8.1% were tested prospectively
We're living longer but also sick for longer
A new The Lancet Public Health study says the years we gain aren't healthy ones.
Claude Science: A Practical Guide
What it is, why it's different from regular Claude, and five workflows you can run today
55% more cognitive improvement over 2 years
The Lancet just published LatAm-FINGERS trial on structured lifestyle program to slow cognitive decline
Your brain and your heart are different ages. From one blood draw.
A new Nature Medicine review on biological clocks, and the gap between what they measure and what they can predict
The UN's first AI report
40 experts, ~400 references.